VSEBINA – CONTENTS
IZVIRNI ZNANSTVENI ^LANKI – ORIGINAL SCIENTIFIC ARTICLES
Experimental verifications and numerical thermal simulations of automobile lamps
Eksperimentalna preverjanja in numeri~ne toplotne simulacije avtomobilskih `arometov
M. Guzej, J. Horsky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Tensile and compressive tests of textile composites and results analysis
Natezni in tla~ni preizkusi tekstilnih kompozitov in analiza rezultatov
K. Kunc, T. Kroupa, R. Zem~ík, J. Krystek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Deformation behaviour of a natural-shaped bone scaffold
Obna{anje naravno oblikovanega ogrodja kosti pri deformaciji
D. Kytýø, T. Doktor, O. Jirou{ek, T. Fíla, P. Koudelka, P. Zlámal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Printed microstrip line-fed patch antenna on a high-dielectric material for C-band applications
Tiskana mikrotrakasta linijsko napajana krpasta antena na visoko dielektri~nem materialu za uporabo v C-pasu
Md. M. Islam, M. R. I. Faruque, M. F. Mansor, M. T. Islam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Compressive properties of auxetic structures produced with direct 3D printing
Stiskanje struktur materialov z negativnim Poissonovim razmerjem, proizvedenih z neposrednim tridimenzionalnim tiskanjem
P. Koudelka, O. Jirou{ek, T. Fíla, T. Doktor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Model of progressive failure for composite materials using the 3D Puck failure criterion
Model postopnega popu{~anja kompozitnega materiala z uporabo Puckovega tridimenzionalnega kriterija poru{itve
L. Bek, R. Zem~ík . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Physicochemical properties of a Ti67 alloy after EO and steam sterilization
Fizikalno kemijske lastnosti zlitine Ti67 po EO in parni sterilizaciji
W. Walke, M. Basiaga, Z. Paszenda, J. Marciniak, P. Karasinski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Surface properties of a laser-treated biopolymer
Lastnosti povr{ine biopolimera, obdelanega z laserjem
I. Michaljani~ova, P. Slepi~ka, S. Rimpelova, P. Sajdl, V. [vor~ík . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Analyzing the heat-treatment effect on the mechanical properties of free-cutting steels
Analiza vpliva toplotne obdelave na mehanske lastnosti avtomatnih jekel
M. K. Kulekci, U. Esme, F. Kahraman, R. Ozgun, A. Akkurt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
Analysis of the cutting temperature and surface roughness during the orthogonal machining of AISI 4140 alloy steel via the
Taguchi method
Analiza temperature rezanja in hrapavosti povr{ine s Taguchi metodo pri ortogonalni strojni obdelavi legiranega jekla AISI 4140
A. R. Motorcu, Y. Isik, A. Kus, M. C. Cakir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Weldability of Ti6Al4V to AISI 2205 with a nickel interlayer using friction welding
Preizku{anje varivosti pri varjenju s trenjem Ti6Al4V in AISI 2205 z vmesno plastjo niklja
I. Kirik . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Effect of activated flux and nitrogen addition on the bead geometry of borated stainless-steel GTA welds
Vpliv aktiviranega topila in dodatka du{ika na geometrijo kopeli pri GTA zvarih boriranega nerjavnega jekla
G. R. Kumar, G. D. J. Ram, S. R. Koteswara Rao . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Microstructural evolution during the transient liquid-phase bonding of dissimilar nickel-based superalloys of IN738LC and
NIMONIC 75
Razvoj mikrostrukture med spajanjem s prehodno teko~o fazo neenakih superzlitin na osnovi niklja IN738LC in NIMONIC 75
M. G. Khakian, S. Nategh, S. Mirdamadi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
Workability behaviour of Cu-TiB2 powder-metallurgy preforms during cold upsetting
Preoblikovalnost Cu-TiB2 predoblik izdelanih z metalurgijo prahov med hladnim kovalnim preizkusom
S. Gadakary, A. Kumar Khanra, M. J. Davidson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
Effects of extrusion shear on the microstructures and a fracture analysis of a magnesium alloy in the homogenized state
Vplivi stri`enja med iztiskanjem homogenizirane magnezijeve zlitine na mikrostrukturo in na analizo preloma
H. J. Hu, Z. Sun, D. F. Zhang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
ISSN 1580-2949
UDK 669+666+678+53 MTAEC9, 50(3)287–466(2016)
MATER.
TEHNOL.
LETNIK
VOLUME 50
[TEV.
NO. 3
STR.
P. 287–466
LJUBLJANA
SLOVENIJA
MAY–JUNE
2016
FSW welding of Al-Mg alloy plates with increased edge roughness using square pin tools of various shoulder geometries
FSW varjenje plo{~ iz Al-Mg zlitine s pove~ano hrapavostjo robov z orodjem s kvadratno konico in razli~no geometrijo bokov
S. Balos, L. Sidjanin, M. Dramicanin, D. Labus Zlatanovic, A. Antic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Improvement of selective copper extraction from a heat-treated chalcopyrite concentrate with atmospheric sulphuric-acid
leaching
Izbolj{anje selektivne ekstrakcije bakra iz toplotno obdelanega koncentrata halkopirita z lu`enjem z `vepleno kislino na zraku
E. Uzun, M. Zengin, Ý. Atýlgan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Homogenization of an Al-Mg alloy and alligatoring failure: alloy ductility and fracture
Homogenizacija Al-Mg zlitine in krokodiljenje: duktilnost zlitine in prelom
E. Romhanji, T. Radeti}, M. Popovi}. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
STROKOVNI ^LANKI – PROFESSIONAL ARTICLES
Assessment of tubular light guides with respect to building physics
Ocena cevastih vodnikov svetlobe glede na gradbeno fiziko
F. Vajkay, D. Be~kovský, V. Tichomirov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Creep behaviour of a short-fibre C/PPS composite
Vedenje kratkih vlaken C/PPS kompozitov pri lezenju
T. Fíla, P. Koudelka, D. Kytýø, J. Hos, J. [leichrt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Increasing micro-purity and determining the effects of the production with and without vacuum refining on the qualitative
parameters of forged-steel pieces with a high aluminium content
Pove~anje mikro~isto~e in dolo~itev u~inka proizvodnje, z vakuumskim rafiniranjem ali brez, na kvalitativne parametre kovanega jekla z
visoko vsebnostjo aluminija
V. Kurka, J. Pindor, J. Kosòovská, Z. Adolf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Use of the ABI technique to measure the mechanical properties of aluminium alloys: effect of heat-treatment conditions on the
mechanical properties of alloys
Uporaba ABI tehnike za merjenje mehanskih lastnosti aluminijevih zlitin: vpliv pogojev toplotne obdelave na mehanske lastnosti zlitin
O. Trudonoshyn, M. Puchnin, O. Prach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Investigation of the effect of holding time and melt stirring on the grain refinement of an A206 alloy
Preiskava vpliva ~asa zadr`evanja in me{anja taline na zmanj{anje velikosti zrn zlitine A206
N. Akar, Z. Tanyel, K. Kocatepe, R. Kayikci . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
Investigating the influence of cutting speed on the tool life of a cutting insert while cutting DIN 1.4301 steel
Preiskava vpliva hitrosti rezanja na zdr`ljivost vlo`ka za rezanje pri rezanju jekla DIN 1.4301
R. Dubovská, J. Majerík, R. ^ep, K. Kouøil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
NiAl intermetallic prepared with reactive sintering and subsequent powder-metallurgical plasma-sintering compaction
Reakcijsko sintranje in zgo{~evanje s plazemskim sintranjem NiAl intermetalne zlitine
A. Michalcová, D. Vojtìch, T. F. Kubatík, P. Novák, P. Dvoøák, P. Svobodová, I. Marek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Microscopic characterization and particle distribution in a cast steel matrix composite
Mikroskopska karakterizacija in razporeditev delcev v kompozitu z matrico litega jekla
A. Kra~un, M. Torkar, J. Burja, B. Podgornik . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
A comparison of as-welded and simulated heat affected zone (HAZ) microstructures
Primerjava mikrostrukture toplotno vplivanega podro~ja varjenega in simuliranih vzorcev
R. Celin, J. Burja, G. Kosec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Degradation of an AISI 304 stainless-steel tank
Degradacija rezervoarja iz AISI 304 nerjavnega jekla
M. Torkar, I. Paulin, B. Podgornik. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
288 Materiali in tehnologije / Materials and technology 50 (2016) 3,
M. GUZEJ, J. HORSKY: EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL SIMULATIONS ...
289–293
EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL
SIMULATIONS OF AUTOMOBILE LAMPS
EKSPERIMENTALNA PREVERJANJA IN NUMERI^NE TOPLOTNE
SIMULACIJE AVTOMOBILSKIH @AROMETOV
Michal Guzej, Jaroslav Horsky
Brno University of Technology, Faculty of Mechanical Engineering, Technicka 2, 616 69 Brno, Czech Republic
guzej@lptap.fme.vutbr.cz
Prejem rokopisa – received: 2014-07-31; sprejem za objavo – accepted for publication: 2015-05-21
doi:10.17222/mit.2014.149
Over the last decade lamps became a design feature of a car body and to preserve their primary purpose they are much more
complex than in the past. Today’s car lamps contain powerful light sources, which utilize new technologies such as
dot-concentrated light (LED chips or xenon discharge tubes). One of the unfavourable properties of these products is a high
amount of thermal energy produced in a small area, causing a high thermal stress on the components. For this reason, it is
important to know the temperature in critical locations to avoid damaging the lamp body or causing a defect in the light source
itself. This paper presents the results of a numerical simulation of the working conditions of a lamp, the measurement inside the
environmental-simulation chamber with a real lamp, providing the same conditions and the verification of the simulation results
with the measurement, which confirmed the precision of the numerical simulation. The verification procedure was developed at
the Heat Transfer and Fluid Flow Laboratory of the BUT in partnership with [koda Auto a.s.
Keywords: automobile lamps, thermocouples, numerical simulation, thermal measurements, verification
Zadnje desetletje so avtomobilski `arometi postali del oblikovanja avtomobilske karoserije in za ohranitev prvotne vloge so, v
primerjavi s preteklostjo, postali bolj kompleksni. Dana{nji avtomobilski `arometi imajo mo~an izvor svetlobe, ki uporablja
nove tehnologije, kot je to~kasto skoncentrirana svetloba (LED ~ip ali ksenonska razelektritvena cev). Ena od ne`eljenih
lastnosti teh proizvodov je velika koli~ina toplotne energije, ki nastane na majhnem podro~ju in zaradi tega pride do ve~je
termi~ne napetosti v komponentah. Zato je pomembno poznati temperature na kriti~nih lokacijah, da se prepre~i po{kodba
ohi{ja `arometa ali da nastane napaka na samem izvoru svetlobe. ^lanek predstavlja rezultate numeri~ne simulacije delovnih
pogojev `arometov, meritve v notranjosti numeri~no simuliranega okolja realnega `arometa z realno `arnico in preverjanje
rezultatov simulacije z meritvami, ki so potrdile natan~nost numeri~ne simulacije. Postopek verifikacije je bil razvit v
Laboratoriju za prenos toplote in tok fluidov na BUT, s sodelovanjem [koda Auto a.s.
Klju~ne besede: avtomobilski `arometi, termoelementi, numeri~na simulacija, toplotne meritve, preverjanje
1 INTRODUCTION
Modern automobile lamps have a complex body
design and contain powerful built-in sources of light.
This can create problems for the materials used for their
construction.1 Manufacturers and consumers, in this case
[koda Auto a.s., need to know the temperature generated
at critical locations and compare it with the temperature
limit for each plastic material. Every material has a
long-term temperature limit and a short-term temperature
limit.2 Electrical losses of dot-concentrated light are
converted into thermal energy. The result of this effect is
a temperature gradient, which produces significant ther-
mal stresses within the structure.3 Solving this problem
with an analytical method is a formidable task. The few
solutions available cover only simple structural shapes.
The only option is to use simulation software such as
ANSYS, COMSOL, etc. Verifying the results of a com-
puter simulation is always necessary when high precision
is required. Experimentation under the same conditions
as in the simulation determines the accuracy of the
results or detects possible errors in the settings of the
simulation model. Wire thermocouples built into control
points are used for the temperature measurement.
Deviations between the simulation and reality are deter-
mined from the heat-settled part of the experiment.
2 EXPERIMENTAL PART
To numerically simulate an automotive lamp (Fig-
ure 1), the ambient temperature during the experiment
must be known. A lamp placed inside an environmental
simulation chamber (type: MK 720 E3.1) manufactured
by Binder was used for this purpose. This device can
maintain the required temperature by forced convection
in a very small range.
2.1 Description of the measurement circuit
Data collection during the experiments consists of
five major parts (Figure 2). First, data is measured
through the thermocouples. Then, information in milli-
volts is transferred into a 7001 switch system manu-
factured by Keithley, an American company. This model
is a half-rack, high-density, two-slot mainframe with a
maximum of 80 signal channels (in this experiment, 12
Materiali in tehnologije / Materials and technology 50 (2016) 3, 289–293 289
UDK 621.9.08:620.179.13:656.057.87 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)289(2016)
wire thermocouples were used). The built-in, non-vola-
tile memory stores up to 100 complete switch patterns.
Then, the signal from the switcher is transferred through
a GPIB cable into a Model 2000 multimeter made by the
same company. Model 2000 is a high-performance digi-
tal multimeter, which can measure DC voltage, AC volt-
age, DC current, AC current, two- and four-wire resist-
ance, and temperature. A thermocouple card is placed at
a built-in cold junction. The final step is to save the data
as a .txt file.
2.2 Assembly of the thermocouples
Measurement points (the locations of the thermo-
couples) were chosen based on the simulation results
from two cross-sections, A and B (Figure 3). At these
points, 1-mm-diameter holes were drilled to a depth of
1 mm. The junction of a wire thermocouple was em-
bedded into to this hole and filled with a high-tempe-
rature-resistant epoxy (Figure 4a). The operational
temperature range of this epoxy is from –55 °C to
180 °C, and its thermal properties are similar to PP
GF30, the plastic material used for the body of the lamp.
If the hole for the thermocouple is made into a trans-
parent material (such as the glass of a lamp), an epoxy
with the same optical properties must be used due to
transmittance (Figure 4b). Using a material with
different properties has a negative effect on the tempe-
rature measurement near the built-in thermocouple. A
K-type thermocouple was used to record the data. This
type is the most common general-purpose thermocouple,
with a sensitivity of approximately 41 μV/°C.4 K-type
thermocouples are recommended for use in oxidizing or
completely inert atmospheres in a temperature range of
–200 °C to 1260 °C. The positive wire consists of 90 %
nickel and 10 % chromium, while the negative wire is
consists of 95 % nickel and 5 % aluminum and silicon.4
290 Materiali in tehnologije / Materials and technology 50 (2016) 3, 289–293
M. GUZEJ, J. HORSKY: EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL SIMULATIONS ...
Figure 3: Top view of cross-sections A and B. Cross-section A is
situated in the axis of the bulb filament, cross-section B is offset by 30
mm from section A, in the Y-axis direction.
Slika 3: Pogled od zgoraj na preseka A in B. Presek A je v osi `arilne
nitke `arnice, presek B je zamaknjen za 30 mm od preseka A, v smeri
Y-osi
Table 1: Thermal properties of lamp materials
Tabela 1: Toplotne lastnosti materialov `arometa
Materials Density(kg m–3)
Thermal
conductivity
(W m–1 K–1)
Specific heat
(J kg–1· K–1)
Body PP GF30 1066 0.3 1351
Glass PMMA 1194 0.33 1288
Figure 2: Data collection in the measurement circuit: 1) environ-
mental simulation chamber with a lamp, 2) switcher, 3) GPIB cable,
4) multimeter and 5) computer
Slika 2: Zbiranje podatkov v merilnem tokokrogu: 1) simulacijska ko-
mora okolja z `arnico, 2) preklopnik, 3) GPIB-kabel, 4) multimeter in
5) ra~unalnik
Figure 1: Lamp with thermocouples, front view
Slika 1: @aromet s termoelementi, pogled od spredaj
2.3 Calibration
Each thermocouple had to be calibrated before the
first measurement to obtain accurate results. This is due
to the fact that thermocouples do not always have exactly
the same characteristics, even if they meet fairly strict
standards and tolerances. Calibration represents a
deviation between the thermocouple measurement and a
precise etalon. For a K-type thermocouple and a tempe-
rature range from –40 °C to 375 °C, the tolerance is
±1.5 °C (the highest class of accuracy). All the thermo-
couples had a deviation under the tolerance value. The
tolerance of thermoelectric voltage U is determined with
the following Equation (1):5
Δ ΔU T
dU
dTT
T= ⋅
⎛
⎝
⎜ ⎞
⎠
⎟ (1)
where UT is the thermoelectric voltage tolerance, T
is the temperature difference (°C), T is the temperature
(°C), and UT is the thermoelectric voltage (V).
3 LAMP SIMULATION
Predictions of the temperatures of the lamps were
computed with computational fluid dynamics (CFD).
The CFD process consists of three important steps:
pre-processing, solving, and post-processing.
3.1 Pre-processing
The geometry of the lamp was imported from raw
CAD data. The next step was to create a computation
mesh. The CutCell meshing method (Figure 5) in
ANSYS TGrid 14.5 was used for this purpose. This
method was chosen because its preparation time is
shorter than the times needed for conventional
approaches. The size of the mesh was about 13 million
polyhedral cells and the average size of these cells was 1
mm.
3.2 Solving
The resulting mesh was exported into the ANSYS
Fluent 14.5 solver, where appropriate boundary condi-
tions were set up. These boundary conditions included
the type of fluid flow (laminar), the material properties,
the emissivity of the surfaces, the heat source, and,
eventually, the forced convection around the lamp
(simulating an environment as similar as possible to the
climate chamber).
The distribution of temperatures inside the lamp
depended mainly on natural convection (incompressible
ideal gas) and radiation of the heat from the bulb. The
heat absorbed by the surfaces was transported through
the materials by heat conduction. The DO (discrete ordi-
nate) model was used to model the radiation, allowing a
calculation for semi-transparent walls.
The DO model uses a discretization technique, which
solves the radiation-transfer equations for a finite num-
ber of angles or directions. The angular discretization is
controlled by either increasing or decreasing the amount
of theta-phi divisions and pixels.6 The number of divi-
sions and pixels should therefore be kept at a minimum
in order to keep computational costs as low as possible,
but the minimum amount will also give the coarsest
discretization. The settings vary between a minimum of
3 × 3 × 3 × 3, which is the coarsest, to 10 × 10 × 10 × 10,
which is the maximum and gives the highest accuracy
and finest discretization. Since the maximum discretiza-
tion requires a lot of computational time and data
storage, the angular discretization used in the radiation
model was determined to be 6 × 6 × 6 × 6, which was
sufficient. The effect of solar radiation could be specified
but was not used for this case.
The solver solves the mass, momentum, energy and
radiative-transport equations throughout the whole
domain. The equations are discretized along the cells of
the domain and solved until convergence is reached.
M. GUZEJ, J. HORSKY: EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL SIMULATIONS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 289–293 291
Figure 5: Example of the CutCell meshing method
Slika 5: Primer mre`e pri CutCell metodi
Figure 4: Thermocouple attachment: a) a normal surface and b) a
transparent surface
Slika 4: Pritrditev termoelementa: a) je normalna povr{ina, b) pro-
zorna povr{ina
3.3 Post-processing
The results from the solution were visualized with
plotting parameters of interest for every part, in our case
focusing mainly on the temperatures. Post-processing
was performed in ANSYS CFD-Post.
4 MEASURED DATA
The following two diagrams (Figures 6 and 7) show
recorded temperatures of the lamp during the experiment
for an ambient temperature of 70 °C, a bulb supply
voltage of 13.2 V and a sampling frequency of 5 s.
5 DISCUSSION
Figures 8 to 10 show a comparison between the
numerical and experimental results (the temperature
distributions were plotted for two cross-sectional points,
A and B), including a bar graph showing the variation
between the measurement and the simulation.
M. GUZEJ, J. HORSKY: EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL SIMULATIONS ...
292 Materiali in tehnologije / Materials and technology 50 (2016) 3, 289–293
Figure 6: Results for cross-section A
Slika 6: Rezultati iz preseka A
Figure 9: Temperatures for cross-section B
Slika 9: Temperature na preseku B
Figure 7: Results for cross-section B
Slika 7: Rezultati iz preseka B
Figure 8: Temperatures for cross-section A
Slika 8: Temperature na preseku A
Figure 10: Comparison of the temperatures between the measurement
and simulation
Slika 10: Primerjava izmerjenih in simuliranih temperatur
The numbers at the top of the columns are percent
deviations between the measured and simulated values
for all the positions of the thermocouples.
6 CONCLUSIONS
This paper demonstrated the importance of a compa-
rison between an experiment and a numerical simulation
due to the unknown boundary conditions at the walls of
real lamps. Determining an acceptable accuracy is
necessary to verify the lamp model. The greatest devi-
ations in the data are found for the T5A thermocouples
(the simulated value is 4 % above the measured value)
and the T4B thermocouples (the simulated value is 7.7 %
above the measured value). Possible reasons for these
inaccuracies are slightly different positions of the drilled
holes and the control point from the ones used in the
simulation, and the air flow around the lamp during the
experiment, which had a higher (or lower) speed than the
value calculated with the simulation.
Acknowledgement
This work is an output of the research and scientific
activities of the NETME Centre, regional R&D centre
built with the financial support from the Operational
Programme Research and Development for Innovations,
carried out within a NETME Centre project (New
Technologies for Mechanical Engineering), Reg. No.
CZ.1.05/2.1.00/01.0002 and, in the follow-up sustain-
ability stage, supported through NETME CENTRE
PLUS (LO1202) with the funds from the Ministry of
Education, Youth and Sports under the National Sustain-
ability Programme I.
7 REFERENCES
1 V. A. Drebushchak, Universality of the emf of thermocouples, Ther-
mochimica Acta, 496 (2009) 1–2, 50–53, doi:10.1016/j.tca.2009.06.
025
2 T. Shiozawa, Thermal air flow analysis of an automotive headlamp:
the PIV measurement and the CFD calculation for a mass production
model, JSAE Review, 22 (2001) 2, 245–252, doi:10.1016/S0389-
4304(00)00111-9
3 K. Sokmen, K. Furkan, E. Pulat, N. Yamankaradeniz, S. Coskun,
Thermal computations of temperature distribution and bulb heat
transfer in an automobile headlamp, Heat and Mass Transfer, 50
(2014) 2, 199–210, doi:10.1007/s00231-013-1229-5
4 Thermocouples, Sponsored by ASTM Committee E-20 on Tempe-
rature Measurement and Subcommittee E20.04 on Thermo-
couples, Manual on the use of thermocouples in temperature
measurement, 6th ed., American Society for Testing and Materials,
Philadelphia 1992
5 J. H. Zhao, B. Zhou, Thermocouple Automatic Verifying Device
Based on Virtual Instrument Technology, Applied Mechanics and
Materials, 303–306 (2013), 588–596, doi:10.4028/www.scienti-
fic.net/AMM.303-306.588
6 ANSYS Fluent Theory Guide, release 14.0., ANSYS Inc., 2011
M. GUZEJ, J. HORSKY: EXPERIMENTAL VERIFICATIONS AND NUMERICAL THERMAL SIMULATIONS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 289–293 293

K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS
295–299
TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES
AND RESULTS ANALYSIS
NATEZNI IN TLA^NI PREIZKUSI TEKSTILNIH KOMPOZITOV IN
ANALIZA REZULTATOV
Kry{tof Kunc, Tomá{ Kroupa, Robert Zem~ík, Jan Krystek
University of West Bohemia in Plzeò, NTIS, Univerzitní 22, 306 14, Plzeò, Czech Republic
ptaeck@kme.zcu.cz
Prejem rokopisa – received: 2014-08-01; sprejem za objavo – accepted for publication: 2015-05-05
doi:10.17222/mit.2014.167
The presented work is focused on an experimental investigation of the behavior of six types of textile composites subjected to
pure tensile, cyclic tensile and compressive loading according to ASTM standards. Each type was loaded in directions forming
angles between 0° and 90°, with a step of 15° with respect to the reference directions of the weaves. Two types of woven fabrics
were tested (plain and quasi-unidirectional plain-woven fabric). Images of specimens taken during the tests were subsequently
used for the calculation of the so-called locking angle of yarns (bundles) just before failure. Force-displacement dependencies
were recorded during the tensile tests. Ultimate forces were obtained from the compressive tests. The second half of the article
is dedicated to the analysis of the experimental data gathered with nearly 1000 experiments. Special software for automatic
calculation of averaged dependencies, maximum forces and maximum displacements was created. Furthermore, the
methodology for calculating the locking angle was proposed and tested. The obtained results will be used for the identification
of the material parameters of the proposed material model in the following research.
Keywords: textile composites, woven fabric, tensile test, compressive test, result analysis, weave locking
Prispevek je usmerjen v eksperimentalno preiskavo obna{anja {estih vrst tekstilnih kompozitov, obremenjenih z natezno,
cikli~no-natezno in tla~no obremenitvijo skladno z ASTM standardi. Vsaka vrsta je bila obremenjena v smeri, ki je tvorila kot
med 0° in 90°, s koraki po 15°, glede na smer tkanja. Preizku{eni sta bili dve vrsti tkanin (obi~ajna in kvazi enosmerna obi~ajna
tkanina). Posnetki vzorcev med preizkusi so bili uporabljeni za izra~un zapornega kota preje (sve`njev) tik pred poru{itvijo.
Odvisnosti sila-raztezek so bile posnete med nateznimi preizkusi. Kon~ne sile so bile dobljene iz tla~nih preizkusov. Naslednji
del prispevka je bil posve~en analizi eksperimentalnih podatkov iz skoraj 1000 preizkusov. Kreirana je bila posebna programska
oprema za avtomatsko ra~unanje povpre~nih odvisnosti: maksimalnih sil in maksimalnih raztezkov. Poleg tega je bila
predlagana in preizku{ena metodologija za izra~un zapornega kota. Dobljeni rezultati bodo uporabljeni pri nadaljevanju raziskav
za dolo~anje parametrov materiala v predlaganem modelu materiala.
Klju~ne besede: tekstilni kompoziti, tkanina, natezni preizkus, tla~ni preizkus, analiza rezultatov, zaklepanje vezave
1 INTRODUCTION
Textile composites made from carbon, glass and
aramid fibers are nowadays commonly used. However, to
be able to simulate the behavior of these modern mate-
rials as in the case of classical metals, it is appropriate to
use complex mathematical models with many more
material parameters. A significantly non-linear behavior
of composite materials is caused by different properties
of its components and by a complicated manufacturing
process. Sophisticatedly gathered data from many
experimental tests are, therefore, required for identifying
material parameters and designing modern tailored
composite structures.1
2 EXPERIMENTAL PART
Experimental tests were focused on three composite
types – glass, carbon and aramid. Each type was tested in
two woven-fabric versions: a) a plain weave with a 1:1
fiber ratio and b) a quasi-unidirectional plain weave with
a 1:9 fiber ratio.2
The following markings are used to describe all the
tested specimens: GP – glass plain weave, GU – glass
quasi-unidirectional weave, CP – carbon plain weave,
CU – carbon quasi-unidirectional weave, AP – aramid
plain weave and AU – aramid quasi-unidirectional
weave.
The tested specimens (coupons) were cut from six
composite plates, manufactured with the RTM techno-
logy, using a water jet to get seven different groups of
specimens with the principal material orientation  of the
weave (between 0° and 90° with a step of 15°) with
respect to each coupon’s longitudinal axis (and load
direction). Average thicknesses of the coupons are shown
in Table 1.
Table 1: Average thicknesses of composite plates
Tabela 1: Povpre~ne debeline kompozitnih plo{~
Material t (mm) Material t (mm)
GP 1.8 GU 1.8
CP 2.0 CU 1.5
AP 2.2 AU 2.0
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 295
UDK 620.172:621.315.614:67.017 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)295(2016)
To ensure the objectivity of the measurement, a
minimum of seven specimens for each material, weave
type and weave orientation was prepared.
Three types of tests were performed on a Zwick/
Roell Z050 machine – pure tensile, cyclic tensile and
compressive loading – according to ASTM standards.3
For the following experiment analysis, data from a
minimum of six identical tests were accepted. The total
number of specimens for one experiment (tensile/cyclic
tensile/compressive) is seven coupons for seven material
orientations multiplied by three different materials and
two types of woven fabrics. In total, nearly 1000 speci-
mens were tested, including the preliminary work.
Typical specimen set-ups before the tensile a) and
compressive b) tests and their dimensions are shown in
Figure 1. Cyclic tensile tests were driven by the pre-
scribed displacement with amplitude l = 1 mm. In the
finished cyclic tests, the number of hysteresis loops for
the same specimens was not always the same. This is
evident from the result graphs, where the tangents of a
few last load/unload cycles are missing.
The tensile tests for the plain-woven fabrics were less
problematic than the ones for the quasi-unidirectional
plain-woven fabrics, which had to be equipped with
aluminium pads in numerous cases to avoid a premature
destruction of a coupon by machine grips and to get
acceptable results from the experiment. All the coupons
for the compressive tests had to be prepared with alumi-
nium pads to achieve the proper attachment to the testing
machine.4 The schemes shown in Figure 2 represent
typical experimental set-ups for the tensile (a) and com-
pressive tests (b).
All the specimens subjected to the tensile tests (load-
ing) were photographed to identify the so-called locking
angle.5
The locking angle (Figure 3) is defined by two
angles, 1 and 2, representing the diversions of the
principal material directions, just before the rupture
(axes , , ), from their initial states (axes e1, e2, e3).
This phenomenon was most notable on the specimens
made of the aramid textile and least notable on the glass
textile. Similarly, for the boundary material orientations
296 Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS
Figure 3: Aramid plain-weave specimen ( = 30°) and locking-angle
interpretation
Slika 3: Vzorec aramida z navadnim valom ( = 30°) in predstavitev
zapornega kota
Figure 2: a) Tensile- and b) compressive-test schemes
Slika 2: Shemi: a) nateznega preizkusa in b) tla~nega preizkusa
Figure 1: Dimensions of specimens for: a) tensile and b) compressive
tests and the principal material orientation
Slika 1: Dimenzije vzorcev za: a) natezne in b) tla~ne preizkuse in
glavna orientacija materiala
Figure 4: Photographs of selected plain-weave specimens after the
tensile test (GP, CP, AP)
Slika 4: Posnetki izbranih vzorcev z navadnim valom po nateznem
preizkusu (GP, CP, AP)
of  = 0° and  = 90° and quasi-unidirectional plain-
weave composites made of carbon and glass, no
significant weave-locking phenomenon was proven. Car-
bon composites with plain weave showed a high strength
during the tests and never ruptured completely (ex-
cluding the boundary orientations). Photographs of
selected specimens taken after the tensile test are shown
on Figure 4 (with plane weave) and Figure 5 (with
quasi-unidirectional plain weave).
A further description of implementing the measured
data into the identification process and the details of the
material model can be found in another paper of the
above co-authors.6
3 ANALYSIS OF THE EXPERIMENTAL DATA
Standalone application ploTra was written in the
Python programming language for the processing of a
large amount of experimental data. The application is
designed to read experimental data from the Zwick/Roell
software (in the TRA format) and execute multiple ope-
rations resulting in the following outputs:
Averaged force-displacement dependencies (dark
curves in the presented graphs). The application accepts
data from one sorted set of measurements and calculates
the average using one of the various available methods,
e.g., 2D averaging, averaging in a given interval, or the
arc length. The results are saved as graphs (PNG/PDF)
and binary files for future usage.
Averaged tangents (slopes) of unload/load cycles
(straight lines in the presented graphs). All the hysteresis
loops from the cyclic tensile tests are identified; their
lowest and highest points are connected to form lines,
the tangents of which are averaged for each experiment
group including the specimens with the same orientation.
It was observed that these tangents were not constant
during the tests.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 297
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS
Figure 7: Averaged force-displacement dependencies for GU compo-
site
Slika 7: Povpre~na odvisnost sila-raztezek pri GU kompozitu
Figure 5: Photographs of selected quasi-unidirectional plain-weave
specimens after the tensile test (GU, CU, AU)
Slika 5: Posnetki izbranih vzorcev s kvazi-enosmerno, obi~ajno
vezavo po nateznem preizkusu
Figure 6: Averaged force-displacement dependencies and tangent of
unload/load cycle for GP composite
Slika 6: Povpre~na odvisnost sila-raztezek in tangenta na neobreme-
njen/obremenjen cikel pri GP kompozitu
Figure 8: Averaged force-displacement dependencies and tangent of
unload/load cycle for CP composite
Slika 8: Povpre~je odvisnosti sila-raztezek in tangenta cikla neobre-
menjeno/obremenjeno pri CP kompozitu
Tables with averaged statistical data – the maximum
forces and displacements including standard deviations.
4 RESULTS
The tables and graphs below represent the outputs
from the ploTra application. The maximum tensile and
ultimate forces of the compressive tests are shown in
Tables 2 and 3. Averaged force-displacement dependen-
cies (Figures 6 to 11) are suitable for the use of the com-
plex material model including damage. The weave-
locking angles shown in Tables 4 and 5 were collected
using the common tools available in graphics editors.
Table 2: Tensile tests – average maximum forces in (kN)
Tabela 2: Natezni preizkusi – povpre~je maksimalnih sil (kN)
 GP GU CP CU AP AU
0° 9.50 18.45 20.20 39.71 16.01 29.50
15° 5.68 5.63 9.33 3.66 8.25 7.03
30° 4.33 2.78 5.96 1.80 7.33 3.66
45° 3.63 1.91 5.28 0.99 6.63 2.73
60° 3.66 1.67 5.77 0.82 7.01 2.30
75° 4.23 1.39 9.18 0.60 8.37 2.67
90° 6.40 1.36 19.09 0.57 17.37 2.87
Table 3: Compressive tests – average ultimate forces (kN)
Tabela 3: Tla~ni preizkusi – povpre~je kon~nih sil (kN)
 GP GU CP CU AP AU
0° 8.32 13.21 12.01 11.70 4.89 5.36
15° 6.71 4.71 7.32 5.74 4.69 5.37
30° 4.03 4.25 4.91 3.75 3.77 3.99
45° 3.67 4.01 4.51 3.16 3.70 3.46
60° 3.96 3.96 4.75 2.81 4.01 3.37
75° 4.85 4.04 7.13 2.55 4.68 3.45
90° 4.20 4.05 11.6 2.54 4.70 3.52
Table 4: Averaged weave-locking angles 1 and 2 for plain-weave
composites in (°)
Tabela 4: Povpre~je kota tkanja 1 in 2 za obi~ajno tkane kompozite
v (°)
 GP CP AP
0° 0.00 0.00 0.00 0.00 0.00 0.00
15° 2.30 2.00 8.60 1.30 17.80 2.60
30° 5.00 3.50 12.80 5.80 26.70 12.30
45° 7.50 7.10 13.30 12.50 17.30 18.00
60° 3.10 3.30 8.50 20.60 9.80 22.50
75° 1.10 3.20 1.50 11.70 1.30 17.00
90° 0.00 0.00 0.00 0.00 0.00 0.00
5 CONCLUSION
All the materials showed a complex non-linear
behavior.
• Force-displacement dependencies are non-linear even
for the plain-weave material orientations of  = 0°
and  = 90°. Hardening was noticed in the case of
carbon and aramid textiles (exhibiting a convex
298 Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS
Figure 11: Averaged force-displacement dependencies and tangent of
unload/load cycle for AU composite
Slika 11: Povpre~je odvisnosti sila-raztezek in tangenta cikla neobre-
menjeno/obremenjeno pri AU kompozitu
Figure 10: Averaged force-displacement dependencies and tangent of
unload/load cycle for AP composite
Slika 10: Povpre~je odvisnosti sila-raztezek in tangenta cikla razbre-
menjeno/obremenjeno pri AP kompozitu
Figure 9: Averaged force-displacement dependencies for CU compo-
site
Slika 9: Povpre~je odvisnosti sila-raztezek pri CU kompozitu
load-displacement curve when zoomed) and soften-
ing (a concave curve) was noticed in the case of glass
textile.
• Specimens made of aramid fibers reached the highest
strength during the tests; on the other hand, glass-
fiber specimens reached the lowest strength.
• Weave-locking phenomenon has significant impacts
on the plain-weave orientations of  = 45° – plastic
behavior of the tested materials was observed.
• Unsymmetrical results for the plain-weave compo-
sites are probably caused by an imperfect technology
of manufacturing the textiles or by the preparation of
the specimens.
Acknowledgements
This publication was supported by the project
L01506 of the Czech Ministry of Education, Youth and
Sports and by the project of University of West Bohemia
SGS-2016-038.
6 REFERENCES
1 R. Böhm, W. Hufenbach, Experimentally based strategy for damage
analysis of textile-reinforced composites under static loading,
Composites Science and Technology, 70 (2010) 9, 1330–1337,
doi:10.1016/j.compscitech.2010.04.008
2 T. Kroupa, P. Janda, R. Zem~ík, Linear two scale model for
determination of mechanical properties of textile composite material,
Mater. Tehnol., 46 (2012) 2, 97–101
3 ASTM Standard D 3039 / D 3039M – 08, 2008, Standard Test Me-
thod for Tensile Properties of Polymer Matrix Composite Materials,
ASTM International, West Conshohocken, PA 19428–2959, 2003,
www.astm.org
4 ASTM Standard D 3410 / D 3410M – 03, 2008, Standard Test
Method for Compressive Properties of Polymer Matrix Composite
Materials, ASTM International, West Conshohocken, PA
19428–2959, 2003, www.astm.org
5 N. Hamila, P. Boisse, Locking in simulation of composite rein-
forcement deformations, Analysis and treatment, Composites Part A,
Applied Science and Manufacturing, 53 (2013), 109–117
doi:10.1016/j.compositesa.2013.06.001
6 T. Kroupa, K. Kunc, R. Zem~ik, T. Mandys, Non-linear finite
elements simulations of the tensile tests of textile composites, Mater.
tehnol., 49 (2015) 4, 509–515, doi:10.17222/mit.2014.117
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS

D. KYTÝØ et al.: DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE SCAFFOLD
301–305
DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE
SCAFFOLD
OBNA[ANJE NARAVNO OBLIKOVANEGA OGRODJA KOSTI PRI
DEFORMACIJI
Daniel Kytýø1,2, Tomá{ Doktor1,2, Ondøej Jirou{ek1, Tomá{ Fíla1,2,
Petr Koudelka1,2, Petr Zlámal2
1Czech Technical University in Prague, Faculty of Transportation Sciences, Department of Mechanics and Materials, Konviktská 20,
110 00 Prague 1, Czech Republic
2Institute of Theoretical and Applied Mechanics, v.v.i., Academy of Sciences of the Czech Republic, Prosecká 76, 190 00 Prague 9,
Czech Republic
kytyr@fd.cvut.cz
Prejem rokopisa – received: 2014-08-11; sprejem za objavo – accepted for publication: 2015-05-07
doi:10.17222/mit.2014.190
The study aims at mechanical testing of an artificial bone structure in the form of a scaffold for the application in the repairs of
trabecular bones after wounds or degenerative diseases. Such artificial construct has to conform to many requirements including
biocompatibility, permeability properties and bone-integration characteristics. Recently, self-degradable bone scaffolds suitable
for natural-bone-tissue ingrowth optimized with respect to mechanical properties and body-fluid flow have been considered as
an alternative to allografts and autografts. Here, an analysis of deformation behaviour of a scaffold with a morphology identical
to the natural bone is the first step in this task. In this work, the geometry and morphology of scaffold specimens produced with
direct 3D printing were based on a 3D model derived from the X-ray-computed micro-tomography measurement of a real
trabecular bone. The geometrical model was upscaled four times in order to achieve the optimum ratio between its resolution
and the resolution of the 3D printer. For its biocompatibility and self-degradability, polylactic acid was used as the printing
material. The mechanical characteristics were obtained from a series of uniaxial compression tests, with an optical evaluation of
the strain field on the surfaces of the specimens. The acquired stress-strain curves were compared with the characteristics of a
real trabecular bone obtained with time-lapse microtomography measurements, evaluated with the digital volumetric correlation
method. The results show good correspondence of the stiffness values for both the natural and artificial bone specimens.
Keywords: bone scaffold, polylactic acid, additive manufacturing, compression loading, microtomography
Namen {tudije je mehansko preizku{anje umetnega ogrodja kosti za obnovo trabekularnih kosti po po{kodbah ali degenerativnih
boleznih. Tako umetno ogrodje mora ustrezati mnogim zahtevam, kot so biokompatibilnost, prepustnost in mo`nost vra{~anja
kostnega tkiva. Samo razgradljivo ogrodje kosti, primerno za naravno vra{~anje kostnega tkiva, optimirano glede na mehanske
lastnosti in toka telesnih teko~in, je bilo prou~evano kot nadomestek za presajanje tujega ali lastnega tkiva. Prvi korak pri tej
nalogi je analiza obna{anja ogrodja, z morfologijo podobno naravni kosti. V tem delu je bila geometrija in morfologija
vzor~nega ogrodja izdelana z neposrednim tridimenzionalnim tiskanjem, na osnovi tridimenzionalnega modela, dobljenega z
meritvami s pomo~jo rentgenske tomografije realne trabekularne kosti. Geometrijski model je bil pove~an {tirikrat, da bi dobili
optimalno razmerje med njegovo resolucijo in resolucijo tridimenzionalnega tiskalnika. Za biokompatibilnost in sâmo
razgradljivost, je bila za tiskanje uporabljena polilakti~na kislina. Mehanske zna~ilnosti so bile dobljene z vrsto enoosnih tla~nih
preizkusov in z opti~no oceno napetostnega polja na povr{ini vzorcev. Pridobljene krivulje napetost-raztezek so bile primerjane
z zna~ilnostmi realne trabekularne kosti, ki so bile dobljene z zaporednimi mikrotomografskimi meritvami in ocenjene z digital-
no volumetri~no metodo korelacije. Rezultati ka`ejo dobro ujemanje togosti obeh vzorcev, tako naravne kot umetne kosti.
Klju~ne besede: ogrodje kosti, polilakti~na kislina, tridimenzionalnega tiskanje, tla~no obremenjevanje, mikrotomografija
1 INTRODUCTION
The change in the lifestyle during the past decades
(the so-called modern lifestyle characteristic caused by a
lack of sufficient physical activities coupled with an
improper type of nutrition) coupled with an increasing
life expectancy in many countries caused a significant
increase in health-care costs. Globally more than 30 % of
women and 20 % of men in elderly age suffer from bone
disorders. The increase is expected to double by 2020,1
partially also due to an increased occurrence of obesity.
Here, bone-tissue engineering and specifically designed
implants with functionally graded properties represent a
modern approach to the bone-repair process2 with the
goal to create an implant that is anatomically and
functionally compatible with the surrounding tissue.3 In
the recent orthopaedic practice, repairs of defective
bones have been most commonly carried out using
autografts and allografts. Although used for many years,
natural grafts possess several limitations and potentials
for complications due to various influences including the
donor-site morbidity, a loss of bone inductive factors,
resorption during healing, anatomical variations, etc.4 To
overcome these problems, implants in the form of an
artificial bone represent an attractive alternative with a
significant, though not yet fully exploited potential. In
order to engineer such a synthetic scaffold with optimum
characteristics, several factors and parameters have to be
taken into account.5 Among such factors, the type of the
cancellous bone, mechanical properties, the geometry6
Materiali in tehnologije / Materials and technology 50 (2016) 3, 301–305 301
UDK 620.173:620.11:67.017 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)301(2016)
and the expected transport phenomena7 are the most
important for the design of a synthetic porous bone.
Thus, with respect to the selected application (i.e., the
location within the skeletal microarchitecture)8 stiffness
and permeability properties have to be properly selected
to precisely match the properties of the natural can-
cellous bone. Identification of the trabecular microstruc-
ture and comparison of deformation behaviour of the
natural bone with the synthetic replication is the first
step in the artificial-bone-structure optimization. A com-
bination of radiological imaging methods and additive
manufacturing was used for preparing a natural-shaped
bone scaffold. A set of uniaxial compression tests were
performed to obtain the scaffold loading response repre-
sented by stress-strain diagrams. The strain calculation
was based on image-registration techniques, digital
image correlation (DIC) for optical data and digital
volume correlation (DVC) for tomographic data.
2 MATERIALS AND METHODS
2.1 Natural bone sample
Trabecular bone tissue is a biological material con-
sisting of basic structural elements called trabeculae.
Individual trabeculae are joined into a network with a
high porosity (50–90 %) naturally adapted to distribute
and dissipate the energy from external loads. For the
purpose of this study, original natural bone samples were
prepared by drilling them from the caput femoris of
42-year and 72-year-old male donors. The samples with
a diameter of 5 mm and a height of approximately 10
mm were put into an ultrasonic bath to remove the
remaining bone marrow. Formaldehyde solution was
used to avoid biological degradation and to ensure
sample preservation. The upper and bottom parts of the
samples were embedded in a 2-component epoxy resin
and after the hardening, two plan-parallel baseplates
were cut with a precise saw. This preparation procedure
minimises the risk of a rapid collapse of the bone
samples under loading due to a crack damage of the
microstructure.
2.2 Microtomography measurement
In order to acquire a geometrically precise model of
the internal trabecular structure, a custom-designed
microtomography device (Figure 1) was employed. The
irradiation of the sample was performed using a micro-
focus X-ray tube with a high-resolution transmission
target (XWT 160, X-Ray Worx, Germany). Radiograms
were acquired using a large (410 mm × 410 mm) X-ray
flat-panel scintillating detector (XRD 1622, PerkinElmer
Inc., USA) with an effective resolution of 2048 pixels ×
2048 pixels with a 200 μm pitch. A custom-designed
loading device composed of a stiff frame with a low
absorption of X-rays was designed specially for the
microtomography under loading. Force was gradually
applied on the specimens using the translation stage
(7T173-20, Standa Ltd., Lithuania) with a 2 μm tracking
accuracy and the applied force was measured using a
small-scale force transducer (U9b, HBM, Germany) with
a nominal capacity of 500 N. Before the loading proce-
dure, a sample was scanned in 720 projections (acquisi-
tion 2 × 0.5 s) with an angular step of 0.5° to obtain a
detailed geometrical model of the microstructure. Then a
time-lapse tomography (tomography under loading)
measurement was performed to obtain the information
about the spatial strain distribution during the loading.9
The specimen was incrementally loaded with 1 %
increment up to the total deformation of 6 %. After each
loading step, microtomography was performed to capture
the deformed microstructure. Acquisitions with 180 pro-
jections at 2° were used for this purpose to reduce the
computational costs of the reconstruction procedure. The
spatial strain distribution was assessed utilizing DVC,
which is an extension of the image-registration techni-
ques to three dimensions.10
Linearization of the attenuation range (a beam-
hardening correction) was applied to reduce the noise
and improve the contrast of individual projections.11 Due
to a high porosity of the samples and a small thickness of
the trabeculae (approximately 200 μm), the cone-beam
reconstruction algorithm12 was used to eliminate the
distortion of the reconstructed data caused by the
divergent nature of the X-ray beam.
2.3 Polylactic acid
Polylactic acid (C3H4O2)n (PLA) is a biodegradable
thermoplastic polyester derived from biomass and pro-
duced from the starch of various crop plants (e.g., corn,
cereals, potatoes, etc.). The material properties of PLA
are comparable with those of synthetic plastics, but they
are achieved at significantly lower energetic require-
ments.13 PLA is highly biocompatible14 and suitable for
the use in 3D printers making an attractive solution for
endoprosthesis.
D. KYTÝØ et al.: DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE SCAFFOLD
302 Materiali in tehnologije / Materials and technology 50 (2016) 3, 301–305
Figure 1: Microtomography set-up: 1) X-ray source, 2) detector, 3)
rotary stage, 4) loading device, 5) load-cell controller, 6) linear stage,
7) anti-vibration table and 8) cooling
Slika 1: Naprava za mikrotomografijo: 1) vir rentgenskih `arkov, 2)
detektor, 3) nosilec za rotiranje, 4) naprava za obremenjevanje, 5) nad-
zor pri obremenjevanju, 6) linearni nosilec, 7) antivibracijska mizica
in 8) hlajenje
2.4 PLA bulk testing
The production of the samples was carried out using
a Profi3Dmaker (Aroja, Czech Republic) 3D printing
device equipped with a 200 μm printing nozzle, operated
in the rapid additive-manufacturing mode with a slice
resolution of 250 μm. The material properties of the PLA
printer filament are not guaranteed by the producer and
the influence of the printing process on the material
characteristics is unknown; therefore, cylinder samples
with dimensions of 20 mm × 28 mm (diameter, height)
were tested in different printing modes to choose the
ideal printing set-up for the trabecular-bone replica
manufacturing. The samples were produced at 100 %, 70
% and 50 % filling levels. With all the filling levels, the
products consisted of a 500 μm shell and a core with a
defined filling amount. In the case of a reduced filling
amount, the core was created using a random-oriented
fibre meshwork with a defined overall porosity. To verify
the declared values of the filling level, the samples were
weighed with a laboratory scale.
2.5 Bone scaffold development
The reconstructed tomographic-image data were sub-
jected to the standard post-processing methods (includ-
ing thresholding and identification of the connected
components) using custom segmentation and modelling
software to obtain binary spatial-image data without
isolated fragments. Then the marching cube algorithm15
was used to extract the polygonal mesh of an isosurface
from the three-dimensional scalar-image data (as
depicted in Figure 2). Subsequently, this polygonal mesh
was carefully smoothed and decimated in an iterative
manner to obtain a surface suitable for 3D printing.
Because of the technical parameters of the available 3D
printer, the model was upscaled three times to ensure a
proper geometry of the printed replica.
2.6 Compression tests
The compression test of both bulk and porous speci-
mens was performed in a displacement-controlled
loading mode using a custom-designed uniaxial loading
device with a loading capacity of up to 2 kN. The
load-bearing frame was designed as an open cylinder
with expanded ends mounted on a stiff metallic plate
manufactured from polyamide-imide thermoplastic. The
displacement of the loading platens was controlled by
stepper motor SX17-1705 (Microcon, Czech Republic)
attached to a CPU 17A 100 harmonic drive (Harmonic
Drive, USA) with a transmission ratio of 250 : 1 leading
to an accuracy of displacement in the order of micro-
meters. High-accuracy load cell U9b (HBM, Germany)
was connected to an OM 502T (Orbit Merret, Czech
Republic) programmable indicator. The samples were
loaded up to a 20 % deformation at a constant loading
rate of 20 μm s–1. Strains were derived from optically
measured deformations, evaluated with the DIC algo-
rithm. For this purpose, images of the loaded specimens
were acquired with a high-resolution CCD camera
(Manta G-504B, AVT, Germany) attached to telecentric
zoom lens TCZR 072 (Opto Engineering, Italy). An
in-house software based on the GNU/Linux real-time
operation software and LinuxCNC open-source project
was used to control the experiments. For the DIC pro-
cedure, a custom Matlab tool16 based on the Lucas-Kana-
de algorithm17 was used.
3 RESULTS
The material properties of bulk PLA were measured
based on the compression tests of the PLA solid cylin-
drical samples at different levels of filling. Stress-strain
curves obtained from the force record and optically
measured strain are depicted in Figure 3. From the slope
of the linear parts of the stress-strain diagrams, Young’s
moduli presented in Table 1 were estimated.
D. KYTÝØ et al.: DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE SCAFFOLD
Materiali in tehnologije / Materials and technology 50 (2016) 3, 301–305 303
Figure 3: Stress-strain diagram of the PLA bulk material with
different levels of filling
Slika 3: Obremenitveni diagram PLA materiala z razli~nimi stopnjami
polnjenja
Figure 2: Model of trabecular-bone microstructure used for additive
manufacturing
Slika 2: Model mikrostrukture trabekularne kosti, uporabljen pri tri-
dimenzionalnem tiskanju
Table 1: Properties of PLA used for additive manufacturing
Tabela 1: Lastnosti PLA, ki se uporabljajo pri tridimenzionalnem
tiskanju
Sample Filling (%) Weight (g) E (GPa)
1 100 15.300 2.396
2 70 11.937 1.864
3 50 9.160 1.380
Because of a low PLA elastic modulus (compared to
the bone tissue) and a rod- or shell-like shape of the
replicated inner structure, the PLA scaffold was manu-
factured only in the full filling mode. Four identical PLA
specimens representing the natural-bone-shaped
upscaled scaffold were used for the compression test. For
the optical measurement of the deformation of the com-
plex sample surface, the data acquired from four diffe-
rent loading scenes (rotated by 90°) were evaluated. A
sample collapsed after the compression test is depicted in
Figure 4.
Deformation behaviour of the PLA models of the
trabecular structure in comparison with the stress-strain
curves of the natural bone structure is depicted in Figure
5. Good agreement between the elastic parts of both
materials was observed. The sample with a natural bone
volume fraction (bone volume/total volume = 25 %)
exhibits a stiffness similar to the real-bone structure.
Therefore, PLA is suitable for other bone-scaffold deve-
lopments based on regular cells.
4 CONCLUSION
Compression tests of the natural-shaped additive-
manufactured PLA samples of the trabecular-bone struc-
ture were performed to assess deformation behaviour of
the homogenous solid phase of the synthetic material.
The results will be used for designing an artificial bone
structure with effective mechanical properties, close to
the ones of a real, healthy bone. A low bone volume/total
volume ratio of the real-bone structure provides a suffi-
cient reserve for increasing the strength of the scaffold
by increasing its relative density. A higher cross-section
of the basic structure elements (which would still enable
the tissue ingrowth) may reduce the required resolution
of manufacturing devices. Additive manufacturing of
biodegradable complex structures is a promising way for
the bone-scaffold development. Based on these findings,
the artificial bone structure used for the replacements of
trabecular bones will be optimized with respect to
structural, mechanical and permeability properties.
Acknowledgements
The research was supported by the Grant Agency of
the Czech Technical University in Prague (grant No.
SGS15/225/OHK2/3T/16) and by institutional support
RVO: 68378297.
5 REFERENCES
1 U.S. Department of Health and Human Services, Bone Health and
Osteoporosis: A Report of the Surgeon General, Rockville 2004
2 H. L. M. Bao, E. Y. Teo, M. S. K. Chong, Y. Liu, M. Choolani, J. K.
Y. Chan, Advances in Bone Tissue Engineering, Regenerative
Medicine and Tissue Engineering, InTech, Rijeka 2013,
doi:10.5772/55916
3 H. Razi, S. Checa, K. Schaser, G. N. Duda, Shaping scaffold struc-
tures in rapid manufacturing implants: A modeling approach toward
mechano-biologically optimized configurations for large bone defect,
Journal of Biomedical Materials Research, Part B: Applied
Biomaterials, 100 (2012) 7, 1736–1745, doi:10.1002/jbm.b.32740
4 A. R. Vaccaro, K. Chiba, J. G. Heller, T. C. Patel, J. S. Thalgott, E.
Truumees, J. S. Fischgrund, M. R. Craig, S. C. Berta, J. C. Wang,
Bone grafting alternatives in spinal surgery, Spine Journal, 2 (2002)
3, 206–215, doi:10.1016/S1529-9430(02)00180-8
5 A. Gantar, L. P. Da Silva, J. M. Oliveira, A. P. Marques, V. M. Corre-
lo, S. Novak, R. L. Reis, Nanoparticulate bioactive-glass-reinforced
gellan-gum hydrogels for bone-tissue engineering, Materials Science
and Engineering C, 43 (2014), 27–36, doi:10.1016/j.msec.2014.
06.045
6 V. P. W. Shim, L. M. Yang, J. F. Liu, V. S. Lee, Characterisation of
the dynamic compressive mechanical properties of cancellous bone
from the human cervical spine, International Journal of Impact Engi-
neering, 32 (2005) 1–4, 525–540, doi:10.1016/j.ijimpeng.2005.
03.006
D. KYTÝØ et al.: DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE SCAFFOLD
304 Materiali in tehnologije / Materials and technology 50 (2016) 3, 301–305
Figure 5: Stress-strain diagrams for the PLA-additive-manufactured
and natural bone
Slika 5: Diagram napetost-raztezek za 3D tiskano in naravno kost
Figure 4: Example of a PLA scaffold collapsed after the compression
test
Slika 4: Primer poru{enega PLA kostnega ogrodja po tla~nem preiz-
kusu
7 S. S. Kohles, J. B. Roberts, M. L. Upton, C. G. Wilson, L. J.
Bonassar, A. L. Schlichting, Direct perfusion measurements of
cancellous bone anisotropic permeability, Journal of Biomechanics,
34 (2001) 9, 1197–1202, doi:10.1016/S0021-9290(01)00082-3
8 E. F. Morgan, H. H. Bayraktar, T. M. Keavenyemail, Trabecular bone
modulus-density relationships depend on anatomic site, Journal of
Biomechanics, 36 (2003) 7, 897–904, doi:10.1016/S0021-9290
(03)00071-X
9 O. Jirousek, P. Zlamal, D. Kytyr, M. Kroupa, Strain analysis of trabe-
cular bone using time-resolved X-ray microtomography, Nuclear
Instruments and Methods in Physics Research, Section A: Accele-
rators, Spectrometers, Detectors and Associated Equipment, 633
(2011) 1, 148–151, doi:10.1016/j.nima.2010.06.151
10 O. Jirousek, I. Jandejsek, D. Vavrik, Evaluation of strain field in mi-
crostructures using micro-CT and digital volume correlation, Journal
of Instrumentation, 6 (2011) 1, C01039, doi:10.1088/1748-0221/6/
01/C01039
11 D. Vavrik, J. Jakubek, Radiogram enhancement and linearization
using the beam hardening correction method, Nuclear Instruments
and Methods in Physics Research Section A: Accelerators,
Spectrometers, Detectors and Associated Equipment, 607 (2009) 1,
212–214, doi:10.1016/j.nima.2009.03.156
12 L. A. Feldkamp, L. C. Davis, J. W. Kress, Practical cone-beam
algorithm, Journal of the Optical Society of America A, 1 (1984) 6,
612–619
13 A. P. Mathew, K. Oksman, M. Sain, Mechanical properties of bio-
degradable composites from poly lactic acid (PLA) and micro-
crystalline cellulose (MCC), Journal of Applied Polymer Science, 97
(2005) 5, 2014–2025, doi:10.1002/app.21779
14 J. M. Anderson, M. S. Shive, Biodegradation and biocompatibility of
PLA and PLGA microspheres, Advanced Drug Delivery Reviews, 28
(1997) 1, 5–24, doi:10.1016/S0169-409X(97)00048-3
15 W. E. Lorensen, H. E. Cline, Marching cubes: A high resolution 3D
surface construction algorithm, ACM SIGGRAPH Computer
Graphics, 21 (1987) 1, 163–169
16 I. Jandejsek, J. Valach, D. Vavrik, Optimization and Calibration of
Digital Image Correlation method, Proc. of Experimental Stress
Analysis, Olomouc 2010, 121–126
17 B. D. Lucas, T. Kanade, An iterative image registration technique
with an application to stereo vision, Proc. of the 7th International
Joint Conference on Artificial Intelligence – Volume 2, San Fran-
cisco 1981, 674–679
D. KYTÝØ et al.: DEFORMATION BEHAVIOUR OF A NATURAL-SHAPED BONE SCAFFOLD
Materiali in tehnologije / Materials and technology 50 (2016) 3, 301–305 305

Md. M. ISLAM et al.: PRINTED MICROSTRIP LINE-FED PATCH ANTENNA ON A HIGH-DIELECTRIC MATERIAL ...
307–310
PRINTED MICROSTRIP LINE-FED PATCH ANTENNA ON A
HIGH-DIELECTRIC MATERIAL FOR C-BAND APPLICATIONS
TISKANA MIKROTRAKASTA LINIJSKO NAPAJANA KRPASTA
ANTENA NA VISOKO DIELEKTRI^NEM MATERIALU ZA
UPORABO V C-PASU
Md. Moinul Islam1, Mohammad Rashed Iqbal Faruque1, Mohd Fais Mansor2,
Mohammad Tariqul Islam2
1Universiti Kebangsaan Malaysia, Complex Penyelidikan Building, Centre for Space Science Angkasa, 43600 UKM,
Bangi, Selangor D. E., Malaysia
2Universiti Kebangsaan Malaysia, Deparment of Electrical, Electronic & Systems Engineering, 43600 UKM,
Bangi, Selangor D. E., Malaysia
mmoiislam@yahoo.com
Prejem rokopisa – received: 2014-08-15; sprejem za objavo – accepted for publication: 2015-03-11
doi:10.17222/mit.2014.199
A printed microstrip line-fed patch antenna for C-band applications is presented, using a high-dielectric material. The proposed
antenna dimensions are 0.53  × 0.53  × 0.02  and it is fed by a microstrip line. The antenna outline and electromagnetic
analysis were done with the help of a commercially available computer-aided EM simulator. This antenna initiates three
resonances at 4.64 GHz, 5.52 GHz, and 6.34 GHz with the average gains of 2.68 dBi, 6.02 dBi and 4.83 dBi, respectively,
covering the entire frequency bands. The overall performance analysis and a nearly omnidirectional radiation pattern prove that
the proposed antenna is promising for C-band applications.
Keywords: C-band, dielectric material, microstrip line feeding
Predstavljena je tiskana mikrotrakasta, linijsko napajana, krpasta antena za uporabo v C-pasu, z uporabo visoko dielektri~nega
materiala. Predlagana dimenzija antene je 0,53  × 0,53  × 0,02 , ki je napajana z linijo mikrotraku. Oris antene in
elektromagnetska analiza sta bili izvr{eni s pomo~jo komercialno razpolo`ljivega in ra~unalni{ko podprtega EM simulatorja. Ta
antena spro`i tri resonance pri 4,64 GHz, 5,52 GHz in 6,34 GHz, s povpre~no sposobnostjo 2,68 dBi, 6,02 dBi in 4,83 dBi pri
pokrivanju vseh frekven~nih pasov. Analiza zmogljivosti in skoraj vsesmerna slika sevanja ka`eta, da predlagana antena obeta
dobro uporabo v C-pasu.
Klju~ne besede: C-pas, dielektri~ni material, linijsko napajanje z mikrotrakom
1 INTRODUCTION
Currently, the microstrip patch antenna is a milestone
in the wireless communication system and it continues to
fulfill the changing requirements of the new-generation
antenna technology. Microstrip patch antennas are
widely utilized in the present wireless communication
system because of their low profile, light weight, confor-
mal design, low cost, and because they are easy to fabri-
cate and integrate. Advances in wireless communications
have initiated remarkable demands. Antennas are used
for a wide range of cellular mobile phones in the current
society, causing concerns about their harmful radi-
ation.1–5 Many researches were done, covering the entire
C-band and many techniques and methods are stated in
the reference literature.
A hexagonal scrimp-horn antenna with different
aperture sizes was proposed for operating in C-band
applications.6 A modified dual-band CPW-fed antenna
was proposed for a WLAN-band application on a thin
substrate.7 A broadband planar monopulse antenna was
presented to increase the impedance bandwidth for
C-band applications, where a monopulse comparator was
used as the sum-difference feed network.8 A rectangular
slot antenna with a U-shaped strip was proposed for a
dual broadband operation in WLAN applications.9 A
compact broadband slot antenna with a circular polari-
zation was proposed for C-band applications, where two
rectangular stubs are embedded to excite two orthogonal
E vectors in the feedline structure.10
In this paper, a printed microstrip line-fed patch an-
tenna with a high-dielectric material that attains a com-
pact triple-resonant profile due to a nearly omnidirec-
tional radiation, high gain and a reasonable current
distribution is proposed. This line-fed antenna is made of
circular radiating patches with a partial ground plane
generating three resonances for C-band applications. The
antenna is smooth, with a simple design and comfortable
fabrication. The proposed line-fed antenna generates
three resonances to cover C-band applications. The
results are impedance bandwidth values of (160, 100 and
160) MHz at three resonances on the C-band. Due to a
double -shaped radiating patch with a partial ground,
nearly omnidirectional radiation properties are realized
over the entire operating bands with a reasonable gain.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 307–310 307
UDK 621.396.67:621.315.61 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)307(2016)
This line-fed antenna with a high-dielectric material is
very effective for C-band applications.
2 ANTENNA STRUCTURES
The design of the proposed antenna is indicated in
Figure 1. The antenna comprises of a double -shaped
patch and a partial ground. The design procedure begins
with a radiating patch with a substrate, a ground plane
and a feed line. It is printed on a ceramic-filled bio-
plastic substrate with a relative permittivity of 15 and a
relative permeability of 1. The overall antenna dimen-
sions are 40 mm × 40 mm × 1.2 mm. An SMA (Sub-
Miniature version A) connector is used for providing a
50  impedance and it is attached at the end of the
antenna feeding.
Figure 2 exhibits the structure of the substrate
material. This sandwich-structured substrate material
was generated using ceramic powder and bioplastic. The
selected ceramic powder was sintered with a polymeric
binder using the polymeric sponge method. A 9.8 ml
(0.25) bioplastic sheet was included. This bioplastic
sheet was obtained from organic biomass sources, such
as cornstarch, vegetable oil and palm oil, and used as the
ceramic cover. The three-layer bioplastic-ceramic-bio-
plastic sandwich structure was laminated using 35 μm of
copper foil. The characteristics of this substrate material
are low cost, ease of fabrication, design flexibility and
availability. For this reason, a high-dielectric material is
preferred for the antenna design.
Two -shaped circular slots were cut from the
copper patch with a partial ground. In this way, the
proposed line-fed patch antenna was achieved. Three
resonant frequencies of (4.64, 5.52 and 6.34) GHz were
obtained, continuously adjusting the length, the width
and the slots of the proposed antenna. Here, the
microstrip line is used to provide the feeding to the
proposed antenna. The length and width of the patch
antenna can be calculated from Equations (1) and (2).11 L
and W are the length and width of the patch, c is the
velocity of light, r is the dielectric constant of the sub-
strate, f0 is the target center frequency, and e is the
effective dielectric constant:
W =
c
f2
1
20
 r +
(1)
L =
c
f
l
2
2
0  r
− Δ (2)
Finally, the optimum dimensions were determined as
follows: L = 40 mm, W = 40 mm, P = 20 mm, Mw = 2.5
mm, Ws = 8 mm, and Lg = 19 mm.
3 RESULTS AND DISCUSSION
The simulated return loss of the proposed antenna is
illustrated in Figure 3. Return losses of –20.21 dB,
–19.58 dB and –16.70 dB were acquired at three reso-
nant frequencies of (4.64, 5.52 and 6.34) GHz, respec-
tively.
We obtained the 160 MHz bandwidth with the 1st
resonant frequency, 100 GHz with the 2nd and 1.60 MHz
with the 3rd frequency. The mutual coupling effect was
increased with the lower frequency; as a result, the
bandwidth was small with the 1st and 2nd resonances. On
the other hand, the bandwidth was broadened due to the
suppressed mutual coupling effect. These bandwidths
Md. M. ISLAM et al.: PRINTED MICROSTRIP LINE-FED PATCH ANTENNA ON A HIGH-DIELECTRIC MATERIAL ...
308 Materiali in tehnologije / Materials and technology 50 (2016) 3, 307–310
Figure 1: Proposed C-band antenna: a) front view, b) back view
Slika 1: Predlagana antena za C-pas: a) pogled spredaj, b) pogled
zadaj
Figure 3: Proposed C-band antenna return loss
Slika 3: Povratne izgube, predlagane antene za C-pas
Figure 2: Structure of the substrate material12
Slika 2: Struktura materiala podlage12
were generated at the operating frequencies throughout
the entire C-band application.
The average gain of the proposed antenna is shown in
Figure 4. The average gains of (2.68, 6.02 and 4.83) dBi
are achieved in the operating frequency bands of (4.64,
5.52 and 6.34) GHz, respectively. The used dielectric
substrate material controls the mutual coupling effect
and, as a result, the antenna gain is widened. It can be
observed that the antenna gain was considerably
increased with the incorporation of this high-dielectric
material in the lower and upper bands, compared to the
existing antennas.
The voltage standing wave ratio (VSWR) of the
proposed antenna is plotted in Figure 5. The value of the
VSWR is less than 2, as clearly seen on the graph. It is
the desired value.
Figure 6 exhibits the result of the radiation efficiency
of the proposed patch antenna. The radiation efficiency
is 94 % with the 1st resonance, 90.06 % with the 2nd reso-
nance and 94.08% with the 3rd resonance. This efficiency
is broadly appropriate for C-band applications. It is
considerable in comparison with the existing ones. It is
obtained using a high-dielectric material for the pro-
posed antenna and this antenna is perfect for C-band
applications.
The surface-current distribution of the proposed
patch antenna is demonstrated in Figure 7. The arrow
sign is applied to denote the flow of the current distri-
bution. From the graph, it can be easily observed that the
current flow is maximum at the microstrip line and the
lower -shaped slot, at 4.64 GHz. At 5.52 GHz, the
upper -shaped slot and the microstrip line show the
maximum current. At 6.34 GHz, the parts of the intersec-
tion between double -shaped slots control the maxi-
mum current flow. Due to the high-dielectric substrate
material, the overall surface-current distribution is
smooth and sharp. As a result, the mutual coupling effect
is under consideration and it is controlled in the case of
the proposed patch antenna.
Md. M. ISLAM et al.: PRINTED MICROSTRIP LINE-FED PATCH ANTENNA ON A HIGH-DIELECTRIC MATERIAL ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 307–310 309
Figure 6: Proposed C-band antenna efficiency
Slika 6: U~inkovitost predlagane antene za C-pas
Figure 4: Proposed C-band antenna gain
Slika 4: Sposobnost predlagane antene za C-pas
Figure 7: Surface current of the proposed C-band antenna at: a) 4.64
GHz, b) 5.52 GHz and c) 6.34 GHz
Slika 7: Tok na povr{ini predlagane antene za C-pas pri: a) 4,64 GHz,
b) 5,52 GHz in c) 6,34 GHz
Figure 5: Proposed C-band antenna VSWR
Slika 5: VSWR predlagane antene za C-pas
The radiation patterns of the proposed antenna on the
E-plane and H-plane, at resonant frequencies of (4.64,
5.52 and 6.34) GHz are shown in Figure 8. It is shown
from the results that significant, nearly omnidirectional
radiation patterns are acquired along the H-plane and
E-plane, respectively. The cross-polarization is low on
the E-plane at all the resonances; on the other hand, the
cross-polarization is high on the H-plane. The cross-po-
larization is lower than the co-polarization at all the
resonances, leading to omnidirectional or nearly omni-
directional radiation characteristics. As a result, the
radiation pattern of the proposed patch antenna is almost
durable for C-band applications.
4 CONCLUSION
The article presents a printed line-fed patch antenna
with a high-dielectric material appropriate for C-band
applications. It uses a double -shaped patch instead of
a conventional patch with a view to obtaining a triple
band operation. The microstrip line-fed antenna with a
high-dielectric material was designed and simulated
using the HFSS software, while the current-distribution
plots were made to verify the proposed track. The simu-
lation results indicate good characteristics. Conse-
quently, the proposed microstrip line-fed antenna with a
high-dielectric material can be appropriate for C-band
applications.
Acknowledgement
This work was supported by the Ministry of Educa-
tion (MOE) under Fundamental Research Grant Scheme
Top Down, Code: FRGS TOP DOWN/2014/TK03/
UKM/01/1.
5 REFERENCES
1 M. R. I. Faruque, M. T. Islam, N. Misran, Evaluation of specific
absorption rate (SAR) reduction for PIFA antenna using meta-
materials, Frequenz, 64 (2010) 7–8, 144–149, doi:10.1515/FREQ.
2010. 64.7-8.144
2 M. M. Islam, M. R. I. Faruque, W. Hueyshin, J. S. Mandeep, T.
Islam, A double inverted F-shape patch antenna for dual-band
operation, International Journal of Antennas and Propagation, 8
(2014), doi:10.1155/2014/791521
3 M. M. Islam, M. T. Islam, M. Samsuzzaman, M. R. I. Faruque, N.
Misran, M. F. Mansor, A Miniaturized Antenna with Negative Index
Metamaterial Based on Modified SRR and CLS Unit Cell for UWB
Microwave Imaging Applications, Materials, 8 (2015) 2, 392–407,
doi:10.3390/ma8020392
4 M. R. I. Faruque, M. T. Islam, B. Yatim, M. A. M. Ali, Analysis of
the effects of metamaterials on the radio-frequency electromagnetic
fields in the human head and hand, Mater. Tehnol., 47 (2013) 1,
129–133
5 M. M. Islam, M. R. I. Faruque, M. T. Islam, A Compact 5.5 GHz
Band-Rejected UWB Antenna Using Complementary Split Ring
Resonators, The Scientific World Journal, 2014 (2014) 10,
doi:10.1155/2014/528489
6 S. A. Muhammad, A. Rolland, H. D. Samsul, R. Sauleau, H. Legay,
Hexagonal-shaped broadband compact scrimp horn antenna for
operation in C-band, IEEE Antennas and Wireless Propagation
Letters, 11 (2012), 842–845, doi:10.1109/LAWP.2012.2208259
7 C. Yoon, W. J. Lee, S. P. Kang, S. Y. Kang, H. C. Lee, H. D. Park, A
planar CPW-fed slot antenna on thin substrate for dual-band
operation of WLAN applications, Microwave and Optical
Technology Letters, 51 (2009), 2799–2802, doi:10.1002/mop.24742
8 Z. W. Yu, G. M. Wang, C. X. Zhang, A broadband planar monopulse
antenna array of C-band, IEEE Antennas and Wireless Propagation
Letters, 8 (2009), 1325–1328, doi:10.1109/LAWP.2009.2038077
9 J. W. Wu, H. M. Hsiao, J. H. Lu, S. H. Chang, Dual broadband
design of rectangular slot antenna for 2.4 and 5 GHz wireless
communication, Electronics Letters, 40 (2004) 23, 1461–1463,
doi:10.1049/el:20046873
10 S. Mohammadi, J. Nourinia, J. Pourahmadazar, M. Shokri, Compact
broadband circularly polarized slot antenna using two linked
elliptical slots for C-band applications, IEEE Antennas and Wireless
Propagation Letters, 12 (2013), 1094–1097, doi:10.1109/LAWP.
2013.2280457
11 M. M. Islam, M. T. Islam, M. R. I. Faruque, Dual-band operation of a
microstrip patch antenna on a duroid 5870 substrate for Ku- and
K-bands, The Scientific World Journal, 2013 (2013) 10, doi:10.1155/
2013/378420
12 M. T. Islam, M. H. Ullah, M. J. Singh, M. R. I. Faruque, A new
metasurface superstrate structure for antenna performance enhance-
ment, Materials, 6 (2013) 8, 3226–3240, doi:10.3390/ma6083226
Md. M. ISLAM et al.: PRINTED MICROSTRIP LINE-FED PATCH ANTENNA ON A HIGH-DIELECTRIC MATERIAL ...
310 Materiali in tehnologije / Materials and technology 50 (2016) 3, 307–310
Figure 8: Radiation patterns: a) E-plane at 4.64 GHz, b) H-plane at
4.64 GHz, c) E-plane at 5.52 GHz, d) H-plane at 5.52 GHz, e) E-plane
at 6.34 GHz and f) H-plane at 6.34 GHz
Slika 8: Sevalni diagrami: a) E-ravnina pri 4,64 GHz, b) H-ravnina pri
4,64 GHz, c) E-ravnina pri 5,52 GHz, d) H-ravnina pri 5,52 GHz, e)
E-ravnina pri 6,34 GHz in f) H-ravnina pri 6,34 GHz
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
311–317
COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES
PRODUCED WITH DIRECT 3D PRINTING
STISKANJE STRUKTUR MATERIALOV Z NEGATIVNIM
POISSONOVIM RAZMERJEM, PROIZVEDENIH Z NEPOSREDNIM
TRIDIMENZIONALNIM TISKANJEM
Petr Koudelka1,2, Ondøej Jirou{ek2, Tomá{ Fíla1,2, Tomá{ Doktor1,2
1Institute of Theoretical and Applied Mechanics, Academy of Sciences of the Czech Republic, Prosecká 76, 190 00 Prague, Czech Republic
2Czech Technical University in Prague, Faculty of Transportation Sciences, Konviktská 20, 110 00 Prague 1, Czech Republic
jirousek@fd.cvut.cz
Prejem rokopisa – received: 2014-08-19; sprejem za objavo – accepted for publication: 2015-04-24
doi:10.17222/mit.2014.204
In the presented paper, three types of auxetic structures were produced with direct 3D printing and their compressive mechanical
properties were tested. Samples were prepared from acrylic material suitable for high-resolution direct printing. Three different
structures exhibiting in-plane and volumetric negative strain-dependent Poisson’s ratio were selected for the analysis:
two-dimensional missing-rib cut, two-dimensional inverted (re-entrant) honeycomb and three-dimensional inverted (re-entrant)
honeycomb. The samples were subjected to quasi-static compression, from which stress-strain relationships were established.
For a proper strain evaluation, digital-image correlation was applied to measure full-field displacements on the sample surfaces.
From the displacement fields, true strain/true stress curves were derived for each sample. Furthermore, for each structure a
three-dimensional FE model was developed using beam elements and subjected to identical loading conditions. Then,
experimentally obtained stress-strain relationships were compared with numerically obtained results. For all the tested auxetic
structures, the compressive behaviour was predicted well by the FE models. This demonstrates that parametric FE models can
be used to tune the design parameters of the structures with a negative Poisson’s ratio to optimize their overall properties.
Keywords: auxetics, cellular materials, quasi-static testing, finite-element method
V prispevku so predstavljene tri vrste struktur materialov z negativnim Poissonovim razmerjem, ki so proizvedene z
neposrednim tridimenzionelnim tiskanjem. Preizku{ene so bile njihove mehanske lastnosti pri stiskanju. Vzorci so bili
pripravljeni iz akrilnih materialov, ki so primerni za visoko resolucijsko neposredno tiskanje. Za analizo so bile izbrane tri
razli~ne strukture, ki prikazujejo negativno odvisno Poissonovo razmerje v ravnini in v prostoru: dvodimenzionalni prerez z
manjkajo~im rebrom, dvodimenzionalni obrnjeni (navznoter usmerjeni) vzorec satovja in tridimenzionalni obrnjeni (navznoter
usmerjeni) vzorec satovja. Vzorci so bili izpostavljeni kvazi-stati~nem stiskanju pri katerem smo ugotavljali razmerja sile –
raztezek. Za primerno oceno sile obremenitve je bila uporabljena metoda korelacije digitalne slike in s tem izmerjeni odmiki na
povr{ini vzorcev. Glede na te odmike so bile za vsak vzorec izpeljane dejanske obremenitvene krivulje. Nadalje je bil za vsako
strukturo izdelan tridimenzionalni FE model, z uporabo matemati~nega modela podpornih struktur in izpostavljen identi~nim
pogojem obremenitve. Nato smo primerjali eksperimentalno pridobljena razmerja med silo in obremenitvijo, z ra~unsko
pridobljenimi rezultati. S pomo~jo primerjalnih diagramov sile in raztezka lahko ugotovimo, da FE modeli dobro napovedujejo
obna{anje pri stiskanju vseh preizku{enih struktur z negativnim Poissonovim razmerjem. To prikazuje mo`nost uporabe
parametri~nih FE modelov za prilagoditev zasnovnih parametrov struktur z negativnim Poissonovim razmerjem za optimiziranje
njihovih splo{nih lastnosti.
Klju~ne besede: materiali z negativnim Poissonovim razmerjem, celi~ni materiali, kvazi-stati~no preizku{anje, metoda kon~nih
elementov
1 INTRODUCTION
Porous solids (i.e., open- or closed-cell foams) are
materials suitable for applications requiring a significant
mass reduction and a simultaneous high-impact energy
absorption. This is provided by the foams’ low specific
weight and thus high specific stiffness. However, for
certain applications (including blast protection), it may
be necessary to use materials with a relatively high
compressive strength, which disqualifies the usage of
most foam types including metallic foams.1 To improve
the strength and energy absorption capacity without
increasing the mass of constructional elements, a new
type of material had to be found. One of the possible
approaches to dealing with the absorption of enormous
amounts of deformation energy during blast and impact
loading of structures is to produce a highly optimized
porous structure, taking advantage of the negative
Poisson’s ratio of its skeleton.2
Historically, a structure with a negative Poisson’s
ratio was first reported for single crystals of iron pyrites
and was attributed to crystal twinning.3 This initial
finding was followed only by isolated reports in the
1970s and 1980s showing that negative Poisson’s ratio is
a rather unique phenomenon among natural constructs.
First artificially prepared auxetic polymeric foam was
reported in 1987 by Lakes4 when commercially available
foam was modified in a process involving 30 % volu-
metric compression and heating of the samples to the
polymer’s softening temperature followed by cooling
Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317 311
UDK 620.192.47:66.069.852:519.61/.64 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)311(2016)
whilst remaining under compression. Although research
of the microstructures prepared with various similar
techniques from the existing materials continued, the
investigation of the auxetic-based materials capable of
deformation-energy absorption closely followed the
advances in additive manufacturing (i.e., selective laser
sintering, direct 3D printing, etc.).
The reported auxetic topologies are mostly based on
re-entrant, chiral, double-arrow-head and rotating-rigid
unit cells.5–8 Deformation behaviour (both elastic and
plastic) of such a porous construct is determined by
concurrent effects of intrinsic behaviour of the material
used for the scaffold production, cell topology and
connectivity. To optimize its microstructure to suit the
intended application and to achieve a stable negative
Poisson’s ratio up to high strains, a control over the pore
structure is required. Here, the usage of additive manu-
facturing is favourable as all the intended geometrical
characteristics can be attained deterministically, satis-
fying the need for a high mechanical integrity of the
construct during deformation and the need for precise
tuning of the overall stiffness and plastic properties.
Several unit-cell models having a well-defined
Poisson’s function (strain-dependent Poisson’s ratio)
described analytically have already been proposed.9–12 In
this study, samples of three microstructures utilizing both
the in-plane and volumetric negative Poisson’s ratio were
produced with direct 3D printing to evaluate their
compressive-deformation behaviour in both elastic and
plastic regime up to the densification. Deformation of the
microstructures during quasi-static uni-axial com-
pression tests was optically observed using a CCD
camera and the digital-image-correlation (DIC) method
was used for the strain evaluation. Moreover, mechanical
behaviour of the structures was investigated numerically
with virtual experiments (i.e., simulations of the
compression tests) using finite element (FE) method.
The simulations were carried out using identical loading
and boundary conditions and the numerically obtained
stress-strain curves were compared to the experimental
ones to test the proposed material model for the base
material.
2 EXPERIMENTAL PART
2.1 Specimen geometry
In this study, three different types of unit-cell geo-
metry were used: a) two-dimensional missing-rib cut, b)
two-dimensional inverted (re-entrant) honeycomb and c)
three-dimensional inverted honeycomb. These unit cells
were arranged so that predictable (determinate) in-plane
or volumetric negative strain-dependent Poisson’s ratio
was achieved in the cases of two-dimensional and three-
dimensional geometries, respectively. Constructs of
two-dimensional geometries were generated by extrud-
ing a planar (single-layer) arrangement of the unit cells
whereas a three-dimensional construct was created by
copying a fully three-dimensional unit cell along all the
spatial directions.
The missing-rib-cut model was formed by removing
selected ribs (the elements forming a unit cell) from the
periodical arrangement of squares and by rotating the
construct by 45° to the direction of loading.13 Auxetic
behaviour of such a construct depends on the unit-cell
dimensions and the angles between individual ribs. The
geometry of a unit cell with the dimensions used in this
work and a visualisation of the specimen are depicted in
Figure 1. The dimensions of the produced construct
were 25.05 mm × 25.40 mm × 37.75 mm (width, depth,
height), the overall porosity was 72.8 % and the con-
struct consisted of 10 × 15 cells.
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
312 Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317
Figure 2: a) 2D re-entrant structure – unit cell geometry, b) visu-
alisation of the whole sample, c) 3D re-entrant structure – unit cell
geometry and d) visualisation of the whole sample
Slika 2: a) dvodimenzionalna navznoter usmerjena struktura – geo-
metrija celice, b) vizualizacija celotnega vzorca, c) tridimenzionalna
navznoter usmerjena struktura – geometrija celice in d) vizualizacija
celotnega vzorca
Figure 1: Missing-rib-cut structure: a) unit cell geometry and b) visu-
alisation of the whole sample
Slika 1: Struktura prereza z manjkajo~im rebrom: a) geometrija celice
in b) vizualizacija celotnega vzorca
The re-entrant mesh is generated by changing the
four side angles between the ribs in a six-sided honey-
comb.14 The magnitude of Poisson’s ratio at a given
strain is here primarily given with the length ratio of
individual ribs forming the unit cell. Both two-dimen-
sional and three-dimensional arrangements of the
re-entrant unit cells together with visualisations of the
specimens are shown in Figure 2. Dimensions of the
produced two-dimensional construct were 25.65 mm ×
25.40 mm × 58.89 mm (width, depth, height), the overall
porosity was 73.2 % and the construct consisted of 10 ×
15 cells; dimensions of the produced three-dimensional
construct were 7.87 mm × 7.87 mm × 18 mm (width,
depth, height), the overall porosity was 91.7 % and the
construct consisted of three cells in every spatial direc-
tion.
2.2 Specimen preparation
The specimens were manufactured from VisiJet
EX200 (3D Systems, USA) UV curable acrylic material
suitable for high-resolution 3D printing. The physical
properties of the material are summarised in Table 1.
For the specimen production, a Pro Jet HD3000 (3D
Systems, USA) 3D printer in the high-definition mode
was used. The manufacturing principle is based on a
multi-jet modelling technology where a special printing
head covers the whole working area of 198 mm × 185
mm × 203 mm and builds up the model by adding indivi-
dual layers of the produced geometry. Simultaneously,
while modelling the material, a supporting wax material
is automatically added to the construct to enable a pro-
duction of very complex geometries.
Thanks to its low melting point (approximately
55–65 °C) all the supporting material can be simply
removed from the products by heating it in a water bath
to approximately 80 °C without a potential mechanical
damage to the products. A SolidWorks (Dassault Sys-
tèmes SolidWorks Corp., France) parametric modeller
was used to design the sample geometry that was
exported to the STL format for the 3D printing. The final
samples were produced with a resolution of (328 × 328 ×
606) DPI (x, y, z direction) with a layer thickness of
0.036 mm. In this mode, the accuracy of printing was
approximately 0.025–0.05 mm and the production
process took approximately 11 h.
Table 1: Properties of the VisiJet EX200 material
Tabela 1: Lastnosti materiala VisiJet EX200
Mass density 1.02 g/cm3
Tensile modulus 1.283 GPa
Tensile strength 42.4 MPa
Flexural modulus 1.159 GPa
Glass transition temp. 52.5 °C
2.3 Experimental set-up
The experiments were carried out using an in-house
designed loading set-up based on a novel modular
compression/tension loading device suitable for both
optical and X-ray observation of deformation processes15
equipped with a U9b force transducer (HBM, Germany)
with a nominal force capacity of 2 kN. The signal from
the load cell was read out using an OM502T (Orbit
Merret, CZ) load-cell indicator at a sampling rate of 50
Hz.
Imaging was performed using a Manta G504B
monochromatic GigE vision camera (AVT, Germany).
The camera was equipped with an ICX655 CCD sensor
and its maximum frame rate was 9 fps, achieved at a
resolution of 2452 px × 2056 px. In order to guarantee a
high reliability of the correlation procedure and an
accuracy of the computed strains, the camera was
equipped with a TCZR 072 bi-telecentric zoom lens
(Opto Engineering, Italy). The lens used a stepper-
motor-controlled zoom revolver to set four different
magnifications of the scene in the range of 0.125–1.000,
with a very high image-centre stability, parfocality and
no need for a re-calibration after the zooming. The
specimens were illuminated using a KL2500 high-power
white-light LED source (Schott, Germany).
A detailed description of the loading set-up is shown
in Figure 3.
2.4 Loading procedure and strain measurement
The loading of the samples was performed as a
displacement-driven uniaxial compression. The maxi-
mum displacement was set to 8 mm with a loading rate
of 20 μm s–1. The positioning of the camera as well as the
loading were carried out using in-house developed con-
trol software based on the GNU/Linux real-time ope-
rating software and LinuxCNC open-source project.
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317 313
Figure 3: Loading set-up used in the experiments
Slika 3: Postavitev obremenjevanja, ki je bilo uporabljeno pri preiz-
kusih
The strains were derived from optically measured
displacements using a sequence of the images capturing
a deforming sample. Depending on the sample height,
0.125 or 0.25 magnifications were used and the images
were captured with the maximum resolution of 2452 px
× 2056 px at 2 fps, enabling the identification of a suffi-
cient number of points on the stress-strain curve. The
acquisition of the projections was controlled with
custom-developed software based on the OpenCV library
and Python programming language. The observed faces
of the specimens were sprayed using granite paint to
generate a random pattern for a high reliability of the
optical deformation tracking. A selected loading scene is
depicted in Figure 4.
2.5 Stress-strain diagram evaluation
Displacements were determined from the image
sequences using a custom-developed DIC software tool16
based on the Lucas-Kanade tracking algorithm17 imple-
mented in MATLAB. Two rows of correlation points
(markers) perpendicular to the loading direction were
selected near the upper and lower edges of the observed
surfaces of the specimens. Each marker was in a series of
projections tracked by searching for the highest corre-
lation coefficient between two consequent projections.
Engineering-stress (eng) and strain (
eng) values were
determined from the geometrical properties of the tested
specimens and optically measured displacements of the
markers. Then, true stress (true) and true strain (
true)
were calculated according to Equations (1) and (2):
  
true eng eng= +( )1 (1)

 
true eng= +ln( )1 (2)
3 NUMERICAL PART
Apart from the experimental methods, analytical and
FE models can be used for a description of deformation
behaviour of auxetic constructs, allowing a prediction
and optimization of the effective mechanical characte-
ristics that facilitate the material design for a specific
application.
Most of the analytical models assume small deflec-
tions, neglecting the axial deformation of the struts.19
Thus, analytical approach can be used to prove the con-
cept of negative Poisson’s ratio, optimize the parameters
of a structure (e.g., the re-entrant angle, relative density,
strut thickness) and maximize the effective parameters of
the resulting constructs (i.e., deformation energy per unit
volume, yield strength of the structure, compressive
strength) according to specific requirements. For
instance, using these analytical models, it is possible to
express Poisson’s ratio μ of the re-entrant honeycomb
structure (Figure 2a) with Equation (3):
( )
μ =
− + −
−
sin( / sin(
cos (
90 90
90
2 1
2
 

L L
(3)
where L1 and L2 are the strut lengths and  is the strut
angle (measured in degrees). The value of resulting
Poisson’s ratio is negative and its dependency on the
magnitude of the angle  and the strut length ratio L2/L1
corresponds to the observed experimental results. Using
the Timoshenko beam theory and the elastic-behaviour
assumption, it is possible to express the overall elastic
modulus E and critical-yield compressive force Fm
based on the yield strength y of the bulk solid material.
From the yield compressive force the compressive
strength of the structure can be then expressed.
However, these analytical models are effective only
when simplifying the assumptions such as the small-
deflection theory and linear elastic-material properties
are used. Consequently, these models only give satisfac-
tory results for small deformations and are limited to the
calculation of the overall elastic properties or to the
estimation of the yield point of a structure.
When large strains with non-linear material pro-
perties are to be considered, FE models have to be used
instead. Thus, FE models of the tested auxetic structures
were developed and subjected to the same loading
conditions as during the compression tests. Deformation
behaviour of the tested samples under large strains (up to
10 % or 20 % strain) was then compared with the predic-
tions obtained from the numerical models to verify their
suitability for a representation of such microarchitec-
tures.
Stress-strain curves were inversely assessed from the
FE simulations, i.e., from the reaction forces calculated
at the restrained side of a sample. Using such inverse FE
simulations, it is relatively easy not only to obtain the
stress-strain curves for each considered sample, but also
to establish the stresses and strains arising at individual
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
314 Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317
Figure 4: Loading scene captured using a CCD camera during the
experiment. Missing-rib specimen with sprayed surface for a DIC
strain evaluation.
Slika 4: Trenutek obremenitve, ki je prikazan na CCD kameri med
preizkusom. Vzorec z manjkajo~im rebrom z napr{eno povr{ino za
oceno DIC sile obremenitve.
struts from the deformation of the structure. Hence, these
strains can be easily compared to the values experimen-
tally obtained from the digital-image correlation at the
same positions (i.e., individual markers).
For all the considered auxetic constructs, the FE
model was created using 3D beam elements with 6° of
freedom (three translational and three rotational ones)
defined at two nodal points. The element is based on
Timoshenko beam theory, which includes shear-defor-
mation effects. The material model set in the simulations
was elasto-plastic, combining von Mises yield criteria
and bilinear isotropic work hardening. The material pro-
perties are summarised in Table 2.
Table 2: Material properties used in the FE simulations
Tabela 2: Lastnosti materiala uporabljenega pri FE simulaciji
Young’s modulus 1.159 GPa
Poisson’s ratio 0.2 –
Yield stress 42.4 MPa
Hardening tangent modulus 12.8 MPa
The loading was prescribed to be done in 100 loading
steps, i.e., in each step a 0.1 % or 0.2 % deformation was
applied.
In the case of such a large-strain analysis, a highly
deformed geometry has an important effect on the strain
and, therefore, geometric nonlinearities must be con-
sidered. To consider the post-buckling behaviour of the
thin beams subjected to a large compression, strain
measures have to account for higher-order terms. Thus,
in our analyses, material stress-strain properties were
input in terms of true stress versus logarithmic strain. In
every loading step, reaction forces originating from the
supports were calculated and the true stresses and strains
were established using Equations (1) and (2). Visuali-
zations of the FE models for individual constructs can be
seen in Figure 5.
4 RESULTS AND DISCUSSION
From both the numerical and experimental quasi-
static compression tests true stress/true strain diagrams
for all three considered auxetic microarchitectures were
plotted.
A comparison of the experimental and numerical
stress-strain diagrams of the missing-rib-structure cut is
depicted in Figure 6.
It can be seen that this type of microstructure exhibits
a similar initial compressive behaviour as a typical
closed-cell porous solid. The initial linear elastic part is
followed by an apparent yield point and a compaction
region with a constant stress plateau. These parts are
then, at an approximately 17.5 % strain, followed by a
region of localized densification due to negative
Poisson’s ratio of the unit cell and repeated decreases of
the stress that can be attributed to the ruptures of beams
due to excessive bending. Good correlation of the
numerical and experimental results was obtained in
terms of stiffness, yield point and plateau stress up to a
20 % strain.
In Figure 7, a comparison of the experimentally and
numerically acquired stress-strain diagrams of the two-
dimensional inverted honeycomb structure is presented.
It is clearly apparent that the microstructure of such a
construct exhibits a significantly different deformation
behavior than the missing-rib-structure cut. After the
initial linear elastic region, a 30 % drop in the stress is
followed by cyclic increases and decreases in the stress
levels in the specimens with an apparent progressive
trend. After a visual inspection of individual projections
during the deformation, the occurrence of the cycles can
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317 315
Figure 6: Comparison of experimentally and numerically obtained
stress-strain diagrams of the missing-rib-cut structure
Slika 6: Primerjava eksperimentalno in {tevil~no pridobljenih diagra-
mov sile in raztezka za strukturo prereza z manjkajo~im rebrom
Figure 5: Visualization of FE-model elements used in the numerical
simulations: a) missing-rib-structure cut with a detail of the unit cell,
b) 2D re-entrant structure with a detail of the unit cell and c) detail of
3D re-entrant unit cell
Slika 5: Vizualizacija elementov FE modela, ki je bil uporabljen v
{tevil~nih simulacijah: a) struktura prereza z manjkajo~im rebrom s
podrobnostmi enote celice, b) dvodimenzionalna navznoter usmerjena
struktura s podrobnostmi enote celice in c) podrobnosti tridimenzio-
nalna navznoter usmerjene enote celice
be attributed to the collapse of individual layers of the
unit cells in the microstructure. For this type of
microstructure, FE simulations give a good prediction of
the yield point and strain-softening behaviour up to the
densification of the individual layers in the microstruc-
ture.
A stress-strain diagram showing the experimentally
and numerically assessed behavior of the three-dimen-
sional inverted honeycomb structure is shown in Figure
8.
The mechanical response is similar to that of the
two-dimensional structure but with a more significant
stress drop after the linear elastic region and a lower
number of the stress cycles, caused by a 20 % lower
overall porosity and a lower number of the cells in the
structure. The performance of the FE model is, in this
case, similar to the two-dimensional re-entrant structure
with a numerically well-determined yield point and
strain-softening characteristics.
Only for the missing-rib cut of the auxetic structure,
it was possible to perform the FE analysis up to a 20 %
strain. Larger strain values could not be calculated as the
elements became extremely distorted, yielding instability
and convergence issues of the simulations. Furthermore,
to capture the stiffening during the compaction, it would
have been necessary to include self-contact between
individual struts, which would have significantly
increased the complexity and computational costs of the
simulations.
The FE simulations of the re-entrant structures pre-
dict a smaller overall stiffness, which is apparent from
the comparison between the experimental stress-strain
curves and the numerically obtained responses. For this
reason, a set of three-point-bending experiments was
carried out using prismatic beams with rectangular
cross-sections that were carefully cut from the printed
specimens. Based on the DIC strain evaluation, a
bending modulus of approximatly 1.5 GPa was calcu-
lated. This value is close to the nominal flexural modulus
of 1.2 GPa provided by the manufacturer that was also
used in the FE simulations. Thus, the discrepancies
between the numerically and experimentally evaluated
stiffness might have been caused by properties that were
different from the predicted properties of the joints
between individual struts, influencing the bending
characteristics of individual layers, which formed the
principal mode of deformation of the re-entrant struc-
tures. Here, a combination of a precise inspection of the
geometry and possibly a nanoindentation measurement
of the joints should be applied to obtain an accurate
material model for the FE simulations.
5 CONCLUSIONS
Mechanical behaviour of three different porous
microarchitectures exhibiting in-plane and volumetric
negative Poisson’s ratios was studied both experimen-
tally and numerically. The specimens prepared with
high-resolution direct 3D printing were compressively
loaded up to the densification regions of their mechani-
cal responses. The true stress/true strain diagrams for the
compression were derived from a high-precision force
measurement and an optical DIC evaluation of the strain
field. Based on the experimental results, numerical FE
models of all the considered microarchitectures were
developed and their ability to predict mechanical
responses of the studied constructs was assessed by
comparing the numerically and experimentally obtained
stress-strain diagrams. It was found that the deformation
response of the missing-rib-cut structure was well cap-
tured by the FE model up to a 20 % strain. Simulations
of the re-entrant honeycomb structures showed good
correlation of the yield point and strain-softening charac-
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
316 Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317
Figure 7: Comparison of experimentally and numerically obtained
stress-strain diagrams of the 2D inverted honeycomb structure
Slika 7: Primerjava eksperimentalno in {tevil~no pridobljenih diagra-
mov sile in raztezka za dvodimenzionalno obrnjeno strukturo satovja
Figure 8: Comparison of experimentally and numerically obtained
stress-strain diagrams for the 3D inverted honeycomb structure
Slika 8: Primerjava eksperimentalno in {tevil~no pridobljenih diagra-
mov sile in raztezka za tridimenzionalno obrnjeno strukturo satovja
teristics up to a 10 % strain, while the calculated stiffness
of the models was lower than the stiffness of the
measured specimens. Still, the acquired results demon-
strate that parametric FE models can be used to tune the
design parameters of the structures with negative
Poisson’s ratio and numerically optimize their overall
properties. Therefore, we can conclude that such FE
models can be successfully used in material engineering
to design highly optimized structures for a given range of
strain rates, with maximized strain-deformation energy.
Acknowledgements
The research was supported by the Czech Science
Foundation (research project Nos. P105/12/0824 and
15-15480S) and by RVO: 68378297.
6 REFERENCES
1 Y. Sugimura, J. Meyer, M. Y. He, H. Bart-Smith, J. Grenestedt, A. G.
Evans, On the mechanical performance of closed cell foams, Acta
Materialia, 45 (1997), 5245–5259
2 M. Grujicic, J. Galgalikar, J. S. Snipes, R. Yavari, S. Ramaswami,
Multi-physics modeling of the fabrication and dynamic performance
of all-metal auxetic-hexagonal sandwich-structures, Materials &
Design, 51 (2013), 113–130, doi:10.1016/j.matdes.2013.04.004
3 W. Voigt, Lehrbuch der Kristallphysik, Teubner, Leipzig 1928
4 R. Lakes, Foam Structures with a Negative Poisson’s Ratio, Science,
235 (1987), 1038–1040, doi:10.1126/science.235.4792.1038
5 C. Lira, P. Innocenti, F. Scarpa, Transverse elastic shear of auxetic
multi re-entrant honeycombs, Composite Structures, 90 (2009) 3,
314–322, doi:10.1016/j.compstruct.2009.03.009
6 J. Grima, A. Alderson, K. Evans, Auxetic behaviour from rotating
rigid units, Physica Status Solidi (b), 242 (2005) 3, 561–575,
doi:10.1002/pssb.200460376
7 D. Prall, R. Lakes, Properties of a chiral honeycomb with a Poisson’s
ratio of -1, International Journal of Mechanical Sciences, 39 (1997)
3, 305–314
8 U. D. Larsen, O. Sigmund, S. Bouwstra, Design and fabrication of
compliant micromechanisms and structures with negative Poisson’s
ratio, Journal of Microelectromechanical Systems, 6 (1997) 2,
99–106
9 S. Burns, Negative Poisson’s Ratio Materials, Science, 238
(1987), 551, doi:10.1126/science.238.4826.551
10 R. H. Baughman, S. Stafstram, C. Cui, S. O. Dantas, Science, 279
(1998), 1522–1524, doi:10.1126/science.279.5356.1522
11 K. E. Evans, M. A. Nkansah, I. J. Hutchinson, S. C. Rogers, Mole-
cular network design, Nature, 353 (1991), 124, doi:10.1038/
353124a0
12 L. Rothenburg, A. A. Berlin, R. J. Bathurst, Materials with Negative
Poisson’s Ratio, Nature, 354 (1991), 470–472
13 C. W. Smith, J. N. Grima, K. E. Evans, A novel mechanism for gene-
rating auxetic behaviour in reticulated foams: missing rib foam
mode, Acta Materialia, 48 (2000) 17, 4349–4356, doi:10.1016/
S1359-6454(00)00269-X
14 I. G. Masters, K. E. Evans, Models for the elastic deformation of
honeycombs, Composite Structures, 35 (1996), 403–422,
doi:10.1016/S0263-8223(96)00054-2
15 P. Zlámal, O. Jirou{ek, D. Vavøík, A Novel Compression/Tension
Device for Investigation of Trabecular Bone Failure Using Real-
Time Micro-CT Imaging, 7th Youth Symposium on Experimental
Solid Mechanics, 2008, Wroclaw, Poland, 91
16 I. Jandejsek, J. Valach, D. Vavrik, Optimization and Calibration of
Digital Image Correlation Method, Proceedings of Experimental
Stress Analysis 2010, edited by P. Smid, (2010), Olomouc, Czech
Republic, 121–126
17 B. D. Lucas, T. Kanade, An Iterative Image Registration Technique
with an Application to Stereo Vision, Proceedings of Imaging
Understanding Workshop, 1981, Vancouver, Canada, 121–130
18 T. L. Warren, Negative Poisson’s ratio in a transversely isotropic
foam structure, Journal of Applied Physics, 67 (1990), 7591–7594,
doi:10.1063/1.345826
19 L. Yang, O. Harrysson, H. West, D. Cormier, Compressive properties
of Ti–6Al–4V auxetic mesh structures made by electron beam
melting, Acta Materialia, 60 (2012) 8, 3370–3379, doi:10.1016/
j.actamat.2012.03.015
P. KOUDELKA et al.: COMPRESSIVE PROPERTIES OF AUXETIC STRUCTURES PRODUCED WITH ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 311–317 317

L. BEK, R. ZEM^ÍK: MODEL OF PROGRESSIVE FAILURE FOR COMPOSITE MATERIALS ...
319–322
MODEL OF PROGRESSIVE FAILURE FOR COMPOSITE
MATERIALS USING THE 3D PUCK FAILURE CRITERION
MODEL POSTOPNEGA POPU[^ANJA KOMPOZITNEGA
MATERIALA Z UPORABO PUCKOVEGA TRIDIMENZIONALNEGA
KRITERIJA PORU[ITVE
Luká{ Bek, Robert Zem~ík
University of West Bohemia, European Centre of Excellence, Faculty of Applied Sciences, NTIS – New Technologies for the Information
Society, Univerzitní 22, 306 14 Pilsen, Czech Republic
lukasbek@kme.zcu.cz
Prejem rokopisa – received: 2014-09-16; sprejem za objavo – accepted for publication: 2015-05-07
doi:10.17222/mit.2014.233
A model for the progressive failure of composite materials that considers the materials’ non-linearity was developed and
implemented with the Abaqus FE software. An extended Puck failure criterion for the 3D stress state was used for the failure
prediction. Furthermore, a simplified approach for the simulation of the delamination was considered. For the progressive
failure simulation, the stiffness matrix degradation was used and the degradation parameters were a function of the fracture
angle. The model was tested on problems of a pin-loaded composite plate and of a composite tube subjected to compressive
loading perpendicular to the tube axis.
Keywords: progressive failure, composite, Puck criterion, finite-element analysis
Razvit je bil model postopnega popu{~anja kompozitnega materiala z upo{tevanjem nelinearnosti materiala, ki je bil uporabljen
v Abaqus FE programski opremi. Raz{irjeni Puckov kriterij poru{itve za tridimenzionalno napetostno stanje je bil uporabljen za
napoved poru{itve. Poleg tega je bil uporabljen tudi poenostavljen pribli`ek za simulacijo delaminacije. Za simulacijo napredo-
vanja popu{~anja je bila uporabljena degradacija togosti matrice. Degradacijski parametri pa so bili funkcija kota poru{itve.
Model je bil preizku{en na problemu obremenjevanja kompozitne plo{~e s konico in kompozitne cevi, izpostavljene tla~ni
obremenitvi pravokotno na os cevi.
Klju~ne besede: postopno popu{~anje, kompozit, Puckov kriterij, analiza kon~nih elementov
1 INTRODUCTION
Composite materials are frequently used in the aero-
space, automotive and marine industries, where extrem-
ely strong components and structures are necessary. Due
to the complex loading, finite-element (FE) analyses are
frequently used for the investigation of the stress state
and the failure of structures.1 Commercial FE software
systems are usually able to predict only the first failure,
which can occur at 20 % of the total strength of com-
posite structures. Some new releases of FE systems are
able to perform progressive failure analyses. However,
the analyses are often not sufficiently precise or have
problems with numerical stability. Therefore, new mo-
dels of progressive failure are developed and imple-
mented into the FE systems using a user-defined material
subroutine.2
The development, implementation and testing of the
progressive failure model for the 3D stress state based on
the Puck failure criterion and considering the material’s
non-linearity in the Abaqus FE software using the
UMAT material subroutine was the aim of this investi-
gation.
2 NON-LINEAR MATERIAL BEHAVIOUR
For the simulation of the non-linear material beha-
viour of composite materials, a non-linear function with
a constant asymptote was used for the calculation of the
shear modulus G12 and G13:3
G
G
G
n n
12 12
12
0
12
0
12
12
0
1
1
12
1( )


=
+
⋅⎛
⎝
⎜⎜
⎞
⎠
⎟⎟
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
+
12
(1)
G
G
G
n n
13 13
12
0
12
0
13
12
0
1
1
12
1( )


=
+
⋅⎛
⎝
⎜⎜
⎞
⎠
⎟⎟
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
+
12
(2)
where G120 is the initial shear modulus, 12 and 13 are
the shear strains, 120 is the asymptote value of the shear
stress and n12 is the shape parameter.
3 FAILURE CRITERION
The failure criterion determines the occurrence of
failure and indicates the failure’s propagation. The Puck
Materiali in tehnologije / Materials and technology 50 (2016) 3, 319–322 319
UDK 621.785.7:620.168:620.17 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)319(2016)
criterion for the 3D stress state, described in 2,4, was
selected for this model because it provides the fracture
angle fr, later used for the stiffness degradation. Further-
more, the influence of the fibre parallel-stress extension
and the influence of the non-fracture plane extension
were used with this criterion.4
4 PROGRESSIVE FAILURE IN THE CASE OF
INTER-FIBRE FAILURE
The stiffness-matrix degradation method was used to
simulate the progressive failure. In order to simplify the
determination of the degradation parameters, the stiff-
ness matrix C, in UMAT, called DDSDDE, was trans-
formed from the material coordinate system (1, 2, 3) to
the crack coordinate system (x, y, z) described in Fig-
ure 1.
The transformation of the C matrix in the (1, 2, 3)
system to the C’ in the (x, y, z) system was carried out
using the Equation (3):
C T C T' ( )  fr = ⋅ ⋅
−1 (3)
where
T
c s sc
s c sc
c s
s c
sc sc
 =
−
−
−
1 0 0 0 0 0
0 0 0 2
0 0 0 2
0 0 0 0
0 0 0 0
0 0
2 2
2 2
0 2 2c s−
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
⎥
⎥
(4)
is the transformation matrix for the stress vector and
T
−1 is the inverted transformation matrix
T
c s sc
s c sc
c s
s c
sc sc
 =
−
−
−
1 0 0 0 0 0
0 0 0
0 0 0
0 0 0 0
0 0 0 0
0 2 2 0
2 2
2 2
0 2 2c s−
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
⎥
⎥
(5)
for the strain vector. In Equations (4) and (5), c repre-
sents cos fr and s represents sin fr.
The non-zero components of the C’ matrix Cij’ are
multiplied by (1 – dij) terms. The degradation parameters
dij ∈ 01, are constant values and differ for tensile and
compressive failure.
Afterwards, the C’ matrix is transformed back from
the (x, y, z) system to the (1, 2, 3) system using the trans-
formation matrices:
C d T C d T' ' ( ) ' ( , )ij fr ij= ⋅ ⋅
−
 
1 (6)
5 PROGRESSIVE FAILURE IN THE CASE OF
FIBRE FAILURE
The transformation of the C matrix is not necessary.
Therefore, the non-zero components of the C matrix Cij
are only multiplied by (1 – dij) terms, as in the case of
inter-fibre failure.
6 DELAMINATION
During the testing it was observed that delamination
must be considered because after the initial fibre or
inter-fibre failure, the crack often propagates in the form
of a delamination. Therefore, an approach for the simula-
tion of the delamination was also implemented.
A thin isotropic layer of brittle matrix was inserted
between each of the orthotropic layers in the FE model.
For the prediction of the matrix failure, the maximum
stress criterion, originally used for orthotropic materials,
was considered because it provides information about
which stress component suffered failure. The normal
stress components were compared to the compressive
and tensile strengths of the matrix, while the shear com-
ponents were compared to the shear strength of the
matrix.
In the case of the failure, the non-zero components of
the C matrix Cij are again multiplied by (1 – dij) terms as
in the case of inter-fibre failure.
320 Materiali in tehnologije / Materials and technology 50 (2016) 3, 319–322
L. BEK, R. ZEM^ÍK: MODEL OF PROGRESSIVE FAILURE FOR COMPOSITE MATERIALS ...
Figure 1: Description of the material coordinate system (1, 2, 3) and
the crack coordinate system (x, y, z)
Slika 1: Opis koordinatnega sistema materiala (1, 2, 3) in koordinat-
nega sistema razpoke (x, y, z)
Figure 2: Geometric properties of the pin-loaded plate
Slika 2: Geometrijske lastnosti s konico obremenjene plo{~e
7 CASE STUDY 1 – PIN JOINT
First, in order to test the model, the failure simula-
tions of pin-loaded carbon composite plates were
compared with the experiments. Two types of specimens
with different failure modes (shear-out and net-tension5)
were selected for the failure simulation. The geometric
properties of the specimens are described in Figure 2,
where the 0° layup orientation is parallel to the y axis
and the pin diameter D = 8 mm.
The failure simulation for the first type of specimens
with the shear-out failure mode, a composite layup
[0°|45°|–45°|90°] s, ratios E/D = 1 and W/D = 3, and a
thickness t = 2.32 mm, is illustrated in Figure 3. The
black colour indicates the elements with a degraded
stiffness matrix and represents the failure of the material.
All the layers representing the isotropic matrix were also
degraded. The error for the ultimate load F was 6.8 %
(compared to the average value from the experiments).
The error for the ultimate load F investigated using
the failure simulation of the second type of specimens
with a net-tension failure mode, a composite layup
[90°|45°|–45°|0°] s, ratios E/D = 4 and W/D = 2, and a
thickness t = 2.32 mm was 10.9 % (compared to the ave-
rage value from the experiments as well).
L. BEK, R. ZEM^ÍK: MODEL OF PROGRESSIVE FAILURE FOR COMPOSITE MATERIALS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 319–322 321
Figure 4: Load-displacement diagrams of the experiment and the
numerical simulation: a) first failure investigated using experiment, b)
first failure investigated using numerical simulation and c) loss of
numerical stability
Slika 4: Diagram obremenitev-raztezek eksperimenta in numeri~ne
simulacije: a) prva preiskovana poru{itev pri eksperimentu, b) prva
preiskovana poru{itev pri numeri~ni simulaciji in c) izguba numeri~ne
stabilnosti
Figure 3: Numerical simulation of the final shape of failure in the
case of specimens with a shear-out failure mode; different layers
displayed
Slika 3: Numeri~na simulacija kon~ne oblike poru{itve v primeru
vzorca s poru{itvijo z izstri`enjem; prikazane so razli~ne plasti
Figure 5: Comparison of the position and shape of the first failure
investigated using the experiment and the numerical simulation
Slika 5: Primerjava polo`aja in oblike prve poru{itve pri preizkusu in
pri numeri~ni simulaciji
8 CASE STUDY 2 – COMPOSITE TUBE
In addition, the testing was carried out on a thin-
walled composite tube subjected to compressive loading
perpendicular to the tube’s axis. The tube consisted of
carbon fibres with a composite layup [45°|–45°], a
wall-thickness of 1 mm and an outer diameter of 42 mm.
The length of the tested tube was 200 mm.
A stiffness comparison of the experiment and the
numerical simulation is illustrated in Figure 4. A com-
parison of the position and the shape of the first failure
investigated using the experiment (Figure 4a) and the
numerical simulation (Figure 4b) is illustrated in Figure
5. Unfortunately, the numerical model was not able to
simulate the whole specimen failure due to a loss of
numerical stability. In Figure 6, the failure just before
the loss of numerical stability in both layers is illu-
strated. The error of the simulation at this point (Figure
4c) is 13.8 %.
9 CONCLUSION
Our model of progressive failure using the extended
Puck failure criterion for the 3D stress state and con-
sidering the simplified approach for the simulation of
delamination and the material’s non-linearity showed
very good agreement between the numerical simulation
and the experiments. The error for all the simulations
was below 14 %. In future work, the problem of nume-
rical stability will be further investigated.
Acknowledgements
This work was supported by the European Regional
Development Fund (ERDF), project "NTIS – New Tech-
nologies for Information Society", European Centre of
Excellence, CZ.1.05/1.1.00/02.0090, by the research
project GACR_P101/11/0288 and by the grant project
SGS-2013-036.
10 REFERENCES
1 V. La{ová, P. Bernardin, Numerical modelling of glued joints bet-
ween metal and fibre composites using cohesive elements, Applied
Mechanics and Materials, 611 (2014), 156–161, doi:10.4028/
www.scientific.net/AMM.611.156
2 H. M. Deuschle, 3D Failure Analysis of UD Fibre Reinforced Com-
posites: Puck’s Theory within FEA, Institut für Statik und Dynamik
der Luft- und Raumfahrtkonstruktionen, Universität Stuttgart,
Stuttgart 2010
3 J. Krystek, T. Kroupa, R. Kottner, Identification of mechanical pro-
perties from tensile and compression tests of unidirectional carbon
composite, 48th International Scientific Conference proceedings:
Experimental Stress Analysis 2010, Palacky University, 2010,
193–200
4 A. Puck, Festigkeitsanalyse von Faser-Matrix-Laminaten: Modele
für die Praxis, Carl Hanser Verlag, München, Wien 1996
5 H. Schürmann, Konstruieren mit Faser-Kunststoff Verbunden, Sprin-
ger Verlag, Berlin, Heidelberg 2007, doi:10.1007/978-3-540-72190-1
L. BEK, R. ZEM^ÍK: MODEL OF PROGRESSIVE FAILURE FOR COMPOSITE MATERIALS ...
322 Materiali in tehnologije / Materials and technology 50 (2016) 3, 319–322
Figure 6: Numerical simulation of the shape and the position of
failure just before the loss of numerical stability
Slika 6: Numeri~na simulacija oblike in polo`aja po{kodbe tik pred
izgubo numeri~ne stabilnosti
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
323–329
PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO
AND STEAM STERILIZATION
FIZIKALNO KEMIJSKE LASTNOSTI ZLITINE Ti67 PO EO IN
PARNI STERILIZACIJI
Witold Walke1, Marcin Basiaga1, Zbigniew Paszenda1, Jan Marciniak,
Pawe³ Karasinski2
1Silesian University of Technology, Faculty of Biomedical Engineering, Roosevelta 40, 41-800 Zabrze, Poland
2Silesian University of Technology, Faculty of Electrical Engineering, Boles³awa Krzywoustego 2, 44-100 Gliwice, Poland
witold.walke@polsl.pl
Prejem rokopisa – received: 2014-09-23; sprejem za objavo – accepted for publication: 2015-05-15
doi:10.17222/mit.2014.242
The techniques of surface modification play a significant role in forming the physical and chemical properties of titanium and
its alloys. Among many techniques for the layers’ application, chemical and electrochemical methods are particularly intere-
sting, as they make it possible to control the process of depositing thin layers of the material and modifying their properties
through a change of reagents and the parameters of the deposition process. A special advantage the methods bring is the possibi-
lity to obtain layers that offer a perfect coating for geometrically complex surfaces. Apart from improved haemocompatibility, a
significant issue related to the creation of the layers is also a proper set of physicochemical properties. Therefore, the study
comprised tests of the physicochemical properties of oxide layers deposited on the surface of samples taken from a Ti-6Al-7Nb
alloy. The samples were subject to various surface modifications, i.e., grinding, electrolytic polishing, a SiO2 layer was applied
using the sol-gel method and TiO2 by means of an anodic oxide and medical sterilisation methods (EO and steam). The
corrosion-resistance tests were performed on the basis of registered anodic polarisation curves and the Stern method.
Electrochemical Impedance Spectroscopy (EIS) was also used in order to evaluate the phenomena taking place on the surface of
the tested alloys. As a part of the evaluation of the mechanical properties of surface layers created in such a way, hardness tests
and tests of the adhesion of those layers to a metallic substrate were made. Measurements of the instrumental hardness were
made with the Oliver & Pharr method, whereas the adhesion of the layers to the substrate was measured by means of a scratch
test. The suggestion of proper surface treatment variants has perspective significance and will help to develop the technological
conditions with specified parameters of the oxide coating’s creation on the surface of metallic implants.
Keywords: Ti-6Al-7Nb (Ti67) alloy, sol-gel, anodic oxide, scratch-test, nanohardness, EIS, potentiodynamic method
Tehnike modifikacije povr{ine igrajo pomembno vlogo pri doseganju fizikalnih in kemijskih lastnosti titana in njegovih zlitin.
Med mnogimi tehnikami uporabe tankih plasti so {e posebej zanimive kemijske in elektrokemijske metode, ker omogo~ajo
kontrolo postopka depozicije tanke plasti materiala in spreminjanje njihovih lastnosti, z zamenjavo reagentov in parametrov
procesa depozicije. Posebna prednost, ki jo omogo~ajo metode, je mo`nost izdelave plasti, ki omogo~ajo popolno prevleko pri
geometrijsko kompleksnih povr{inah. Poleg izbolj{ane hemokompatibilnosti je pomembno vpra{anje povezano z nastankom
plasti in tudi z doseganjem ustreznih fizikalno kemijskih lastnosti. Zato je {tudija obsegala preizkuse fizikalno kemijskih last-
nosti oksidnih plasti, po depoziciji na povr{ini vzorcev iz Ti-6Al-7Nb zlitine. Na vzorcih so bile opravljene razli~ne modifika-
cije povr{ine: bru{enje, elektrolitsko poliranje, SiO2 plast, izdelana po sol-gel postopku in TiO2 plast, izdelana z anodno
oksidacijo in metodo medicinske sterilizacije (EO in para). Preizkusi korozijske odpornosti so bili izvr{eni na osnovi zabele-
`enih polarizacijskih krivulj in metode Oliver & Pharr. Za oceno pojavov na povr{ini preizku{enih zlitin je bila uporabljena tudi
elektrokemijska impedan~na spektroskopija. Kot del ocene mehanskih lastnosti nastalih plasti na povr{ini, so bile izvr{ene tudi
meritve trdote in preizkusi oprijemljivosti plasti na kovinski podlagi. Meritve trdote so bile izvr{ene z metodo Oliver & Phar,
medtem ko je bila oprijemljivost izmerjena s preizkusom razenja. Predlog primernih na~inov obdelave povr{ine je pomemben,
ker bo v prihodnje pomagal pri razvoju tehnolo{kih pogojev za dolo~ene parametere oksidne plasti, nastale na povr{ini
kovinskih vsadkov.
Klju~ne besede: Ti-6Al-7Nb (Ti67) zlitina, sol-gel, anodni oksid, preizkus razenja, nanotrdota, EIS, potenciodinami~na metoda
1 INTRODUCTION
The haemocompatibility of titanium alloys, including
the Ti-6Al-7Nb alloy, is increased by, e.g., modification
of the surface layer of cardiovascular implants with sur-
face-engineering methods. The methods used to modify
the surface layers must ensure the repeatability and
uniformity of their physical and chemical properties.1
The structure and chemical composition of the titanium
and titanium-alloy implant layer may be modified with
the use of various methods, among which the main ones
are mechanical, chemical, electrochemical and thermal
methods. Mechanical treatment techniques are used to
modify the surface topography. The properties of the
oxide layer after the application of these techniques are
difficult to control.2 Chemical methods include primarily
etching and passivation, which result in the formation of
a thin (<10 nm) oxide layer composed mostly of TiO2
and oxides of the alloying elements, as well as impurities
from chemical reagents.3–6 Repeatable layers with a fully
controlled thickness, microstructure and chemical com-
position are obtained with high-temperature treatments,
immersion in H2O2, alkaline etching, electropolishing,
anodic oxidation and vacuum treatments. However, the
Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329 323
UDK 67.017:669.295:669.148 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)323(2016)
method for increasing the haemocompatibility of tita-
nium and titanium-alloy surfaces, which is increasingly
often applied, involves using the sol-gel technique to
produce thin oxide coatings based on Si. The advantage
of this method is the low temperature at which the
coating is produced, which guarantees unchanged me-
chanical properties of the metal base. Moreover, this
method ensures the sol’s homogeneity, the possibility to
regulate polycondensate molecules, a large number of
metalorganic and inorganic metal salt compounds, used
as precursors, as well as the possibility to obtain multi-
ingredient coatings of high purity on different bases.7,8
Another important factor affecting the final quality of the
products that come into contact with blood is the proper
resistance of the modified surface to medical sterilisa-
tion. Presently, cardiovascular implants are usually steri-
lised with ethylene oxide (EO) and with pressurised-
water steam in an autoclave. The positive results of
already-published papers by the authors regarding the
assessment of the usefulness of the surface-layer modi-
fication processes involving anodic oxidation, as well as
the creation of a SiO2 layer with the sol-gel method,
enabled the selection of the most beneficial parameters
for the process.1,9–11 An analysis of the literature data
indicates that reducing the number of failed blood-
system disease treatments depends to a large extent on
the electrochemical stability of the surface layer under
the medical sterilisation conditions. Therefore, this
article evaluates the effects of steam and ethylene oxide
sterilisation processes on the physical and chemical
properties of the surface layer of the Ti-6Al-7Nb (Ti67)
alloy.
2 MATERIALS AND METHODS
The tested material was a titanium alloy Ti-6Al-7Nb
(Ti67) in the form of discs of diameter d = 14 mm and
thickness g = 2 mm. A number of surface-treatment
methods were applied to the samples, including the
following processes: grinding with the use of 1000- and
1200-grit sand paper, electrolytic polishing, the appli-
cation of layers with the sol-gel technique, and anodic
oxidation. The electrolytic polishing was conducted in a
solution based on chromic acid z (E-395 by POLIGRAT
Gmbh), with a current density i = 10–30 A/cm2. The
final stage of the surface treatment consisted of applying
layers with two different techniques: sol-gel and anodic
oxidation. In the case of the sol-gel method, a layer of
SiO2 was applied with the following process parameters:
v = 2.5 cm/min, T = 430°C, t = 60 min. The silica pre-
cursor used in the test was tetraethoxysilane Si(OC2H5)4,
TEOS, and tetramethoxysilane Si(OCH3)4, TMOS. The
remaining starting ingredients contained ethyl alcohol
(EtOH) and water.1,9 In the case of anodic oxidation a
layer of TiO2 was applied in an electrolyte based on
phosphoric acid and sulphuric acid (TitanColor by
POLIGRAT GmbH) at a potential of 90 V. Previous
studies conducted by the authors made it possible to
select the most favourable parameters, both for the
sol-gel technique and for the anodic oxidation.10,11
Next, the prepared samples were sterilised with ethy-
lene oxide and steam. The sterilisation with ethylene
oxide was conducted in a 12-h cycle of exposure to ethy-
lene oxide at 30 °C. After the process was completed, the
samples were ventilated for 2 h with the use of an
EOGas series steriliser from the Andersen Products com-
pany. The sterility assurance level (SAL) obtained during
the cycle was 10–6. The process was controlled with a
chemical and biological indicator, as well as an indicator
of exposure to the ethylene oxide control. The steam
sterilisation was conducted in a Basic Plus autoclave at
T = 134 °C, under a pressure of p = 2.1 bar for t = 12
min.
To evaluate the effect of ethylene oxide sterilisation
and steam sterilisation on the mechanical and electroche-
mical properties of the proposed Ti67 surface modifi-
cation, the authors suggested the following tests of the
mechanical properties: measurements of the adhesion of
the analysed layers to the base and their hardness.
Electrochemical tests included potentiodynamic and
impedance measurements.
First, as part of the mechanical properties’ tests, the
measurement of a layer’s adhesion to the base was
performed using the scratching method, with the use of
an open platform equipped with a Micro-Combi-Tester
from the CSM company, in accordance with the stan-
dard.12 The test consisted of making a scratch using a
penetrator – a Rockwell diamond cone – with a gra-
dually increasing normal force weighting the penetrator.
To assess the value of the critical force Lc, records of
variations in the acoustic emission signals, the friction
force and the friction coefficient were used, as well as a
microscopic observation with the use of an optical
microscope, and an integral component of the platform.
The tests were performed with an increasing weighting
force of 0.03–20 N, and with the following parameters:
weighting speed 10N/min, table movement speed 1.5
mm/min, and length of the scratch ~3 mm.
Later, measurements of the nanohardness of the
layers applied with the sol-gel method, and with the
anodic oxidation method, were conducted. The instru-
mental hardness measurement was performed with the
Oliver and Pharr method, using a Berkovich penetrator.
The speed of the increasing weighting and relieving
force was 0.40 mN/min. The measurement of the layer’s
nanohardness was made with the Micro-Combi-Tester
open platform from the CSM Instruments company,
where the weighting force of the penetrator was 0.20
mN.13
Subsequently, as part of the electrochemical proper-
ties testing, the resistance to pitting erosion was tested
with the potentiodynamic method, recording the polari-
sation curves. They were used as a base to determine the
values of specific parameters: the corrosion potential
324 Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
Ecorr (V) and the polarisation potential Rp (	cm2). At the
beginning of the testing, the value of the opening po-
tential EOCP was determined without any electric current.
Then, the anode polarisation curves were recorded. The
measurements started for the potential of Estart = EOCP –
100 mV, and the change of potential in the anodic
direction was at a speed of 0.16 mV/s, until the anodic
current density reached a value of i = 1 mA/cm2, or the
measurement range of 4V was reached.14
As part of the EIS testing, the impedance spectra of
the analysed corrosive systems were determined, and
then the obtained measurement data were adjusted to the
corresponding substitute systems. The impedance spectra
of the systems tested are presented on Nyquist diagrams
for the different frequencies as well as on Bode
diagrams. Also, the numerical values of the resistance R
were established, as well as the capacities C of the
analysed corrosive systems. The resulting spectra were
interpreted after being adjusted with the least-squares
method to the substitute electric systems. Based on the
results obtained, it was possible to characterise the impe-
dance of the phase boundaries, i.e., Ti-6Al-7Nb (Ti67) –
surface layer – blood plasma, with an approximation of
the impedance data using an electric model of a sub-
stitute circuit. The testing environment was an artificial
blood-plasma solution of T = 37±1 °C. The measure-
ments were performed using the AutoLab PGSTAT
302N measurement system, equipped with a FRA2 (Fre-
quency Response Analyser) module. The reference elec-
trode was a saturated calomel electrode SCE, type
KP-113, whereas the supporting electrode was a plati-
num electrode type PtP-201. The system used made it
possible to conduct tests within the frequency range 104
to 10–3 Hz. The voltage amplitude of the sinusoid stimu-
lating signal was 10 mV.15, 16
Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329 325
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
Figure 1: Results of the adhesion tests of the sample Ti67+TiO2 (steam)
Slika 1: Rezultati preizkusov oprijemljivosti vzorca Ti67+TiO2 (para)
Table 1: The results of the adhesion of the layer on the Ti67 substrate
Tabela 1: Oprijemljivost plasti na podlagi iz Ti67
Failure of the layer
The value of registered indenter load Fn, N
Ti67+SiO2 Ti67+TiO2
inital state steam EO inital state steam EO
Measure-
ment 1
Delamination Lc1 2.01 1.57 1.54 6.64 2.98 3.14
Complete break Lc2 3.57 2.24 3.21 8.01 6.54 5.03
Measure-
ment 2
Delamination Lc1 2.48 1.33 1.89 3.06 3.14 2.36
Complete break Lc2 3.89 2.45 3.22 6.88 6.18 4.27
Measure-
ment 3
Delamination Lc1 3.55 1.41 2.41 4.33 3.52 2.86
Complete break Lc2 5.02 2.12 4.31 7.37 6.88 4.81
Average
Delamination Lc1 2.68 1.43 1.94 4.67 3.21 2.78
Complete break Lc2 4.16 2.27 3.58 7.42 6.53 4.70
Standard
deviation
Delamination Lc1 ±0.78 ±0.12 ±0.43 ±1.81 ±0.59 ±0.39
Complete break Lc2 ±0.76 ±0.16 ±0.63 ±0.56 ±0.35 ±0.39
3 RESULTS AND DISSCUSION
The test results for the adhesion of the analysed
layers to the base made of Ti-6Al-7Nb (Ti67) alloy are
presented in Table 1, Figures 1 and 2. It was found that
in the case of samples in the initial state the critical value
that caused the layer delamination, the external and the
internal delamination, was Lc2 = 4.16 N – Ti67+SiO2 and
Lc2 = 7.42 N – Ti67+TiO2. While using both steam and
ethylene oxide sterilisation, the critical force value was
reduced and for Ti67+SiO2 it was Lc2 = 2.27 N (steam),
Lc2 = 3.58 N (EO), whereas for Ti67+TiO2 it was Lc2 =
6.53 N (steam), Lc2 = 4.70 N (EO). Regardless of the
analysed sample type, an acoustic emission signal did
not occur during the test, which indicates that the
binding energy between the coating and the base was too
low. Moreover, no significant differences between using
steam sterilisation or ethylene oxide sterilisation were
found.
Furthermore, the hardness of the analysed layers was
tested. The test results are presented in Figures 3 and 4.
On the basis of the results obtained, an increase in the
hardness value following the steam sterilisation, as well
as the ethylene oxide sterilisation, compared to the initial
state was observed. The polarisation curves determined
for the samples with a Ti67(TiO2) layer are presented in
Figure 5, and for the samples with a Ti67(SiO2) layer in
Figure 6.
Regardless of the surface-preparation method or the
sterilisation technique, a hysteresis loop was not present
in the anodic range up to 4 V, which is a positive pheno-
menon, indicating the absence of pitting erosion. The deter-
mined values of the corrosive potential Ecorr and the polari-
sation resistance Rp for the individual variant of the samples
tested were as follows: Ti67+TiO2 – Ecorr = –112 mV,
326 Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
Figure 2: Results of the adhesion tests of the sample Ti67+SiO2 (steam)
Slika 2: Rezultati preizkusov oprijemljivosti vzorca Ti67+SiO2 (para)
Figure 4: Results of the nanohardness tests of the sample Ti67+TiO2
Slika 4: Nanotrdota vzorca Ti67+TiO2
Figure 3: Results of the nanohardness tests of the sample Ti67+SiO2
Slika 3: Nanotrdota vzorca Ti67+SiO2
Rp = 7480 k	 cm2; Ti67+TiO2 (steam) – Ecorr = –67 mV,
Rp = 3340 k	 cm2; Ti67+TiO2 (EO) – Ecorr = –217 mV,
Rp = 8340 k	 cm2; Ti67+SiO2 – Ecorr = –108 mV, Rp =
1460 k	 cm2; Ti67+SiO2 (steam) – Ecorr = –118 mV, Rp =
1850 k	 cm2; Ti67+SiO2 (EO) – Ecorr = –219 mV, Rp =
4290 k	 cm2, respectively. Figures 7 and 8 present the
impedance spectra recorded for the samples before and
after the sterilisation process for different variants of the
surface preparation. To analyse the experimentally deter-
mined impedance spectra for the corrosive system of
Ti67+SiO2, Ti67+SiO2 (steam), Ti67+SiO2 (EO), a sub-
stitute electric system was used, which indicates the
presence of a double layer (two time invariables visible
in the diagram), where Rs signifies the electrolyte resis-
tance, Rp is the electrolyte resistance in pores, and CPEp
is the capacity of the double layer (porous layer, surface
layer), while Rct and CPEdl are the resistance and capa-
city of the oxide layer. Using two constant phase
elements in the electric substitute circuit had an advan-
tageous effect on the quality of the adjustment of the
experimentally determined curves (Figure 9a and Table
2).
The impedance spectra obtained for the Ti67+TiO2
and Ti67+TiO2 (steam) samples were interpreted by
comparing them to the substitute electric system, which
indicates the presence of an anodic layer composed of
two sublayers:9–11 a compact internal layer and a porous
external one, composed primarily of titanium oxide TiO2
(Figure 9b). It is indicated by the presence of the War-
burg impedance, which in this case represents probable
oxygen transport to the alloy surface. In addition, the
CPEp element models the capacity of the surface mate-
rial sphere with a significant surface extension, while Rp
Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329 327
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
Figure 7: Impedance spectra for the sample Ti67+SiO2: a) Nyquist
plot, b) Bode diagram
Slika 7: Impedan~ni spekter vzorca Ti67+SiO2: a) diagram Nyquist,
b) diagram Bode
Figure 6: Anodic polarisation curves of Ti67+SiO2
Slika 6: Krivulje anodne polatizacije Ti67+SiO2
Table 2: EIS analysis results
Tabela 2: Rezultati EIS-analize
Surface
Rs,
	 cm2
Rpore,
	 cm2
CPEpore Rp
k	 cm2
CPEp Rct
M	 cm2
CPEdl W
μ	 cm2Y0,	–1 cm–2 s–n n
Y0,
	–1 cm–2 s–n n
Y0,
	–1 cm–2 s–n n
Ti67(TiO2)
inital state 17 – – – 68 0,3580E-4 0,92 25,00 0,2306E-4 0,90 3
EO 18 54 0,5407E-6 0,93 1830 0,6381E-6 0,83 20,88 0,1455E-5 0,81 –
steam 17 – – – 42 0,2534E-6 0,96 0,96 0,2504E-6 0,90 35
Ti67(SiO2)
inital state 17 53 0,9823E-5 0,98 9,44 0,5294E-5 0,93 –
EO 18 – – – 52 0,7975E-5 0,92 11,87 0,1068E-4 0,87 –
steam 18 – – – 87 0,1213E-4 0,89 4,84 0,1524E-6 0,96 –
Figure 5: Anodic polarisation curves of Ti67+TiO2
Slika 5: Krivulje anodne polarizacije Ti67+TiO2
reflects the electrolyte resistance in this sphere of the
material (Table 2).
The impedance spectra obtained for the Ti67+TiO2
(EO) sample were adjusted to the substitute system,
which indicates the presence of three time invariables
(Figure 9c). The symbols in Figure 9c signify the
following: CPEpore – capacity of the surface sphere of the
material with a high level of surface extension (porous),
Rpore – electrolyte resistance in pores, CPEdl – capacity of
the double layer, Rct – charge-transfer resistance at the
phase boundary (it characterises the speed of the corro-
sive process), Cdl – capacity of the double layer, CPEP –
capacity of the passive layer (oxide), RP – passive (oxide)
layer resistance (Table 2).9–11
4 CONCLUSIONS
An important problem in the process of modelling
the performance properties of the implants used in car-
diology is the proper selection of the physical and
chemical characteristics of their surface. The physico-
chemical properties of the implant surface should be
adjusted to the characteristics of the human tissue
environment – in this case, to the blood environment.
The safety of the device’s use is also associated with the
need to follow the proper procedures preventing the
transfer of pathogenic microorganisms into the human
organism. The aim of these procedures is to remove and
effectively destroy the microorganisms, i.e., to obtain
sterile devices that meet the quality requirements defined
in the standards. For medical devices that come into
contact with blood, sterilisation with ethylene oxide or
pressurised-water steam are the most frequently used.
Therefore, accounting for the effect of these sterilisation
processes on the properties of the analysed surface layer
will enable their complete characterisation.17
The conducted impedance tests revealed that on the
surface of the Ti-6Al-7Nb alloy, modified through ano-
dic oxidation, a porous layer (TiO2) is found, in which
parallel channels with an ionic conduction are formed. It
is a layer that forms as a result of diffusion processes,
which is indicated by the presence of the Warburg impe-
dance. These processes intensify during the activity of
pressurised-water steam in the sterilisation process. As a
consequence, it leads to the partial dissolution of TiO2,
as indicated by the lower value of the ionic transfer
resistance Rct. Sterilisation with ethylene oxide positively
increases the Rct value. This phenomenon may be caused
by the increased oxygen concentration near the surface
during the process, and the formation of an additional
oxide layer. Diffusion processes associated with the
partial dissolution of oxide in the solution were not
observed in the samples with an applied SiO2 layer. This
layer appears to be more compact than the TiO2. Regard-
less of the sterilisation method used, no changes in the
properties of the layer were found. Only, as was the case
with the samples undergoing anodic oxidation, a reduc-
tion in the ion-transfer resistance Rct was observed. The
conducted tests for layer adhesion to the base revealed a
slight reduction in the adhesive force of the sterilised
layers versus the samples in the initial state. The tests de-
monstrated the better adhesion of TiO2 than SiO2 to the
Ti-6Al-7Nb alloy base. The hardness testing conducted
in the study revealed that pressurised steam does not
cause significant changes to the hardness of the TiO2 or
SiO2, whereas sterilisation with ethylene oxide results in
an increased hardness of the SiO2 and a significant
reduction in the hardness of the TiO2. This phenomenon
328 Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION
Figure 9: Physical models of an electrical equivalent system of the
corrosion system metal – solution9–11
Slika 9: Fizikalni model elektri~nega ekvivalentnega sistema korozij-
skega sistema kovina-raztopina9–11
Figure 8: Impedance spectra for the sample Ti67+TiO2: a) Nyquist
plot, b) Bode diagram
Slika 8: Impedan~ni spekter vzorca Ti67+TiO2: a) diagram Nyquist,
b) diagram Bode
may cause increased porosity of the TiO2 layer, resulting
from the effect of ethylene oxide, as demonstrated in the
EIS tests. The SiO2 layer also reacts when in contact
with ethylene oxide. Its hardness significantly increases,
which may be the cause of the formation on its surface of
an additional oxide layer, based on Ti and Si, revealing
better mechanical properties. To sum up, the conducted
study of the modified surfaces of the Ti-6Al-7Nb alloy
samples with the TiO2 and SiO2 layers demonstrated that
the medical sterilisation process affects the physical and
chemical properties of these layers. The selection of the
proper surface layers should also depend on the manner
and the method of sterilisation.
Acknowledgements
The project was funded by the National Science
Centre, allocated on the basis of decision No. 2011/03/
B/ST8/06499.
5 REFERENCES
1 W. Walke, Z. Paszenda, M. Basiaga, P. Karasiñski, M. Kaczmarek,
EIS study of SiO2 oxide film on 316L stainless steel for cardiac
implants. Information Technologies in Biomedicine, 4, Advances in
Intelligent Systems and Computing, 284 (2014), 403–410
doi:10.1007/978-3-319-06596-0_38
2 N. Ibris, J. C. M. Rosca, EIS study of Ti and its alloys in biological
media, Journal of Electroanalytical Chemistry, 526 (2002), 53–62,
doi:10.1016/S0022-0728(02)00814-8
3 J. Y. Park, J. E. Davies, Red blood cell and platelet interactions with
titanium implant surfaces, Clinical Oral Implants Research, 11
(2000) 6, 530–539, doi:10.1034/j.1600-0501.2000.011006530.x
4 J. E. Ellingsen, Pre-treatment of titanium implants with fluoride im-
proves their retention in bone, Biomaterials, 12 (1991) 6, 593–596,
doi:10.1016/ 0142-9612(91)90057-H
5 M. Basiaga, Z. Paszenda, A. Kajzer, The effect of EO and steam ste-
rilization on mechanical and electrochemical properties of titanium
Grade 4, Mater. Tehnol., 50 (2016) 1, 153–158 doi:10.17222/mit.
2014.241
6 C. Sitting, G. Hahner, A. Marti, M. Textor, N. D. Spencer, R. Hauert,
Mechanoelectrochemical synthesis of porous Ti-based nanocompo-
site biomaterials, Journal of Materials Science: Materials in Medi-
cine, 10 (1999), 191–198
7 S. Shibli, S. Mathai, The role of calcium gluconate in electroche-
mical activation of titanium for biomimetic coating of calcium
phosphate, J. Mat. Sci., 19 (2008), 2971–2981
8 P. Karasiñski, J. Jaglarz, M. Reben, E. Skoczek, J. Mazur, Porous
silica xerogel films as antireflective coatings. Fabrication and
characterization, Optical Materials, 33 (2011), 1989–1994
9 M. Basiaga, Z. Paszenda, W. Walke, P. Karasiñski, J. Marciniak,
Electrochemical impedance spectroscopy and corrosion resistance of
SiO2 coated cpTi and Ti-6Al-7Nb alloy, Information Technologies in
Biomedicine, 284 (2014), 411–420, doi:10.1007/978-3-319-06596-
0_39
10 W. Walke, M. Basiaga, Z. Paszenda, Tests of mechanical properties
of SiO2 layers for implants used in vascular system, Engineering of
Biomaterials, 124 (2014), 36–41
11 A. Kajzer, W. Kajzer, J. Dzielicki, D. Matejczyk, The study of physi-
cochemical properties of stabilizing plates removed from the body
after treatment of pectus excavatum. Acta of Bioengineering and
Biomechanics, 2 (2015), 35–44 doi:10.5277/ABB-00140-2014-02
12 J. von Stebut, Multi-mode scratch testing – a European standards,
measurements and testing study, Surface and Coatings Technology,
200 (2005), 346–350, doi:10.1016/j.surfcoat.2005.02.055
13 EN ISO 14577-1 Metallic materials-Instrumented indentation test for
hardness and materials parameters-Part1: Test method
14 ASTM F2129-08 Standard Test Method for Conducting Cyclic
Potentiodynamic Polarization Measurements to Determine the
Corrosion Susceptibility of Small Implant Devices
15 M. Kaczmarek, W. Walke, Z. Paszenda, Application of electroche-
mical impedance spectroscopy in evaluation of corrosion resistance
of Ni-Ti alloy, Electrical Review, 12b (2011), 74–78
16 M. Kiel-Jamrozik, J. Szewczenko, W. Walke, J. Marciniak, Zastoso-
wanie EIS do oceny w³asnoœci fizykochemicznych modyfikowanego
powierzchniowo stopu Ti-6Al-4V ELI, Electrical Review, 12b
(2012), 232–235
17 W. Walke, J. Przondziono, J. Szade, Electrochemical and XPS
studies of AISI 316L after Development of warm forging method for
magnesium alloyEuropean Cells and Materials, 26 (2013) 6, 117
Materiali in tehnologije / Materials and technology 50 (2016) 3, 323–329 329
W. WALKE et al.: PHYSICOCHEMICAL PROPERTIES OF A Ti67 ALLOY AFTER EO AND STEAM STERILIZATION

I. MICHALJANICOVA et al.: SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER
331–335
SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER
LASTNOSTI POVR[INE BIOPOLIMERA, OBDELANEGA Z
LASERJEM
Iva Michaljani~ova1, Petr Slepi~ka1, Silvie Rimpelova2, Petr Sajdl3, Václav [vor~ík1
1University of Chemistry and Technology, Department of Solid State Engineering, Technická 5, 166 28 Prague, Czech Republic
2University of Chemistry and Technology, Department of Biochemistry and Microbiology, Technická 5, 166 28 Prague, Czech Republic
3Institute of Chemical Technology, Department of Power Engineering, Technická 5, 166 28 Prague, Czech Republic
iva.michaljanicova@vscht.cz
Prejem rokopisa – received: 2014-10-14; sprejem za objavo – accepted for publication: 2015-05-04
doi:10.17222/mit.2014.260
Structured surfaces allow the application of commonly used polymers to be extended into specialized fields. This paper
describes the construction of surface structures on biopolymer poly(L-lactide) (PLLA), with a method combining krypton
fluoride laser (KrF), excimer laser exposure and thermal annealing. PLLA is a commonly used substrate for medical purposes,
such as implants and tissue matrices, but it still has a number of limitations, which can be eliminated with its modification. This
work is focused on morphological studies and roughness measurements of a structured PLLA substrate using atomic-force
microscopy (AFM) and chemical changes investigated with UV-Vis spectroscopy and X-ray photoelectron spectroscopy (XPS).
Finally, the biocompatibility of the material was tested using a model cell line of mouse embryonic fibroblasts (NIH 3T3).
Using the laser treatment in combination with thermal annealing, we prepared surface layers with various patterns dependent on
the chosen input parameters.
Keywords: biopolymer, excimer laser, nanostructuring, thermal annealing, characterization
Strukturirane povr{ine omogo~ajo raz{iritev uporabe obi~ajnih polimerov tudi na posebna podro~ja. ^lanek opisuje pripravo
povr{inske zgradbe biopolimera poly(L-lactide) (PLLA), s kombinirano metodo, z izpostavitvijo excimer laserju (KrF) in
toplotno obdelavo. PLLA je obi~ajno uporabljena osnova za medicinske namene kot so vsadki in osnove tkiv, vendar so {e
{tevilne omejitve, ki se jih da odpraviti z njihovim modificiranjem. ^lanek je osredoto~en na {tudij morfologije in meritve
hrapavosti strukturirane PLLA podlage, z mikroskopijo na atomsko silo (AFFM) in preiskave kemijskih sprememb z UV-Vis
spektroskopijo in rentgensko fotoelektronsko spektroskopijo (XPS). Preizku{ena je bila tudi biokompatibilnost materiala z
uporabo modelne celi~ne linije embrionskih fibroblastov mi{i (NIH 3T3). Z lasersko obdelavo, v kombinaciji s toplotno
obdelavo, so bile pripravljene plasti na povr{ini z razli~nimi vzorci odvisno od izbranih vhodnih parametrov.
Klju~ne besede: biopolimer, ekscimer laser, nanostrukturiranje, postopek `arjenja, karakterizacija
1 INTRODUCTION
Poly(L-lactic acid) or poly(lactide) is a biodegradable
and bioabsorbable thermoplastic polyester, produced
from renewable sources.1 In medical applications, it is
used in the matrices for tissue engineering, stents,
sutures or in drug delivery systems, but it still has a lot of
limitations, which can be eliminated with its appropriate
modification.
Polymer structuring allows a utilization of ordinary
materials in highly specialized fields. Nanostructured
materials find different applications, e.g., in DNA and
protein sequencing2, in the creation of a suitable synthe-
tic environment for cell growth3 or in the solar-cell tech-
nology.4 Self-organized structures are prepared with a
bottom-up method. A typical example of nanostructuring
is a ripple or dot formation caused by laser irradiation.
The ripples arise due to the interference pattern forma-
tion at a surface and the subsequent response of the
surface.5
Another example of a self-organizing mechanism is
wrinkling instability, which exhibits a variety of surface
patterns. Wrinkles are produced by the residual stress,
which exceeds the critical value.6 Wrinkle patterns are
included, e.g., in tunable optical devices7 or flexible
electronics.6–8
This paper deals with the surface modification of
poly(L-lactic acid) using laser treatment in combination
with thermal annealing. Treated samples were studied
with atomic-force microscopy (AFM), UV-Vis spectro-
scopy, and ARXPS (angle-resolved photoelectron
spectroscopy). Tests of biocompatibility were carried out
with mouse embryonic fibroblasts (NIH 3T3). With this
modification, we prepared various surface patterns with a
wrinkle-like structure.
2 MATERIALS AND METHODS
2.1 Materials and modification
We used biopolymer poly(L-lactic acid) (PLLA, a
density of 1.25 g cm–3, Tg = 60 °C, a crystallinity of
60–70 %, 50-μm-thick foils, supplied by Goodfellow
Ltd., Cambridge, Great Britain).
For the irradiation of PLLA we used a KrF excimer
laser (Coherent Compex Pro 50, a wavelength of 248
nm, a pulse duration of 20–40 ns, a repetition rate of 10
Materiali in tehnologije / Materials and technology 50 (2016) 3, 331–335 331
UDK 621.793:669.058:66.088 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)331(2016)
Hz). The beam of the KrF laser was polarized linearly
with a cube of UV-grade fused silica 25 mm × 25 mm ×
25 mm with an active polarization layer. For a homo-
geneous illumination of the samples, we used only the
central part of the beam profile by means of an aperture
0.5 cm × 1.0 cm. The samples were mounted onto a
translation stage, being perpendicular to the laser beam.
The chosen pulses were in a range of 100–6000, having
laser fluences in an interval of 6–30 mJ cm–2.
The thermal treatment of the polymers was accom-
plished with a BINDER thermostat. The samples were
heated to 60 °C (the glass-transition temperature of
PLLA) immediately after the laser treatment. After 30
min of thermal annealing, the samples were cooled down
to room temperature (RT).
2.2 Measurement techniques
The surface morphology and roughness of the pri-
stine and modified polymer samples were examined with
the atomic-force-microscopy (AFM) technique using a
VEECO CP II device in the tapping mode. The tapping
mode was chosen to minimize the damage to the sample
surface. A RTESPA-CP Si probe with a spring constant
of 20–80 N m–1 was used. The mean roughness value
(Ra) represents the arithmetic average of the deviations
from the center plane of a sample.
The presence of oxygen and carbon in the PLLA
surface layer was determined with X-ray photoelectron
spectroscopy (XPS). An Omicron Nanotechnology
ESCAProbeP spectrometer was used. The exposed and
analyzed area had a dimension of 2 mm × 3 mm. The
X-ray source was monochromated at 1486.7 eV.
Characteristic O(1s) and C(1s) peaks were searched for.
Atomic concentrations of the elements were determined
with the CASA XPS program using an integrated area of
spectrum lines and relative sensitivity factors, quoted in
the database of CASA XPS.
UV-Vis spectra were measured using a Perkin Elmer
Lambda 25 spectrometer in a spectral range of 190–1100
nm with a bandwidth of 1 nm (fixed).
2.3 Cytocompatibility tests
For cell-culture experiments, we used an adherent
model cell line of mouse embryonic fibroblasts (NIH
3T3) (ATCC, USA). NIH 3T3 cells were cultivated on a
regular basis in high-glucose Dulbecco’s modified Eagle
medium (DMEM, Sigma, USA) supplemented with
stable 2 mM L-glutamine, 10 % fetal bovine serum and
1 % MEM vitamin solution (Invitrogen, USA). The cells
were maintained at standard conditions (37 °C, a 95 %
humidified atmosphere, 5 % CO2). The cells were main-
tained in exponential growth.
The bio-response of individual PLLA samples was
tested. The polymers were first sterilized in 70 % ethanol
for 1 h, air dried, inserted into 12-well plates for cell cul-
tures (VWR, Ø 2.14 cm) and weighted with poly(methyl
methacrylate) cavus cylinders. The samples were seeded
with the NIH 3T3 cells, with a density of 14,000 cells
per cm–2 in 800 μL of a complete DMEM. An identical
batch of cells growing on a polystyrene Petri dish (PS)
was used as a control. The experiments were done in
triplicates.
The cells intended for a fluorescence-microscopy
analysis were fixed and stained. They were washed twice
with phosphate-buffered saline (PBS, pH = 7.4) and
fixed with 1 mL of a 4 % formaldehyde (Thermo Scien-
tific, USA) solution in PBS at 37 °C for 20 min. A
phalloidin-tetramethylrhodamine B isothiocyanate (Sig-
ma, USA) solution in PBS (0.5 μg·mL–1, 10 min) was
used to visualize F-actin; cellular nuclei were stained
with a solution of 4’,6-diaminido-2-phenylindole dihy-
drochloride (DAPI, Sigma, USA) in PBS (0.5 μg·mL–1,
5 min). During and after the staining, the cells were
rinsed twice with PBS to remove the excess of unbound
dyes.
3 RESULTS
Because the laser itself has just a small effect on the
surface morphology and the roughness of PLLA, we
investigated the influence of the laser treatment in a
combination with thermal annealing. We used excimer
radiation followed by thermal annealing, which was
proven to significantly influence the morphology. For the
reverse order of the applied methods, we observed just
insignificant morphological deviations on the samples in
comparison to the samples treated only by a laser beam.
On the contrary, for those samples where the thermal
annealing was the second step of the treatment, the sam-
ple surfaces were rapidly changed. Moreover, for a low
laser fluence and a high number of pulses (6–15 mJ cm–2
and 6000 pulses) the surface roughness was also signifi-
cantly increased.
Figure 1 shows the influence of the applied methods
on the sample morphology. The modification of the sam-
ples at the top of Figure 1 shows a structure practically
identical with the pristine PLLA and the roughness is
also similar to the Ra value of the pristine PLLA, which
was determined as 6.9 nm. By applying different laser
parameters, we prepared different patterns with various
roughness values. An important parameter affecting the
sample roughness is the combination of the laser fluence
and the number of pulses. If the laser energy is too high,
the surface is flattened (30 mJ cm–2, 6000 pulses, Ra =
0.7 nm), but with a lower energy, the surface prepared
can be extremely rough, with a spongious structure
(15 mJ cm–2, 3000 pulses, Ra = 44.7 nm).
The concentration of the surface oxygen was deter-
mined with XPS and the results for the selected samples
are listed in Table 1. According to these results, it is not
possible to conclude that the oxygen concentration
decreases or increases after the modification because the
differences in the concentration were within the stati-
332 Materiali in tehnologije / Materials and technology 50 (2016) 3, 331–335
I. MICHALJANI^OVA et al.: SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER
stical error (about 2 %), which could have taken place
during the measurement. It is possible to conclude that
the oxygen concentration decreases towards the surface
of the spongious structure (the plane-surface oxygen
concentration is higher in comparison with the pattern at
the very top). This decrease can be contributed to the
reorientation of dipoles (oxygen-containing groups)
toward the polymer surface.
Table 1: Element concentration of the PLLA surface. The values were
determined, with XPS method, for pristine and laser-treated
(9 mJ cm–2, 6000 pulses) samples, further for samples treated with
laser followed by thermal annealing (9 mJ cm–2, 6000 pulses, 60 °C,
30 min). The samples were measured at angles of 0° and 80°.
Tabela 1: Koncentracija elementov na povr{ini PLLA. Vrednosti so
bile dolo~ene z metodo XPS na originalnih in z laserjem obdelanih
(9 mJ cm–2, 6000 pulzov) vzorcih, nato na vzorcih obdelanih z
laserjem, ki mu je sledilo `arjenje (9 mJ cm–2, 6000 pulzov, 60 °C, 30
min). Vzorci so bili merjeni pri kotih 0° in 80°.
Sample Angle (°) C (amountfractions, x/%)
O (amount
fractions, x/%)
Pristine
0 64.4 35.6
80 67.7 32.3
Laser
treatment
0 63.2 36.8
80 65.4 34.6
Laser
treatment +
annealing
0 62.6 37.4
80 67.7 32.3
The UV-Vis absorption spectra of the samples
exposed to laser fluences of 9 and 30 mJ cm–2 with 1000
and 6000 pulses and subsequently treated with thermal
Materiali in tehnologije / Materials and technology 50 (2016) 3, 331–335 333
I. MICHALJANI^OVA et al.: SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER
Figure 1: Surface morphology of the samples treated with a combination of an excimer laser (6 mJ cm–2, 6.000 pulses) and thermal annealing
(60 °C, 30 min). The samples at the top were first annealed and then modified with the laser; the samples at the bottom were treated in the reverse
order. Ra represents the arithmetic mean surface roughness in nm.
Slika 1: Morfologija povr{ine vzorcev obdelanih s kombinacijo ekscimer laserja in `arjenja. Morfologija povr{ine vzorcev, obdelanih s
kombinacijo ekscimer laserja (6 mJ cm–2, 6,000 pulzov) in `arjenjem (60 °C, 30 min). Vzorci zgoraj so bili najprej `arjeni in nato modificirani z
laserjem, vzorci spodaj so bili obdelani v obratnem vrstnem redu. Ra predstavlja aritmeti~no sredino hrapavosti povr{ine v nm.
Figure 2: UV-Vis spectra of PLLA. UV-Vis spectra of: A) pristine
PLLA and samples exposed to laser fluence 9 mJ cm–2 and 30 mJ
cm–2 (1000 and 6000 pulses) and subsequently treated by thermal
annealing (60 °C, 30 min), B) 9 mJ cm–2, 1000 pulses + annealing, C)
9 mJ cm–2, 6000 pulses + annealing, D) 30 mJ cm–2, 1.000 pulses +
annealing and E) 30 mJ cm–2, 6.000 pulses + annealing.
Slika 2: UV-Vis spekter PLLA. UV-Vis spekter: A) prvoten PLLA in
vzorci izpostavljeni laserju pri 9 mJ cm–2 in 30 mJ cm–2 (1000 in 6000
pulzov) in nato obdelani z `arjenjem (60 °C, 30 min), B) 9 mJ cm–2,
1000 pulzov + `arjenje, C) 9 mJ cm–2, 6000 pulzov + `arjenje, D) 30
mJ cm–2, 1000 pulzov + `arjenje in E) 30 mJ cm–2, 6000 pulzov +
`arjenje.
annealing (60 °C, 30 min) are shown in Figure 2. In the
case of the spectra of the modified samples, there are
significant peaks at a position of approximately 275 nm.
The curve with the most significant peak represents the
treatment with a laser fluence of 9 mJ cm–2 and 6000 pul-
ses, belonging to a sample with an interesting spongious
morphology. The other modified samples also show at
least a "small" peak. After the combination of the laser
treatment and thermal annealing, a diagonal shift of the
absorbance curve was observed.
The cell adhesion represents the first stage of the
cell-substrate interaction, thus the quality of adhesion
influences cell ability to proliferate and differentiate in
the contact with a substrate. After successful adhesion,
the adaptation of the cells to the new environment (the
lag phase) occurs. For adhesion and proliferation studies
NIH 3T3 cells were chosen. The selected samples of the
modified PLLA substrate (9 or 30 mJ cm–2, 100 pulses,
annealing) were tested and compared with PLLA pristine
and control samples of polystyrene used for tissue cul-
tures (PS). Samples with 100 pulses were chosen, be-
cause with the increasing number of pulses the material
became brittle. From the Figure 3, it is apparent that the
best results were obtained on PS, which is commonly
considered as a model substrate. The PLLA pristine is a
material which biocompatibility can be improved by
plasma modification. It was shown that modified mate-
rial with 30 mJ cm–2 was slightly less cytocompatible,
but with 9 mJ cm–2 moderately improved its cytocompa-
tibility. Immediately after seeding cell adhesion was
found unaffected. In the Figure 4, there are introduced
selected pictures of cells growing on treated and pristine
PLLA samples.
4 DISCUSSION
By the combination of excimer laser and thermal
annealing, it is possible to prepare different surfaces of
PLLA with a wide range of surface roughness. The most
interesting part of this work is a wrinkle pattern creation.
Wrinkles could appear on polymer after annealing as a
thin bi-layer film (prepared by treatment and by anneali-
ng). In this case, the surface laser treatment and follow-
ing thermal annealing has created the wrinkle pattern.
We suggest that the structure is influenced also during
cooling. With less counts of pulses, the staminate
334 Materiali in tehnologije / Materials and technology 50 (2016) 3, 331–335
I. MICHALJANI^OVA et al.: SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER
Figure 3: Tests of surface cytocompatibility. Dependence of the
number of adhered and proliferated NIH 3T3 cells (6, 24, and 72) h
after seeding on pristine PLLA (PLLA) and PLLA modified by laser
beam (9 mJ cm–2 or 30 mJ cm–2, 100 pulses) and subsequent thermal
annealing (60 °C, 30 min). The values for tissue polystyrene (PS) are
also shown for comparison.
Slika 3: Preizkus citokompatibilnosti povr{ine. Odvisnost {tevila
oprijetih in razmno`enih NIH 3T3 celic, (6, 24, 72) h po sejanju na
prvotni PLLA (PLLA) in na PLLA obdelan z laserskim `arkom (9 mJ
cm–2 ali 30 mJ cm–2, 100 pulzov), ki mu je sledilo `arjenje (60 °C, 30
min). Za primerjavo so prikazane tudi vrednosti za tkivni polistiren
(PS).
Figure 4: Mouse embryonic fibroblasts (NIH 3T3) growing on different substrates. Proliferated NIH 3T3 cells 72 h after seeding on various
substrates: A) modified PLLA (30 mJ cm–2, 100 pulses and subsequent thermal annealing (60 °C, 30 min); B) pristine PLLA, and C) tissue
polystyrene for comparison.
Slika 4: Mi{ji embrionski fibroblasti (NIH 3T3), ki so zrasli na razli~nih podlagah. Razmno`ene NIH 3T3 celice po 72 h po sejanju na razli~ne
podlage: A) modificiran PLLA (30 mJ cm–2, 100 pulzov in nato toplotna obdelava (60 °C, 30 min), B) prvotni PLLA in C) tkivni polistiren za
primerjavo.
patterns were built up. We propose that difference in the
structure was caused by the thickness of laser treated
layer, which was insufficient to create wrinkles. Instead
of that surface cracking occurred. We suggest that “rods”
on the PLLA surface were produced by crystallization of
the newly formed material from chopped polymer
strings. By reverse order of treatment, when the samples
were exposed to annealing and subsequently treated by
laser beam, the samples showed the same structure as the
samples treated just by laser without annealing. This
supports the theory, how the structure was formed.
The peak around area of 275 nm is typical for transi-
tions of n electrons to the * excited state, and represents
the presence of C=O group. This type of transition needs
an unsaturated group in the molecule to provide
the  electrons. The diagonal shift could be explained by
increasing concentration of double bonds.
Cytocompatibility tests show slight decrease of
ability to support cell proliferation for the PLLA treated
by high laser fluence, the biocompatibility increases for
lower laser fluence. The application of lower energy has
therefore similar effect as plasma treatment which we
studied in our previous experiments.
5 CONCLUSIONS
Various types of surface structures on modified sam-
ples of PLLA were produced by exposure of KrF
excimer laser beam and subsequent thermal annealing:
• by UV-Vis spectroscopy we observed new C=O
groups and creation of double bonds;
• by choosing optimal input parameters, it is possible
to prepare structures from porous and spongeous to
flat biopolymer surface with staminate structures;
• the roughness is significantly dependent on laser
treatment and annealing input values;
• low laser fluence has a positive effect on cytocom-
patibility, but high laser fluence loses this effect.
Acknowledgements
This work was supported by the GACR under project
13-06609S.
6 REFERENCES
1 R. M. Rasal, A. V. Janorkar, D. E. Hirt, Progress in Polymer Science,
3 (2010) 35, 38–356, doi:10.1016/j.progpolymsci.2009.12.003
2 B. L. Hancock-Hanser, A. Frey, M. S. Leslie, P. H. Dutton, F. I.
Archer, P. A. Morin, Targeted multiplex next-generation sequencing:
advances in techniques of mitochondrial and nuclear DNA
sequencing for population genomics, Molecular Ecology Resources,
2 (2013) 13, 254–268, doi:10.1111/1755-0998.12059
3 E.-K. Yeong, S.-H. Chen, Y.-B. Tang, The Treatment of Bone
Exposure in Burns by Using Artificial Dermis, Annals of Plastic
Surgery, 6 (2012) 69, 607–610, doi:10.1097/SAP.0b013e318273f845
4 R. Lecover, N. Williams, N. Markovic, D. H. Reich, D. Q. Naiman,
H. E. Katz, Next-Generation Polymer Solar Cell Materials: Designed
Control of Interfacial Variables Acs Nano, 4 (2012) 6, 2865–2870,
doi:10.1021/nn301140w
5 M. Bolle, S. Lazare, Large scale excimer laser production of sub-
micron periodic structures on polymer surfaces, Applied Surface
Science, 1–4 (1993) 69, 31–37, doi:10.1016/0169-4332(93)90478-T
6 C.-M. Chen, S. Yang, Wrinkling instabilities in polymer films and
their applications, Polymer International, 7 (2012) 61, 1041–1047,
doi:10.1002/pi.4223
7 H. S. Kim, A. J. Crosby, Solvent-Responsive Surface via Wrinkling
Instability, Advanced Materials, 36 (2011) 23, 4188–4192,
doi:10.1002/adma.201101477
8 D. Y. Khang, H. Q. Jiang, Y. Huang, J. A. Rogers, A Stretchable
Form of Single-Crystal Silicon for High-Performance Electronics on
Rubber Substrates, Science, 5758 (2006) 311, 208–212, doi:10.1126/
science.1121401
Materiali in tehnologije / Materials and technology 50 (2016) 3, 331–335 335
I. MICHALJANI^OVA et al.: SURFACE PROPERTIES OF A LASER-TREATED BIOPOLYMER

M. K. KULEKCI et al.: ANALYZING THE HEAT-TREATMENT EFFECT ON THE MECHANICAL PROPERTIES ...
337–341
ANALYZING THE HEAT-TREATMENT EFFECT ON THE
MECHANICAL PROPERTIES OF FREE-CUTTING STEELS
ANALIZA VPLIVA TOPLOTNE OBDELAVE NA MEHANSKE
LASTNOSTI AVTOMATNIH JEKEL
Mustafa Kemal Kulekci1,2, Ugur Esme1,2, Funda Kahraman1,2, Rahim Ozgun3,
Adnan Akkurt4
1Mersin University, Institute of Applied and Natural Sciences, Manufacturing Engineering Department, 33480 Tarsus, Mersin, Turkey
2Mersin University, Faculty of Tarsus Technology, 33480 Tarsus, Mersin, Turkey
3Mersin University, Faculty of Tarsus Technical Education, 33480 Tarsus, Mersin, Turkey
4Gazi University, Faculty of Technology, 06500 Ankara, Turkey
mkkulekci@yahoo.com
Prejem rokopisa – received: 2015-01-14; sprejem za objavo – accepted for publication: 2015-06-02
doi:10.17222/mit.2015.011
In this research, the heat-treatment effect on the mechanical properties of free-cutting steels was investigated. Free-cutting steels
(FCSs) are used where high degrees of machining are required as they increase the machining speed and reduce the tool wear.
The effect of heat treatment on mechanical properties was identified using tensile and fatigue tests, and microstructure images
taken with a scanning electron microscope (SEM). The studied material included commercially available AISI 12L14
cylindrical bars of free-cutting steel. FCS was heated to 900 °C and held at this temperature for different time spans. The
degradation of the mechanical properties of free-cutting steel due to the elevated temperature was assessed. At a microscopic
level, more mechanical damage was observed between the steel matrix and the second phase of the heat-affected specimens.
Keywords: fatigue, tensile strength, free-cutting steel, heat treatment
V prispevku je bil raziskan vpliv toplotne obdelave na mehanske lastnosti avtomatnih jekel. Avtomatna jekla (FCS) se
uporabljajo v primerih, ko so zahtevane velike stopnje obdelave, saj dopu{~ajo ve~je hitrosti rezanja ob manj{i obrabi orodja.
Vpliv toplotne obdelave na mehanske lastnosti je bil dolo~en z rezultati nateznega preizkusa in preizkusa utrujenosti. Z
vrsti~nim elektronskim mikroskopom (SEM) so bili narejeni posnetki mikrostrukture. Preiskovane so bile komercialno
dosegljive palice iz avtomatnega jekla AISI 12L14. FCS je bil segret na 900 °C in razli~no dolgo zadr`an na tej temperaturi.
Dolo~eno je bilo zmanj{anje mehanskih lastnosti avtomatnega jekla zaradi povi{ane temperature. V toplotno obdelanih vzorcih
je bilo z mikroskopom opa`enih ve~ mehanskih po{kodb.
Klju~ne besede: utrujenost, nateg, avtomatna jekla, toplotna obdelava
1 INTRODUCTION
A significant increase in mechanical-machining costs
has led to a reappraisal of the importance of steel ma-
chinability.1 In order to achieve higher automation and
cost competitiveness, a series of steels commonly known
as free-cutting steels (an S minimum of 0.10 %) are
being increasingly used. Special lead-containing steels
differ from the normal structural, quenched-and-
tempered and case-hardened steels because of the
presence of Pb (approximately 0.15–0.35 %) in order to
improve their machinability. These steels allow an
excellent chip removal and they are particularly suitable
for large productions.2,3 Lead has a very good lubricating
effect and, combined with the heating produced by the
tools, breaks the chips, thus allowing for a higher
productivity resulting in more advantageous production
runs. It also guarantees a lower tool wear.
Free-cutting steels are used where high degrees of
machining are required as they increase the machining
speed and reduce the tool wear. These steels contain
manganese, lead, sulphur and phosphorous, which im-
prove the machining performance. The additives reduce
the coefficient of friction between the tool and chip,
thereby reducing the cutting force, the temperature and
the built-up edge formation, allowing higher feeds and/or
speeds.4–6 Free-cutting steels are preferred in manu-
facturing mechanical components for their improved
machinability.6–8
AISI 12L14 is a Pb-added low-carbon resulphurised
free-cutting steel containing 0.3 % Pb and 0.3 % S. It is
used in large quantities in automotive applications such
as transmission oil hydraulic control valves and oil
hydraulic hose connectors, in printer shafts and other
parts of office automation equipment. Lead is insoluble
in free-cutting steel and during the cutting process lead
particles are sheared and smeared over the tool-chip
interface.9,10 Lead improves the machinability with little
effect on mechanical properties. Due to its low shear
strength, lead acts as a solid lubricant. Manganese and
sulphides assist the chip formation and reduce the fric-
tion and wear of a cutting tool. Free-cutting steels are
used where high degrees of machining are required as
they increase the machining speed and reduce the tool
wear.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 337–341 337
UDK 621.78:67.017:691.714 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)337(2016)
In this study, the heat-treatment effect on the me-
chanical properties of FCS was investigated. FCS was
heated to 900 °C and held at this temperature for diffe-
rent time spans. Degradation of the mechanical proper-
ties of the free-cutting steel due to the elevated tempera-
ture was assessed. The effect of heat treatment on
mechanical properties was investigated using tensile and
fatigue test results, and microstructure images taken with
a scanning electron microscope (SEM).
2 EXPERIMENTAL PROCEDURE
The studied material included commercially available
AISI 12L14 cylindrical bars of free-cutting steel with a
diameter of 12 mm. The bars were machined into tensile
and fatigue specimens with dimensions as shown in Fig-
ure 1. After the machining, the specimens were grinded
with 4000 emery paper. Two samples were used for each
stress level and the results were averaged for the as-
received specimens and the specimens exposed to 900 °C
for (3, 6, 9, 12 and 15) h. The percent chemical com-
position (% of mass fractions) and mechanical properties
of the FCS are given in Tables 1 and 2, respectively.
Tensile-test specimens were machined according to EN
ISO 6892-1 standard. Fatigue tests were conducted on an
R. R. Moore type rotating-beam fatigue-testing machine
with a frequency of 25 Hz. Microstructural changes in
the FCS exposed to high temperature were investigated
using SEM images. Tensile tests were conducted to inve-
stigate the heat-treatment effect on the ultimate tensile
strength (Rm), the yield strength (Re), the ductility and the
toughness. FCS specimens were heated to 900 °C and
held at this temperature for different time spans. Degra-
dation of the mechanical properties of free-cutting steel
due to elevated temperature was investigated. More
mechanical damage was observed between the steel
matrix and the phase in the heat-affected specimens at a
microscopic level. The effect of heat treatment on the
mechanical properties was investigated using tensile and
fatigue tests, and microstructure images taken with SEM.
The Rm and Re values of the FCS decreased as the
toughness and ductility increased due to the effect of
elevated heat.
Table 2: Mechanical properties of free-cutting steel (AISI 12L14)
Tabela 2: Mehanske lastnosti avtomatnega jekla (AISI 12L14)
Yield
strength
(MPa)
Tensile
strength
(MPa)
Elongation
(%)
Brinell hardness (HB)
(10 mm steel ball and
500 kg load)
465 587 12 150
3 RESULTS AND DISCUSSION
Fatigue test results for the FCS specimens with
different dwell times and the as-received specimens are
given in Figure 2. From this figure, it is seen that the
fatigue strength of the FCS exposed to elevated tempe-
rature decreased. There was a limited effect of the time
span affecting the fatigue strength of the tested speci-
mens. Under the cyclic load of 300 MPa, the heat-
affected and as-received specimens fractured at 50,000
and 300,000 cycles, respectively. The fatigue strength of
the FCS reduced in the ratio of 83 % because of the
effect of high temperature. The regression lines of the
fatigue strength given in Figure 2 are expressed as:
y = –a ln(x) + b (1)
M. K. KULEKCI et al.: ANALYZING THE HEAT-TREATMENT EFFECT ON THE MECHANICAL PROPERTIES ...
338 Materiali in tehnologije / Materials and technology 50 (2016) 3, 337–341
Figure 1: Dimensions and images of test specimens: a) fatigue, b) tensile test
Slika 1: Mere in posnetki preizku{ancev: a) utrujanje, b) natezni preizkus
Table 1: Chemical composition of free-cutting steel (AISI 12L14) in mass fractions, (w/%)
Tabela 1: Kemijska sestava avtomatnega jekla (AISI 12L14) v masnih dele`ih, (w/%)
C Si Mn P S Cr Mo Ni Cu Pb Fe
0.074 0.005 1.203 0.045 0.294 0.04 0.02 0.08 0.12 0.30 balance
where y is the loaded stress range, x is the number of
cycles to failure, a and b are the fitting constants. The
constants of the equations of regression lines, the
R-square, which indicates how well data points fit a
statistical model (the coefficient of determination: R2)
and the residual variance (Rv), which is a measure of the
variation of the y values about the regression line (Rv =
1 – R2) of the fatigue tests are given in Table 3.
Table 3: R-square, residual variance and fitting constants of the tests
Tabela 3: R-kvadrat, preostale variance in ustrezne konstante preizku-
sov
Experimental
variables
Fitting constants of
regression lines R-square
(R2)
Residual
variance
(Rv = 1 – R2)a b
Base material –89.19 1433.9 0.9841 0.0159
Heat treated:
15h –83.61 1202.9 0.9824 0.0176
Heat treated:
12h –78.819 1176.1 0.9885 0.0115
Heat treated:
9h –79.957 1178.6 0.9805 0.0195
Heat treated:
6h –81.643 1188.4 0.9816 0.0184
Heat treated:
3h –71.404 1093.6 0.9367 0.0633
The reduction in the fatigue strength of the specimens
exposed to high temperature can be explained with
microstructural changes. The grain size of the base
material and the heat-affected specimens were measured
as 10 μm and 25 μm, respectively, as seen from Figure 3.
From this figure, it is seen that the grain size of the FCS
increased by 250 % with the heat effect and 3–15 h time
spans. The grain growth reduced the load-carrying
capacity of the specimens. At the elevated temperature of
900 °C and under a long dwell time, the precipitated
phase coalesced and grew in the microstructure as seen
in Figure 3. Kalpakjian states that the grain growth
reduces the grain boundaries per volume unit of a
grain.11,12 Smaller grains have a higher strength and a
higher contact surface area.
The increase in the size of the grains reduced the
fatigue strength of the FCS. The ultimate tensile strength
(Rm) and the yield strength (Re) of the heat-effected FCS
were reduced as seen in Figure 4. The toughness of the
heat-affected FCS specimens was calculated with the
following Equation (2):
Toughness
e m=
+
⋅
R R
e
2
(2)
where Re and Rm are the yield and tensile strengths,
respectively, and e is the engineering strain. The Re and
Rm values of the initial FCS specimens were 616 MPa
and 512 MPa, respectively. For the specimens heat-
treated at 900 °C for 3 h, these values were reduced to
441 MPa and 318 MPa. The reduction in Rm and Re was
28.4 and 37.8 %, respectively. The amounts of energy
per volume (toughness) that the FCS absorbed before
rupturing were 100.51 MPa and 143.67 MPa for the
specimens in the as-received state and heat treated at
900 °C for 15 h, respectively, as seen in Figure 4. The
toughness of the FCS increased in the ratio of 42.94 %
due to the effects of heat and time. This increase in the
toughness of FCS can be explained with the disposal of
the residual stress induced during manufacturing stages.
The ductility of the FCS specimens was determined
M. K. KULEKCI et al.: ANALYZING THE HEAT-TREATMENT EFFECT ON THE MECHANICAL PROPERTIES ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 337–341 339
Figure 3: SEM images of grains: a) as received and b) heat treated at
900 °C for 15 h
Slika 3: SEM-posnetek zrn: a) stanje ob dobavi in b) `arjena 15 h na
900 °C
Figure 2: Effect of the heat exposure and time on the fatigue test
results for FCS
Slika 2: Vpliv ~asa ogrevanja na rezultate utrujenostnega preizkusa
FCS
with the elongation calculated with the following
Equation (3):
Elongation
f 0
0
=
+
⋅
l l
l
100 (3)
where l0 and lf are the gage lengths of the original and
fractured sample, respectively. The ductility of the FCS
changed with the effect of heat treatment, as shown in
Figure 5. The ductility was 16.78 % and 39.06 % for
the as-received specimens and the specimens heat-
treated at 900 °C for 15 h, respectively. SEM images of
the fatigue-fractured surfaces of the as-received and
heat-affected specimens are given in Figure 6. The heat
effect increased the voids between the matrix and the
second phase of the FCS under cyclic load during the
fatigue test, as seen in Figure 6. From this figure, it is
clearly seen that the voids between the matrix and the
M. K. KULEKCI et al.: ANALYZING THE HEAT-TREATMENT EFFECT ON THE MECHANICAL PROPERTIES ...
340 Materiali in tehnologije / Materials and technology 50 (2016) 3, 337–341
Figure 6: SEM images of fatigue-fractured surfaces of FCS: a) as received, b) heat treated at 900 °C for 15 h (x and y are voids)
Slika 6: SEM-posnetek povr{ine utrujenostnega loma FCS: a) stanje ob dobavi, b) `arjeno 15 h na temperaturi 900 °C (x in y so praznine)
Figure 5: Heat-treatment effect on the elongation of free-cutting steel
Slika 5: Vpliv `arjenja na raztezek avtomatnega jekla
Figure 4: Effect of heat treatment on ultimate tensile strength (Rm),
yield strength (Re) and toughness of FCS
Slika 4: Vpliv `arjenja na natezno trdnost (Rm), mejo plasti~nosti (Re)
in `ilavost FCS
Figure 7: Macrostructures (49X) and microstructures (1000X) of
fractured specimens: a) as received and b) heat treated at 900 °C for
15 h
Slika 7: Makrostruktura in mikrostruktura prelomljenega vzorca: a)
dobavljeno in b) `arjeno 15 ur na 900 °C
second phase are larger in the structure of the heat-
affected specimens (Figure 6b). The heat-treatment
effect made the material more ductile and, as a result,
the bonding between the matrix and the second phase
was weakened, while the fatigue and tensile strengths of
the FCS decreased as well.
Macro- and microstructure images of the fractured
specimens in the as-received and heat-affected states are
given in Figures 7a and 7b, respectively. The base
material exhibited brittle fracture, while the heat-treated
specimens exhibited ductile fracture. The fatigue values,
Rm and Re, of the base material were higher than those of
the heat-treated FCS, as seen in Figures 2 and 4. This
situation can be explained with the residual stress
included in the as-received specimens. The grain growth
due to the heat treatment of the FCS material also
reduced these values.
4 CONCLUSION
From the above results, the following conclusions
can be drawn:
• The heat effect made the material more ductile and
weakened the bonding between the matrix and the
second phase.
• The microstructure of the FCS changed at 900 °C.
The elevated temperature and a long dwell time
resulted in the grain growth and precipitation, which
caused a decrease in the fatigue and tensile strengths.
• The grain size of the FCS increased by up to 250 %
with the heat treatment at 900 °C and a dwell time of
15 h. The grain growth reduced the load-carrying
capacity of the specimens.
• The fatigue strength of the FCS was reduced in the
ratio of 83 % due to the effect of the high temperature
and long dwell time.
• The heat treatment at 900 °C, for a period of less than
3 h reduced Rm and Re in the ratio of 28.4 and 37.8 %,
respectively, when compared to the base material.
• The toughness of the FCS increased in the ratio of
42.94 % due to the effect of heat treatment and dwell
time.
• The heat effect led to increased ductility and larger
voids between the matrix and the second phase of the
FCS under a cyclic load. At a microscopic level,
more mechanical damage was observed between the
steel matrix and the second phase of the heat-affected
specimens.
• The effect of time on the fatigue strength of the tested
specimens is less significant.
Acknowledgement
This work was financially supported by the Scientific
Research Foundation of the Mersin University under
contract MAE-1-2010. This support is gratefully ackno-
wledged.
5 REFERENCES
1 M. Toshiyuki, T. Kunikazu, S. Tetsuo, Development of free cutting
steel without lead addition to replace AISI12L14, JFE Technical
Report, 15 (2010), 10–16
2 L. D. Chiffre, of metal cutting and cutting fluid action, International
Journal of Machine Tool Design and Research, 17 (1977) 4,
225–234, doi:10.1016/0020-7357(77)90016-6
3 A. J. Deardo, C. I. Garcia, U. S. Patent Nr. 6200395, 2001
4 P. L. B. Oxley, Mechanics of Machining: An Analytical Approach to
Assessing Machinability, John Wiley & Sons, New York 1989
5 P. S. Sreejith, B. K. A. Ngoi, Dry machining: Machining of the
future, Journal of Materials Processing Technology, 101 (2000) 1–3,
287–291, doi:10.1016/S0924-0136(00) 00445-3
6 H. Roelofs, A. M. G. Boeira, R. Margot, J. T. Gomes, M. Eglin,
Machinability of Inclusion Engineered Free Cutting Steel Under
Built Up Edge Conditions, 8th Int. Conf. on Advanced Manu-
facturing Systems and Technology, Udine 2008, 12–13
7 Z. Li, D. Wu, Effect of Free-cutting Additives on Machining Charac-
teristics of Austenitic Stainless Steels, Journal of Materials Science
& Technology, 26 (2010) 9, 839–844, doi:10.1016/S1005-0302(10)
60134-X
8 Y. Y. Wei, Z. Q. Liu, Q. L. An, M. Chen, Study on the Machinability
of Free-Cutting Steels 1214 and 12L15 with Coated Tool, Advanced
Materials Research, 426 (2012), 151–154, doi:10.4028/www.scienti-
fic.net/AMR.426.151
9 A. D. Foster, J. Lin, D. C. J. Farrugia, T. A. Dean, An Investigation
into Damage Nucleation and Growth for a Free-Cutting Steel at Hot
Rolling Conditions, The Journal of Strain Analysis for Engineering
Design, 42 (2007) 4, 227–235, doi:10.1243/03093247JSA230
10 J. Polak, M. Petrenec, J. Man, Cyclic plasticity, cyclic creep and
fatigue life of duplex stainless steel in cyclic loading with positive
mean stress, Metallic Materials, 49 (2011) 5, 347–354, doi:10.4149/
km_2011_5_347
11 R. Özgün, Investigation of the Effect of High Temperature and Boric
Acid Environment on the Mechanical Properties of Free Cutting
Steels, University of Mersin, 2014
12 S. Kalpakjian, S. Schimid, Manufacturing Engineering and Tech-
nology, 6th Ed., Pearson Hall, New York 2010
M. K. KULEKCI et al.: ANALYZING THE HEAT-TREATMENT EFFECT ON THE MECHANICAL PROPERTIES ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 337–341 341

A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
343–351
ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE
ROUGHNESS DURING THE ORTHOGONAL MACHINING OF AISI
4140 ALLOY STEEL VIA THE TAGUCHI METHOD
ANALIZA TEMPERATURE REZANJA IN HRAPAVOSTI POVR[INE
S TAGUCHI METODO PRI ORTOGONALNI STROJNI OBDELAVI
LEGIRANEGA JEKLA AISI 4140
Ali Riza Motorcu1, Yahya Isik2, Abdil Kus2, Mustafa Cemal Cakir3
1Çanakkale Onsekiz Mart University, Engineering Faculty, Department of Industrial Engineering, 17100 Çanakkale, Turkey
2Uludað University, Vocational School of Technical Science Machinery Program, 16059 Bursa, Turkey
3Uludað University, Engineering Faculty, Department of Mechanical Engineering, 16059 Bursa, Turkey
armotorcu@comu.edu.tr
Prejem rokopisa – received: 2015-01-22; sprejem za objavo – accepted for publication: 2015-06-08
doi:10.17222/mit.2015.021
In this research, the tool-chip interface temperature (TCTI), the tool temperature (TT) and the average surface roughness (Ra) were
measured experimentally during the turning of AISI 4140 alloy steel with TiAlN-TiN, PVD-coated, WNVG 080404-IC907
tungsten carbide inserts using an IR pyrometer technique, a K-type thermocouple and a portable surface-roughness
measurement device, respectively. The workpiece material was heat treated by an induction-hardening process and hardened up
to a value of 50 HRC. The Taguchi method L18 (21 × 37) was used for the determination of the optimum control factors. The
depth of cut, the cutting speed and the feed rate were taken as control factors. The analysis of variance was applied in order to
determine the effects of the control factors on the tool-chip interface temperature, the tool temperature and the surface
roughness. The optimum combinations of the control factors for TCTI, TT and Ra were determined as a2v1f3, a1v3f2 and a2v3f1,
respectively. Second-order predictive models were developed with a linear-regression analysis, and the coefficients of
correlation for TCTI, TT and Ra were calculated as R2 = 92.8, R2 = 68.1 and R2 = 82.6, respectively.
Keywords: tool temperature, thermocouple, pyrometer, machining, Taguchi method
V raziskavi so bile eksperimentalno izmerjenene tempetratura na stiku orodje-ostru`ek (TCTI), temperatura orodja (TT) in
povpre~na hrapavost povr{ine (Ra) pri stru`enju legiranega jekla AISI 4140, z volfram karbidnimi vlo`ki WNVG 080404-IC907
s PVD prevleko iz TiAlN-TiN, z uporabo IR pirometra, termoelementi vrste K in s prenosnim merilnikom hrapavosti.
Obdelovanec je bil toplotno obdelan z indukcijskim ogrevanjem in hlajenjem na trdoto 50 HRC. Za dolo~anje optimalnih
kontrolnih faktorjev je bila uporabljena Taguchi metoda L18 (21 × 37). Globina rezanja, hitrost rezanja in hitrost podajanja so
bile vzete kot kontrolni faktorji. Analiza variance je bila uporabljena za dolo~anje vpliva kontrolnih faktorjev na temperaturo
prehoda orodje-ostru`ek, temperaturo orodja in hrapavost povr{ine. Dolo~ene so bile optimalne kombinacije kontrolnih
faktorjev za TCTI, TT in Ra , kot a2v1f3, a1v3f2 and a2v3f1. Z linearno regresijsko analizo so bili razviti modeli drugega reda za
napovedovanje in izra~unani so bili koeficienti korelacije za TCTI, TT in Ra kot R2 = 92,8, R2 = 68,1 in R2 = 82,6.
Klju~ne besede: temperatura orodja, termo~len, pirometer, strojna obdelava, Taguchi metoda
1 INTRODUCTION
In order to overcome the difficulties in terms of
efficiency and the quality of production encountered in
the metal-cutting industries, all the stages of the
machining process need to be monitored. During the
metal-cutting processes, one of the key factors is the
cutting temperature, which directly affects the surface
quality, the tool wear, the tool life, and the cost of
production. The amount of heat generated varies with the
type of material being machined and the cutting
parameters (especially the cutting speed, which had the
biggest influence on the temperature).1
Temperature monitoring is one of the most difficult
and complicated procedures in metal-cutting operations.
It is extremely complex to develop a model for
measuring the temperature due to the complexity of the
different events at the point of contact between the tool
and the workpiece. Therefore, an accurate and repeatable
temperature prediction still remains as a challenge due to
this complexity of the contact phenomenon.2 It is quite
difficult to measure the temperature since the heat in the
region is very close to the cutting edge. Due to a lack of
sufficient experimental data, it is not possible to verify a
mathematical model. Numerous attempts have been
made to measure the temperature during machining ope-
rations.3
Amongst the many experimental methods to measure
the temperature directly, only a few systems have used
the temperature as an indicator of machine performance
and for industrial applications.4 Therefore, the tempe-
rature can be controlled using the appropriate cutting
parameters to design and develop the system and it will
be beneficial to increase the efficiency in production.
In recent years, experimental studies related to
metal-cutting processes have made use of the Taguchi
Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351 343
UDK 620.181.4:621.9.015:669.15 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)343(2016)
method. This method has been used successfully for a
determination of the appropriate cutting parameters and
in the optimization of parameters related to tool wear,
tool life, and the surface quality. The Taguchi method
and Analysis of Variance (ANOVA) can conveniently
optimize the cutting parameters with several experimen-
tal runs that are well designed. Taguchi parameter design
can optimize the performance characteristics through the
settings of the design parameters and reduce the sen-
sitivity of the system’s performance to the source of
variation.5 On the other hand, ANOVA is used to identify
the most significant variables and interaction effects.6,7
In the Taguchi method, quality is measured by the
deviation of a quality characteristic from its target value.
Therefore, the objective is to create a design that is
insensitive to all possible combinations of uncontrollable
factors and is at the same time effective and cost efficient
as a result of setting the key controllable factors at their
optimum levels.8 Taguchi’s parameter design offers a
simple and systematic approach that can reduce the num-
ber of experiments to optimize the design for perfor-
mance, quality and cost. The signal-to-noise (S/N) ratio
and the orthogonal array (OA) are two major tools used
in robust design.9
A lot of research has been conducted for determining
the optimal cutting parameters. W. H. Yang and Y. S.
Tarng10 employed the Taguchi method, and the optimal
cutting parameters for the turning of S45C steel bars
were successfully obtained. B. M. Gopalsamy et al.11
applied the Taguchi method to find the optimum machin-
ing parameters while machining hard steel and used the
L18 orthogonal array. The S/N ratio and ANOVA were
used to study the performance characteristics of the
machining parameters. F. Ficici et al.12 used the Taguchi
method to study the wear behaviour of boronized AISI
1040 steel. They used the S/N ratio to investigate the
optimum setting parameters.
M. Adinarayana et al.13 presented the multi-response
optimization of the turning parameters for the turning of
AISI 4340 alloy steel. The experiments were designed
and conducted based on Taguchi’s L27 orthogonal array
design. They discussed an investigation into the use of
Taguchi parameter design to predict and optimize the
surface roughness, the metal removal rate and the power
consumption during turning operations. E. D. Kirby14
discussed an investigation into the use of Taguchi para-
meter design for optimizing the surface roughness gene-
rated by a CNC turning operation. He used a standard
orthogonal array for determining the optimum turning
parameters with an applied noise factor. The controlled
factors include the spindle speed, the feed rate, and the
depth of cut.
In this paper, the measurement of temperature during
the turning of AISI 4140 alloy steel was performed using
various cutting parameters. The tool-chip interface
temperature TCTI was measured by infrared thermometer,
the tool temperature TT was measured with a K-type
thermocouple in the cutting zone, and the average
surface roughness Ra was measured using a portable
surface-roughness measurement device. The Taguchi
design was selected to find the relationships between the
control factors. The depth of cut (ap), the cutting speed
(vc), and the feed rate (f) were taken as the control
factors.
2 TEMPERATURES DURING METAL CUTTING
In the cutting process, nearly all of the energy
dissipated during plastic deformation is converted into
heat, which in turn raises the temperature in the cutting
zone. Since the heat generation is closely related to the
plastic deformation and friction, we can specify three
main sources of heat when cutting:
• plastic deformation by shearing in the primary shear
zone;
• friction on the cutting face and friction between the
chips;
• tool on the tool flank.
Temperature results in dimensional errors on the
machined surface. The cutting tool elongates as a result
of the increased temperature, and the position of the
cutting tool edge shifts towards the machined surface,
resulting in a dimensional error of about 0.01–0.02 mm.
Since the processes of thermal generation, dissipation,
and solid-body thermal deformation are all transient,
some time is required to achieve the steady-state con-
dition.
Heat is mostly dissipated by: the discarded chip that
carries away about 60–80 % of the total heat, the
workpiece acts as a heat sink drawing away 10–20 % of
heat, while the cutting tool draws away ~10 % of the
heat. The balance between heat generation and heat
dissipation during metal cutting is shown in Figure 1.
3 MATERIALS AND METHOD
3.1 Workpiece and cutting tool
The workpiece material is AISI 4140 alloy steel. The
chemical composition of the workpiece material (in
volume fractions) is shown in Table 1. The machining
process was performed using a NR 2020K-08 tool holder
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
344 Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351
Figure 1: The balance of heat generation and heat dissipation during
metal cutting
Slika 1: Izravnava med spro{~eno in odvedeno toploto pri rezanju
kovin
and a TiAlN-TiN, PVD-coated, WNVG 080404-IC907
solid carbide insert. Figure 2 and Table 2 show a
schematic of the tip geometry and the specifications of
the insert.
Table 1: Chemical composition of AISI 4140 alloy steel (in volume
fractions, x/%)
Tabela 1: Kemijska sestava legiranega jekla AISI 4140 (v volumen-
skih odstotkih, x/%)
C Cr Ni Mn P S Si Mo
0.38 0.80 9.58 0.75 0.035 0.04 0.15 0.15
Table 2: The specifications of the insert
Tabela 2: Specifikacije vlo`ka za rezanje
TiAlN-TiN PVD-coated WNVG 080404-IC907
d1 S I r HRA TRS d
12.70 4.83 8.70 0.40 92.80 560 4.70
Property Value
ISO Range – P/M/K (P10-P30)(M05-M20)
ISO Range – H/S/N (H05-H15)(S05-S20)
Grade or coating type PVD
Coating layers TiAlN-TiN
3.2 Experimental conditions, temperature and sur-
face-roughness measurements
In this study, two methods of tool-temperature eva-
luation are presented:
• the placement of the K-type thermocouple on the
tool,
• the infrared pyrometer.
A schematic view of the experimental setup is shown
in Figure 3. Cylindrical workpieces (Ø45 × 300 mm)
were fixed between the chuck and the tailstock and were
pre-machined using a separate insert. The workpiece
samples were heat treated by induction hardening and a
hardness of 50 HRC was maintained. The samples were
then solution heat treated and oil quenched in order to
achieve the proper hardness.
In this study, an Optris CF4 infrared thermometer
was used to measure TCTI. The maximum temperature
(which was about 525 °C) was recorded around the
cutting zone. A total of 18 trials were conducted
throughout these experiments and brand new inserts
were used for each temperature measurement. Hence, the
cutting temperature increased with the cutting speed, the
feed rate and the depth of cut. The experiments were
repeated three times for the same cutting conditions and
the measured values were averaged. TT was measured
using a K-type thermocouple. The thermocouple
measurements were recorded every five seconds.
The Ra surface roughness was measured to charac-
terize the surface quality. The Ra measurements were
carried out using a Time TR 200 device by obtaining
values from different points that were parallel to the
workpiece axis at a cut-off length of 5.6 mm. According
to the experimental design, three measurements were
made on the surfaces at the specified values of the con-
trol factors, and the Ra values were determined by taking
the average of the measurement results.
3.3 Experimental design using the Taguchi method
The Taguchi design was selected to find the relation-
ships between the control factors and the quality charac-
teristics. The cutting speed (vc), feed rate (f) and depth of
cut (ap), whose levels are given Table 3, were selected as
the control factors. The quality characteristics were the
tool-chip interface temperature (TCTI), the tool tempe-
rature (TT) and the average surface roughness (Ra). As
the total degree of freedom of the factor group was 5, a
standard Taguchi experimental plan with the notation
L18 (21 × 37) was chosen as the orthogonal array. The
rows in the L18 orthogonal array used in the experiment
corresponded to each trial and the columns contained the
factors to be studied. The first column consists of the
depth of cut; the second and the third columns contain
the cutting speed and the feed rate, respectively. In the
Taguchi method, the experimental results are trans-
formed into a S/N ratio. The S/N ratio is used while
approaching or moving away from the desired value and
measuring the quality characteristics.15-18 The smaller-is-
better (SB), the nominal-best (NB) and the larger-is-
better (LB) approaches are found according to the results
of the S/N ratio.15-18 As the tool-chip interface tempera-
ture (TCTI), the tool temperature (TT) and the surface
roughness (Ra) values were required to be the lowest, the
S/N ratios of these quality characteristics were calculated
in dB using Equation (1) according to the SB option in
the study.15–18
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351 345
Figure 3: Thermocouple and IR pyrometer connections to the lathe
Slika 3: Povezava termo~lena in IR-pirometra na stru`nici
Figure 2: Schematic of tip geometry
Slika 2: Shema geometrije rezilne konice
S/N y i
i
n
SB lg= − ⋅
⎛
⎝
⎜ ⎞
⎠
⎟
=
∑10 1
2
2
1
(1)
In the Equation (1), n is the number of the experi-
ment and yi is the ith data point obtained.15–18 ANOVA
was applied in order to determine the percentage effects
of the control factors on TCTI, TT and Ra.
Table 3: Control factors and their levels
Tabela 3: Kontrolni faktorji in njihovi nivoji
Sym-
bol
Control
factors Unit
Level
1
Level
2
Level
3
Degree of
freedom
(DoF)
ap
Depth of
cut mm 0.40 0.60 – 1
vc
Cutting
speed m/min 76 114 170 2
f Feed rate mm/rev 0.05 0.08 0.12 2
3.4 Predictive models for temperature and surface
roughness with multiple regression analysis
Equations were developed for the prediction of TCTI,
TT and Ra using the experimental results in a multiple
regression analysis. The second-order linear models
containing the main effects of the control factors and
their interactions are signified with the Equation (2):
Y y b x b x b x
b x b x b x b x
1 0 0 1 1 2 2
3 3 4 12 5 13 6 23
= − = + + +
+ + + +

 (2)
where Y1 is the estimated answer of the second-order
equation and y is the tool-chip interface temperature
(TCTI), tool temperature (TT) or surface roughness (Ra)
measured on a the logarithmic scale, x0 = 1 is the fixed
variable, the x1, x2 and x3 control factors are the
logarithmic transformations of the depth of cut, the
cutting speed and the feed rate and, the x12, x13 and x23
interactions of the control factors are the logarithmic
transformations of the depth of cut–cutting speed, the
depth of cut–feed rate and cutting speed–feed rate. The
coefficient of the experimental error is 
, and the b
values (b0, b1, b2, b3, b4, b5 and b6) are the coefficients
of related factors.
4 ANALYSIS OF THE RESEARCH RESULTS
The present study was performed to understand and
evaluate the infrared- and thermocouple-based tempera-
ture measurements during metal cutting and to consider
the practical difficulties. TCTI, TT and Ra were used as the
quality characteristics. The experimental results are
shown in Table 4.
The TCTI, TT and Ra measurement results from the
turning of the quenched and tempered AISI 4140 steel
with coated carbide tools were resolved and analyzed by
means of the Minitab 16.0 package software. From
Table 4 it is clear that the overall means for TCTI, TT and
Ra were calculated as 446.11 °C, 70.78 °C and 0.578 μm,
respectively.
Table 4: The experimental results for the quality characteristics and
S/N ratios
Tabela 4: Rezultati preizkusov za opis kvalitete in S/N razmerja
Exp.
no
Control
factors Measured values S/N Ratios (dB)
ap vc f
Tool-
chip in-
terface
tempera-
ture, TCTI
(°C) (IR
Pyro-
meter)
Tool
tempera-
ture, TT
(°C)
(Thermo-
couple)
Surface
rough-
ness, Ra
(μm)
S/N
TCTI
S/N
TT
S/N
Ra
1 0.4 76 0.05 410 57 0.295 –52.26 –35.12 10.60
2 0.4 76 0.08 405 66 0.483 –52.15 –36.39 6.32
3 0.4 76 0.12 410 72 0.958 –52.26 –37.15 0.37
4 0.4 114 0.05 460 65 0.484 –53.26 –36.26 6.30
5 0.4 114 0.08 465 61 0.579 –53.35 –35.71 4.75
6 0.4 114 0.12 425 67 0.988 –52.57 –36.52 0.10
7 0.4 170 0.05 520 65 0.410 –54.32 –36.26 7.74
8 0.4 170 0.08 500 67 0.492 –53.98 –36.52 6.16
9 0.4 170 0.12 475 71 0.872 –53.53 –37.03 1.19
10 0.6 76 0.05 400 72 0.489 –52.04 –37.15 6.21
11 0.6 76 0.08 390 80 0.530 –51.82 –38.06 5.51
12 0.6 76 0.12 395 76 0.720 –51.93 –37.62 2.85
13 0.6 114 0.05 430 80 0.429 –52.67 –38.06 7.35
14 0.6 114 0.08 435 75 0.547 –52.77 –37.50 5.24
15 0.6 114 0.12 420 83 0.722 –52.46 –38.38 2.83
16 0.6 170 0.05 485 81 0.354 –53.71 –38.17 9.02
17 0.6 170 0.08 525 67 0.406 –54.40 –36.52 7.83
18 0.6 170 0.12 480 69 0.643 –53.62 –36.78 3.84
Overall mean of TCTI = 446.11 °C, S/N ratio of TCTI = –52.95 dB
Overall mean of TT = 70.78 °C, S/N ratio of TCTI = –36.95 dB
Overall mean of Ra = 0.578 ìm, S/N ratio of Ra = 5.24 dB
The variation of the tool temperature and the tool-
chip interface temperature with the cutting parameters
are shown in Figures 4a and 4b. Obviously, it is clear
that the tool-chip interface temperature and the tool tem-
perature increase with an increase in the cutting speed
(Figure 4a). The influence of the tool temperature and
the feed rate on the surface roughness is shown in Figure
5a. It was observed that the lowest feed rate produced a
better surface quality. The experiments showed that the
cutting speed and the feed rate are the main factors
affecting the surface roughness (Figure 5b).
4.1 Analysis of the control factors for the temperature
and surface roughness
The responses for the S/N ratios (smaller is better) of
TCTI, TT and Ra are presented in Table 5 and the
responses for the means in Table 6. While the signal
value represents the real desired value that the system
gives and which is to be measured, the noise factor
represents the portion of the undesired factors in the
measured value. The S/N ratio analysis provided signi-
ficant information about the nature of the process of
turning hardened AISI 4140 steel with coated carbide
cutting tools under selected conditions. The fact that the
differences between the highest and the lowest S/N ratio
values of each control factor calculated at different levels
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
346 Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351 347
Figure 4: The influence of cutting speed and feed rate on the tempera-
ture: a) cutting speed, b) feed rate
Slika 4: Vpliv hitrosti rezanja in hitrosti podajanja na temperaturah:
a) hitrost rezanja, b) hitrost podajanja
Figure 5: Influence of tool temperature and feed rate on surface
roughness: a) tool temperature, b) feed rate
Slika 5: Vpliv temperature orodja in hitrosti podajanja na hrapavost
povr{ine: a) temperatura orodja, b) hitrost podajanja
Table 5: Response table for S/N ratios (smaller is better) of TCTI, TT and Ra
Tabela 5: Razpredelnica odgovorov za S/N razmerja (manj{e je bolj{e) za TCTI, TT in Ra
Level Tool-chip interface temperature, TCTI(dB) Tool temperature, TT (dB) Surface roughness, Ra (dB)
ap vc f ap vc f ap vc f
1 –53.07 –52.08 –53.04 –36.33 –36.91 –36.84 4.838 5.313 7.873
2 –52.83 –52.85 –53.08 –37.58 –37.07 –36.78 5.632 4.429 5.969
3 – –53.93 –52.73 – –36.88 –37.24 – 5.963 1.864
 0.25 1.85 0.35 1.25 0.19 0.46 0.793 1.534 6.008
Rank 3 1 2 1 3 2 3 2 1
Table 6: Response table for means of TCTI, TT and Ra
Tabela 6: Tabela odgovorov za pomen TCTI, TT in Ra
Level Tool-chip interface temperature, TCTI(°C) Tool temperature, TT (°C) Surface roughness, Ra (μm)
ap vc f ap vc f ap vc f
1 452.2 401.7 450.8 65.67 70.50 70.00 0.6179 0.5792 0.4102
2 440.09 439.2 453.3 75.89 71.83 69.33 0.5378 0.6248 0.5062
3 – 497.5 434.2 – 70.00 73.00 – 0.5295 0.8172
 12.2 95.8 19.2 10.22 1.83 3.67 0.0801 0.0953 0.4070
Rank 3 1 2 1 3 2 3 2 1
are higher or lower was used in the determination of the
factors effective on TCTI, TT and Ra. The most effective
parameters on TCTI were the cutting speed, the feed rate
and the depth of cut because there were (1.85, 0.35 and
0.25) dB differences between their levels (Table 5). The
most effective parameters on TT were determined to be
the depth of cut, the feed rate and the cutting speed, with
differences of (1.25, 0.46 and 0.19) dB, respectively (Table
5). The most effective parameters on Ra were determined
to be the feed rate, the cutting speed and the depth of cut,
with differences of (6.008, 1.534 and 0.793) dB, res-
pectively (Table 5). The optimum values for the surface
roughness and the dimensional accuracy were reported to
be a2v1f3, a1v3f2 and a2v3f1, respectively (Table 6).
The main effects of the control factors on the perfor-
mance characteristics during the turning of the quenched
and tempered AISI 4140 steel with coated carbide
cutting tools were demonstrated using the "Graphical
Representation of Factor Effects" and evaluated.8–11 The
main effect graphs showing the effects of the control
factors on TCTI, TT and Ra are given in Figures 6 and 7,
respectively.
In Figure 6, the optimum levels of the control factors
for the tool-chip interface temperature are a2 (ap = 0.6
mm), v1 (vc = 76 m/min) and f3 (f = 0.12 mm/rev),
respectively. TCTI increases depending on the increase of
the cutting speed and the decrease of the depth of cut and
the feed rate. From the same graphic it is clear that the
most effective control factor on TCTI is the cutting speed.
In Figure 7, the optimum levels of the control factors for
the tool temperature are a1 (ap = 0.4 mm), v2 (vc = 114
m/min) and f2 (f = 0.08 mm/rev), respectively.
In Figure 7, when the effects of the control factors on
tool temperature were examined, a significant increase
was observed on TT, depending on the increase in the
depth of cut. With an increase of the cutting speed from
76 m/min to 114 m/min and an increase of the feed rate
from 0.08 mm/rev to 0.12 mm/rev the tool temperature
was increased (Figure 7). Similarly, the optimum levels
for the minimum Ra surface roughness were observed to
be a2 (ap = 0.6 mm), v3 (vc = 170 m/min) and f1 (f = 0.05
mm/rev), respectively (Figure 5). The most effective
parameter on Ra was the feed rate (Figure 7). With a
further increase in the feed rate value the Ra surface
roughness value increased.
ANOVA is a statistically based, objective, decision-
making tool used for determining any difference in the
average performance of a group of items being
tested.15–18 In the case when the F value of a process
parameter is greater than the tabulated F ratio, it shows
that the control factor has a significant effect on the per-
formance characteristic. An analysis of variance
(ANOVA) with a 95 % confidence interval was carried
out for each experiment using the L18 orthogonal array in
order to determine the effects of the control factors and
their interactions on selected performance/quality cha-
racteristics. The results of the ANOVA carried out for
TCTI, TT and Ra are presented in Tables 7, 8 and 9. The
cutting speed became the most effective factor for the
tool-chip interface temperature, with a contribution of
86.57 % followed by the feed rate with 4.03 % (Table 7).
The effects of other control factors and their interactions
on TCTI became insignificant with a smaller 5 % contri-
bution (Table 7). The results of the ANOVA for the tool
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
348 Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351
Figure 6: Mean effect plots for temperatures: a) tool-chip interface
temperature, b) tool temperature
Slika 6: Diagram srednjega vpliva na temperature: a) temperatura
stika orodje-ostru`ek, b) temperatura orodja
Figure 7: Mean effect plots for Ra surface roughness
Slika 7: Diagrami srednjega vpliva na hrapavost povr{ine Ra
temperature (TT) indicate that the depth of cut (ap) has
more influence on the tool temperature, with a contri-
bution of 52.65 % and a cutting speed–feed rate (vxf)
16.75 %, depth of cut–cutting speed (axv) 9.12 %, depth
of cut–feed rate (axf) 7.51 %, feed rate (f) 5.13 % and
cutting speed (vc) 1.21 % followed by a contribution %,
respectively (Table 8). Finally, from Table 9, it is
concluded that the feed rate with a contribution of
76.63 % has more influence on the surface roughness
(Ra) followed by the depth of cut–feed rate (axf), the
depth of cut (ap) and the cutting speed (vc) to obtain the
minimum surface roughness (Table 9).
Table 7: Results of ANOVA for tool-chip interface temperature (TCTI)
Tabela 7: Rezultati ANOVA za temperaturo stika orodje-ostru`ek
(TCTI)
Source DoF SS V F-Ratio Prob.>F Contr.(%)
ap 1 672.3 672.2 3.44 0.137 2.08
vc 2 27986.1 13993.1 71.71 0.001 86.57
f 2 1302.8 651.4 3.34 0.140 4.03
axv 2 302.8 151.4 0.78 0.519 0.94
axf 2 369.4 184.7 0.95 0.461 1.14
vxf 4 913.9 228.5 1.17 0.441 2.83
Res.Err. 4 780.6 195.1 2.41
Total 17 32327.8 100.00
R2 = 97.6, R2 (adj) = 89.7 (significant at 95 % confidence level)
Table 8: Results of ANOVA for tool temperature (Tt)
Tabela 8: Rezultati ANOVA za temperaturo orodja (Tt)
Source DoF SS V F-Ratio Prob.>F Contr.(%)
ap 1 470.22 470.22 27.57 0.006 52.65
vc 2 10.78 5.389 0.32 0.746 1.21
f 2 45.78 22.889 1.34 0.358 5.13
axv 2 81.44 40.722 2.39 0.208 9.12
axf 2 67.11 33.556 1.97 0.254 7.51
vxf 4 149.56 37.389 2.19 0.233 16.75
Res.Err. 4 68.22 17.056 7.64
Total 17 893.11 100.00
R2 = 92.4, R2 (adj) = 67.5 (significant at 95 % confidence level)
Table 9: Results of ANOVA for surface roughness (Ra)
Tabela 9: Rezultati ANOVA za hrapavost povr{ine (Ra)
Source DoF SS V F-Ratio Prob.>F Contr.(%)
ap 1 0.028880 0.028880 10.73 0.031 4.18
vc 2 0.027281 0.013641 5.07 0.080 3.95
f 2 0.543172 0.271586 100.89 0.000 78.63
axV 2 0.014830 0.007415 2.75 0.177 2.15
axf 2 0.062656 0.031328 11.64 0.022 9.07
Vxf 4 0.003192 0.000798 0.30 0.867 0.46
Res.Err. 4 0.010767 0.002692 – – 1.56
Total 17 0.690779 – – – 100.00
R2 = 98.4, R2 (adj) = 93.4 (significant at 95 % confidence level)
4.2 Developed second-order predictive equations for
the temperature and surface roughness
The equations that were developed with multiple
linear regression analysis to predict TCTI, TT and Ra in the
turning of quenched and tempered AISI 4140 steel with
coated carbide cutting tools and the equations that
contain the main effects of the control factors and their
interaction effects are presented in Equations (3) to (5),
respectively.
T a v f av vfCTI = − + + + −382 146 0 966 179 0 708 358. . . (4)
T a v f av vfT = − + + − −2 0 968 0328 255 0381 174. . . . . (5)
R a v f
av
a = − − + + −
− −
0114 034 0 00332 715
0 00617
. . . .
. 0 0102. vf
(6)
These equations were developed according to the
un-coded values of the control factors (i.e., 0.4, 0.6 mm,
etc. for ap; i.e., 76, 114, 170 m/min, etc. for vc; i.e., 0.05,
0.08, 0.012 mm/rev, etc. for f). af is highly correlated
with other variables, so af has been removed from all of
the equations. The correlation coefficients (R2) and the
adjusted correlation coefficients (R2(adj)) of the second-
order equations developed for the predictive tool-chip
interface temperature (TCTI) measured with an IR pyro-
meter, the tool temperature (TT) measured with a thermo-
couple and the surface roughness (Ra) were calculated as
R2 = 92.8 %, R2(adj) = 89.8 %, R2 = 68.1 %, R2(adj) =
54.8 % and R2 = 82.6 % R2(adj) = 75.3 %, respectively.
R2(adj) determines the amount of deviation about the
mean that is described by the model. The predicted R2
value and the R2(adj) value were found to be in good
agreement. These values show that the equations deve-
loped are sufficient to determine all the response values
at a confidence interval of 95 %. The regression models
can be successfully adopted for estimating TCTI, TT and
Ra. Moreover, as seen in these equations, vc and f have
additive effects, while ap has a negative effect on TCTI, TT
and Ra.
The comparisons of the results of TCTI, TT and Ra
measured experimentally (Table 4) with the fits for TCTI,
TT and Ra estimated via the Taguchi method and fits for
TCTI, TT and Ra estimated via the Regression model
(Equation (3) to (5)) are given in Table 8. As can be seen
from this table, the TCTI results obtained from the
Taguchi method and the linear regression analysis were
found to be very close. The mean of the % error ratios of
the estimated results obtained by the Taguchi method
and the predictive equations were less than 14 %. This
reflects the reliability of the statistical analyses (Tables
10 and 11).
5 CONCLUSIONS
In this study, the Taguchi design was selected to
determine the effects of the control factors. The effects
of the depth of cut, the cutting speed and the feed rate on
the tool-chip interface temperature (TCTI), the tool
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351 349
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
350 Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351
Table 10: The comparisons of measured of TCTI, and Tt experimentally with fits estimated via the Taguchi method and regression models
Tabela 10: Primerjava izmerjenih TCTI in eksperimentalnih Tt, z ujemanji, dolo~enimi po Taguchi metodi in z regresijskimi modeli
Exp.
no
Tool-chip interface temperature, TCTI (°C) Tool temperature, Tt (°C)
Meas-
ured
TCTI
Fits for TCTI
estimated via
Taguchi method
Error
%
Fits for TCTI
estimated via
Regression model
Error
%
Meas-
ured
Tt
Fits for Tt
estimated via
Taguchi method
Error
%
Fits for Tt
estimated via
Regression model
Error
%
1 410 418 2.0 414 0.9 57 56 1.0 60 5.7
2 405 401 0.9 411 1.4 66 68 2.9 64 3.2
3 410 406 1.1 407 0.7 72 71 1.9 69 4.5
4 460 462 0.5 454 1.3 65 62 3.9 64 2.1
5 465 458 1.5 447 3.8 61 61 0.1 65 7.0
6 425 430 1.1 438 3.1 67 70 3.9 68 0.8
7 520 510 2.0 514 1.1 65 68 4.8 69 5.5
8 500 511 2.1 501 0.2 67 65 2.8 67 0.5
9 475 475 0.1 484 1.9 71 70 1.7 66 7.5
10 400 392 2.0 395 1.2 72 73 0.8 74 2.5
11 390 394 0.9 392 0.6 80 78 2.4 77 3.1
12 395 399 1.1 389 1.6 76 77 1.8 82 8.4
13 430 428 0.5 441 2.6 80 83 3.2 74 7.1
14 435 442 1.6 434 0.2 75 75 0.1 76 1.3
15 420 415 1.1 425 1.2 83 80 3.1 78 5.8
16 485 495 2.1 509 4.9 81 78 3.8 75 7.4
17 525 514 2.0 496 5.5 67 69 2.8 74 10.1
18 480 480 0.1 479 0.2 69 70 1.8 72 4.5
Min 390 392 0.1 389 0.2 57 56 0.1 60 0.5
Max 525 514 2.1 514 5.5 83 83 4.8 82 10.1
Mean 446 – 1.3 – 1.8 71 – 2.4 – 4.8
Table 11: The comparisons of the measured Ra with fits estimated via the Taguchi method and the regression models
Tabela 11: Primerjava izmerjene Ra z ujemanji, dolo~enimi po Taguchi metodi in z regresijskimi modeli
Exp.no Measured Ra Fits for Ra estimated viaTaguchi method Error %
Fits for Ra estimated via
Regression model Error %
1 0.295 0.338 14.4 0.406 37.5
2 0.483 0.478 1.1 0.597 23.5
3 0.958 0.921 3.9 0.851 11.1
4 0.484 0.461 4.7 0.418 13.5
5 0.579 0.594 2.5 0.598 3.3
6 0.988 0.996 0.8 0.837 15.3
7 0.410 0.390 4.9 0.438 6.7
8 0.492 0.483 1.9 0.600 21.9
9 0.872 0.901 3.4 0.816 6.4
10 0.489 0.446 8.7 0.380 22.3
11 0.530 0.535 1.0 0.571 7.7
12 0.720 0.757 5.2 0.826 14.7
13 0.429 0.452 5.3 0.346 19.4
14 0.547 0.532 2.7 0.525 4.0
15 0.722 0.714 1.1 0.765 5.9
16 0.354 0.374 5.7 0.296 16.5
17 0.406 0.415 2.3 0.458 12.8
18 0.643 0.614 4.6 0.674 4.9
Min 0.295 0.338 0.8 0.296 3.3
Max 0.988 0.996 14.4 0.851 37.5
Mean 0.578 – 4.1 – 13.7
temperature (TT) and the surface roughness (Ra) were
investigated in the turning of the quenched and nor-
malized AISI 4140 steel workpieces that were machined
using TiAlN-TiN, PVD-coated, carbide tools and the
obtained results are as follows:
The most effective parameter on the tool-chip inter-
face temperature was the cutting speed with a contribu-
tion ratio of 86.57 %. The effective parameters for tool
temperature were the depth of cut, the cutting speed-feed
rate, the depth of cut–cutting speed, the depth of cut-feed
rate, the feed rate and the cutting speed with contribu-
tions of 52.65 %, 16.75 %, 9.12 %, 7.51 %, 5.13 % and
1.21 %.
The feed rate with a contribution of 76.63 % has
more influence on the surface roughness (Ra) followed
by the depth of cut–feed rate (axf), the depth of cut (ap)
and the cutting speed (vc) to obtain the minimum surface
roughness.
The optimum levels of the control factors were ap =
0.6 mm, vc = 76 m/min and f = 0.12 mm/rev for the
minimum tool-chip interface temperature; ap = 0.4 mm,
vc = 114 m/min and f = 0.08 mm/rev for the minimum
tool temperature, and ap = 0.6 mm, vc = 170 m/min and
f = 0.05 mm/rev for the minimum Ra surface roughness.
The tool-chip interface temperature increased signi-
ficantly depending on the increase of the cutting speed.
The depths of cut and feed rate do not have a significant
effect on the tool-chip interface temperature.
The tool temperature increased significantly depend-
ing on the increase of the depth of cut.
The surface roughness increased depending on the in-
crease of the feed rate, while the same tendency was not
observed for the depth of cut and the cutting speed.
The correlation coefficients of the predictive equa-
tions developed for the estimation of the minimum
tool-chip interface temperature, the tool temperature and
the surface roughness by multiple linear regression anal-
ysis were calculated as 0.928, 0.681 and 0.826, respec-
tively. Higher correlation coefficients reflect the reliabi-
lity of the developed equations.
The mean of the % error ratios of the estimated re-
sults obtained by Taguchi method and the predictive
equations were less than 14 %. This reflects the reliabi-
lity of the statistical analyses.
6 REFERENCES
1 X. L. Liu, D. H. Wen, Z. J. Li, L. Xiao, F. G. Yan, Cutting tempe-
rature and tool wear of hard turning hardened bearing steel, Journal
of Materials Processing Technology, 129 (2002), 200–206,
doi:10.1016/S0924-0136(02)00651-9
2 N. A. Abukhshim, P. T. Mativenga, M. A. Sheikh, Investigation of
heat partition in high speed turning of high strength alloy steel, Inter-
national Journal of Machine Tools and Manufacture, 45 (2005),
1687–1695, doi:10.1016/j.ijmachtools.2005.03.008
3 N. A. Abukhshim, P. T. Mativenga, M. A. Sheikh, Heat generation
and temperature prediction in metal cutting: a review and impli-
cations for high speed machining, International Journal of Machine
Tools and Manufacture, 46 (2006) 7–8, 782-800, doi:10.1016/
j.ijmachtools.2005.07.024
4 A. H. Suhail, N. Ismail, S. V. Wong, N. A. Abdul Jalil, Optimization
of cutting parameters based on surface roughness and assistance of
workpiece surface temperature in turning process, American Journal
of Engineering and Applied Sciences, 3 (2010) 1, 102–108,
doi:10.3844/ajeassp.2010.102.108
5 P. D. Berger, R. E. Maurer, Experimental design with applications in
management, Engineering and The Sciences, 1th ed., Duxbury Press,
USA 2001
6 G. R. Henderson, Six sigma: Quality improvement with MINITAB,
2nd ed., John Wiley and Sons, England 2006, 452–460
7 T. P. Ryan, Statistical methods for quality improvement, 2nd ed.,
John Wiley and Sons, USA 2000
8 A. Mishra, A. Gangele, Application of Taguchi method in optimi-
zation of tool flank wear width in turning operation of AISI 1045
steel, Industrial Engineering Letters, 2 (2012) 8, 11–18
9 M. S. Phadke, Quality engineering using design of experiment, quali-
ty control, Rebust design and Taguchi method, 1st ed., Warsworth
and Books, California 1998
10 W. H. Yang, Y. S. Tarng, Design optimization of cutting parameters
for turning operations based on Taguchi method, Journal of Materials
Processing Technology, 84 (1998), 122–129, doi:10.1016/S0924-
0136(98)00079-X
11 B. M. Gopalsamy, B. Mondal, S. Ghosh, Taguchi method and
ANOVA: An approach for process parameters optimization of hard
machining while machining hardened steel, Journal of Scientific and
Industrial Research, 68 (2009) 8, 686-695
12 F. Ficici, M. Kapsiz, M. Durat, Applications of Taguchi design
method to study wear behaviour of boronized AISI 1040 steel,
International Journal of Physical Sciences, 6 (2011) 2, 237–243
13 M. Adinarayana, G. Prasanthi, G. Krishnaiah, Parametric analysis
and multi objective optimization of cutting parameters in turning
operation of AISI 4340 alloy steel with CVD cutting tool, Inter-
national Journal of Research in Engineering and Technology, 3
(2014) 2, 449–456
14 E. D. Kirby, A parameter design study in a turning operation using
the Taguchi method, The Technology Interface, (2006), 1–14
15 P. J. Ross, Taguchi techniques for quality engineering: Loss function,
orthogonal experiments, parameter and tolerance design, McGraw-
Hill, New York 1988
16 D. C. Montgomery, Taguchi’s contributions to experimental design
and quality engineering, design and analysis of experiment, Wiley,
Canada 1991
17 E. Canyýlmaz, F. Kutay, An alternative approach to analysis of vari-
ance in Taguchi method, Journal of the Faculty of Engineering &
Architecture of Gazi University, 18 (2003) 3, 51–63
18 R. K. Roy, A primer on the Taguchi method, Van Nostrand Reinhold,
New York 1990
A. R. MOTORCU et al.: ANALYSIS OF THE CUTTING TEMPERATURE AND SURFACE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 343–351 351

I. KIRIK: WELDABILITY OF Ti6Al4V TO AISI 2205 WITH A NICKEL INTERLAYER ...
353–356
WELDABILITY OF Ti6Al4V TO AISI 2205 WITH A NICKEL
INTERLAYER USING FRICTION WELDING
PREIZKU[ANJE VARIVOSTI PRI VARJENJU S TRENJEM Ti6Al4V
IN AISI 2205 Z VMESNO PLASTJO NIKLJA
Ihsan Kirik
Batman University, Faculty of Engineering and Architecture, Department of Metallurgical and Materials Engineering, 72060 Batman, Turkey
alihsankirik@gmail.com
Prejem rokopisa – received: 2015-02-12; sprejem za objavo – accepted for publication: 2015-05-04
doi:10.17222/mit.2015.039
The aim of this study was to friction weld dissimilar metals, i.e., Ti6Al4V to duplex stainless steel, with and without a nickel
interlayer using a new method. The metallographic examinations of the weld were carried out and the strength of the joints was
determined with tensile tests. The experimental results indicate that the Ti6Al4V and duplex stainless steel could be joined with
a nickel interlayer. The highest tensile strength (605 MPa) was obtained and the tensile strength of the joint was significantly
increased with an increase of the rotation speed and the friction pressure.
Keywords: friction welding, titanium alloy, duplex stainless steel, nickel interlayer
Namen te {tudije je varjenje s trenjem razli~nih materialov Ti6Al4V in dupleks nerjavnega jekla, z in brez vmesne plasti niklja,
z uporabo nove metode. Izvr{eni so bili metalografski preizkusi zvara. Rezultati preizkusov ka`ejo, da je z vmesnim slojem
niklja mogo~e spajati Ti6Al4V in dupleks nerjavno jeklo. Z nara{~anjem hitrosti vrtenja in pritiska pri trenju je mogo~e dose~i
najvi{jo natezno trdnost (605 MPa), hkrati pa nara{~a tudi natezna trdnost spoja.
Klju~ne besede: varjenje s trenjem, titanova zlitina, dupleks nerjavno jeklo, vmesni sloj niklja
1 INTRODUCTION
Materials such as low-carbon steels, ceramics and
composites that are problematic and difficult to join
using traditional welding methods can be joined by
friction welding (FW). One of the biggest advantages of
FW is the production of a new material from pairs of
dissimilar materials.1–2 FW is an important manu-
facturing technique used in the machine construction and
hydraulic industry, the automotive industry, and the
industry for cutting and drilling tools. This technique has
an important application area in welding technology, as
it can join materials with different compositions if their
sizes and shapes are appropriate, it does not have a
limited melting event and it has a very low welding
failure.3–5
The joinability of Ti and its alloys to steels is
extremely important because more and more of these
metals are used together.6 When Ti alloys and stainless
steels join mechanically, many intermetallic phases and
different tension concentration areas occur at the inter-
mediate of the joint, which then causes embrittlement
and cracking.7 When Ti directly welds to the stainless
steel, intermetallic compounds such as FeTi and Fe2Ti
can occur at the joining area due to Ti and Fe having
very little fusion capabilities. Besides, TiC may form
since Ti is a strong carbide-forming element and also
occurrences of these compounds cause a crispiness in the
joining area. On the other hand, cracks may occur in the
welding area due to the difference in the thermal
conductivity between the Ti alloy and the steel. To avoid
these negatives, the Ni interlayer diffusion welding
method was used to combine the TiC4, the Ti alloy and
the stainless steel, and sound joints were obtained, which
had high strengths.8 The Ti-6Al-4V alloy and a micro-
duplex stainless steel (AVESTA 2205) were joined using
the diffusion-welding technique, and good results were
obtained at temperatures as low as 800 °C in 30 min.9
Aleman et al.10 studied pure Ti and 316L stainless steel
using the diffusion-welding technique and they stated
that the -phase was observed on the stainless-steel side,
Fe2Ti and FeTi in the inner side, and Fe2Ti4O oxide on
the Ti side. Muralimohan et al.11 welded Ti and 304L
stainless steel by FW and through a nickel interlayer,
which is deposited by electroplating on stainless-steel
substrates with a range of 100±3 μm. The joining of
dissimilar materials such as aluminium, titanium, magne-
sium and their alloys to stainless steels was reported in 1–14,
but very limited studies are reported for titanium and its
alloy to stainless steel using FW. Moreover, using an
interlayer in FW is very limited, because it is difficult to
keep the intermediate layer in the intermediate zone.
TiAl and AISI 4140 steel were joined through FW with a
copper insert layer and two-step joining of the FW was
carried out to complete the joints by W. B. Lee et al.12
Madhusudhan and Venkata13 investigated the role of the
nickel insert layer in the FW of maraging steel to low-
alloy steel. To incorporate nickel as an interlayer, marag-
Materiali in tehnologije / Materials and technology 50 (2016) 3, 353–356 353
UDK 621.791:669.295:669.055:669.24 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)353(2016)
ing steel and nickel were welded first. To maintain the
5-mm length of interlayer the nickel was cut. Sub-
sequently, low-alloy steel was welded on the nickel side
of the joint.
The literature examined for using an insert layer in
FW joining revealed two stages or the interlayer is kept
at the interface using a different method. For this reason,
the goal of this study is the successful joining of the
Ti6Al4V titanium alloy and the AISI 2250 duplex stain-
less using FW single stage and with a nickel interlayer
using a new technique. The strengths of the joints were
determined by tensile tests and compared with those of
base materials. Then, the microstructures in the welding
zone were obtained and investigated using macro- and
micro-photographs.
2 EXPERIMENTAL PROCEDURES
In this study, three materials with different properties
were used in order to be joined using friction welding.
The analysed chemical compositions of the materials are
illustrated in Table 1. The duplex stainless steel and the
Ti6Al4V alloy bars were provided commercially with a
12-mm diameter and were processed on a turning lathe
machine according to the sizes shown in Figure 1a.
Firstly, during the FW the most difficult stage while
using the insert layer is to keep it at the interface. Thus,
to avoid the escape of the interlayer, the surface of the
duplex stainless steel is processed with a 10 mm dia-
meter and a 3 mm depth, as shown in Figure 1a. Later,
the nickel interlayer was placed on the forehead of the
duplex stainless steel with the help of a press and then
the welding was performed in a continuously driven FW
machine using the parameters given in Table 2. After the
FW, in order to determine the microstructural changes to
the samples that were friction welded, the axial cross-
section of the joint was achieved by abrasive cutting and
attached for polishing, and then the samples were etched
in a chemical solution for the metallographic examina-
tion. The microstructure analyses of the joints were
carried out with an optical microscope, and the scanning
electron microscopy (SEM) and quantitative chemical
analyses were performed with an energy-dispersive
spectrometer (EDS). The friction-welded joints, with and
without an interlayer, were tested for their tensile
strength.
I. KIRIK: WELDABILITY OF Ti6Al4V TO AISI 2205 WITH A NICKEL INTERLAYER ...
354 Materiali in tehnologije / Materials and technology 50 (2016) 3, 353–356
Table 1: Chemical compositions of the test materials, in mass fractions (w/%)
Tabela 1: Kemijska sestava preiskovanih materialov, v masnih dele`ih (w/%)
Materials Alloying elements, in mass fractions (w/%)
Ti C Mn P Si Cr Al Ni Cu V Fe
Ti6Al4V Bal = 0.08 – – 0.15 – 5–6.50 – – 3.5–4.5 = 0.40
AISI 2205 – 0.01–0.03 1.68–2.00 0.026 – 21–23 – 3.37 – – Bal
Nickel – – 0.007 – – – – Bal 0.01 – –
Table 2: The process parameters, used in the FW and the tensile test results according to these parameters
Tabela 2: Procesni parametri, uporabljeni pri FW in rezultati nateznega preizkusa pri teh parametrih
Sample
No.
Welding parameters
Rotation speed
(m–1)
Fric. time
(s)
Fric. press.
(MPa)
Forging
pressure (MPa)
Forging time
(s)
Without
Interlayer
With Interlayer
Ten. Stren.
(MPa)
N1 1500 6 150 200 4 Failed 380
N2 1500 6 125 200 4 Failed 326
N3 1500 6 100 200 4 Failed –
N4 1800 6 150 200 4 Failed 420
N5 1800 6 125 200 4 Failed 605
N6 1800 6 100 200 4 Failed 180
Nickel – – – – – – 450
AISI 2205 – – – – – – 956
Ti6Al4V – – – – – – 870
Figure 1: Schematic illustration of friction welded Ti6Al4V and AISI
2205 using: a) nickel interlayer, b) macrograph of cross-section of
friction welded N5 sample
Slika 1: Shematski prikaz zvara pri trenju Ti6Al4V in AISI 2205: a) z
uporabo vmesnega sloja niklja, b) makroposnetek preseka vzorca N5,
zvarjenega s trenjem
3 RESULTS AND DISCUSSION
3.1 Examinations of microstructure
A macro image of the FW joint N5 sample is shown
in Figure 1b. It is evident that the joint is unsymmetrical
and the flash dimensions of the Ti alloy are much larger
than the stainless steel due to the decrease of the tensile
yield strength of the Ti alloy with temperature being
much more significant than that of duplex stainless steel,
although at room temperature the situation is exactly
reversed.15
The optical photographs taken from the FW joint N5
sample shows that the Ti6Al4V/AISI 2205 materials
were successfully friction welded using a nickel
interlayer (Figure 2). It is clear that there is a structural
disorder and differently directed grains on the Ti6Al4V
side near the nickel interlayer. Moreover, the FW of
dissimilar metals, Ti6Al4V to duplex stainless steel,
without an interlayer were tried so many times with
different parameters, but the joints were not achieved due
to brittle phase reaction and volume expansion.
The SEM micrograph for the N4 sample joined at a
rotation of 1800 min–1, a 6-s friction time, a 150 MPa
friction pressure, a 200 MPa forging pressure and a 4-s
forging time is illustrated in Figure 3. From the micro-
graph it is clear that at the nickel and Ti6Al4V interface
there were three different zones: (-Ti + -Ti) of
Ti6Al4V, -phase and intermetallic phase in interface
zone, nickel zone and invisible stainless steel zone.
From 14, when Ti alloys and the stainless-steel joint FeTi
and Fe2Ti intermetallic phase occurred. Also, these
phases adversely affect the quality of the joint. To elimi-
nate the effects of this occurrence a nickel interlayer was
used and the results confirmed that this is necessary. The
concentrations of the elements across the interface zone
are shown in Table 3 and the EDS graph is presented in
Figure 3. From the EDS results of the N4 sample diffe-
rent amounts of Al, Si, Ti, V, Ni and Mo were obtained,
but in the nickel some oxygen is obtained. As shown in
Table 3, Ti and Ni have a high degree of diffusion in the
approximately 15-μm region. Also, Ti diffused from
Ti6Al4V to the nickel interlayer, the nickel diffused into
the Ti alloy over the same distance. The FW of dissimilar
materials with a nickel interlayer exhibits a wider pla-
sticized zone in the middle of the welding interface. The
variation of the alloying elements at the mutual interfe-
rence is on account of the thermo-plastic stirring and the
diffusion mechanism.
Table 3: Elementary variation rates from EDS analyses across the
welding interface of the N4 sample
Tabela 3: Spreminjanje vsebnosti elementov dolo~ena z EDS analizo
preko zvara vzorca N4
Alloying
elements
(w/%)
EDS points
1 2 3
O – – 12.47
Al 6.799 5.755 –
Si – 0.799 –
Ti 88.087 60.583 0.181
V 4.025 1.950 –
Ni 0.910 30.705 87.160
Mo 0.189 0.208 0.188
I. KIRIK: WELDABILITY OF Ti6Al4V TO AISI 2205 WITH A NICKEL INTERLAYER ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 353–356 355
Figure 3: SEM photograph of the interface of the N4 sample and the
EDS analysis
Slika 3: SEM-posnetek vmesnega sloja pri vzorcu N4 in EDS-analiza
Figure 2: Optical microstructure of the interface of the N5 sample
Slika 2: Mikrostruktura vmesnega sloja pri vzorcu N5
3.2 Tensile testing
The tensile strength of the joints was determined, as
seen in Table 2, based on two different rotation speeds
(1500 min–1 and 1800 min–1), three different friction
pressures (100, 125 and 150) MPa, a constant friction
time (6 s), a forging time (4 s) and a forging pressure
(200 MPa). The highest tensile strength (605 MPa) was
obtained for the N5 sample. The tensile strength was
increased by increasing the rotation speed with a
constant friction time and pressure, and increasing the
tensile strength results in heat input and a high plastic
deformation. The fracture surface of the tensile test
sample N5 is characterized with SEM to understand the
failure mechanism. Figure 4 illustrates the fracture-
surface morphologies taken from the centre of N5. It is
clear that the fracture surface of the joint indicates a
brittle cleavage fracture for Ti6Al4V/AISI 2205 stainless
steel for different materials welding by FW using a
nickel interlayer.
4 CONCLUSION
The effects of a nickel interlayer and the process
parameters on the microstructure and tensile strength of
friction welds between a Ti6Al4V alloy and AISI 2205
duplex stainless steel with and without an interlayer were
studied. The following results were achieved:
Ti6Al4V and AISI 2205 were successfully joined by
FW using a nickel interlayer.
In the event of FW between the Ti6Al4V and AISI
2205 the soundness of the joints was increased with a
nickel interlayer and the rotation speed; but without an
interlayer cracks occurred and it was not joined due to a
brittle phase reaction and the volume expansion.
The highest tensile strength (605 MPa) was obtained
for the N5 sample joined at 1800 min–1, a 6-s friction
time, 125 MPa of friction pressure, 200 MPa of forging
pressure and a 4-s forging time. The tensile strength was
increased when the rotation speed and the friction
pressure increased.
5 REFERENCES
1 N. Özdemir, Investigation of the mechanical properties of fric-
tion-welded joints between AISI 304L and AISI 4340 steel as a
function rotational speed, Materials Letters, 59 (2005), 2504–2509,
doi:10.1016/j.matlet.2005.03.034
2 I. Krik, N. Özdemir, Weldability and joining characteristics of AISI
420/AISI 1020 steels using friction welding, Int. J. of Mat. Research,
8 (2013), 769–776, doi:10.3139/146,110917
3 M. Sahin, Characterization of properties in plastically deformed
austenitic-stainless steels joined by friction welding, Materials and
Design, 30 (2009), 135–144, doi:10.1016/j.matdes.2008.04.033
4 V. V. Satyanarayana, G. M. Reddy, T. Mohandas, Dissimilar metal
friction welding of austenitic-ferritic stainless steels, J. of Mat. Proc.
Tech., 160 (2005) 2, 128–137, doi:10.1016/ j.jmatprotec.2004.05.017
5 H. Ates, M. Turker, A. Kurt, Effect of friction pressure on the pro-
perties of friction welded MA956 iron-based superalloy, Materials
and Design, 28 (2007) 3, 948–953, doi:10.1016/j.matdes.2005.
09.015
6 M. Ghosh, S. Chatterjee, Effect of interface microstructure on the
bond strength of the diffusion welded joints between titanium and
stainless steel, Materials Characterization, 54 (2005) 4–5, 327–337,
doi:10.1016/j.matchar.2004.12.007
7 M. Ghosh, S. Chatterjee, Characterization of transition joints of
commercially pure titanium to 304 stainless steel, Materials Charac-
terization, 48 (2002) 5, 393–399, doi:10.1016/S1044-5803(02)
00306-6
8 P. He, J. C. Feng, B. G. Zhang, Y. Y. Qian, A new technology for
diffusion bonding intermetallic TiAl to steel with composite barrier
layers, Materials Characterization, 50 (2003) 1, 87–92, doi:10.1016/
S1044-5803(03)00122-0
9 N. Orhan, T. I. Khan, M. Eroðlu, Diffusion bonding of a microduplex
stainless steel to Ti–6Al–4V, Scripta Materialia, 45 (2001) 4,
441–446, doi:10.1016/S1359-6462(01)01041-7
10 B. Aleman, I. Gutiérrez, J. J. Urcola, The use of kirkendall effect for
calculating intrinsic diffusion coefficients in a 316L/Ti6242 diffusion
bonded couple, Scripta Materialia, 36 (1997) 5, 509–515,
doi:10.1016/S1359-6462(96)00414-9
11 C. H. Muralimohan, V. Muthupandi, K. Sivaprasad, Properties of
friction welding titanium-stainless steel joints with nickel interlayer,
Procedia Materials Science, 5 (2014), 1120–1126, doi:10.1016/
j.mspro.2014.07.406
12 W. B. Lee, Y. J. Kim, S. B. Jung, Effects of copper insert layer on the
properties of friction welded joints between TiAl and AISI 4140
structural steel, Intermetallics, 12 (2004) 6, 671–678, doi:10.1016/
j.intermet.2004.02.004
13 G. Madhusudhan Reddy, P. Venkata Ramana, Role of nickel as an
interlayer in dissimilar metal friction welding of maraging steel to
low alloy steel, J. of Mat. Process. Tech., 212 (2012), 66–77,
doi:10.1016/j.jmatprotec.2011.08.005
14 B. Kurt, N. Orhan, E. Evin, A. Çalik, Diffusion bonding between
Ti–6Al–4V alloy and ferritic stainless steel, Materials Letters, 61
(2007) 8–9, 1747–1750, doi:10.1016/j.matlet.2006.07.123
15 P. Li, J. Li, M. Salman, L. Liang, J. Xioeng, F. Shang, Effect of
friction time on mechanical and metallurgical properties of
continuous drive friction welded Ti6Al4V/SUS321 joints, Materials
and Design, 56 (2014), 649–656, doi:10.1016/ j.matdes.2013.11.065
I. KIRIK: WELDABILITY OF Ti6Al4V TO AISI 2205 WITH A NICKEL INTERLAYER ...
356 Materiali in tehnologije / Materials and technology 50 (2016) 3, 353–356
Figure 4: SEM photograph of fractured surface of N5 sample
Slika 4: SEM-posnetek povr{ine preloma vzorca N5
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
357–364
EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON
THE BEAD GEOMETRY OF BORATED STAINLESS-STEEL GTA
WELDS
VPLIV AKTIVIRANEGA TOPILA IN DODATKA DU[IKA NA
GEOMETRIJO KOPELI PRI GTA ZVARIH BORIRANEGA
NERJAVNEGA JEKLA
Guttikonda Raja Kumar1, Gabbita Durga Janaki Ram2, Sajja Rama Koteswara Rao3
1Swarna Barathi Institute of Science and technology, Khammam, T. S., 507002, India
2IIT Madras, Department of Metallurgical & Materials Engineering, Chennai 600036, India
3S.S.N. College of Engineering, Department of Mechanical Engineering, Chennai 603110, India
rajkumardotcom@yahoo.com
Prejem rokopisa – received: 2015-03-05; sprejem za objavo – accepted for publication: 2015-05-14
doi:10.17222/mit.2015.052
Borated stainless steels (304B) are used in nuclear power plants as control rods, shielding material, spent-fuel storage racks and
transportation casks as they have a high capacity to absorb thermal neutrons. In this study, bead-on-plate welds were made on
10-mm-thick 304B plates using gas tungsten arc welding with Ar and Ar+2% nitrogen as the shielding gases, activated-flux
GTA and electron-beam welding processes. The effects of the activated flux and nitrogen addition to the weld metal through the
shielding gas, on the microstructure, bead geometry and mechanical properties were investigated. Activated-flux GTA welding
and electron-beam welding substantially enhanced the depth of penetration and the aspect ratio compared to the other processes.
Full-penetration welds were obtained in a single pass using activated-flux GTA and EB welding. The fusion-zone (FZ)
microstructure of an activated GTA weld exhibits a columnar dendritic structure with eutectic borides in interdendritic regions,
while a fine equiaxed dendritic structure was noticed in EB welds. GTA, nitrogen-added GTA and activated-flux GTA welds
exhibited a partially melted zone adjacent to the fusion zone, with the activated-flux GTAW process resulting in a significantly
thinner partially melted zone (PMZ). No PMZ was noticed in the EB welds. All the welds exhibited a high joint efficiency and
impact toughness equal to those of the base material. It is concluded that the activated-flux GTA and EB welding processes are
advantageous due to the use of a low heat input and failure location.
Keywords: borated stainless steels, bead geometry, activated flux, GTAW, partially melted zone, nitrogen
Borirana nerjavna jekla (304B) se uporabljajo v nuklearnih elektrarnah kot kontrolne palice, za{~itni material, nosilci shranjeval
za izrabljene palice in za transportne sode, saj imajo sposobnost velike absorbcije termi~nih nevtronov. V tej {tudiji so bili
izdelani okrogli in plo{~ati zvari na 10 mm debelih plo{~ah iz 304B, z uporabo TIG varjenja v za{~itni atmosferi iz Ar + 2 %
du{ika, z aktiviranim talilom GTA in varjenjem z elektronskim curkom. Preiskovan je bil vpliv aktiviranega talila in dodatka
du{ika v zvar s pomo~jo za{~itnega plina, na mikrostrukturo, geometrijo kopeli in mehanske lastnosti. Varjenje z aktiviranim
talilom GTA in varjenje z elektronskim curkom, ob~utno pove~ata globino penetracije in razmerje, v primerjavi z drugimi
procesi. Z uporabo aktiviranega talila GTA in z varjenjem z elektronskim curkom je bila dose`ena popolna penetracija zvarov.
Mikrostruktura cone zlivanja (FZ) pri zvaru z aktiviranim talilom GTA, ka`e stebrasto dendritno strukturo z evtekti~nimi boridi
v meddendritnih podro~jih, medtem ko je bila pri zvarih z elektronskim curkom opa`ena struktura z drobnimi enakoosnimi
dendriti. Zvari GTA, GTA z dodanim du{ikom in GTA z aktiviranim talilom, so kazali delno raztaljena podro~ja okrog podro~ja
zlivanja ter z mo~no stanj{ano delno staljeno podro~je (PMZ) pri postopku z aktiviranim talilom GTA. V zvarih z elektronskim
curkom ni bilo opaziti PMZ. Vsi zvari so pokazali veliko u~inkovitost spoja in udarno `ilavost, ki je enaka kot pri osnovnem
materialu. Ugotovljeno je, da imata varjenje z aktiviranim talilom GTA in varjenje z elektronskim snopom, prednost zaradi
uporabe majhnega vnosa toplote in zaradi mo`nosti lokacije napak.
Klju~ne besede: borirano nerjavno jeklo, geometrija kopeli, aktivirano talilo, GTAW, delno staljeno podro~je, du{ik
1 INTRODUCTION
Borated stainless steel 304B is an austenitic-type
stainless steel containing 0.2 % to 2.25 % boron; it is
widely used in nuclear industries for various applica-
tions. These applications involve storage of spent nuclear
fuel in the forms of long-term storage tanks or caskets,
transportation baskets and control rods.1,2 The objective
of developing these steels is to absorb radiation from the
spent fuel. Due to the presence of the 10B isotope, these
stainless steels have superior neutron-absorption capabi-
lities.3 Spent fuel rods are stored in dry casks, made up
of borated stainless steel for long-term thermal neutron
irradiation. Despite the fact that boron allows an ade-
quate neutron absorption, it has an adverse influence on
mechanical properties, particularly on the fracture
toughness. Very low solubility of boron in an austenitic
matrix results in the formation of hard, brittle (Cr, Fe)2B
precipitating phases, which are low-melting eutectics.4
The size, shape and distribution of these borides also
deteriorate the mechanical properties.5 Park et al.6 con-
ducted post-weld heat treatments and reported that sphe-
roidized accicular eutectic phases were found to enhance
the ductility.
Earlier, borated SSs were typically used as bolts on
additions to a structural framework. However, due to the
Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364 357
UDK 62-4:621.791.053:669.14.018.8 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)357(2016)
slow and non-automatic riveting process, welding was
adopted. The solidification cracking susceptibility is
high at 0.2 % of mass fractions of boron due to a wider
solidification range.7 In later studies, it was found that a
boron content of 0.5 % of mass fractions or more, i.e.,
above its solid-solubility limit causes a reduction in the
coefficient of thermal expansion and constricts the
solidification range, which, in turn, causes the
crack-healing phenomenon and decreases the cracking
susceptibility.8,9 The lower solidification-cracking
susceptibility of high-boron steels is mainly due to the
healing of cracks by the abundant amounts of the
low-melting eutectic liquid of (Cr,Fe)2B.10
ASTM A887-89 covers eight different types of
borated stainless steels, and each type has two grades,
i.e., A and B with different boron contents.11 Specifi-
cation designates Grade A for the materials produced
with powder metallurgy containing finer and more
uniformly distributed borides compared to Grade B made
with ingot metallurgy. The borated SSs fabricated with
the hot-rolling technique, employing the powder-sinter-
ing method have sufficient ductility and other
properties.12
The GTAW is the welding process normally used in
the fabrication of borated stainless steels. However, its
disadvantage is a relatively shallow penetration during a
single-pass welding operation on thicker plates. The
thickness of the austenitic stainless steels that can be
welded in a single pass with argon as the shielding gas is
normally restricted to 3 mm. In this connection, a novel
variant of the GTA welding process called the acti-
vated-flux GTA welding process was initially developed
for the welding of titanium at the Paton Institute of
Electric Welding.13,14 Several investigations were carried
out by various researchers to develop suitable flux com-
ponents to extend this technique to other alloys. In the
activated-flux GTA welding process, a thin coating of the
activated flux is applied onto the surface of a joint and
the electrons outside the arc are captured by the flux
when it is vaporized during welding. Thus, the arc gets
constricted which, in turn, leads to an increase in the
penetration. Sun and Pan reported that the penetration
capability can be increased by as much as 300 % due to
the activated-flux GTA welding when compared with the
conventional GTA welding.15,16
The main reasons for the significant improvement in
the penetration in the steels are reported to be the arc
constriction and the reversed Marangoni convection.17
However, several researchers stated that the increase in
the penetration is due to the positive surface-tension
gradient in a molten-weld pool, which causes the fluid to
flow towards the bottom of the weld.18,19 Thus, based on
the productivity, the activated GTA welding process is
more viable than the conventional GTA welding process.
The activated GTA welding of austenitic stainless steels
was investigated by various authors to understand the
effect of the flux on the depth of penetration, the aspect
ratio, the microstructure and the mechanical beha-
vior.20–26
Furthermore, the effect of the activated flux on the
GTA welding of austenitic stainless steel 316L was
studied by Kuang Hung et al.27 It was reported that the
activated TIG welding increases the joint penetration and
the aspect ratio, which, in turn, reduce the angular dis-
tortion of the weldments. A great deal of research was
carried out to develop various flux combinations and
understand their beneficial influence on the GTA weld-
ing of austenitic stainless steels. Several investigations
showed that an addition of nitrogen to the shielding gas
influences the weld-bead shape, the depth of penetration
and the weld metal microstructure.28,29
However, not much information pertaining to the
effect of the activated flux and nitrogen addition on the
GTA welding of borated SS has been reported so far.
Therefore, this work presents the microstructure and
mechanical behavior of GTAW, activated-flux GTAW,
nitrogen-added GTAW and electron-beam welding
(EBW) of 10-mm-thick borated stainless steel.
2 EXPERIMENTAL WORK
The base material used in this study was borated
stainless steel (BSS) 304B4 and its chemical compo-
sition is given in Table 1. As-received rolled plates of 40
mm were cut, using an electrical discharge machine to
prepare plates of 300 mm × 110 mm × 10 mm. The
plates were cleaned using acetone to remove surface con-
tamination before the welding.
Table 1: Chemical composition of the base material
Tabela 1: Kemijska sestava osnovnega materiala
Material Cr Ni Mn B Si P C
304B4 19.3 13.4 2.0 1.05 0.74 0.045 0.08
Bead-on-plate welds were produced using the auto-
matic GTAW process. A tungsten electrode (AWS classi-
fication EWTh-2) with a diameter of 3.2 mm and the
shielding gas of pure argon were used in the conven-
tional GTAW process with direct-current electrode
negative (DCEN) polarity. In addition to argon-shielded
bead-on-plate welds, welds were produced using a
shielding-gas mixture containing 2 % (volume) nitrogen
along with argon. Also, autogenous full-penetration
electron-beam welds were also made.
For the activated GTA welding, the activated flux was
prepared, with its constituents in the required propor-
tions, and then mixed with ethanol and the resulting
mixture was stirred until it turned into paste. The flux
used in this process is commercially available and its
composition was reported in the patent by M.Vasu-
devan.30 A thin layer of this paste was applied manually
on the surface to be welded and dry powder remained
there after the evaporation of ethanol. Using this acti-
vated flux, several bead-on-plate experiments were
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
358 Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364
carried out in order to achieve a single-pass full-pene-
tration weld in the activated GTAW process.
The parameters used in various welding processes are
presented in Table 2. Using standard metallographic pro-
cedures, specimens were prepared for a microstructural
investigation. After polishing, the specimens were etched
with Kalling’s 1 solution containing 5 g of cupric chlo-
ride, 100 mL of hydrochloric acid and 100 mL of
ethanol. The welds were initially examined with a stereo
microscope and then bead-geometry measurements were
taken using image-analysis software. The microstruc-
tures of different zones of interest such as FZ, PMZ and
the base metal (BM) were observed with a light micro-
scope. Micro-hardness tests were carried out using a
Vickers digital micro-hardness tester along the weld
joint. A load of 500 g was applied for a duration of 10 s.
In accordance with ASTM E8, tensile specimens were
machined with an EDM-wire-cutting machine so that the
weld metal was located in the center of gauge length.
The tensile properties of weldments, i.e., the ultimate
tensile strength, the proof strength and the percent
elongation were evaluated on the basis of the results of
the tensile tests conducted at room temperature. The
tensile-test results are listed in Table 3 and all the values
presented are average values of at least three specimens.
Charpy-impact test specimens were prepared in accor-
dance with ASTM standard E23. Identical V-notches
were machined on three specimens, at the weld center,
using a broaching machine. Impact testing was con-
ducted at room temperature using a pendulum-type
machine with the maximum capacity of 300 J.
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364 359
Table 2: Common used welding parameters
Tabela 2: Uporabljeni splo{ni parametri varjenja
Cross-sections of the
welds
Depth of
penetration (mm)
Aspect ratio
(d/w)
Weld area
(mm2) Welding parameters
Heat input
(KJ/mm)
GTAW
3.8 0.27 51.45
Weld current: 235 A
Arc voltage: 17 V
Travel speed:
120 mm/min
Gas flow rate:
25,3 Pa m3 s–1
2
Nitrogen-added GTAW
5.77 0.4 57.54
Weld current: 235 A
Arc voltage: 17 V
Travel speed:
120 mm/min
Gas flow rate:
25,3 Pa m3 s–1
2
Activated GTAW
10 0.8 92.01
Weld current: 235 A
Arc voltage: 17 V
Travel speed:
120 mm/min
Gas flow rate:
25,3 Pa m3 s–1
2
EBW
10 2.82 17.02
Acceleration voltage:
60KV
Beam current: 70 mA
Travel speed:
700 mm/min
0.36
3 RESULTS
Cross-sections of the welds made using different
welding processes are presented in Table 2, along with
the bead-profile parameters and corresponding welding
parameters. The microstructure of the base metal is pre-
sented in Figure 1. The microstructures of FZ and PMZ
for the welds made using various welding processes are
shown in Figures 2a to 2d and Figures 3a to 3d, res-
pectively. The microhardness surveys carried out along
the various weld zones such as FZ, PMZ and BM of
different welds are shown in Figures 4a to 4d. The ten-
sile properties for the different welds considered in the
present study are presented in Table 3. The results of the
Charpy-impact test conducted at room temperature for
the welds produced using various welding processes are
presented in Table 3.
4 DISCUSSION
4.1 Weld bead geometry
From Table 2, it can be seen that the three variants of
the GTA process use the same welding parameters and
heat input, i.e., 2 kJ/mm. The use of Ar+2%N shielding
gas in place of Ar significantly increases the depth of
penetration, from 3.8 mm to 5.7 mm; and the aspect ratio
of the weld bead also increases. This was observed by
earlier researchers for other materials.28,29 For the same
heat input, the effect of the activated flux on the bead
geometry is much more significant and a full penetration
of 10 mm is observed. It can also be seen from the
macrograph presented in Table 2 that the activated flux
causes the aspect ratio of the weld bead to reach 0.8,
which is a significant change. Electron-beam welds
exhibit the keyhole type of the bead geometry and a full
penetration can be achieved using a much smaller heat
input of 0.36 kJ/mm, showing the process to be greatly
enhanced.
It was observed in this study that the bead shape
becomes relatively wider and shallower with an increase
in the welding current for both GTA and nitrogen-added
GTA welding. Generally, the surface tension () on the
pool surface, formed by the cohesive forces of liquid me-
tal, decreases with an increase in the temperature. Thus,
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
360 Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364
Figure 2: Microstructures of the FZs after various welding processes: a) GTAW, b) activated GTAW, c) nitrogen-added GTAW, d) EBW
Slika 2: Mikrostrukture FZ pri razli~nih postopkih varjenja: a) GTAW, b) aktiviran GTAV, c) du{ik dodan GTAW, d) EBW
Figure 1: Microstructure of the base metal showing a boride network
in an austenitic matrix
Slika 1: Mikrostruktura osnovnega materiala, ki ka`e mre`o boridov v
avstenitni osnovi
the temperature gradient becomes negative, i.e., d/dT < 0
which, in turn, generates centrifugal Marangoni convec-
tion in the molten pool. This constitutes one of the main
reasons for the shallow penetration and lower aspect
ratio in the conventional GTA welding. However, in the
case of activated GTA welds, the presence of the acti-
vated flux changes the temperature gradient to a positive
value, i.e., d/dT > 0. This positive temperature gradient
causes centripetal Marangoni convection, which, in turn,
directs the flow toward the pool center resulting in a
deeper and narrower weld pool. By and large, it was
noticed that the activated flux beneficially influenced the
Marangoni convection mode by changing the tempera-
ture gradient. Furthermore, the aspect ratio was in-
creased due to the variation in the temperature gra-
dient.21,22
4.2 Microstructure
As it can be noticed, irregular boride particles
(Fe,Cr)2B, seen as a dark phase, are dispersed in the
austenitic matrix. Boron is insoluble in austenite vir-
tually at all temperatures. The insolubility is more
significant in the case of steels with high boron levels,
which, in turn, results in a continuous network of boride
eutectics such as Fe2B and Cr2B in an austenitic matrix.
The boride eutectics strengthen the austenitic matrix but
adversely affect the toughness and ductility of these
steels. Furthermore, the shape of these eutectic phases is
also one of the factors affecting the mechanical behavior
of a weld.6
In order to have full-penetration welds for micro-
structural studies and an evaluation of mechanical
properties, the welds were made using two passes in the
case of the GTA and nitrogen-added GTA processes and
a single pass in the case of the activated-flux and EB
processes. The fusion zone of the activated-flux-GTA
weld was characterized by a columnar, austenite den-
dritic structure with eutectic borides solidified in the
interdendritic regions (Figure 2b). However, the GTA
and nitrogen-added-GTA welds exhibited both equiaxed
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364 361
Figure 3: Microstructures of the weldments after various welding processes: a) GTAW, b) activated GTAW, c) nitrogen-added GTAW, d) EBW
Slika 3: Mikrostrukture zvarov pri razli~nih procesih varjenja: a) GTAW, b) aktiviran GTAW, c) GTAW z dodanim du{ikom, d) EBW
Table 3: Mechanical properties of 304B stainless-steel welds made with various GTA welding processes
Tabela 3: Mehanske lastnosti zvarov nerjavnega jekla 304B, izdelanih po razli~nih GTA varilnih postopkih
Process Proof strength Ultimate tensilestrength % Elongation
Joint efficiency
in terms of
tensile strength
Failure location Impacttoughness
Base metal 384 576 12 – – 7
GTAW 379 545 10 94.62 PMZ 7
Nitrogen added GTAW 390 550 11 96.32 PMZ 8
Activated GTAW 400 569 14 98.78 BM 7
EBW 415 570 12 98.95 BM 10
and columnar dendritic structures, as can be seen in Fig-
ures 2a and 2c. A short interaction span with an acute
energy density causes a low heat input in EBW. As a
result, the fusion zone in EBW cools faster resulting in a
finer equiaxed dendritic structure.
Typical appearances of the PMZ for various weld
types are presented in Figure 3. As it can be noticed, the
representative PMZ consists of localized regions of
austenite that remain solid during the welding,
surrounded by irregular boride eutectics. However, the
width of PMZ was found to be larger in the case of the
GTA and nitrogen-added-GTA welds than that of the
activated-flux-GTA welds. This is attributed to slow
cooling rates associated with high heat inputs prevailing
in the GTA and nitrogen-added-GTA welds. The low
heat input associated with high cooling rates almost
eliminated the PMZ in electron-beam welding and no
localized region of austenite was noticed in Figure 3d.
4.3 Mechanical properties
4.3.1 Microhardness
Hardness profiles revealed that FZ exhibited a higher
hardness for all the welds. The increase in the hardness
of FZ is attributed to the presence of a dendritic micro-
structure with boride eutectics in the interdendritic
regions (Figure 2). It was observed that the nitrogen
addition in the GTA welds significantly enhances the FZ
hardness. It was also noticed that there is a sudden fall in
the hardness of PMZ in the case of both the GTA and
nitrogen-added GTA welds, while there is no such trend
in the activated GTA and EB welding. The significant
reduction in the PMZ hardness is attributed to the
difference in the cooling rate, which results in a variation
in the size and shape of the eutectic borides formed in
PMZ as can be seen from Figures 3a to 3d. It can be
seen from Figure 5 that the fracture during the tensile
test in the case of the GTA and nitrogen-added GTA
welds, occurs in PMZ, whereas in the case of the other
two welds the fracture occurs in the base materials far
away from the weld metal.
4.3.2 Tensile properties
In order to obtain joint properties, automatic GTAW
was carried out in two passes, one from each side.
Though there is no significant variation in the joint
efficiencies of the welds, a marginal improvement in the
yield strength for the electron-beam and activated-GTA
welds was noticed. The GTA and nitrogen-added-GTA
welds were found to fracture at PMZ, as shown in Fig-
ure 5, the region where a loss in the hardness can be
clearly noticed due to an irregular distribution of boride
eutectics. It was also observed that the activated-GTA
and EB welds failed in BM. Low-heat-input welding
processes exhibited a significant improvement in the
tensile strength compared to high-heat-input welding
process.
4.3.3 Impact toughness
It was observed that the welds exhibit the same
toughness as that of BM, irrespective of the welding
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
362 Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364
Figure 4: Microhardness profiles for various welding processes
Slika 4: Profil mikrotrdote pri razli~nih procesih varjenja
process employed. The lower toughness of BM and FZ is
mainly due to brittle (Cr,Fe)2B borides.
5 CONCLUSIONS
1. Defect-free welds of 304B4 borated stainless steels
can be easily made using GTA, activated-flux GTA,
nitrogen-added GTA and EB welding processes.
2. The activated-flux GTA and EB welding processes
substantially enhance both the depth of penetration
and the aspect ratio.
3. The fusion-zone microstructure of the activated-flux
GTA welds reveals a columnar, austenite dendritic
structure with eutectic borides solidified in the
interdendritic regions. However, the EB welds exhibit
a fine equiaxed dendritic structure.
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364 363
Figure 5: Fracture locations of the welds made with different welding processes: a) tensile-test samples, b) cross-sections
Slika 5: Lokacija preloma zvarov, izdelanih z razli~nimi postopki varjenja: a) vzorci za natezne preizkuse, b) preseki
4. Low-heat-input EBW inhibits the formation of PMZ.
A significant loss in the hardness occurs in the PMZs
of the GTA and nitrogen-added GTA welds, which is
mainly due to the slow cooling rates associated with
high heat inputs.
5. A tensile failure occurs in the PMZ area of the GTA
and nitrogen-added GTA welds, while the
activated-flux GTA and EB welds failed in the base
material.
Acknowledgments
The authors would like to acknowledge the technical
support of IIT Madras, Defence Research & Develop-
ment Laboratory (DRDL), Hyderabad and Ador Welding
Pvt. Ltd., Chennai.
6 REFERENCES
1 C. V. Robino, M. J. Cieslak, Fusion welding of a modern borated
stainless steel, Welding Journal-Including Welding Research Supple-
ment, 76 (1997) 1, 12–23, doi:10.2172/10183280
2 C. V. Robino, M. J. Cieslak, High-temperature metallurgy of ad-
vanced borated stainless steels, Metallurgical and Materials
Transactions A, 26 (1995) 7, 1673–1685, doi:10.1007/BF02670754
3 D. A. Moreno, B. Molina, C. Ranninger, F. Montero, J. Izquierdo,
Microstructural characterization and pitting corrosion behavior of
UNS S30466 borated stainless steel, Corrosion, 60 (2004) 6,
573–583, doi:10.5006/1.3287761
4 H. J. Goldschmidt, Effect of boron additions to austenitic stainless
steels – Part I – effect of boron additions to 20%Cr, 25%Ni austenitic
steel with and with6ut Nb, Mn and Si, Journal of the Iron and Steel
Institute, 209 (1971) 11, 900–909
5 B. Shanmugarajan, J. N. Chary, G. Padmanabham, B. Arivazhagan,
S. K. Albert, A. K. Bhaduri, Studies on autogenous laser welding of
type 304B4 borated stainless steel, Optics and Lasers in Engineering,
51 (2013) 11, 1272–1277, doi:10.1016/j.optlaseng.2013.05.004
6 T. D. Park, K. K. Baek, D. S. Kim, PWHT effect on the mechanical
properties of borated stainless steel GTA weldments for nuclear
shield, Metals and Materials, 3 (1997) 1, 46–50, doi:10.1007/
BF03026105
7 T. Shinoda, H. Miyake, T. Matsuzaka, T. Matsumoto, H. Kanai, Hot
cracking susceptibility of boron modified AISI 304 austenitic
stainless steel welds, Materials Science and Technology, 8 (1992) 10,
913–921, doi:10.1179/026708392790169777
8 W. R. Apblett, W. S. Pellini, Factors which influence weld hot
cracking, Welding Journal, 33 (1954) 2, 83–90
9 J. C. Borland, R. N. Younger, Some aspects of cracking in welded
Cr-Ni austenitic steels, British Welding Journal, 7 (1960), 22–59
10 T. Matsumoto, T. Shinoda, H. Miyake, T. Matsuzaka, H. Kanai,
Effect of low-melting point eutectic on solidification cracking sus-
ceptibility of boron-added AISI 304 stainless steel welds, Welding
Journal-Including Welding Research Supplement, 74 (1995),
397–405
11 ASTM A887-89 (2009) ASM International, West Conshohocken,
PA, USA 2009
12 S. Yamamoto, M. Honda, Y. Kobayashi, T. Sakiyama, Y. Ehira, N.
Takashi, Boronated stainless steels for thermal neutron shielding, Jr.
Thermal & Nuclear Power Generation, 41 (1990) 9, 1149–1157
13 W. Lucas, D. Howse, Activating flux – increasing the performance
and productivity of the TIG and plasma processes, Welding and
Metal fabrication, 64 (1996), 11
14 P. C. J. Anderson, R. Wiktorowicz, Improving productivity with
A-TIG welding, Welding and Metal Fabrication, 64 (1996) 3,
108–109
15 M. Vasudevan, A. K. Bhaduri, Baldev Raj, Development and eva-
luation of activated flux for TIG welding of type 304 LN and type
316 LN stainless steels, Proc. of the Inter. Institute of Welding,
International Congress, Chennai 2008
16 Z. Sun, D. Pan, Welding of titanium alloys with activating flux,
Science and Technology of Welding and Joining, 9 (2004) 4,
337–344, doi:10.1179/136217104225021571
17 P. Vasantharaja, M. Vasudevan, Studies on A-TIG welding of low
activation ferritic/martensitic (LAFM) steel, Journal of Nuclear
Materials, 421 (2012), 117–123, doi:10.1016/j.jnucmat.2011.11.062
18 C. R. Heiple, J. R. Roper, Effect of selenium on GTAW fusion zone
geometry, Welding Journal, 60 (1981) 8, 143–145
19 C. R. Heiple, R. Roper, Mechanism for minor element effect on GTA
fusion zone geometry, Welding Journal, 61 (1982) 4, 97–102
20 P. Modenesi, E. R. Apolinario, TIG welding with single-component
fluxes, Journal of Materials Processing Technology, 99 (2000) 1–3,
260–265, doi:10.1016/S0924-0136(99)00435
21 S. Lu, H. Fujii, H. Sugiyama, M. Tanaka, K. Nogi, Marangoni
convection and welding penetration in A-TIG welding, Transactions
of the Japan Welding Research Institute, 32 (2003) 79–82,
doi:10.1016/j.tafmec.2007.05.004
22 H. Fujii, T. Sato, S. Lua, K. Nogi, Development of an advanced
A-TIG (AA-TIG) welding method by control of Marangoni con-
vection, Materials Science and Engineering A, 495 (2008), 296–303,
doi:10.1016/j.msea.2007.10.116
23 H. Y. Huang, S. W. Shyu, K. H. Tseng, C. P. Chou, Evaluation of
TIG flux welding on the characteristics of stainless steel, Science and
Technology of Welding and Joining, 10 (2005) 5, 566–573,
doi:10.1179/174329305X48329
24 A. Rodrigues, A. Loureiro, Effect of shielding gas and activating flux
on weld bead geometry in tungsten inert gas welding of austenitic
stainless steels, Science and Technology of Welding and Joining, 10
(2005) 6, 760–765, doi:10.1179/174329305X68769
25 S. Leconte, P. Paillard, P. Chapelle, G. Henrion, J. Saindrenan, Effect
of oxides fluxes on activation mechanisms of tungsten inert gas
process, Science and Technology of Welding and Joining, 11 (2006)
4, 389–397, doi:10.1179/174329306X129544
26 L. Sanbao, Y. Chunli, L. Fengyao, W. Lin, S. Sheng, Effect of
activating fluxes on weld mechanical properties in TIG welding,
China Welding, 10 (2001) 2, 170–174
27 K. H. Tseng, C. Y. Hsu, Performance of activated TIG process in
austenitic stainless steel welds, Journal of Materials Processing and
Technology, 211 (2011), 503–512, doi:10.1016/j.jmatprotec.2010.
11.003
28 D. D. Nage, V. S. Raja, R. Raman, Effect of nitrogen addition on the
microstructure and mechanical behaviour of 317L and 904L
austenitic stainless steel welds, Journal of Materials Science, 41
(2006) 7, 2097–2112, doi:10.1007/s10853-006-3150-5
29 K. H. Tseng, C. P. Chou, The study of nitrogen in argon gas on the
angular distortion of austenitic stainless steel weldments, Journal of
Materials Processing and Technology, 142 (2003), 139–44,
doi:10.1016/S0924-0136(03)00593-4
30 An international patent Tungsten inert gas welding flux for en-
hancing penetration in single pass for austenitic stainless steels,
patent no. PCT/IN06/00442/dt. 8/11/06
G. R. KUMAR et al.: EFFECT OF ACTIVATED FLUX AND NITROGEN ADDITION ON THE BEAD GEOMETRY ...
364 Materiali in tehnologije / Materials and technology 50 (2016) 3, 357–364
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
365–371
MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT
LIQUID-PHASE BONDING OF DISSIMILAR NICKEL-BASED
SUPERALLOYS OF IN738LC AND NIMONIC 75
RAZVOJ MIKROSTRUKTURE MED SPAJANJEM S PREHODNO
TEKO^O FAZO NEENAKIH SUPERZLITIN NA OSNOVI NIKLJA
IN738LC IN NIMONIC 75
Meysam Khakian Ghomi, Saeid Nategh, Shamsoddin Mirdamadi
Islamic Azad University, Science and Research Branch, Department of Materials Engineering, Tehran, Iran
khakian-meysam@yahoo.com, nategh@sharif.ir
Prejem rokopisa – received: 2015-04-01; sprejem za objavo – accepted for publication: 2015-05-08
doi:10.17222/mit.2015.072
The joining of dissimilar nickel-based superalloys, i.e., IN738LC to Nimonic 75, using transient liquid-phase bonding with a
Ni-15Cr-3.5B interlayer (MBF-80) was carried out at temperatures of 1080 °C to 1180 °C for different bonding times of 30–150
min. A joint cross-section was surveyed using optical and scanning electron microscopy. Microstructural examinations showed
that after short bonding times, the joint microstructure consists of continuous eutectic intermetallic phases and longer bonding
times lead to a eutectic-free microstructure. It was clear that for all the bonding times and temperatures, boride phases were
precipitated at the interface of the base metal and the interlayer due to boron diffusion in to the base metals. The results also
showed that the morphology of the precipitates in the diffusion-affected zone (DAZ) varies from globular to acicular by
increasing the bonding time. Completion of the isothermal solidification, which prevents the formation of continuous
intermetallic phases at the joint centerline, was studied at different temperatures.
Keywords: IN738LC, Nimonic 75 superalloy, TLP bonding, isothermal solidification, microstructure
Spajanje razli~nih superzlitin na osnovi niklja IN738LC na Nimonic 75, z uporabo prehodno teko~e faze z vmesnim slojem
Ni-15Cr-3,5B (MBF- 80), je bilo izvedeno pri temperaturah od 1080 °C do 1180 °C in pri razli~nih ~asih spajanja od 30 min do
150 min. Pre~ni prerez spoja je bil pregledan s svetlobno in z vrsti~no elektronsko mikroskopijo. Mikrostrukturne preiskave so
pokazale, da pri kratkih ~asih spajanja mikrostruktura sestoji iz zvezne evtekti~ne intermetalne faze, dolgi ~asi spajanja pa
povzro~ijo mikrostrukturo brez evtektika. Izkazalo se je, da se pri vseh ~asih in temperaturah spajanja, boridna faza zaradi
difuzije bora izlo~a na meji z osnovnim materialom v osnovni material. Rezultati so tudi pokazali, da morfologija izlo~kov v
difuzijsko vplivani coni (DAZ) z nara{~anjem ~asa spajanja varira, od globularne do acikularne. Dokon~anje izotermnega
strjevanja, ki prepre~i nastanek zvezne intermetalne faze na liniji spajanja, je bilo prou~evano pri razli~nih temperaturah.
Klju~ne besede: IN738LC, superzlitina Nimonic 75, TLP spajanje, izotermno strjevanje, mikrostruktura
1 INTRODUCTION
IN738LC is one of the most practical casting
nickel-based superalloys. IN738LC is strengthened by
both solid-solution and precipitation-hardening mecha-
nisms.1 Due to the presence of coherent 
’ precipitates in
the alloy with a complex chemical composition and a
high stability at high temperatures, the alloy can keep its
excellent strength in high-temperature operations and
severe conditions, such as gas turbine components.
Because of the mentioned properties, the alloy is a good
choice for utilization in the first rows of gas-turbine hot
sections. Also, this alloy has outstanding strength
accompanied by excellent creep, fatigue, oxidation and
corrosion resistance at high temperatures.2–4
Nimonic 75 is an 80/20 nickel-chromium alloy with
controlled additions of titanium and carbon. Nimonic 75
first introduced in the 1940s for turbine blades as sealing
components and is now mostly used for sheet applica-
tions calling for oxidation and scaling resistance coupled
with medium strength at high operating temperatures. It
is still used in gas-turbine engineering and also for
industrial thermal processing, furnace components and
heat-treatment equipment.5
The weldability of superalloys is very much depend-
ent on the amount of Al and Ti in their chemical compo-
sition. The high precipitation rate of 
’ in alloys with a
large amount of Al and Ti make them susceptible to
crack propagation during welding.6–8 Fusion welding,
diffusion bonding and brazing are the three main me-
thods of joining and repairing superalloys that are widely
used in power plants and aerospace industries. Solid-
state diffusion bonding needs high pressure, time and
temperature. Therefore, the process encounters a lot of
practical and economic limitations, like growth of grains
and precipitates during bonding and lengthy processing
causes it to be uneconomic.9 Brazing is the other method
of joining that uses an interlayer, containing a melting-
point depressant like B, Si and P, as the joining media.
The joining temperature must be precisely arranged to
melt only the interlayer. Brittle boride, silicide and phos-
phate continuous phases may be developed in the joint
Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371 365
UDK 67.017:621.791/.792:622.785 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)365(2016)
area during the brazing process and can degrade the
joint’s properties.10,11
Transient liquid-phase (TLP) bonding is a hybrid pro-
cess of brazing and diffusion bonding.1 This method
possesses the advantages of brazing and diffusion
bonding and makes the joint structure very similar to the
base material. In contrast to diffusion bonding, the
process does not need any pressure due to the use of a
liquid interlayer.12,13 The superiority of TLP in compa-
rison with brazing is the completion of isothermal
solidification in the bonding temperature.10 Also, TLP is
widely used in the joining of metals and ceramics.14
In the standard models of TLP, there are three indi-
vidual stages: dissolution, isothermal solidification and
solid-state homogenization. Because of the short-range
diffusion of melting-point depressant elements in the
base metal, the dissolution of the base metal usually
needs several minutes. In the isothermal solidification
stage, long-term diffusion occurs and therefore the pro-
cess needs much more time, depending on the bonding
temperature and the interlayer type and the time varies
from minutes to hours.15 By optimizing the time and
temperature of the bonding, a joint microstructure that is
very similar to the base metal with no discontinuity in
the microstructure can be achieved.16
In this paper, the effect of bonding time and tem-
perature on the microstructure evolution of the transient
liquid-phase bonding of IN 738LC and Nimonic 75
using MBF-80 filler metal was investigated.
2 MATERIALS AND EXPERIMENTAL
PROCEDURES
IN738LC and Nimonic 75 nickel-based superalloys
were used as the base metals in this study. Both super-
alloys were used in the standard solution heat-treated
condition. Commercial Ni-Cr-B in the form of amor-
phous foil (MBF-80) with a 75-μm thickness was used as
the interlayer. The chemical compositions of the base
materials and the interlayer are given in Table 1.
TLP bonding test coupons were cut using EMD
wirecut to a size of 10 mm × 10 mm × 5 mm, then they
were grounded with a pendulum grinding machine, in
order to remove the recast oxide layer from the faying
surface of the coupons, and finally cleaned ultrasonically
in a acetone bath. The interlayer was cut with a size of
11 mm × 6 mm and placed between two faying surfaces
of the base alloys. In order to prevent the movement of
the samples, the assembly was placed in a fixture with-
out any pressure on it. Green stop-off type 1 was used to
prevent the flowing out of the molten interlayer. The
solidus and liquidus temperatures of the interlayer are
1050 °C and 1090 °C, respectively.17 Therefore, the
bonding temperatures were selected above 1050 °C. TLP
bonding was performed in a furnace under a vacuum of
10–5 mbar pressure at different temperatures of (1080,
1120, 1150 and 1180) °C and various holding times of
(30, 60, 120 and 150) min. In order to investigate the
phases formed during the athermal solidification, one
test coupon was bonded at 1120 °C for 5 min. The heat-
treatment cycle of the bonding is illustrated in Figure 1.
The heating rate from 950 °C to the bonding temperature
was set to 20 °C/min. In order to show the effect of time
on the bond strength, a shear test was carried out on all
the samples TLP bonded at 1080 °C.
The TLP-bonded test coupons were cut perpendicular
to the bonding surface, due to brittleness of the inter-
metallic phases formed in the bonding centerline, by
EDM wirecut and then prepared for metallographic
examinations. In order to reveal the microstructure
constituent, the samples were etched with Kalling’s solu-
tion (5 gr CuCl2 + 100 cc HCl +100 cc Ethyl alcohol).18
Metallographic examinations were carried out using
optical and scanning electron microscopy.
3 RESULTS AND DISCUSSION
3.1 Microstructure of the joint
The microstructure of the TLP-bonded specimen at
1120 °C and 5 minutes is shown in Figure 2. As can be
seen, the bonding area can be divided into six distinct
zones:
• Base metal - Nimonic side
• Isothermally Solidified Zone – Nimonic 75 Side (ISZ
– Nimonic Side)
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
366 Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371
Figure 1: Heat-treatment cycle of the TLP bonding test coupons
Slika 1: Potek cikla toplotne obdelave pri TLP spajanju preizku{ancev
Table 1: Chemical composition of base materials and interlayer in mass fractions, (w/%)
Tabela 1: Kemijska sestava osnovnih materialov in vmesne plasti, v masnih odstotkih (w/%)
Alloys Cr Co Al Ti W Mo Ta Nb Fe C B Ni
IN738LC 16.23 8.56 3.41 3.45 3.05 1.73 1.57 0.67 0.08 0.09 0.0063 Bal.
Nimonic 75 20.50 – 0.29 0.55 – – – – 4.78 0.10 – Bal.
MBF–80 14.89 – – – – – – – – – 3.72 Bal.
• Athermally Solidified Zone (ASZ)
• Isothermally Solidified Zone – Inconel 738 Side (ISZ
– Inconel Side)
• Diffusion Affected Zone – Inconel 738 Side (DAZ –
Inconel Side)
• Base metal – Inconel side
As can be seen in Figure 2a, the ISZ zone on both
sides of the ASZ contains a thin layer of 
 solid solution
that formed at the solid/liquid interface during isother-
mal solidification from the base metal to the joint center-
line. The chemical composition of the solid solutions
formed on the ISZ – Inconel Side and ISZ - Nimonic
Side are shown in Table 2. A significant difference bet-
ween these two compositions is related to the chemical
compositions of the Nimonic 75 and the IN738LC. Also,
in Figure 2b, the different phases that are formed during
bonding are specified with A to F.
As the interlayer melts, because the liquid and solid
phases are not in the equilibrium condition, base-metal
dissolution commences. The diffusion of boron into the
base metal and the diffusion of base-metal elements like
Cr, Fe, Co, Al and Ti into the molten interlayer leads to
the equilibrium condition between the liquid and the
solid. A chemical analysis of ISZ shows the diffusion of
the base-alloy elements that were not present in the
primary interlayer. The alloying elements’ diffusion from
the base metal to the interlayer and also boron diffusion
from the interlayer to the base metal increases the
liquidus temperature of the interlayer. The dissolution
stage continues until the liquidus temperature of the
molten interlayer reaches the bonding temperature, and
then the isothermal solidification starts from the
solid/liquid interface to the joint centerline.2, 19
As seen in Table 2, the chemical composition of the
ISZ regions adjacent to the base alloys (Inconel Side and
Nimonic Side) are not similar; therefore, we can
conclude that the time required to reach equilibrium on
each side are not equal. In contrast with the TLP bonding
of two similar alloys, in which the isothermal solidifi-
cation starts simultaneously and the eutectic centerline
forms symmetrically, in dissimilar alloys the TLP
bonding eutectic centerline lies closer to the alloy so that
its equilibrium starts earlier.
A variation of the chemical composition that is due to
the interdiffusion of elements of the base alloy and the
interlayer is the main driving force of solidification for
the ISZ.19 Since the solid and liquid are in an equilibrium
condition during isothermal solidification, the secondary
phases cannot form at this stage.3,20 The solidification
behavior of the remaining liquid during the TLP bonding
of pure nickel using the Ni-Cr-B interlayer, before
completion of the isothermal solidification, was modeled
by K. Ohsasa et al.21 The ternary phase diagram of
Ni-Cr-B, in which the chemical composition of the
interlayer is specified, is shown in Figure 3.22 During the
cooling of the remaining liquid, which is the main
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371 367
Table 2: EDS analysis (x/%) of phase formed in bond line shown in Figure 2
Tabela 2: EDS analiza (x/%) faze, ki je nastala na liniji stika, prikazani na Sliki 2
Al Ti Cr Fe Co Ni Nb Mo Ta W
ISZ - Inconel Side 3.12 0.94 19.41 1.23 3.01 70.43 0.64 0.52 0.28 0.41
ISZ - Nimonic Side 0.57 0.46 22.12 4.26 0.35 71.21 0.34 0.40 0.0 0.28
A (Ni-rich Boride) 1.08 3.83 9.18 0.71 2.28 80.70 0.85 0.56 0.80 0.0
B (
 Eutectic) 0.97 0.53 16.27 1.64 2.47 75.84 0.60 0.66 0.30 0.71
C (Cr-rich Boride) 0.41 0.08 89.72 0.15 0.47 5.37 0.46 2.02 0.0 1.31
D (Cr-rich Boride) 0.04 0.05 87.50 0.40 0.69 5.36 1.37 1.97 1.72 0.89
E 0.75 0.36 73.05 2.34 0.0 22.04 0.63 0.44 0.13 0.25
F 0.56 0.84 40.06 4.11 0.03 52.66 0.66 0.52 0.25 0.30
Figure 3: Liquidus projection of ternary phase diagram Ni-Cr-B
Slika 3: Projekcija likvidusa v ternarnem faznem diagramu Ni-Cr-B
Figure 2: a), b) SEM micrograph of sample bonded at 1120 °C for
5 min
Slika 2: a), b) SEM-posnetek vzorca, spajanega pri temperaturi 1120 °C,
5 min
driving force for ASZ solidification, 
 dendrite is the
first phase that forms at 1100 °C 21 and grows from the
solid/liquid interface towards the joint centerline. Solute
elements with a partition coefficient less than unity (k <
1) are rejected into the liquid. Continuous enrichment of
the liquid from the solute elements leads to the solute
concentration becoming more than the solubility limit of
the solute in the 
 and then the secondary solidification
constituents are formed between the dendrites (Phases E
and F).20 O. Ojo et al.23 and A. Egbewande et al.10 ob-
served the formation of these phases in the TLP bonding
of IN738 and IN600, by use of a Ni-Cr-B interlayer,
respectively.
During the cooling of the liquid and solid-solution
growth additional boron is rejected to the liquid, as a
result of poor solubility and the low partition coefficient
of boron in Ni (0.3 % of amount fractions and ~0.008),24
and then at 1042 °C the solidification path coincides
with the eutectic line that separates the 
 phase stability
region from Ni3B (e8 line in Figure 3) and the 
 solid
solution and nickel boride eutectic phases are formed
simultaneously through a binary eutectic transformation
(L
+Ni3B).
More cooling leads to the growth of the 
 solid
solution and the rejection of boron and the formation of
nickel borides lead to a reduction of the Ni concentration
and an increment in the Cr in the remaining liquid (Cr
solubility in Ni boride is relatively low (10.11 % of
amount fractions).19 Also, the Cr content of the liquid
increases by temperature decrement until at 997 °C the
liquid transforms into three phases of 
 solid solution,
nickel boride and chromium boride through a ternary
eutectic transformation (L
+Ni3B+CrB).21 These
phases are shown in Figure 2b, and its chemical com-
positions are shown in Table 2.
3.2 Effect of time and temperature on the microstruc-
ture of the TLP bonds
In order to investigate the effect of time on the micro-
structure of the TLP bonds, bonding was performed at
1080 °C for 30, 60, 120 and 150 min. It was clear that
the thickness of the intermetallic phases decreases with
the increasing time (Figure 4). The extent of isothermal
solidification during TLP bonding depends on the
amount and rate of boron diffusion into the base alloys.
The thickness of remaining interlayer, which can trans-
form into a continuous eutectic centerline during cool-
ing, decreases for longer bonding times. It was clear that
with an increasing time from 30 to 60 min, the thickness
of the centerline eutectic decreases. After 120 min there
is no continuous eutectic and after a longer time (150
min) there is a negligible amount of intermetallic phases.
The results showed that 150 min is not enough for the
completion of isothermal solidification at 1080 °C.
Figure 5 shows the ASZ thickness and the shear strength
of the bond at different times. It is clear that a dominant
increment in strength can be associated with a decrement
of the ASZ thickness. The shear strengths of Nimonic 75
and IN738 at room temperature are 650 MPa and 860
MPa, respectively. The results showed that the strength
of the TLP-bonded specimen at 1080 °C is less than the
strength of the base metals.
Figure 6 shows the optical microstructure of TLP
bonding at 1120 °C, 1150 °C after 120 min and 1180 °C
after 150 min. It was clear that at bonding temperatures
of 1120 °C and 1150 °C there was no centerline eutectic
micro-constituents, and isothermal solidification was
complete. It can be seen that with increasing bonding
temperatures from 1120 °C to 1150 °C, significant
microstructural changes occurred in the DAZ of IN738,
but no changes happened in the bond line. It is expected
that increasing the bonding temperature and con-
sequently increasing the diffusion rate of boron,
decreases the time required for completing the isother-
mal solidification. But as shown in Figure 6c, at 1180 °C
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
368 Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371
Figure 5: Shear strength and ASZ thickness versus bonding time at
1080 °C
Slika 5: Odvisnost stri`ne trdnosti in debeline ASZ od ~asa spajanja
na temperaturi 1080 °C
Figure 4: Light micrograph of TLP-bonded specimen at 1080 °C:
a) 30 min, b) 60 min, c) 120 min, d) 150 min
Slika 4: Svetlobni posnetek vzorcev, spajanih s TLP, na temperaturi
1080 °C: a) 30 min, b) 60 min, c) 120 min, d) 150 min
there are significant amounts of centerline eutectic in the
bonding after 150 min and isothermal solidification was
not completed yet. O. Idouw et al.25 reported that in TLP
bonding of IN738, increasing the bonding temperature
from 1170 °C leads to the diffusion of aluminum and
titanium into the molten interlayer and a reduction of the
isothermal solidification rate may happen. They reported
that the diffusion of titanium to the molten layer leads to
the formation of a Ni-Ti rich phase, M2SC-type sulpho-
carbide, chromium-rich boride and a 
-
’ centerline
eutectic. These phases are very stable and hard to solute.
Therefore, they can decrease the rate of isothermal
solidification and the completion of the process will be
postponed.
3.3 Precipitation in the diffusion-affected zone (DAZ)
Boride precipitation is expected in the base metal
where the concentration of boron is more than its solu-
bility limit in a 
 solid solution.23 The solubility limit of
boron in nickel is 0.3 % of amount fractions over the
range of 1065–1110 °C, according to the Ni–B binary
phase diagram.26 Therefore, the diffusion of boron into
the base metal during the holding at bonding temperature
leads to the formation of boride precipitates in the DAZ.
These precipitates are shown in Figures 7 and 8 in two
morphologies: globular and acicular.
When the distance between the precipitates and the
interface increases the morphology of the precipitates
changes from globular to acicular. As mention before,
dissolution is the first stage of TLP bonding where the
base-alloys solute elements diffuse into the base alloy
until the equilibrium condition between solid/liquid is
achieved. Therefore, the alloying-elements concentration
in the base metal layer adjacent to the bond is poorer
than the other layers. As shown in Table 3, the amount
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371 369
Figure 7: Variation of precipitates’s morphology in DAZ at 1120 °C
with time from a) 5 min to b) 120 min. The morphology changes from
globular to globular/acicular.
Slika 7: Spreminjanje morfologije izlo~kov v DAZ na temperaturi
1120 °C, v odvisnosti od ~asa od a) 5 min do b) 120 min. Morfologija
se spreminja od globularne do globularno/acikularne.
Figure 6: Light micrograph of TLP-bonded specimens at: a) 1120 °C
for 120 min, b) 1150 °C for 120 min, c) 1180 °C for 150 min
Slika 6: Svetlobni posnetek vzorcev, spajanih s TLP: a) 1120 °C –
120 min, b) 1150 °C – 120 min, c) 1180 °C – 150 min
of alloying elements in the precipitates closer to the
bonding interface (globular precipitates) is less than the
ones further than the bond line (acicular precipitates). As
seen in Table 3, the high chromium content of the
globular precipitates leads to a significant depletion of
the chromium around them. Since the Cr is the main
element for resistance to corrosion of the base alloy, the
depletion of Cr around these precipitates leads to a re-
duction of the matrix’s corrosion resistance.27 Moreover,
a high concentration of 
’-formers in the precipitates
leads to the depletion of Al and Ti in the matrix around
the precipitates, and as seen in Figure 8, there is no 
’
particle around them.
The morphology of the precipitates varies by in-
creasing the bonding temperature from the globular to
the acicular morphology, and at 1150 °C the acicular
morphology is dominant (Figures 6a and 6b). The
reason is probably related to more diffusion of the
alloying elements at the higher temperature and greater
homogeneity of the chemical composition in the DAZ.
The other point that can be gained from Figure 6c is that
the amount of precipitates in the DAZ decreases by
raising the temperature, such that at 1180 °C there is no
acicular and globular precipitate. It can be concluded that
these precipitates may be dissolved at temperatures
lower than 1180 °C.
Figure 7 shows that at 1120 °C, the morphology of
the precipitates varies from globular to globular/acicular
by increasing the holding time from 5 to 120 min. This
phenomenon is related to more diffusion of boron into
the base alloy at longer times and implies that the forma-
tion of acicular precipitates need more alloying elements
than globular precipitates (Table 3).
Figure 8 shows the bond interface in a specimen TLP
bonded at 1120 °C for 150 min. Boron was detected in
all precipitates, but their accurate concentration could
not be reported quantitatively due to X-ray absorption of
the EDS analyzer window. The compositional analyses
shown in Table 3 suggests that globular precipitates are
borides rich in Ni, Cr, Mo, W and Al and acicular
precipitates are borides rich in Ti, Ta, Mo, Nb and W. O.
Ojo et al.22 and O. Idowu et al.24 observed boride
precipitates rich in Cr, W and Mo in TLP bonding of
IN738 by a Ni-Cr-B interlayer at 1130 °C. Also, it can be
seen that there are lots of very fine 
’ around acicular
precipitates. The reason is related to the absorption of Ti
by these precipitates so that the concentration of the 
’
former element is less than normal. Since the Nimonic
75 does not have any boride formers in its chemical
composition, no precipitate was observed in its diffu-
sion-affected zone. Figures 8b and 8c show the different
zones specified in Figure 8a.
4 CONCLUSION
The effect of time and temperature on the TLP bond-
ing of the dissimilar nickel-based super alloys IN738LC
and Nimonic 75 using an MBF-80 interlayer was inve-
stigated. The following conclusions can be drawn from
this study:
• Before completion of the isothermal solidification,
the microstructure of the joint centerline consists of
three eutectic phases: 
 solid solution, Ni-rich boride
and Cr-rich boride.
• DAZ precipitates are formed due to boron diffusion
from the interlayer to the base metal during a
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
370 Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371
Figure 8: a) SEM micrograph of specimen TLP-bonded at 1150 °C for 150 min, b) and c) specified area in a) showed by b) and c) respectively
Slika 8: SEM-posnetek vzorcev, spajanih s TLP, na temperaturi 1150 C, 150 min, b) in c) dolo~eno podro~je a) prikazano na b) oz. c)
Table 3: EDS analysis of phases shown in Figure 8 (x/ %)
Tabela 3: EDS analiza faz prikazanih na Sliki 8 (x/%)
Al Ti Cr Co Ni Nb Mo Ta W
Acicular Precipitates 2.25 39.45 7.99 2.81 18.78 7.49 5.07 11.46 4.69
Globular Precipitates 8.44 3.53 38.70 4.06 34.68 0.57 5.28 0.61 4.12
B 3.71 0.46 25.65 6.32 60.61 0 0.60 1.15 1.49
solid-state transformation. Boride precipitates were
observed in the DAZ at bonding temperatures up to
1150 °C. However, at a bonding temperature of 1180
°C, the boride precipitates were not observed.
• At a given bonding temperature, the morphology of
the DAZ precipitates changes from globular to glo-
bular-acicular with an increasing bonding time.
• Boride precipitation in Nimonic 75 DAZ did not
occur due to a lack of boride-forming elements in its
chemical composition.
• A reduction in the isothermal solidification rate was
observed when TLP bonding at 1180 °C.
5 REFERENCES
1 M. Mosallaee, A. Ekrami, K. Ohsasa, K. Matsuura, Microstructural
evolution in the transient-liquid-phase bonding area of IN-738LC/
BNi-3/IN-738LC, Metallurgical and Materials Transactions A, 39
(2008) 10, 2389–2402, doi:10.1007/s11661-008-9588-0
2 M. Abdelfatah, O. Ojo, On the extension of processing time with
increase in temperature during transient-liquid phase bonding,
Metallurgical and Materials Transactions A, 40 (2009) 2, 377–385,
doi:10.1007/s11661-008-9726-8
3 M. Pouranvari, A. Ekrami, A. Kokabi, Aging response of transient
liquid phase bonded wrought IN718 superalloy: influence of
post-bond heat treatment, Science and Technology of Welding and
Joining, 19 (2014) 2, 105–110, doi:10.1179/1362171813Y.
0000000170
4 M. Pouranvari, A. Ekrami, A. Kokabi, Diffusion Brazing Metallurgy
of IN718/Ni-Cr-Si-B-Fe/IN718, Welding journal, 93 (2014) 2,
60s–68s, doi:10.1179/1743284713Y.0000000320
5 O. Kalvenes, Studies of the reaction of Nimonic 75 with Co2, Corro-
sion Science, 4 (1964), 211–220, doi:10.1016/0010-938X(64)
90020-4
6 O. Ojo, N. Richards, M. Charturvedi, Effect of gap size and process
parameters on diffusion brazing of Inconel 738, Science and Tech-
nology of Welding & Joining, 9 (2004) 3, 209–220, doi:10.1179/
136217104225012175
7 M. Pouranvari, A. Ekrami, A. H. Kokabi, Effect of the bonding time
on the microstructure and mechanical properties of a transient liquid
phase bonded 718 using a Ni-Cr-B filler alloy, Mater. Tehnol., 47
(2013) 593–599
8 N. Wikstrom, A. Egbewande, O. Ojo, High temperature diffusion
induced liquid phase joining of a heat resistant alloy, Journal of
Alloys and Compounds, 460 (2008) 1, 379–385, doi:10.1016/
j.jallcom.2007.06.066
9 X. Wu, R. Chandel, H. Li, Evaluation of transient liquid phase
bonding between nickel-based superalloys, Journal of Materials
Science, 36 (2001) 6, 1539–1546, doi:10.1023/A:1017513200502
10 A. Egbewande, C. Chukwukaeme, O. Ojo, Joining of superalloy
Inconel 600 by diffusion induced isothermal solidification of a
liquated insert metal, Materials Characterization, 59 (2008) 8,
1051–1058, doi:10.1016/j.matchar.2007.08.023
11 R. Saha, T. Khan, Microstructural developments in TLP bonds using
thin interlayers based on Ni–B coatings, Materials Characterization,
60 (2009) 9, 1001–1007, doi:10.1016/j.matchar.2009.04.002
12 W. Gale, D. Butts, Transient liquid phase bonding, Science and
Technology of Welding & Joining, 9 (2004) 4, 283–300,
doi:10.1179/136217104225021724
13 F. Jalilian, M. Jahazi, R. Drew, Microstructural evolution during
transient liquid phase bonding of Inconel 617 using Ni–Si–B filler
metal, Materials Science and Engineering: A, 423 (2006) 1,
269–281, doi:10.1016/j.msea.2006.02.030
14 M. Pouranvari, A. Ekrami, A. Kokabi, TLP bonding of cast IN718
nickel based superalloy: Process–microstructure–strength characte-
ristics, Materials Science and Engineering: A, 568 (2013), 76–82,
doi:10.1016/j.msea.2013.01.029
15 Dffusion Brazing, in ASM Handbook, ASM Internation, 2005,
207–221
16 M. Pouranvari, A. Ekrami, A. Kokabi, Solidification and solid state
phenomena during TLP bonding of IN718 superalloy using Ni–Si–B
ternary filler alloy, Journal of Alloys and Compounds, 563 (2013),
143–149, doi:10.1016/j.jallcom.2013.02.100
17 A. Rabinkin, Brazing with (NiCoCr)–B–Si amorphous brazing filler
metals: alloys, processing, joint structure, properties, applications,
Science and Technology of Welding & Joining, 9 (2004) 3, 181–199,
doi:10.1179/136217104225012300
18 T. A. Christman, K. B. Small, Guide to etching specialty alloys,
Advanced materials & processes, (2008) 10192/1881463/
amp16602p032.pdf/4d43b19b-45b7-4893-84a7-d6babb9505e5
19 M. Pouranvari, A. Ekrami, A. Kokabi, Phase transformations during
diffusion brazing of IN718/Ni-Cr-B/IN718, Materials Science and
Technology, 29 (2013) 8, 980–984, doi:10.1179/1743284713Y.
0000000238
20 N. Wikstrom, O. Ojo, M. Chaturvedi, Influence of process para-
meters on microstructure of transient liquid phase bonded Inconel
738LC superalloy with Amdry DF-3 interlayer, Materials Science
and Engineering: A, 417 (2006) 1, 299–306, doi:10.1016/
j.msea.2005.10.056
21 K. Ohsasa, T. Narita, T. Shinmura, Numerical modeling of the
transient liquid phase bonding process of Ni using Ni-B-Cr ternary
filler metal, Journal of Phase Equilibria, 20 (1999) 3, 199–206,
doi:10.1361/105497199770335721
22 A. Bondar, B-Cr-Ni Boron–Chromium–Nickel, Landolt-Börnstein,
Numerical Data and Functional Relationships in Science and
Technology (New Series). Group IV, Physical Chemistry, 2006
153–167, doi:10.1007/978-3-540-47004-5_19
23 O. Ojo, N. Richards, M. Chaturvedi, Isothermal solidification during
transient liquid phase bonding of Inconel 738 superalloy, Science and
Technology of Welding & Joining, 9 (2004) 6, 532–540,
doi:10.1179/174329304X8702
24 Alloy Phase Diagrams, 10 ed, ASM Handbook, 3, ASM Interna-
tional, 1992
25 O. Idowu, N. Richards, M. Chaturvedi, Effect of bonding tempe-
rature on isothermal solidification rate during transient liquid phase
bonding of Inconel 738LC superalloy, Materials Science and
Engineering: A, 397 (2005) 1, 98–112, doi:10.1016/j.msea.2005.
01.055
26 K. Tokoro, N. Wikstrom, O. Ojo, M. Chaturvedi, Variation in
diffusion-induced solidification rate of liquated Ni–Cr–B insert
during TLP bonding of Waspaloy superalloy, Materials Science and
Engineering: A, 477 (2008) 1, 311–318, doi:10.1016/j.msea.
2007.05.069
27 M. Pouranvari, Isothermal solidification during transient liquid-phase
bonding of GTD-111/Ni-Si-B/GTD-111, Mater. Tehnol., 48 (2014)
1, 113–118
M. KHAKIAN et al.: MICROSTRUCTURAL EVOLUTION DURING THE TRANSIENT LIQUID-PHASE BONDING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 365–371 371

S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
373–380
WORKABILITY BEHAVIOUR OF Cu-TiB2
POWDER-METALLURGY PREFORMS DURING COLD
UPSETTING
PREOBLIKOVALNOST Cu-TiB2 PREDOBLIK IZDELANIH Z
METALURGIJO PRAHOV MED HLADNIM KOVALNIM
PREIZKUSOM
Saikumar Gadakary1, Asit Kumar Khanra1, Maharajan Joseph Davidson2
1National Institute of Technology, Department of Metallurgical and Materials Engineering, Warangal, India
2National Institute of Technology, Department of Mechanical Engineering, Warangal, India
sai.gadakary@gmail.com
Prejem rokopisa – received: 2015-04-07; sprejem za objavo – accepted for publication: 2015-06-17
doi:10.17222/mit.2015.074
An investigation was carried out to find the workability behaviour of a Cu-TiB2 composite under triaxial stress-state conditions.
Initially, the TiB2 powder was prepared by using a self-propagating high-temperature synthesis (SHS) technique and the same
was added to a Cu matrix in order to make Cu-TiB2 composites. Cylindrical preforms with three different TiB2 weight
percentages (2 %, 4 % and 6 %) with aspect ratios of 0.50, 0.75 and 1 were prepared using a uniaxial load. Then the preforms
were pressureless sintered in a tubular furnace with a continuous flow of pure argon gas at 950 °C for a period of 1 h. The cold
upsetting test was carried out on the sintered specimens. The relationships between the various stresses, strains and the relative
density were determined. The results for the various stress-ratio parameters, namely (/eff) and (m/eff), the formability stress
index () under triaxial stress-state conditions were systematically analysed. The formability stress index was found to increase
with the increase in preform fractional density and it decreased with the aspect ratios. This was because the preform contains
more pores and the porous bed height is high. A statistical fitting method was performed on the curve drawn between the axial
strain and the stress-formability index. The compacts with a higher value of the aspect ratio and the initial preform density
showed a very high fracture strain.
Keywords: SHS, powder metallurgy, TiB2, workability, relative density, fracture strain
Izvr{ena je bila preiskava preoblikovalnosti Cu-TiB2 kompozita pri triosnem napetostnem stanju. Najprej je bil pripravljen prah
TiB2; s pomo~jo napredujo~e visoko temperaturne sinteze (SHS), ki je bil dodan Cu osnovi, da bi napravili Cu-TiB2 kompozit.
Z enoosnim stiskanjem so bili pripravljeni vzorci cilindri~ne oblike s tremi razli~nimi vsebnostmi TiB2 v masnih dele`ih (2 %,
4 % and 6 %) in z razmerjem 0,50, 0,75 in 1. Nato so bile predoblike sintrane v cevasti pe~i pri kontinuirnem pretoku ~istega
argona na temperaturi 950 °C in trajanju 1 h. Kovni preizkus v hladnem je bil izveden na sintranih vzorcih. Ugotovljena je bila
odvisnost med razli~nimi napetostmi, raztezki in relativne gostote. Sistemati~no so bili analizirani rezultati razli~nih parametrov
(/eff) in (m/eff) ter indeks preoblikovalnih napetosti () pri triosnem napetostnem stanju. Ugotovljeno je, da indeks
preoblikovalne napetosti nara{~a z nara{~anjem gostote predoblike in se zmanj{uje z razmerjem {irina-vi{ina. Razlog za to je
ve~je {tevilo por v predobliki in zato je vi{ina poroznega vzorca vi{ja. Izvedena je bila tudi statisti~na obdelava krivulje,
narisane med osno napetostjo in indeksom preoblikovalne napetosti. Stiskanci z vi{jo vrednostjo razmerja med {irino in vi{ino
ter ve~jo gostoto predoblike, so pokazali veliko napetost pri poru{itvi.
Klju~ne besede: SHS, metalurgija prahov, TiB2, preoblikovalnost, relativna gostota, napetost pri poru{itvi
1 INTRODUCTION
Powder metallurgy (P/M) is one of the most actively
researched manufacturing processes capable of deliver-
ing near-net-shaped precision metal parts. This process
has delivered a large number of industrial components,
such as connecting rods in engines, self-lubricating bear-
ings, gear sets in automobile transmissions, etc.1–2 Near-
net-shape components can be made and, the process has
the capability to greatly reduce machining costs, and can
improve material utilization.3–4 A series of upsetting,
bending, rolling and plane strain tests to assess the frac-
ture behaviour of porous materials was carried out.5 P/M
components involving copper are a highly researched
composite materials as alloys with copper as one of the
constituents will be stronger and durable.6-8
Copper P/M parts are used extensively in both struc-
tural and non-structural applications because of the high
corrosion resistance, high thermal and electrical conduc-
tivity. The corrosion resistance can be further improved
by the application of chemical conversion coatings or
anodizing treatment. In general, the physical and mecha-
nical properties of near-full (theoretical) density copper
and copper alloy P/M structural parts are comparable to
cast and wrought copper-based materials of a similar
composition. However, P/M copper parts vary in density
from the low-density self-lubricating bearings or filters
to the near-full density of the electrical parts.
TiB2, due to its high melting temperature, hardness,
elastic modulus, electro-conductibility and thermal
diffusivity, and excellent refractory properties and chem-
ical inertness has been widely used in many industrial
Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380 373
UDK 621.763:67.017:621.7.073 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)373(2016)
fields. It has applications in rocket nose cones for atmo-
spheric re-entry, ballistic armour, cathodes for Hall-Heroult
cells, crucibles for molten metals, metal evaporation
boats, and as acoating on cutting tools.9–10 It is widely
used as cutting-tool composites and wear-resistant
parts.11 The TiB2 powder is synthesized using the SHS
method. The main feature of the SHS process is that, it
utilizes the high energy released during the exothermic
chemical reaction of the reactants to yield a variety of
inorganic materials. Once the reactants are ignited by an
external source, the reaction front propagates within the
solid with a certain velocity to complete the chemical
reaction.
The extent of deformation possible without failure is
defined by the term "workability". It is the ability of a
material to withstand the induced internal stresses of
forming before any failure occurs. It is the extent to
which a material can be deformed in a specific metal
working process without the initiation of cracks.12–14
Workability depends on both the material and the
process parameters. The workability of dense material is
better than with P/M material. The workability can be
calculated by interpreting the value of hydrostatic stress
and effective stress for a tri-axial state of compression,
and the hydrostatic stress can be evaluated from the axial
and hoop stresses. The evaluation of different stresses
and the failure strain will reveal the workability limits of
the P/M composites.15 M. Abdel-Rahman and E. Sheikh16
explored the effect of the relative density on the forming
limit of P/M compacts during upsetting. J. J. Park et al.17
developed a constitutive relation involving the Poisson’s
ratio, relative density and flow stress to predict the
plastic deformation behaviour of porous metals. A
mathematical equation for the calculation of the flow
stress in the case of a simple upsetting of P/M sintered
performs was proposed by R. Narayanasamy et al.18 Fur-
thermore, the authors developed a new equation for the
determination of the hydrostatic stress in the case of the
simple upsetting of sintered P/M compacts. Equations
for the determination of the flow stress and the hydro-
static stress depending upon two factors, i.e., (i) the
value of Poisson’s ratio and (ii) the relative density of the
P/M preform in the case of the simple compression test
were proposed in the literature.18 However, copper-based
materials are hard to form as they offer resistance to the
forming load due to the formation of intermetallic com-
pounds. Thus, it is essential to investigate the defor-
mation behaviour of the Cu-TiB2-based composite
developed in the present work.
The deformation behaviours of Al matrix composites
have been studied extensively. There are, however, few
research reports on the deformation behaviour of
Cu-TiB2 composites. In the present paper efforts were
made to make composites of Cu-TiB2. The TiB2 used is
synthesized by using self-propagating high-temperature
synthesis (SHS). Until now there is no report of work-
ability studies on Cu-TiB2 composites. The workability
studies of the composites using a cold upsetting test are
evaluated.
2 EXPERIMENTAL DETAILS
Cu-TiB2 composite sintered preforms were selected
in order to provide a reasonably wide range of study,
namely, workability and work-hardening behaviour dur-
ing cold upset operation. The commercially available
copper powder was obtained from Alfa Aeser and the
TiB2 powder was produced using self-propagating
high-temperature synthesis (SHS) in our lab, igniting the
stoichiometric mixture of 20 g according to Equation (1),
in a tubular furnace, maintaining an argon atmosphere.
To investigate the particle size, shape and its distribution,
copper, TiB2 powders were studied using a scanning
electron microscope (SEM) (Figure 1). The Cu-TiB2
powders with different weight percentages of TiB2,
namely, 2 %, 4 % and 6 %, blend in a mortar mixer in
order to obtain a homogeneous mixture.
The powders were compacted in a 25-ton manual pel-
let press with the closed die set assembly technique.
Compacts of 15-mm diameter were prepared with aspect
ratios of 0.50, 0.75 and 1. The aspect ratio is the ratio of
the height to the diameter of the sample. The approxi-
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
374 Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380
Figure 1: a) Upsetting test setup, b) deformed preforms
Slika 1: a) Sestav za kr~ilni preizkus, b) predoblike po deformaciji
mate initial preform density is 70 % of the theoretical
density. These densities were achieved after sintering.
Then the preforms were sintered in a tubular furnace at a
temperature of 950 °C for a period of 1 h. To avoid
oxidation the preforms were heated in an inert argon
atmosphere. After the sintering schedule, the compacts
were cooled in the furnace itself. The sintered preforms
were cleaned and the dimensional measurement was
made before the deformation.
The upsetting tests (Figures 1a and 1b) were con-
ducted on a hydraulic press having a capacity of 50 tons.
Extreme care was taken to place the cylindrical specimen
within the platens, concentric with the central axis of the
hydraulic press (loading direction). Cylindrical preforms
were cold upset between the flat platens. Each preform
was subjected to an incremental compressive loading in
steps until the appearance of visible cracks on the free
surface.
Immediately after each incremental loading, the con-
tact diameter at the top (DCT), the contact diameter at the
bottom (DCB), the bulged diameter (DB), the height of the
preforms (hf) and the density (f) were recorded. Before
upsetting, the initial diameter (Do), the initial height (ho)
and the initial preform density (o) of the specimens
were measured. Moreover, the density measurements of
the preforms were carried out using the Archimedes
principle. Using the load, the dimensional parameters
and density, the different true stresses (i.e., z, , m and
eff) and the different true strains, (i.e., 
z and 
) and the
workability parameters () were determined using the
expressions specified below for the triaxial stress-state
condition.
For the present investigation, the TiB2 powders were
synthesized in-house as explained by the authors in a
previous study.19 The mixture of titanium oxide (TiO2),
boric acid (H3BO3) and magnesium was taken as per the
stoichiometric reaction (Equation 1). The powders were
mixed in a mortar mixer for about 20 min. A mixture of
20 g was then taken in a stainless-steel boat and was kept
in a tubular furnace (Systems control, Chennai). The
complete process was carried out in a highly pure argon
atmosphere in order to maintain an inert atmosphere.
The furnace was then heated up to 800 °C with a con-
stant heating rate. It was observed that the reaction was
taking place with an explosive sound at an approximate
temperature of 680 ± 15 °C. The furnace is then left to
cool to room temperature.
TiO2 (s) + 2H3BO3 (s) + 5Mg (s) 
TiB2 (s) + MgO (s) + 3H2O (g) + H (1)
After cooling, the synthesized powder was taken out.
It was observed that the reacted mixture is formed of
black lumps, and some amount of white surface layer
was seen on the lumps. The powder is then taken out and
was crushed into fine powder before going to the leach-
ing process, in order to make the leaching process
effective. The leaching process was carried out in diluted
HCl, with normality of 2 N. The solution was mixed
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380 375
Figure 4: XRD patterns of samples: a) pure Cu and b) Cu-6 (6 % of
mass fractions of TiB2)
Slika 4: Rentgenograma vzorcev: a) ~isti Cu in b) Cu-6 (6 % masnega
dele`a TiB2)
Figure 2: XRD pattern of TiB2 synthesized powder
Slika 2: Rentgenogram sintetiziranega prahu TiB2
Figure 3: TEM image of TiB2 synthesized powder
Slika 3: TEM-posnetek sintetiziranega prahu TiB2
with the crushed powder and heated up to 120 °C. The
process was continued while the solution boils for about
10 min, and then the solution was separated using filter
paper. The resulting powder, which was taken after the
leaching process, was then dried in an oven for 1 h. The
resulting powder is used in the present study.
The XRD patterns of the sample produced by the
SHS process after leaching shows the presence of TiB2
as major phase with TiO2 as minor phase in Figure 2.
The TEM image of the synthesized powder is shown in
Figure 3. The TEM images show the formation of spher-
ical and hexagonal TiB2 particles.
The XRD pattern of pure Cu and Cu-6 (6 % of mass
fractions of TiB2) is shown in Figure 4. The pattern
shows the presence of TiB2 as small peaks and Cu as a
major phase. This indicates there is no interaction bet-
ween the Cu and TiB2 during the pressureless sintering.
It is because of the smaller weight percentage of TiB2 in
the Cu matrix.
The scanning electron microscope images of the
Cu-TiB2 samples are shown in Figures 5a and 5d. Pure
Cu is shown in Figure 5a. Cu-2 (2 % of mass fractions
of TiB2), Cu-4 (4 % of mass fractions of TiB2) and Cu-6
(6 % of mass fractions of TiB2) are shown in Figures 5b
to 5d, respectively. The SEM images reveal the surface
morphology of the sintered samples. The images show
the porosity, the distribution of the powder particles and
the sintering behaviour.
3 THEORETICAL ANALYSIS
In the upsetting of P/M parts, the height decreases,
the average density increases, and the various stresses
increase.20 The expressions for the normal stress (z),
normal strain (
z), hoop stress (), hoop strain (
),
hydrostatic stress (m), effective stress (eff), and effective
strain (
eff) were taken from N. Selvakumar et al.21 and
Narayanasamy et al.22
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
376 Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380
Figure 6: a) Relative density (R) versus axial strain (
z) for triaxial stress state condition, b) relative density (R) versus axial strain (
z) for triaxial
stress-state condition (power-law curve-fitting results) and c) relative density (R) versus axial strain (
z) for triaxial stress-state condition
(parabolic curve-fitting results)
Slika 6: a) Odvisnost relativne gostote (R) od osne napetosti (
z) pri triosnem napetostnem stanju, b) odvisnost relativne gostote (R) od osne
napetosti (
z) pri triosnem napetostnem stanju (rezultati urejanja poten~ne krivulje) in c) odvisnost relativne gostote (R) od osne napetosti (
z) pri
pogoju triosnega napetostnega stanja (rezultati urejanja paraboli~ne krivulje)
Figure 5: SEM images of: a) pure Cu, b) Cu-2 (2 % of mass fractions
of TiB2), c) Cu-4 (4 % of mass fractions of TiB2), d) Cu-6 (6 % of
mass fractions of TiB2)
Slika 5: SEM-posnetek: a) ~isti Cu, b) Cu-2 (2 % masnega dele`a
TiB2), c) Cu-4 (4 % masnega dele`a TiB2), d) Cu-6 (6 % masnega
dele`a TiB2)
Triaxial Stress State Condition:
 =
A
B
(1)
A R R z= + − +( ) ( )2 2
2 2    (2)
B R Rz z= + − +( ) ( )2 2
2 2    (3)
Hoop stress, 


 =
+
− +
⎡
⎣⎢
⎤
⎦⎥
2
2 2
2
2 2
R
R R z
(4)
Hydrostatic stress, 
 


m =
+z 2
(5)
Effective stress,

      
eff =
+ − +
−
⎡
⎣⎢
⎤
⎦⎥
z zR
R
2 2 2 2
2
0 5
2 2
2 1
( )
.
(6)
Relative density, R =


f
th
(7)
f is the final density of the compact after deformation
and th is the theoretical density of the compact.
Formability Stress Index, 


= m
eff
(8)
4 RESULTS AND DISCUSSION
Figures 6a to 6c show the relationship between the
relative densities attained and the axial strain for the
Cu-TiB2 preforms. The compaction load was kept con-
stant for all the samples compacted with different pro-
portions of TiB2 and copper. It is observed that the initial
densification achieved is better for the preforms prepared
with copper alone and its relative density is around 75 %.
This reduces as the percentage of TiB2 addition in-
creases. The strain to failure was found to be low for the
preforms with 6 % TiB2 and it was found to increase as
the TiB2 decreases. Moreover, it can also be inferred that
the strain to failure is low for the low initial relative den-
sities.
A statistical curve-fitting technique was adopted for
the drawn curves and the prediction equation developed
from the curves was checked for its applicability by
comparing the correlation coefficient 'R2' values. These
values can be used for modelling purposes and can also
serve as prediction equations. In the present study, two
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380 377
Figure 8: a) axial strain (
z) versus formability stress index () triaxial stress-state condition, b) axial strain (
z) versus formability stress index
() triaxial stress-state condition (power-law curve-fitting results), c) axial strain (
z) versus formability stress index () triaxial stress-state
condition (parabolic curve-fitting results)
Slika 8: a) odvisnost osne napetosti (
z) od indeksa preoblikovalne napetosti () pri triosnem napetostnem stanju, b) odvisnost osne napetosti (
z)
od indeksa napetosti preoblikovanja () pri triosnem napetostnem stanju (rezultati urejanja poten~ne krivulje), c) odvisnost osne napetosti (
z) od
indeksa preoblikovalne napetosti () pri triosnem napetostnem stanju (rezultati urejanja paraboli~ne krivulje)
Figure 7: Fracture strain versus formability stress index ()
Slika 7: Napetost loma v odvisnosti od indeksa preoblikovalne sile ()
different curve-fitting techniques, i.e., the power law and
parabolic curve fitting, were used.
As the aspect ratio increases, the fracture strain in-
creases (Figure 7). The fracture strain decreases with the
addition of TiB2. Irrespective of the TiB2 content, the
fracture strain is less for 0.5 aspect ratio preforms. The
decrease in the fracture strain indicates that the
composite has attained a higher strength level with the
addition of TiB2, with less sacrifice in the strain values.
The addition of TiB2 to a preform with an aspect ratio of
1 has increased the strength with very little loss of frac-
ture strain.
Figures 8a to 8c show the plot drawn between the
axial strain and the formability stress index (b). A statis-
tical fit is made using the polynomial function and the
power-law function. It is found that the power law
related the parameters with higher accuracy. The addi-
tion of TiB2 decreased the strain further.
For preforms with a higher aspect ratio and a lower
initial relative density, the formability stress-index value
moves closer to the minimum value. The reason is that
this preform contains more pores and the porous bed
height is larger or greater. The increase in relative den-
sity with increasing deformation is less in this case com-
pared to lower aspect ratio preform. A parabolic curve-
fitting technique was applied to relate the formability
stress index and the axial strain for a varying aspect ratio
and relative density. The polynomial equations obtained
for each aspect ratio and relative density along with its
regression co-efficient value are presented in Table 1,
where it is observed that the constant value decreases
with a decreasing amount of relative density, irrespective
of the aspect ratio.
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
378 Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380
Figure 10: Axial stress (z) versus relative density (R): a) 1 ASPR, b) 0.75 ASPR and c) 0.5 ASPR
Slika 10: Odvisnost osne napetosti (z) od relativne gostote (R): a) 1 ASPR, b) 0,75 ASPR in c) 0,5 ASPR
Figure 9: a) Stress ratio (m/eff) versus relative density (R), b) stress ratio (z/eff) versus relative density (R), c) stress ratio (/eff) versus
relative density (R)
Slika 9: a) Odvisnost razmerja napetosti (m/eff) od relativne gostote (R), b) odvisnost razmerja napetosti (z/eff) od relativne gostote (R), c)
odvisnost razmerja napetosti (/eff) od relativne gostote (R)
Figures 9a to 9c give the plot of the relative density
with the stress ratio. The change of density along the ax-
ial and hoop stress directions was analysed. Figure 8a
shows the variation of the relative density with the mean
stress ratio. It is found that Cu with 4 % TiB2 and an
aspect ratio of 0.5 yielded high density values with a
high load-bearing capacity. The same was true for the
axial stress ratio and the hoop stress ratio. The hoop
stress is responsible for the initiation of cracks in the
preforms. Thus, it is clear that the addition of 4 % TiB2
improves the density of the preforms and postpones the
initiation of cracks. As the relative density increases the
stress ratio parameter also increases.
It was found that the relative density increases as the
stress-ratio parameter increases. The effect of the aspect
ratio on the stress-ratio parameter is found to be minimal
for the lower initial preform density preforms. However,
as the initial preform density increases, a higher stress
ratio parameter is observed for higher initial preform
densities with a lower aspect ratio. This shows that the
formability increases for the preforms with lower aspect
ratios and higher initial preform densities.
The Figures 10a to 10c show plots of the axial stress
(z) against the relative density R. The experiment was
done with preforms that have initial densities ranging
from 0.6 to 0.75 and aspect ratios ranging from 0.5 to 1.
The axial stress is found to increase rapidly during the
initial stage of densification, and thereafter continue to
increase with a lesser rate. The increase in stress due to
the forming load is followed by the closure of pores in
the preform, leading to its densification. This densifica-
tion is attributed to the combined effect of the geometric
and the matrix work-hardening. The preforms with a
lower TiB2 content were found to attain a higher stress
value than the TiB2 preforms. Along with the densifi-
cation, the load-bearing capability of the preforms also
increases, as is evident from the higher stress values in
the plot (Figures 10a to 10c).
It was found that the preform with 6 % TiB2 and a 0.5
aspect ratio densified more. Preforms with a high initial
preform density had a higher load-bearing capacity and a
longer strain to failure. This is due to the presence of a
smaller number of pores. At the same time, the disloca-
tion density increases rapidly during plastic deformation,
thereby resulting in a steep axial stress regime with a
smaller increase in the corresponding relative density.
5 CONCLUSION
The formability behaviours of sintered Cu-TiB2 com-
posite preforms was studied. The formability stress index
increased with an increase in the initial preform frac-
tional density and decreased with the aspect ratios. A
statistical fitting method was performed on the curve
drawn between the axial strain and the stress formability
index, and the parabolic curve fitting was found to give
better predictive results. For the compacts with a higher
value of the aspect ratio and initial preform density, the
initiation of the crack appeared at a very high fracture
strain.
Acknowledgement
This research work has been funded by the Council
of Scientific and Industrial Research (CSIR) (sanction
letter no. 22/597/12-EMR-II, dated 25/03/2012).
6 REFERENCES
1 J. Mascarenhas, Powder metallurgy: a major partner of the sustain-
able development, Materials Science Forum, 455–456 (2004),
857–860, doi:10.4028/ www.scientific.net/ MSF.455-456.857
2 G. E. Dieter, Mechanical Metallurgy, 3rd ed., McGraw-Hill, New
York 1981
3 T. J. Griffiths, R. Davies, M. B. Bassett, Compatibility equations for
the powder forging process, Powder Metallurgy, 4 (1977), 214–220
4 E. Bilgi, H. Erdem Çamurlu, B. Akgün, Y. Topkaya, N. Sevinç, For-
mation of TiB2 by volume combustion and mechanochemical
process, Mater. Res. Bull., 43 (2008), 873–881, doi:10.1016/
j.materresbull.2007.05.001
5 A. Nekahi, S. Firoozi, Effect of KCl, NaCl and CaCl2 mixture on
volume combustion synthesis of TiB2 nanoparticles, Materials
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380 379
Table 1: Parabolic curve-fitting equations and R2 values
Tabela 1: Ena~be urejanja paraboli~ne krivulje in vrednosti R2
Sample Aspect ratio Equation R2 value
Pure Cu
1
y = –0.2603x2 + 0.4053x + 0.6916 0.9863
Cu-2%TiB2 y = –0.0045x2 + 0.2031x + 0.6646 0.9915
Cu-4%TiB2 y = 0.0187x2 + 0.2652x + 0.5934 0.9671
Cu-6%TiB2 y = –0.307x2 + 0.3572x + 0.5642 0.9506
Pure Cu
0.75
y = –0.2038x2 + 0.312x + 0.7167 0.9804
Cu-2%TiB2 y = 0.3857x2 + 0.1301x + 0.6268 0.9724
Cu-4%TiB2 y = –2.0138x2 + 0.645x + 0.5534 0.8919
Cu-6%TiB2 y = –4.1476x2 + 0.4749x + 0.593 0.9105
Pure Cu
0.5
y = 0.3627x2 + 0.079x + 0.7022 0.9525
Cu-2%TiB2 y = –0.0329x2 + 0.1792x + 0.7117 0.8193
Cu-4%TiB2 y = –0.4124x2 + 0.2648x + 0.72 0.9889
Cu-6%TiB2 y = –4.6848x2 + 0.9689x + 0.702 0.9444
Research Bulletin, 46 (2011) 9, 1377–1383, doi:10.1016/
j.materresbull.2011.05.013
6 T. S. Srivatsan, G. Guruprasad, D. Black, R. Radhakrishnan, T. S.
Sudarshan, Influence of TiB2 content on microstructure and hardness
of TiB2 – B4C composite, Powder Technology, 159 (2005), 161–167,
doi:10.1016/j.powtec.2005.08.003
7 X. K. Sun, S. J. Chen, J. Z. Xu, L. D. Zhen, K. T. Kim, Analysis of
cold compaction densification behaviour of metal powders, Materials
Science and Engineering, A267 (1999), 43–49, doi:10.1016/S0921-
5093(99)00052-0
8 H. A. Kuhn, C. L. Downey, How flow and fracture affect design of
preforms for powder forging, International Journal of Powder
Metallurgy and Powder Technology, 10 (1974) 1, 59–66
9 I. Pokorska, Deformation of powder metallurgy materials in cold and
hot forming, Journal of Materials Processing Technology, 196 (2008)
1–3, 15–32, doi:10.1016/ j.jmatprotec.2007.08.017
10 P. W. Taubenblat, Importance of copper in powder metallurgy, Inter-
national Journal of Powder Metallurgy, 39 (2003), 25–28
11 R. Narayanasamy, V. Anandakrishnan, K. S. Pandey, Some aspects
on plastic deformation of copper and copper–titanium carbide pow-
der metallurgy composite preforms during cold upsetting, Interna-
tional Journal of Material Forming, 1 (2008) 4, 189–209,
doi:10.1007/s12289-008-0383-7
12 M. Sumathi, N. Selvakumar, R. Narayanasamy, Workability studies
on sintered Cu-10SiC preforms during cold axial upsetting, Materials
and Design, 39 (2012), 1–8, doi:10.1016/j.matdes.2012.02.004
13 S. Narayan, A. Rajeshkannan, Workability studies in forming of
sintered Fe-0.35C powder metallurgy preform during cold upsetting,
Journal of Iron and Steel Research International, 18 (2011), 71–78,
doi:10.1016/S1006-706X(12)60012-0
14 G. Dieter, Handbook on workability analysis and applications, ASM
publications 1, 2002
15 R. Narayanasamy, V. Anandakrishnan, K. S. Pandey, Effect of carbon
content on workability of powder metallurgy steels, Materials
Science and Engineering A, 494 (2008), 337–342, doi:10.1016/
j.msea.2008.04.022
16 M. Abdel-Rahman, E. Sheikh, Workability in Forging of Powder
Metallurgy Compacts, Journal of Materials Processing Technology,
54 (1995), 97–102
17 J. J. Park, Constitutive relations to predict plastic deformations in
porous metal during Compaction, International Journal of Mecha-
nical Sciences, 37 (1995), 709–719
18 R. Narayanasamy, R. Ponalagusamy, A Mathematical theory of
plasticity for compressible powder metallurgy materials – Part II,
Journal of Materials Processing Technology, 97 (2000), 110–113,
doi:10.1016/S0924-0136(99)00357-X
19 S. Gadakary, A. K. Khanra, R. Veerababu, Production of nano-
crystalline TiB2 powder through self-propagating high temperature
synthesis (SHS) of TiO2–H3BO3–Mg mixture, Advances in Applied
Ceramics, 113 (2014) 7, 419-426, doi:10.1179/1743676114Y.
0000000188
20 R. Narayanasamy, T. Ramesh, K. S. Pandey, S. K. Pandey, Effect of
particle size on new constitutive relationship of aluminium–iron
powder metallurgy composite during cold upsetting, Material and
Design, 29 (2008), 1011–1026, doi:10.1016/j.matdes.2006.06.004
21 N. Selvakumar, R. Narayanasamy, Phenomenon of strain hardening
behaviour of sintered aluminium preforms during cold axial forming,
Journal of Materials Processing Technology, 142 (2003), 347–54,
doi:10.1016/S0924-0136(03)00605-8
22 R. Narayanasamy, T. Ramesh, K. S. Pandey, Some aspects on
workability of aluminium–iron powder metallurgy composite during
cold upsetting, Materials Science and Engineering, A391 (2005),
418–426, doi:10.1016/j.msea.2004.09.018
S. GADAKARY et al.: WORKABILITY BEHAVIOUR OF Cu–TiB2 POWDER-METALLURGY PREFORMS ...
380 Materiali in tehnologije / Materials and technology 50 (2016) 3, 373–380
H. J. HU et al.: EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES AND A FRACTURE ANALYSIS
381–385
EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES
AND A FRACTURE ANALYSIS OF A MAGNESIUM ALLOY IN
THE HOMOGENIZED STATE
VPLIVI STRI@ENJA MED IZTISKANJEM HOMOGENIZIRANE
MAGNEZIJEVE ZLITINE NA MIKROSTRUKTURO IN NA
ANALIZO PRELOMA
Hongjun Hu1, Zhao Sun1, Dingfei Zhang2
1Chongqing University of Technology, College of Material Science and Engineering, 400050, P.R.China
2Chongqing University, National Engineering Research Center for Magnesium Alloys, 400044, P.R.China
48516686@qq.com
Prejem rokopisa – received: 2015-04-16; sprejem za objavo – accepted for publication: 2015-05-21
doi:10.17222/mit.2015.081
A new type composite extrusion method has been explored that combines extrusion and two continuous shears (referred to as
Extrusion-Shear (ES)). To study the effects of extrusion-shear on the microstructures and the mechanical properties of a
magnesium alloy for homogenization state, the ES process has been performed on extrusion equipment using an ES die.
Grain-size measurements and X-ray diffraction and compression experiments were conducted. The experimental results and
fracture analyses were described and discussed. It was found that the microstructures could be refined gradually from part 1 to 4
in the ES die. But a higher temperature could improve the grain growth and coarsen the microstructures. The results showed that
fine and uniform microstructures can be achieved using the (ES) process and various types of texture can also be found in the
microstructures. From the X-Ray diffraction (0002) the basal plane texture intensity was decreased, and the ES process could
weaken the dominant base texture for (0002). The main fracture mechanism for the specimens extruded at 420 °C is
transgranular fracture, while it is primarily caused by twins when the extrusion temperature was 450 °C.
Keywords: magnesium alloy, microstructure, texture, grain size, fracture analysis
Uporabljena je bila nova vrsta sestavljene metode iztiskovanja, ki kombinira iztiskovanje in dvojno kontinuirano stri`enje (ES).
Za {tudij vpliva stri`enja na mikrostrukturo pri iztiskanju in na mehanske lastnosti magnezijeve zlitine v homogeniziranem
stanju, je bil izvr{en ES postopek na napravi za iztiskovanje, z uporabo ES orodja. Izvedene so bile meritve velikosti zrn in
rentgenska difrakcija stisnjenih vzorcev. Opisani in ocenjeni so bili rezultati preizkusov in analize preloma. Ugotovljeno je, da
le-to postaja mikrostruktura bolj drobnozrnata od 1 do 4 v ES orodju. Vi{je temperature lahko olaj{ajo rast zrn, kar napravi
mikrostrukturo bolj grobozrnato. Rezultati so pokazali, da je mogo~e drobnozrnato in enakomerno mikrostrukturo dose~i s
postopkom stri`enja pri iztiskovanju (ES), v mikrostrukturi pa je mo~ najti razli~ne vrste tekstur. Iz rentgenske difrakcije se vidi,
da intenziteta teksture osnovne ravnine (0002) slabi in da ES postopek zmanj{a prevlado osnovne teksture za (0002). Glavni
mehanizem preloma vzorcev, iztiskanih pri 420 °C, je transkristalni prelom. Slednjega povzro~ajo predvsem dvoj~ki, ~e je
temperatura iztiskanja 450 °C.
Klju~ne besede: magnezijeva zlitina, mikrostruktura, tekstura, velikost zrn, analiza preloma
1 INTRODUCTION
A magnesium alloy is increasingly being specified
for automotive, transportation, electronics, aerospace,
and general engineering applications.1 Magnesium has
the ability to combine a high strength and a lighter
weight.2–4 But the use of magnesium alloys in aerospace
and military applications has begun to decline due to the
poor service performance.5,6 A new type of composite
extrusion method has been explored that combines extru-
sion and shear (referred to as ES). The ES process and
the die design and optimization in production practice
were obtained by repeated testing on industrial extruders.
To study the effects of ES on microstructures and the
mechanical properties of a magnesium alloy for the
as-cast and homogenized state of magnesium alloys
during the ES process, the ES process was performed on
extrusion equipment with an ES die. The grain size
measurement and the X-ray diffraction analysis and
compression experiments and fracture analysis were
made and discussed.
2 EXPERIMENTAL PART
A homogenizing treatment was conducted for AZ31
magnesium alloys (Mg3.02% Al1.01% Zn0.30%Mn),
the microstructures of the homogenized state were ob-
served, and the ES processes were performed with diffe-
rent preheated temperatures. The metallurgical experi-
ment, grain size measurement, mechanical property tests
and X-ray diffraction analysis were conducted on an
extruded AZ31 magnesium alloy. The AZ31 magnesium
alloys in the homogenized state were chosen as the
materials for the experiment. The material used in this
study is the AZ31B Mg alloy. The proof stress of the
AZ31 wrought Mg alloy is typically 160–240 MPa, and
Materiali in tehnologije / Materials and technology 50 (2016) 3, 381–385 381
UDK 67.017:669.721.5:620.1 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)381(2016)
the Mg alloy has a lower density and can withstand
greater column loading per unit weight. The billets were
machined into rounds with a diameter of 80 mm. The
process parameters in the manufacturing process are cri-
tical to the microstructure evolution and the mechanical
properties.
After the homogenization annealing treatment, the
specimen and the die were heated to the desired tempera-
tures and kept for 1 h. The experiments were made by
pressing with a force of 500 tons on the horizontal
extrusion machines with a container diameter of 85 mm.
The ES process was conducted with an extrusion speed
of 0.5 m/min. The microstructures of the AZ31 magne-
sium alloy for the homogenized state with preheated
temperatures 450 °C and 420 °C for the magnesium
alloy, respectively, were observed using a metallographic
microscope.
Immediately after the ES processes the rods and die
were water quenched in order to fix the metallurgical
microstructures, and then the ES die and formed rods, in-
cluding the billet remaining in ES die, were cut along the
plane of symmetry. In order to study the changes of the
microstructures at all stages of the ES process, samples
taken from products in different parts of the ES die were
observed, including microstructural observations and an
analysis of the planes that are parallel to the extrusion
direction.
All samples were grinded, first coarsely, then finely,
with waterproof abrasive paper, no.120, no.280, no.400,
no.600, no.800, no.1000, and no.1200. The etchant
employed in this experiment was a picric-acid-based
etchant including 5 g picric acid, 5 g acetic acid, 100 mL
alcohol and 10 mL distilled water. The microstructures
were observed under the electron microscope. The
specimens used for the metallographic observation and
the X-ray diffraction analysis were taken from the parts
shown in Figure 1. Deformation zone 1 is located within
the compression cone, deformation zone 2 is through the
compression cone, deformation zone 3 is the first
shearing zone, and deformation zone 4 is the second
shearing zone.
An Olympus light microscope was employed to
observe the microstructure. The average grain size (d) of
the microstructure was measured with the cut straight-
line method, which is a test method using the "average
intercept length" ( L) to represent the grain size, where
the L can be determined on the polishing plane. The
average intercept length of the grains filling the space is
as follows:7
L = 1/NL = LT/PM
where NL is the number of grains per unit test line
length, M is the macrostructure magnification, LT is the
total length of any test lines that get through the
microstructure image and P is the intersection number
of test lines and the grain boundaries.
The mechanical properties experiments were per-
formed on a CMT-5150 universal electronic testing ma-
chine, including compression and tensile failure experi-
ments that were conducted at a speed of 1 mm/min.
Before the experiments, deformed specimens were ma-
chined into standard compression specimens and tensile
specimens with dimensions of  8 × 16 mm and 5,
respectively. The tensile and compression fracture tests
were conducted to obtain the yield strength, tensile
strength, compressive strength, elongation after fracture
and the compression ratio of the extruded magnesium
alloy. The crystallographic orientation of the homo-
genized and extruded AZ31 magnesium alloy rod wasen
analyzed qualitatively by X-ray diffraction, which was
used to determine the crystallographic orientation of the
rods.8 Each diffraction peak was calibrated according to
the standard X-ray diffraction chart to determine the
Miller indices corresponding to the diffraction peaks.
The experiments were conducted on a D/Max-1200X
X-ray diffractometer, using a scanning angle of 20–80°,
a scanning speed of 1° min–1, a Cu target, an acceleration
voltage of 40 kV, a filament current of 30 mA and
graphite monochromatic as the filter.
3 RESULTS AND DISCUSSION
By comparing the grain sizes from the cross-sections
of the magnesium alloy billet after the ES process, it is
clear that the grains were significantly refined because
the dynamic recrystallization happens during the ES pro-
cess. In addition, it can also be seen from Figure 2 that
the grains are constantly refined from part 1 to 4 in
Table 1, and the grain sizes are compared in the same
part with different extrusion temperatures. It is well
known that the grain size increases with the temperature,
which is because a higher temperature would improve
the grain growth.8,9
The lamellar microstructures in the first shear zone
begin to reduce for the large shear strain caused by the
H. J. HU et al.: EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES AND A FRACTURE ANALYSIS
382 Materiali in tehnologije / Materials and technology 50 (2016) 3, 381–385
Figure 1: Schematic diagram of ES die with two 120° corner angles,
1 – upsetting zone, 2 – sizing zone, 3 – the first shearing zone, 4 – the
second shearing zone
Slika 1: Shematski prikaz ES z dvema zamikoma s kotom 120°:
1 – podro~je kr~enja, 2 – podro~je dimenzioniranja, 3 – prvo podro~je
stri`enja, 4 – drugo podro~je stri`enja
shear deformation in this zone and the deformed grains
turn into recrystallization grains. In the second shear
zone, DRX occurs, but there is still a small amount of
fine-grain strips in the center of the rod. The average
grain size decreased from about 188 μm to 2–5 μm by
using the ES process. The microstructures are not only
greatly refined but also relatively uniform because the
ES process includes two simple shear steps. The defor-
mation degree of the central parts of the rods increases,
allowing more DRX to occur. Therefore, the microstruc-
ture becomes smaller and more homogeneous.
From the microstructures of the billet after the ES
process it can be observed that the number of dendrites
in the -Mg matrix is greatly reduced, while the second
phase and the dendrites’ segregation were mostly elimi-
nated in the microstructures. The sizes of the grains
before and after the ES process were measured using the
cut line method, as shown in Table 1.
The relationship between the average recrystalliza-
tion grain size (d) and the Zener-Hollomon parameter (Z)
during the dynamic recrystallization is given by –ln d =
A + B ln Z. Based on the present ES process, with the
extrusion temperature, the accumulative strain increase
with the extrusion advancing, the grains will be
refined.10,11
The relationship between the average recrystalliza-
tion grain size (d) and the Zener-Hollomon parameter (Z)
during dynamic recrystallization is given by Equation
(1):
d
d
Z
n
0
3 1 310
⎛
⎝
⎜
⎞
⎠
⎟ = ⋅− − / (1)
The temperature-corrected strain rate Z is given by
Equation (2):
Z
Q
RT
= ⎛⎝
⎜ ⎞
⎠
⎟
 exp (2)
where 
 is the strain rate, Q is the activation energy for
the deformation, T is the temperature and R is the gas
constant.
It can be seen from Equation (1) and Equation (2)
that the preheated temperature of the billets has a signi-
ficant impact on the dynamic recrystallization of the fine
grains if the structure parameters of the ES dies are the
same. So the sizes and the volume fraction of the dyna-
mic recrystallization of the fine grains are inversely
proportional to the preheated temperature of the billets.
It is obvious that the average sizes of the grains for the
lower preheated temperature are finer than those of the
higher preheated temperature. It can be observed that the
lower temperature could inhibit any further grain growth.
To obtain the orientations of the deformed grains,
XRD tests were carried out. The baseline orientation was
obtained by using samples that were taken from the lon-
gitudinal section in a direction parallel to the extrusion
direction of the ES process. The strength of the diffrac-
tion lines represents the relative amount of the crystal
plane that is parallel to the distribution surface. As can
be seen from Figure 3, the strongest diffraction peak is
the (0002) crystal plane, and the (1010) plane is the
second-strongest diffraction peak before the ES process.
In Mg alloys, a strong basal texture is always found,
which resulted in poor ductility at room temperature.
Therefore, it is important to control the orientations
during the processing. In the upsetting zone I, the stron-
gest diffraction peak is still the (0002) crystal plane. But
in the sizing zone, the intensity of the (1010) and (1011)
planes increases very obviously. The crystal plane (1011)
is the strongest diffraction peak in the last two conti-
nuous shearing zones, while the strength of the basal
plane (0002) peak clearly decreases. The two shearing
steps lead to an increase in the peak-intensity ratio. This
shows that the ES process could promote the coexistence
H. J. HU et al.: EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES AND A FRACTURE ANALYSIS
Materiali in tehnologije / Materials and technology 50 (2016) 3, 381–385 383
Figure 2: Microstructures of the cross-section with preheated tempe-
ratures: a) 420 °C, b) 450 °C
Slika 2: Mikrostrukturi preseka pri temperaturi ogrevanca: a) 420 °C,
b) 450 °C
Table 1: Comparison of the grain size during various stages for the ES
process
Tabela 1: Primerjava velikosti zrn med razli~nimi stopnjami ES
postopka
Homogeniza-
tion state
Tempe-
rature Part 1 Part 2 Part 3 Part 4
Grain size
(ìm) 188
420 °C 4.9 2.8 2.4 2.1
450 °C 8.8 8.1 5.6 5.0
of the (0002) basal plane and the non-basal planes. From
the XRD test, it is clear that the basal texture of (0002)
was greatly weakened, and it becomes less dominant
after the ES process. The weakening of the basal texture
could be attributed to the shear deformation during the
ES process. The weakening of the basal texture is
expected to enhance the formability of Mg alloys.
Compression and tensile fracture experiments were
conducted at room temperature. The fracture behaviors
and the regulation of magnesium alloys during the ES
process were studied by observing and analyzing the
fracture phenomenon, which is important in effectively
preventing fractures of the magnesium alloy during the
ES process. Macro images of specimens after com-
pression have been shown in Figure 4.
As can be seen from Figure 5, the main fracture of
the specimen extruded at 420 °C is a transgranular frac-
ture, while it is mainly caused by the twins when the ex-
trusion temperature is 450 °C in Figure 6. The specimen
that was formed at 450 °C would be fractured when the
amount of deformation is 10.5 %, while the specimen
that was formed at 420 °C would fracture when the
amount of deformation is 13.8 %.
The microstructures were refined at an extrusion tem-
perature of 420 °C. Twins were hardly generated in the
part for dynamic recrystallization. The internal micro-
structures of the specimens extruded at 450 °C are
uneven, and there is a portion of elongated grains and a
H. J. HU et al.: EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES AND A FRACTURE ANALYSIS
384 Materiali in tehnologije / Materials and technology 50 (2016) 3, 381–385
Figure 6: Microstructures for fracture caused by compression in the
specimen extruded at 450 °C
Slika 6: Mikrostrukture ob prelomu, ki ga je povzro~ilo stiskanje
vzorca, iztiskanega na 450 °C
Figure 4: Images of the specimens after compression
Slika 4: Posnetek vzorcev po stiskanju
Figure 5: Microstructures of fracture caused by compression in the
specimen extruded at 420 °C
Slika 5: Mikrostrukture ob prelomu, ki je nastal pri stiskanju vzorca,
iztiskanega na 420 °C
Figure 3: X-Ray diffraction of AZ31 Mg alloy in different parts of the
rod in ES dies
Slika 3: Rentgenska difrakcija Mg zlitine AZ31 na razli~nih delih
palice v ES orodju
large amount of fine dynamic-recrystallization grains
distributed among the original grains. The specimens
would be subjected to a compressive stress during the
compression process at room temperature. Once the slip
plane tended to the direction that is parallel to the force
direction, the slip systems in the magnesium alloy would
stop moving. The increase in the external force usually
led to the occurrence of twins. The twins were generated
between the elongated original grain boundaries. Once
the twins occur, due to changes of the crystal orientation
in the twins, the slip plane is no longer parallel to the
direction of force, and the primary slip systems would
begin. The plastic deformation would not stop before the
sample is fractured. When the cracks on the edge of the
fracture encounter twins, the expansion path is forced to
change. Obviously, the hindrance of twins to the crack
propagation leads to an improvement in the material
toughness. Cracks on the fracture would be hindered by
grain boundaries when they encounter small grains. It is
known that the microstructures are inhomogeneous.
Twins would be generated among the elongated original
grains boundaries. However, it is prone to produce cracks
in the twin boundaries. It can be seen from Figure 6 that
a large amount of twins appear near the fracture and the
cracks. A large number of twins were generated along
the elongated grain boundaries near the fracture. There
are interactions between the twins and cracks, and the
cracks could induce twins, and twins would promote
crack nucleation. The twins and cracks develop rapidly,
so a large stress concentration would occur at the tip of
the cracks. Twinning and fracture are two processes that
release stress concentration. Therefore, the factors in
favor of one process are beneficial to the other process.
The microstructures would be elongated under tensile
stress and a large number of fine twins would be gene-
rated because of the severe deformation at the elongated
grain boundaries.
The ES process is actually a combination of extrusion
and ECAP extrusion to achieve a compound of extrusion
and shear processes. In the early extrusion, the micro-
structures appear to be bent and broken under simulta-
neous compression and shear.
4 CONCLUSIONS
To study the effects of extrusion-shear on the micro-
structures and mechanical properties of a magnesium
alloy in the homogenized state during the ES process, the
ES process was performed on extrusion equipment with
the explored ES die. The grain size measurement and the
X-ray diffraction analysis, compression experiments and
fracture analysis were made and discussed. It can also be
seen that the grains are constantly refined from part 1 to
4. The ES process could improve the dynamic recry-
stallization during deformation. The microstructures
become more uniform and finer, when compared to the
original states, during the ES process. A higher pre-
heating temperature could improve the grain growth and
coarsen the grains. The texture analysis showed that after
the ES process there is a variety of types of texture,
which could weaken the dominant position the of the
basal plane texture (0002). The main fracture mechanism
of the specimen extruded at 420 °C is transgranular
fracture, while it is mainly caused by the twins when the
extrusion temperature was 450 °C.
Acknowledgements
This work was supported by the open fund for Key
Laboratory of Manufacture and Test Techniques for
Automobile Parts (Chongqing University of Technology)
Ministry of Education in 2013, and foundation of the
post doctorate in Chongqing city and Project Number is
Xm201327, the China Postdoctoral Science Foundation
funded project, and the Chongqing Natural Science
Foundation Project of cstc2014jcyjA50004.
5 REFERENCES
1 G. Liu, J. Zhou, J. Duszczyk, Finite element analysis of magnesium
extrusion to produce a cross-shaped profile, J. of Manufacturing
Science and Engineering, 129 (2007) 3, 607–614, doi:10.1115/
1.2714590
2 H. Hu, D. Zhang, F. Pan, M. Yang, Analysis of the Cracks Formation
on Surface of Extruded Magnesium Rod Based on Numerical
Modeling and Experimental Verification, Acta Metallurgica Sinica,
22 (2009) 5, 353–364, doi:10.1016/S1006-7191(08)60109-X
3 K. Siegert, S. Jäger, M. Vulcan, C. Wizemann, Forming tubes, extru-
sions and sheet metal from magnesium AZ31, Mater. Sci. Forum,
488–489 (2005), 499–508, doi:10.4028/www.scientific.net/MSF.
488-489.499
4 H. Hongjun, Z. Dingfei, Y. MingBo, D. Ming, 201.Grain refinement
in AZ31 Magnesium alloy rod fabricated by an ES SPD process,
Transactions of Nonferrous Metals Society of China, 21 (2010) 2,
243–249, doi:10.1016/S1003-6326(11)60705-X
5 L. Li, J. Zhou, J. Duszczyk, Determination of a constitute relation-
ship for AZ31B magnesium and validation through comparison
between simulated and real extrusion, Mater. Process. Technol., 172
(2006) 3, 372–380, doi:10.1016/j.jmatprotec.2005.09.021
6 M. Chandrasekaran, Y. M. Shyan John, Effect of materials and tem-
perature on the forward extrusion of magnesium alloys, Mater. Sci.
Eng. A, 381 (2004), 308–319, doi:10.1016/j.msea.2004.04.057
7 N. Ogawa, M. Shiomi, K. Osakada, Forming limit of magnesium
alloy at elevated temperatures for precision forging, Int. J. Mach.
Tools Manu, 42 (2002) 5, 607–614, doi:10.1016/S0890-6955(01)
00149-3
8 R. Ye. Lapovok, M. R. Barnett, C. H. J. Davies, Construction of
extrusion limit diagram for AZ31 magnesium alloy by FE simula-
tion, J. Mater. Process. Technol., 146 (2004) 3, 408–414,
doi:10.1016/j.jmatprotec.2003.12.003
9 S. B. Xu, G. Q. Zhao, X. W. Ma, G. C. Ren, Finite element analysis
and optimization of equal channel angular pressing for producing
ultra-fine grained materials, Journal of Materials Processing
Technology, 184 (2007) 1–3, 209–216, doi:10.1016/j.jmatprotec.
2006.11.025
10 R. Matsumoto, K. Osakada, Development of warm forging method
for magnesium alloy, Materials Transactions, 45 (2004) 9,
2838–2844, doi:10.2320/matertrans.45.2838
11 D. Orlov, G. Raab, T. T. Lamark, M. Popov, Y. Estrin, Improvement
of mechanical properties of magnesium alloy ZK60 by integrated
extrusion and equal channel angular pressing, Acta Materialia, 59
(2011) 1, 375–385, doi:10.1016/j.actamat.2010.09.043
H. J. HU et al.: EFFECTS OF EXTRUSION SHEAR ON THE MICROSTRUCTURES AND A FRACTURE ANALYSIS
Materiali in tehnologije / Materials and technology 50 (2016) 3, 381–385 385

S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
387–394
FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED
EDGE ROUGHNESS USING SQUARE PIN TOOLS OF VARIOUS
SHOULDER GEOMETRIES
FSW VARJENJE PLO[^ IZ Al-Mg ZLITINE S POVE^ANO
HRAPAVOSTJO ROBOV Z ORODJEM S KVADRATNO KONICO IN
RAZLI^NO GEOMETRIJO BOKOV
Sebastian Balos, Leposava Sidjanin, Miroslav Dramicanin, Danka Labus Zlatanovic,
Aco Antic
Faculty of Technical Sciences, Department of Production Engineering, Trg Dositeja Obradovica 6, 21000 Novi Sad, Serbia
danlabus@uns.ac.rs
Prejem rokopisa – received: 2015-04-30; sprejem za objavo – accepted for publication: 2015-06-17
doi:10.17222/mit.2015.088
In the paper, the influence of tool shoulder geometries on the mechanical properties and weld surface roughness of an Al-Mg
alloy was studied. Different types of tools were used: with straight and concave profiles. Three concave-shoulder types were
applied, with volume ratios of 0.5 and 0.9 of the square pin to the shoulder reservoir and one with three concentric semi-toroidal
reservoirs with a volume ratio of 0.5 of the pin to the shoulder reservoir. The tensile and bend properties, hardness profiles and
macro-features of welds were examined. It was found that the optimum tensile and bending properties were obtained when
applying the tool with concentric reservoirs and the lowest welding speed. In this way, the widest nugget zone at the plate axis is
obtained, as well as the thickest nugget-zone layer under the specimen surface, covering the thermomechanical and
heat-affected zones. The overlapping of the nugget zone with the thermomechanical and heat-affected zones enables higher
proof and ultimate tensile strengths compared to the base material. The surface-roughness parameters of the weld face are lower
for the specimens welded with the tools with reservoirs and considerably lower than the base-material edge-surface roughness.
Keywords: friction-stir welding, 5052 aluminum alloy, FSW parameters, joint properties, surface roughness
V ~lanku je prikazana raziskava vpliva geometrije boka orodja na mehanske lastnosti in hrapavost povr{ine zvara Al-Mg zlitine.
Uporabljeni sta bili orodji z ravnim in konkavnim profilom. Uporabljene so bile tri vrste konkavnih bokov, z razmerjem
volumna 0,5 in 0,9 bo~nega rezervoarja ter kvadratne konice in eden s tremi poltoroidnimi koncentri~nimi rezervoarji, z
razmerjem volumnov 0,5 konica-bok. Preiskovane so bile natezne in upogibne lastnosti, profili trdote in makro izgled zvarov.
Ugotovljeno je, da so bile optimalne natezne in upogibne lastnosti dobljene pri uporabi orodja s koncentri~nimi rezervoarji pri
najmanj{i hitrosti varjenja. Na ta na~in se dose`e naj{ir{e podro~je me{anja pod povr{ino vzorca, ki pokriva termomehansko in
toplotno vplivano podro~je. Prekrivanje podro~ja me{anja s termomehanskim in toplotno vplivanim podro~jem, omogo~a vi{jo
mejo plasti~nosti in vi{jo natezno trdnost, v primerjavi z osnovnim materialom. Parametri povr{inske hrapavosti ~ela zvara so
manj{i pri vzorcih zvarjenih z orodji z rezervoarji in so ob~utno ni`ji, kot je osnovna hrapavost roba materiala.
Klju~ne besede: torno vrtilno varjenje, aluminijeva zlitina 5052, parametri FSW, lastnosti spoja, hrapavost povr{ine
1 INTRODUCTION
Friction-stir welding (FSW) is a solid-state metal-
joining process that uses a specialized non-consumable
rotating tool to join work pieces.1 It has been shown that
FSW is a suitable welding method for joining the materi-
als difficult to join using conventional welding tech-
niques. The most notable are aluminium-zinc-magne-
sium and aluminium-copper heat-treated allyos.2–7
Furthermore, Mg–alloys and dissimilar materials have
been successfully welded by FSW.8–13 The main advan-
tages of FSW are related to the fact that no melting oc-
curs and, therefore, gas porosity is avoided. Also, no dis-
tortion occurs and no shielding gases or welding
consumable materials are needed, leading to a relatively
low energy input.14 A decisive influence on the weld per-
formance comes from the welding tool and the parame-
ters such as welding and rotational speeds, as well as the
tilt angle, etc. On the other hand, the FSW tool geometry
can be related to the pin and shoulder geometry and the
relation between the pin and shoulder size.
The tool has three primary functions: heating,
material movement and containment of the heated
material between the tool shoulder and the base
plate.15–16 The tool pin influences deformational and
frictional heating, as well as shearing the material in
front of and moving the material behind the tool.17–20 The
geometry of the FSW tool pin can vary considerably:
round and flat-bottom cylindrical or threaded pins were
found to be adequate for aluminium-alloy plates of up to
12 mm.17 Truncated cone pins were developed to weld
plates thicker than 12 mm at higher welding speeds,
while fluted pins add deformation to the weld line,
increasing the possible welding speeds even further.15
Polygonal pins offer 12–25 % reduced traversing and
forging forces at a comparable strength as fluted pins.21
However, thin metallic plates were reported to be welded
with pinless tools as well.22
Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394 387
UDK 621.791:669.715:669.721.5 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)387(2016)
The shoulder of FSW tools influences a number of
weld features: from the most basic ones such as the weld
appearance and roughness, to microstructural character-
istics influencing the weld strength.23–25 These features
are obtained through the forging action aimed at a proper
containment and consolidation of the base material.25 A
number of shoulder designs emerged. The most basic
shoulder type is straight in the profile, without any cur-
vature. However, now the most common type is the con-
cave shoulder of a certain volume (reservoir), usually re-
quiring tilting of the tool by 2–4°.
Both mentioned types of shoulder enable a relatively
simple fabrication and cleaning after the welding pro-
cess.14–15,19 An alternative is a shoulder with features such
as scrolls, ridges or concentric circles, generally aimed at
increasing the welding speed, the deformation and the
frictional heating.15,26–28 Convex shoulder tools with
scrolls are characterized with an improved ability to
weld curvatures, base material with mismatch tolerances
and different-thickness workpieces.15 Finally, the fric-
tion-stir spot welding (FSSW) of polymers can be used
with a one-piece tool or a tool with a pin and a sleeve to
allow dissimilar polymers to be mixed in lap joints.29–30
Another variable is the tool material, which is tai-
lored to the material to be welded. Aluminium and mag-
nesium alloys can be welded using tool steels, most typi-
cally hot-work tool steel such as H13. However, copper
and copper alloys demand the use of nickel- or tung-
sten-alloy tools, while steel welding is most often done
with polycrystalline cubic boron-nitride (PCBN) or tung-
sten-carbide (WC) material.15
The aim of this paper is to study the influence of dif-
ferent shoulder geometries on mechanical and weld-sur-
face properties. Namely, regardless of what type of
shoulder geometry is applied, a careful optimization of
welding parameters is needed to obtain adequate me-
chanical properties as well as an acceptable weld-face
surface roughness, since rough weld tracks most often
require rework.23 Therefore, the machining of quality
weld tracks is desirable and can be achieved with an effi-
cient FSW tool that combines this outcome with high
mechanical properties, without the need for a tool tilt,
improving the tool life and used on a relatively rough
edge-surface textures of plates.
2 EXPERIMENTAL WORK
In this paper, the base material consisted of Al-Mg
EN-AW5052-0 plates of 5 mm. The chemical composi-
tion of the aluminium alloy determined with an optical
emission spectrometer ARL 3580 is given in Table 1.
The mechanical properties of the workpiece material,
tested with a WPM ZDM 5/91 tensile-testing machine,
on the basis of three specimens, are given in Table 2.
The plates were machined to dimensions of 300 mm
× 65 mm, with the average roughness of the edge to be
welded of Ra = 7.67 μm and the maximum peak rough-
ness (Rz) of 29.8 μm, corresponding to the sawing
process.31 The samples were tightly placed into a steel
fixture into a 130-mm-wide groove and secured by
clamps. The fixture was fitted onto an adapted Prvo-
majska UHG universal milling machine with a power of
5.2 kW. The tool used was made of X38CrMoV5-1
(H11) hot-work tool steel, having had its chemical
composition tested with an ARL 2460 optical emission
spectrometer, as given in Table 3. The hardness of all the
FSW tools was 53 HRC, as measured with a WPM
HPO-250 device. Four different tool geometries were
used, all with four-sided pins of equal dimensions,
Figure 1. It can be seen that three basic geometries were
used: a straight profile without a reservoir (0-type), two
concave shoulders with shoulder-to-pin ratios of 0.5 and
0.9 (5- and 9-type) and a feature shoulder with three con-
centric circles and the overall volume-to-pin ratio of 0.5
(53-type tool). FSW was done without a tool tilt, with a
rotational speed of 925 min–1 and three welding speeds,
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
388 Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394
Table 1: Chemical composition of EN-AW 5052 aluminium alloy (in
mass fractions, w/%)
Tabela 1: Kemijska sestava aluminijeve zlitine EN-AW 5052 (v
masnih odstotkih, w/%)
Cu Mn Mg Si Fe Zn Ti Al
0.09 0.09 2.78 0.24 0.38 0.046 0.015 balance
Table 2: Mechanical properties of EN-AW 5052-0
Tabela 2: Mehanske lastnosti EN-AW 5052-0
Proof strength
RpBM (MPa)
Ultimate ten-
sile strength
RmBM (MPa)
Elongation
ABM (%)
Vickers hard-
ness number
HV5
124±10 193±3 22±1 60±1
Table 3: Chemical composition of X38CrMoV5-1 tool steel (in mass
fractions, w/%)
Tabela 3: Kemijska sestava orodnega jekla X38CrMoV5-1 (v masnih
odstotkih, w/%)
C Si Mn P S Cr Mo V Fe
0.37 1.01 0.38 0.017 0.0005 4.85 1.23 0.32 balance
Table 4: Specimen-group designation system
Tabela 4: Sistem ozna~evanja vzorcev
Shoulder
cavity/pin
volume
ratio
Number of
shoulder
cavities
Area of the
shoulder surface
parallel to the
base metal (mm2)
Tool
designation
Welding
speed
(mm/min)
Specimen
group
designation
0 0
374 0
17 01
0 0 46 04
0 0 91 09
0.5 1
163 5
17 51
0.5 1 46 54
0.5 1 91 59
0.5 3
226 53
17 531
0.5 3 46 534
0.5 3 91 539
0.9 1
163 9
17 91
0.9 1 46 93
0.9 1 91 99
(17, 46 and 91) mm/min. Therefore, a designation sys-
tem was devised, Table 4. The plunge depth of tool
shoulder was 0.3 mm for all the FSW specimens.
The properties of the FSW workpieces were
determined with tensile, bending, hardness testing and
metallographic examination. The tensile and bending
testing was determined with the WPM ZDM 5/91 testing
machine, according to the EN 895 and EN 910
standards, respectively. Hardness was determined with a
VEB HPO-250 Vickers testing machine, with a 5-kg
load. The hardness measurements were done at a 1.5-mm
distance between the indentations to obtain the hardness
profiles. The metallographic examinations were done
after the standard metallographic preparation: grinding
with sandpapers (grit 220 to 2000), polishing with
diamond suspensions (6, 3, 1 and ¼ μm abrasive-grain
sizes) and etching with Keller’s reagent (2 mL HF, 3 mL
HCl, 5 mL HNO3, 190 mL H2O). The obtained metallo-
graphic specimens were then examined with a Leitz
Orthoplan light microscope.
Roughness parameters including the average rough-
ness (Ra), ten-point mean roughness (Ry) and the maxi-
mum peak roughness (Rz) were determined with a
Mitutoyo SJ-301 surface-roughness tester.
3 RESULTS
3.1 Mechanical properties
The results of the tensile and bend testing are shown
in Table 5. The tool design, that is, the tool-shoulder
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394 389
Table 5: Tensile properties and standard deviations, joint efficiencies, fracture locations and angles of bend to the first crack
Tabela 5: Natezne lastnosti in standardni odkloni, u~inkovitost spojev, polo`aj poru{itev in koti pri upogibanju do prve razpoke
Rp
(MPa)
Rm
(MPa)
A
(%)
Joint efficiency Number of
fractures ×
side of
fracture
Number of
fractures ×
fracture location
Angle of
bend to the
first crack-
weld root
(o)*
Angle of bend
to the first
crack-weld
face (o)*
RpFSW/
RpBM
100 (%)
RmFSW/
Rm BM
100 (%)
AFSW/
ABM
100 (%)
01 125±20 171±39 6±2 101 89 27 3×AS 2×NZ/TMAZ1×TMAZ/HAZ 22 (180)
04 134±8 182±22 10±2 108 94 45 3×AS 2×TMAZ/HAZ1×NZ/TMAZ 23 (180)
09 125±8 153±10 4±3 102 80 18 3×AS 3×NZ/TMAZ 17 (180)
51 147±8 191±3 10±2 125 103 45 2×AS1×RS 3×TMAZ/HAZ 26 (180)
54 138±15 160±14 7±2 111 83 32 3×AS 3×NZ/TMAZ 16 (180)
59 120±24 146±23 6±3 97 76 27 3×AS 3×NZ/TMAZ 12 (180)
531 160±2 198±5 13±2 129 103 59 3×RS1×AS
2×TMAZ/HAZ
1×NZ/TMAZ (180) (180)
534 155±5 196±7 10±3 125 102 45 2×AS1×RS 3×TMAZ/HAZ 55 (180)
539 125±3 153±7 7±3 101 80 32 3×AS 2×TMAZ/HAZ1×NZ 18 (180)
91 154±7 198±3 12±3 118 99 55 2×AS1×RS
2×TMAZ/HAZ
1×NZ/TMAZ 66 (180)
94 128±26 151±44 7±3 103 79 32 3×AS 2×NZ/TMAZ1×TMAZ/HAZ 17 (180)
99 123±11 143±29 6±3 99 74 27 3×RS 3×NZ/TMAZ 7 (180)
*Numbers in parentheses indicate that the cracking did not occur until the test was completed (until 180°)
*[tevila v oklepajih ka`ejo, da ni pri{lo do razpok, dokler ni bil test zaklju~en (pri 180°)
Figure 1: FSW tools: a) straight profile without a reservoir (0-type),
b) concave shoulder with the shoulder-to-pin volume ratio of 0.5
(5-type), c) feature shoulder with three concentric circles with the
overall-volume-to-pin ratio of 0.5 (53-type) and d) concave shoulder
with the shoulder-to-pin volume ratio of 0.5 (9-type) 0.9
Slika 1: FSW orodja: a) raven profil brez hranilnika (0-vrsta), b)
konkaven bok z razmerjem volumna na boku in konici 0,5 (5-vrsta), c)
oblikovan bok s tremi koncentri~nimi krogi z razmerjem skupni
volumen-konica 0,5 (53-vrsta) in d) konkavni bok z razmerjem
volumnov bok-konica 0,5 (9-vrsta) 0,9
geometry clearly influences the tensile properties. The
lowest values were obtained with the 0-type tool,
followed by the 9-type tool, the 5-type tool and, finally,
the highest mechanical properties were obtained with the
5- and 53-type tools. Furthermore, a clear correlation
exists between the welding speed and the tensile pro-
perties for the 5-, 53- and 9-type tools. At lower welding
speeds, the proof and tensile strengths, as well as the
elongation increase. In the case of the 0-type tool, the
highest tensile properties were achieved with a welding
speed of 46 mm/min. The same trend can be observed
for joint efficiencies, which are related to the base-
material tensile properties. Thus, the highest joint
efficiencies were obtained with the 531 specimen group
(129 % proof-strength efficiency, 103 % ultimate-ten-
sile-strength efficiency and 59 % elongation efficiency),
followed by the 534 and 91 specimen groups. The lowest
efficiencies were achieved with the 01, 09 and 99
specimen groups.
No clear influence of the welding speed on the
tensile-property standard deviation can be observed.
However, comparing the specimen groups obtained with
different tools, a clear trend can be seen: the highest
standard deviations were achieved with the straight
profile shoulder (the 0-type tool; 01, 04, 09 specimen
groups), while the lowest deviation was obtained with
the three-circular-groove tool (the 53-type tool; 531, 534,
539 specimen groups). For the 0-type tool, the general
trend shows a reduction in the standard deviations with
an increase in the welding speed. An opposite trend can
be observed with the 5-type tool. In the cases of the 53-
and 9-type tools, no clear correlation between the
standard deviation and the welding speed can be made.
Furthermore, no clear correlation between the tool
shoulder design, the welding speed or the tensile/bend
properties, and the side of fracture or the fracture
location can be observed.
The fracture location for all the specimen groups is
either in NZ/TMAZ (nugget zone/thermomechanical
zone) or TMAZ/HAZ (thermomechanical zone/heat
affected zone), with only one NZ fracture. The more
frequent side of fracture was the advancing side (AS), in
contrast to the retreating side (RS). Some cases of the
fracture during the tension test are shown in Figure 2.
Angles of bend to the first crack can be positively
correlated to tensile properties. Namely, as the tensile
properties are higher, the angle of the first crack in the
weld root is also higher. In the case of 531, the specimen
was bent to 180 o without cracking in the weld root. No
cracking occurred in either specimen weld face. Some
cases of bend testing are shown in Figure 3.
Hardness profiles are shown in Figure 4. All the
hardness profiles have a similar general shape, with the
maximum attained hardness values at the middle of the
chart, that is, in the NZ. However, some clear trends can
be observed for all the specimen groups welded with the
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
390 Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394
Figure 3: Bending to the first crack: a) specimen 04, b) specimen 531
Slika 3: Upogibanje do prve razpoke: a) vzorec 04, b) vzorec 531
Figure 2: Fracture locations for tensile specimens: a) NZ on RS
(specimen 539), b) NZ/TMAZ on AS (specimen 54)
Slika 2: Polo`aja preloma pri nateznem preizku{ancu: a) NZ na RS
(vzorec 539), b) NZ/TMAZ na AS (vzorec 54)
same FSW tool. Firstly, the hardness in the NZ is higher
for the specimens welded at higher welding speeds. Sec-
ondly, an increase in the welding speed causes a drop in
the hardness values for the TMAZ and HAZ zones (ap-
proximately 3–10 mm on either side of the NZ).
3.2 Metallographic examinations
The results of the metallographic examinations of the
representative specimens obtained with the 5- and
53-type tools are shown in Figures 5 and 6. It can be
seen that multiple relatively small, tunnel-like defects
occur in Specimen 51, Figure 5a. However, in Speci-
mens 54 and 59, single larger tunnel-like defects appear
at the bottom section of the NZ, on the advancing side.
Also, by increasing the welding speed, a more pro-
nounced tunnelling occurs. A similar trend of the
increasing defect size can be seen in Figure 6, relating to
the specimens welded with the 53-type tool. In Specimen
531, welded at the lowest welding speed, no tunnel-like
defect occurs, while in Specimens 534 and 53,
small-sized closed and open tunnels (root defects) occur.
The variation in the welding speed causes a change in
the transition line between NZ and TMAZ on the
advancing and retreating sides. This becomes closer to a
vertical (normal to the specimen surface) as the welding
speed increases. This means that the thickness of the
refined zone under the specimen surface is higher in the
specimens treated at a lower welding speed.
The macrostructures of the HAZ and TMAZ zones
also vary depending on the type of tool used, as well as
the welding speed. In Figure 5, for the 5-type tool, the
lowest and the highest welding speeds of 17 and 91
mm/min result in a finer microstructure. The same can
be observed for Specimen 534, welded with the 53-type
tool at the medium welding speed, Figure 6.
3.3 Roughness of the weld face
Roughness parameters of the FSW weld faces are
given in Table 6. It can be seen that the obtained results
are generally lower than those for the edges of the base
material (Ra = 7.67 μm; Rz = 29.8 μm). The highest
surface roughness is obtained with the 0-type tool
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394 391
Figure 6: Macrographs of specimens welded with 53-type tool (rota-
tional speed of 925 min–1): a) specimen 531 (17 mm/min welding
speed), b) specimen 534 (46 mm/min), c) specimen 539 (91 mm/min)
Slika 6: Makroposnetki vzorcev zvarjenih z orodjem vrste 53 (hitrost
vrtenja 925 min–1): a) vzorec 531 (hitrost varjenja 17 mm/min), b)
vzorec 534 (46 mm/min), c) vzorec 539 (91 mm/min)
Figure 4: Hardness profiles of specimens welded with a tool rota-
tional speed of 925 min–1: a) 0-type tool, b) 5-type tool, c) 53-type
tool, d) 9-type tool (welding speeds are indicated by different lines,
shown in Figure 4d)
Slika 4: Profili trdote vzorcev zvarjenih s hitrostjo vrtenja 925 min–1:
a) 0 – vrsta orodja, b) 5 – vrsta orodja, c) 53 – vrsta orodja, d) 9 –
vrsta orodja (hitrosti varjenja so prikazane z razli~nimi linijami,
obrazlo`enimi na Sliki 4d)
Figure 5: Macrographs of specimens welded with 5-type tool (rota-
tional speed of 925 min–1): a) specimen 51 (welding speed of 17
mm/min), b) specimen 54 (46 mm/min), c) specimen 59 (91 mm/min)
Slika 5: Makroposnetki vzorcev zvarjenih z orodjem vrste 5 (hitrost
vrtenja 925 min–1): a) vzorec 51 (hitrost varjenja 17 mm/min), b) vzo-
rec 54 (46 mm/min), c) vzorec 59 (91 mm/min)
without a reservoir in the shoulder, followed by the
53-type tool and 5-type tool, while the lowest roughness
was achieved with the 9-type tool with the largest
reservoir. No clear correlation between the roughness
parameters and the welding speed can be observed.
4 DISCUSSION
In this paper, the influence of the shoulder configura-
tion of the FSW tool with a square pin was evaluated in
relation to the weld tensile, bend and hardness pro-
perties, macro-analysis and the surface roughness of the
weld faces of the Al5052 plates with a high edge rough-
ness.
The tensile/bending properties and macro-imaging
are well correlated. Tunnel and root defects have a
considerable influence on the weld properties, causing a
decrease in the weld proof strengths, tensile strengths,
elongations, corresponding efficiencies, as well as the
bending angles to the first crack. Therefore, tunnel-free
specimens had the highest mechanical properties and the
corresponding weld efficiencies.
Such results can be explained with the nature of the
square-pin-tool interaction with the surrounding material
at various welding speeds. At a constant rotational speed,
a relatively low welding speed causes an increase in the
stirring-impulse frequency at a given weld length, lead-
ing to a more effective weld filling and defect avoidance.
These results support the findings from reference20,
where a similar, relatively low welding speed was
applied for FSW, with a square pin tool, of an Al-alloy-
based metal-matrix composite reinforced with TiB2
particles. Furthermore, in reference32, where the influ-
ence of a tunnelling-type defect on the mechanical and
metallurgical properties of an Al-Mg alloy was studied, a
low welding speed was more effective than high welding
speeds, even with the tools having concave shoulders
with large reservoirs. Furthermore, a similar finding was
reported by Balos and Sidjanin in reference32, where a
three-sided pin and an unusually large reservoir were
used to promote the appearance of a tunnel-like defect.
In reference32, the highest mechanical properties were
obtained with the lowest welding speed (17 mm/min)
and the highest rotational speed (1230 min–1). The theory
that refers to the frequency of impulse stirring at a given
weld length is in a marked contrast to the findings of
I. Radisavljevic et al.33, who reported that the avoidance
of a tunnel-like defect depends on the ratio of the
rotation to the welding speed, but with the application of
a threaded-pin tool.
Two specimens welded at the lowest speed, 01 and
51, also developed a tunnel-like defect. This phenome-
non is the result of the tool-shoulder geometry and,
therefore, its influence on the material flow. A
straight-profile pin without a reservoir (the 0-type tool)
provides lower mechanical properties than the 5-type
tool, Table 5, indicating that even a relatively small
reservoir provides a more convenient material flow. This
allows the material to move not only perpendicularly to
the tool axis, as forced by the pin, but also parallelly to
the tool axis, making the tunnel-like defect smaller (the
5-type tool) or eliminating it at a lower welding speed
(the 53- and 9-type tools).
The welding speed also has a marked influence on
the hardness of NZ. It can be seen that the increase in the
welding speed causes a rise in the hardness of NZ for all
the specimen groups. This is the result of the added de-
formation that comes from the increased welding speed
due to the pushing action of the pin while passing
through the material. On the other hand, NZs of the
specimens welded at a lower welding speed are wider
compared to the ones of the specimens welded at a
higher welding speed. The hardness values of the TMAZ
and HAZ zones vary; however, for the majority of the
specimens (welded with the 0-, 5- and 9-type tools), a
lower welding speed results in a higher average hardness
compared to the medium and high welding speeds. This
means that the welding speed of 17 mm/min enables
lower hardness variations throughout the weld. These
results are supported by the macrographs of the welds,
where a change in the NZ is observed.
With the increase in the welding speed, the NZ to
TMAZ transition line, at the advancing side, becomes
closer to a vertical (normal to the specimen surface),
while, at the retreating side, the NZ to TMAZ transition
line gradually diminishes. This observation is supported
by the hardness measurements, which suggest that the
hardness drops more gradually in TMAZ at RS than in
AS. The reason for such results is difficult to determine,
but the major influence may come from the tool-shoulder
geometry, which influences the material flow, causing a
higher amplitude and lower frequency for the specimens
welded with the 5-type tool or a lower amplitude and
higher frequency for the specimens welded with the
53-type tool. Furthermore, this also influences the thick-
ness of the NZ under the specimen surface. Namely, a
thicker refined NZ under the specimen surface and over
the TMAZ and HAZ zones can have a beneficial effect
on the mechanical properties. This elongated layer can
be regarded as very important for achieving higher proof
strength and ultimate tensile strength than those of the
base metal.
The results for the weld face roughness strongly de-
pend on the shoulder contact area and the angle of the
shoulder contact surface with the reservoir. It can be seen
that the 5- and 9-type tool-shoulder contact areas are
equal. This implies that a larger angle found for the
9-type-tool outer/external portion of the reservoir has a
beneficial influence on the surface-profile finishing,
preventing excessive adhesion of the base metal to the
tool material. On the other hand, for the 0-type tool, a
larger contact area (374 versus 163 mm2) proved to have
an adverse effect, probably due to the adhesion of the
base material. For the 53-type tool, a larger angle of the
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
392 Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394
concentric reservoirs has a secondary importance com-
pared to a larger contact area (226 versus 163 mm2) and
the existence of the secondary, tertiary and quaternary
contacts between the tool shoulder and the base material
that have a negative effect on the roughness parameters.
According to the results shown in Table 6, there is no
firm correlation between the roughness parameters and
the welding speed.
Table 6: Roughness parameters obtained with different tools and FSW
speeds
Tabela 6: Parametri hrapavosti, dobljeni z razli~nimi orodji in pri
razli~nih hitrostih FSW
Ra (μm) Ry (μm) Rz (μm)
01 3.24 32.13 21.64
04 2.41 14.51 17.33
09 4.17 20.85 26.14
51 2.48 15.27 13.10
54 1.35 12.69 8.05
59 2.05 10.40 8.94
531 1.46 12.78 8.29
534 2.51 18.30 14.07
539 2.02 12.90 11.59
91 1.95 16.84 12.89
94 1.14 10.94 6.46
99 1.11 7.32 5.84
5 CONCLUSIONS
According to the presented results, some conclusions
can be drawn:
• The tool with a square pin and three concave reser-
voirs, with a reservoir-to-pin volume ratio of 0.5
enables proof and ultimate tensile strengths to sur-
pass those of the base metal. The main reason for
such mechanical properties is the characteristic shape
of NZ that overlaps with TMAZ and HAZ.
• The welding speed of 17 mm/min enables the avoi-
dance of the tunnel-like defect. This way, a full 180°
bending over the weld root can be achieved.
• Low welding speeds are needed for achieving an in-
crease in the stirring-impulse frequency at a given
weld length. This enables a more effective weld
filling and defect avoidance.
• Weld surface-roughness parameters are considerably
lower for the specimens welded with the tools with
reservoirs than with the tools without a reservoir.
• A relatively rough edge-surface texture of the base-
metal specimens can be successfully overcome with a
careful optimization of the tool geometry and
welding speed, providing higher proof and ultimate
tensile strengths compared to the base metal.
Acknowledgement
The authors are very grateful to Mr. Ninkovic Mladen
from the company Unimet DOO in Novi Sad, Serbia, for
his valuable help in determining the roughness parame-
ters of the specimens.
6 REFERENCES
1 W. M. Thomas, E. D. Nicholas, J. C. Needham, M. G. Murch, P.
Templesmith, C. J. Dawes, International Patent Application No.
PCT/GB92/02203, GB Patent Application No. 9125978.8 (Dec.
1991) and U.S. patent No. 5, 460, 317 (Oct. 1995)
2 H. Khodaverdizadeh, A. Mahmoudi, A. Heidarzadeh, E. Nazari,
Effect of friction stir welding (FSW) parameters on strain hardening
behavior of pure copper joints, Materials and Design, 35 (2012),
330–334, doi:10.1016/j.matdes.2011.09.058
3 G. Ypekoglu, B. G. Kiral, S. Erim, G. Cam, Investigation of the
effect of temper condition on the friction-stir weldability of a7075
Al-alloy plates, Mater. Tehnol., 46 (2012) 6, 627–632
4 L. Fratini, G. Buqa, D. Palmeri, J. Hua, R. Shivpuri, Material flow in
FSW of AA7075–T6 butt joints: numerical simulations and experi-
mental verifications, Science and Technology of Welding and
Joining, 11 (2006), 412–421, doi:10.1179/174329306X113271
5 R. S. Mishra, Z. Y. Ma, Friction stir welding and processing, Mate-
rials Science and Engineering R, 50 (2005), 1–78, doi:10.1016/
j.mser.2005.07.001
6 P. L. Threadgill, A. J. Leonard, H. R. Shercliff, P. J. Withers, Friction
stir welding of aluminium alloys, International Materials Review, 54
(2009) 2, 49–93, doi:10.1179/174328009X411136
7 L. Magnusson, L. Kallman, Mechanical Properties of Friction Stir
Welds in Thin Sheet of Aluminium 2024, 6013 and 7475, Proc. of
the 2nd International Symposium on Friction Stir Welding, Gothen-
burg, Sweden 2000, 26–28
8 N. Afrin, D. L. Chen, X. Cao, M. Jahazi, Microstructure and tensile
properties of friction stir welded AZ31B magnesium alloy, Materials
Science and Engineering A, 472 (2008), 179–186, doi:10.1016/
j.msea.2007.03.018
9 G. M. Xie, Z. Y. Ma, L. Geng, Effect of microstructural evolution on
mechanical properties of friction stir welded ZK60 alloy, Materials
Science and Engineering A, 486 (2008), 49–55, doi:10.1016/
j.msea.2007.08.043
10 S. M. Chowdhury, D. L. Chen, S. D. Bhole, X. Cao, Tensile pro-
perties of a friction stir welded magnesium alloy: Effect of pin tool
thread orientation and weld pitch, Materials Science and Engineering
A, 527 (2010), 6064–6075, doi:10.1016/j.msea.2010.06.012
11 T. Debroy, H. K. D. H. Bhadeshia, Friction stir welding of dissimilar
alloys – a perspective, Science and Technology of Welding &
Joining, 15 (2010) 4, 266–270, doi:10.1179/174329310X1272649
607240
12 O. Frigaard, A Process Model for FSW of Age Hardening Alu-
minium Alloys, Ph.D. Thesis, Norwegian University of Science and
Technology, 1999
13 J. Quyang, R. Kovacevic, Material flow and microstructure in the
friction stir butt welds of the same and dissimilar aluminum alloys,
Journal of Materials Engineering and Performance, 11 (2002) 1,
51–63, doi:10.1007/s11665-002-0008-0
14 Z. Barlas, U. Ozsarac, Effects of FWS parameters on joint properties
of AlMg3 alloy, Welding Journal, 91 (2012), 16–22
15 R. S. Mishra, M. W. Mahoney, Friction Stir Welding and Processing,
ASM International, Materials Park, Ohio, USA 2007
16 A. Zivkovic, The influence of tool geometry on FSW process, Ph.D.
Thesis, University of Belgrade, 2010
17 R. Rai, A. De, H. K. D. H. Bhadeshia, T. DebRoy, Review: friction
stir welding tools, Science and Technology of Welding & Joining, 16
(2011) 4, 325–342, doi:10.1179/1362171811y.0000000023
18 Y. N. Zhang, X. Cao, S. Larose, P. Wanjara, Review of tools for
friction stir welding and processing, Canadian Metallurgy Quarterly,
51 (2012) 3, 250–261, doi:10.1179/1879139512y.0000000015
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394 393
19 P. Podrzaj, B. Jerman, D. Klobcar, Welding defects at friction stir
welding, Metalurgija, 54 (2015), 387–389
20 W. Yuan, R. S. Mishra, S. Webb, Y. L. Chen, B. Carlson, D. R.
Herling, G. J Grant, Effect of tool design and process parameters on
properties of Al alloy 6016 friction stir spot welds, Journal of
Materials Processing Technology, 211 (2011) 6, 972–977,
doi:10.1016/j.jmatprotec.2010.12.014
21 P. A. Colegrove, H. R. Shercliff, Development of Trivex friction stir
welding tool, Part 1 – two-dimensional flow modelling and experi-
mental validation, Science and Technology of Welding and Joining, 9
(2004) 4, 345–351, doi:10.1179/136217104225021670
22 D. Klob~ar, J. Tu{ek, A. Smolej, S. Simon~i~, Parametric study of
FSSW of aluminium alloy 5754 using a pinless tool, Weld World, 59
(2015), 269–281, doi:10.1007/s40194-014-0208-x
23 D. A. Burford, B. M. Tweedy, C. A. Widener, Influence of shoulder
configuration and geometric features on FSW track properties, 6th In-
ternational Symposium on Friction Stir Welding, Saint-Sauveur,
Montreal, Canada, Paper No. 36, 2006
24 G. Bussu, P. E. Irving, Fatigue Performance of Friction Stir Welded
2024-T3 Aluminium Joints, 1st International Symposium on Friction
Stir Welding, TWI, United Kingdom 1999
25 M. J. Peel, A. Steuwer, P. J. Withers, T. Dickerson, Q. Shi, H.
Shercliff, Dissimilar friction stir welds in AA5083-AA6082, Part I:
Process parameter effects on thermal history and weld properties,
Metallurgical and Materials Transactions A, 37 (2006) 7, 2183–2193,
doi:10.1007/BF02586138
26 S. J. Vijay, N. Murugan, Influence of tool pin profile on the metallur-
gical and mechanical properties of friction stir welded Al–10 wt.%
TiB2 metal matrix composite, Materials and Design, 31 (2010) 7,
3585–3589, doi:10.1016/j.matdes.2010.01.018
27 S. Brinckmann, A. Strombeck, C. Schilling, J. F. Santos, D. Loh-
wasser, M. Kocak, Mechanical and Toughness Properties of Robo-
tic-FSW Repair Welds in 6061-T6 Aluminum Alloys, Proceedings of
the Second International Conference on Friction Stir Welding,
Gothenburg, Sweden, 2000
28 C. J. Dawes, W. M. Thomas, Development of Improved Tool Designs
for Friction Stir Welding of Aluminum, Proceedings of the First In-
ternational Conference on Friction Stir Welding, Thousand Oaks,
CA, 1999
29 T. Pan, Friction stir spot welding (FSSW) – a literature review, SAE
Technical Paper 2007-01-1702, 2007, doi:10.4271/2007-01-1702
30 P. H. F. Oliveira, S. T. Amancio-Filho, J. F. Dos Santos, E. Hage,
Preliminary study on the feasibility of friction spot welding in
PMMA, Materials Letters, 64 (2010), 2098–2101, doi:10.1016/
j.matlet.2010.06.050
31 E. P. Degarmo, J. T. Black, R. A. Kohser, Materials and Processes in
Manufacturing, 9th ed., John Wiley and Sons, New York 2003, 60
32 S. Balos, L. Sidjanin, Effect of tunneling defects on the joint strength
efficiency obtained with FSW, Mater. Tehnol., 48 (2014) 4, 491–496
33 I. Radisavljevic, A. Zivkovic, N. Radovic, Avoidance of tunnel type
defects in FSW welded Al 5052-H32 plates, Welding & Welded
Structures, 1 (2012) 1, 5–11
S. BALOS et al.: FSW WELDING OF Al-Mg ALLOY PLATES WITH INCREASED EDGE ROUGHNESS ...
394 Materiali in tehnologije / Materials and technology 50 (2016) 3, 387–394
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
395–401
IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A
HEAT-TREATED CHALCOPYRITE CONCENTRATE WITH
ATMOSPHERIC SULPHURIC-ACID LEACHING
IZBOLJ[ANJE SELEKTIVNE EKSTRAKCIJE BAKRA IZ
TOPLOTNO OBDELANEGA KONCENTRATA HALKOPIRITA Z
LU@ENJEM Z @VEPLENO KISLINO NA ZRAKU
Elif Uzun1, Mustafa Zengin2, Ýsmail Atýlgan1
1Karabuk University, Department of Metallurgy and Material Science Engineering, 78050 Karabuk, Turkey
2Kalimakina Company, Department of Mining, 16059 Bursa, Turkey
elifuzun@karabuk.edu.tr
Prejem rokopisa – received: 2015-05-04; sprejem za objavo – accepted for publication: 2015-05-13
doi:10.17222/mit.2015.091
The present work focuses on the extraction of Cu, suitable for production, directly from a leach solution of a chalcopyrite
concentrate as an alternative to the conventional smelting and refining method. Firstly, the main aspects of the kinetics of
chalcopyrite leaching in a sulphuric-acid solution and transformation reactions of chalcopyrite at high temperatures were briefly
reviewed. Secondly, atmospheric-acid leaching experiments were performed as a function of the acid content and temperature
under agitation. Direct leaching experiments resulted in very low and scattered Cu recoveries, hence revealing a sulphide
passivation layer on the chalcopyrite. In order to get rid of this layer, the sulphides in the chalcopyrite were transformed into
sulphates with heat treatment at 500 °C prior to the leaching process, in which a high dissolution of species was obtained.
Finally, the leaching of the pre-heated concentrate resulted in a 99.82 % Cu extraction. Consequently, optimized process
parameters were proposed by comparing the Cu extraction and the increased purity of the pregnant solution without Fe
impurities.
Keywords: chalcopyrite concentrate, leaching, heat treatment, passivation layer, copper extraction, selectivity
Namen tega dela je ekstrakcija bakra (Cu), ki bi bila primerna za neposredno proizvodnjo iz lu`ine koncentrata halkopirita, kot
alternative obi~ajni metodi s taljenjem in rafinacijo. Najprej je bil opravljen pregled glavnih vidikov kinetike izlu`evanja
halkopirita v raztopini solne kisline in reakcij pretvorbe halkopirita pri visokih temperaturah. Nato so bili izvedeni preizkusi
izlu`evanja na zraku v kislini, v odvisnosti od vsebnosti kisline in temperature z me{anjem. Neposredni preizkusi izlu`evanja so
pokazali zelo nizko in raztreseno pridobitev bakra, ker se je halkopirit prekril s sulfidnim pasivacijskim slojem. Da bi se tega
sloja znebili, so bili sulfidi v halkopiritu pretvorjeni v sulfate, s toplotno obdelavo pri 500 °C pred postopkom izlu`evanja, pri
katerem je bilo dose`eno dobro raztapljanje vzorca. Kon~no je bilo dose`eno 99,82 % izlu`evanje bakra (Cu) iz predogretega
koncentrata. Posledi~no so bili predlagani optimalni parametri procesa s primerjavo ekstrakcije bakra (Cu) in pove~ane ~istosti
nosilne raztopine brez ne~isto~ `eleza (Fe).
Klju~ne besede: koncentrat halkopirita, izlu`evanje, toplotna obdelava, pasivacijska plast, ekstrakcija bakra, selektivnost
1 INTRODUCTION
Copper has been one of the most important metals for
over five thousand years.1 Production of copper was easy
for high-grade copper ores, for which traditional batch-
smelting techniques were used before their grades were
degraded. Afterwards, beneficiation of lower-grade ores
was realised with the flotation technique prior to the
smelting process. However, the traditional smelting tech-
nique became costly as the ore grades further decreased.
On the other hand, there is currently an imbalance bet-
ween the copper supply and the world demand.2 In
dealing with this problem, nowadays researchers are
working hard to decrease process costs.
Chalcopyrite (CuFeS2) is one of the most abundant
and widely spread copper-bearing minerals,3 accounting
for approximately 70 % of the Earth’s copper.4
In traditional smelting processes, the chalcopyrite
concentrate of a desired grade is obtained with a multi-
stage flotation of a sulphide ore, having fine grains of
chalcopyrite dispersed in a matrix of various sulphide
minerals and quartz.1,4 The chalcopyrite concentrate is
then smelted in reverberatory flash furnaces.5,6 A major
problem with smelting is the pollution of the environ-
ment, especially with sulphur dioxide.1 At the same time,
there is a decline in the copper-ore grades, often
remarked upon as a future challenge in the production of
copper for industry.7
Hydrometallurgical processes, an alternative to
smelting, offer a high potential for treating chalcopyrite
concentrate, apart from heap leaching, since they result
in increased metal recoveries and reduced air pollution.4
In acidic media (sulfuric, hydrochloric and nitric acid),
the concentrate can be leached under atmospheric or
elevated pressures.8–14 Among them, the most promising
one is sulfuric acid, since it can be readily produced at
lower costs as the ore and, hence, the concentrate contain
considerable amounts of sulphur.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401 395
UDK 621.78:669.333.7:66.063.4 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)395(2016)
The kinetics of leaching in sulfuric-acid media have
been extensively analysed as chalcopyrite can be dis-
solved in a strong sulfuric acid.15 The main reactions are
as follows:
CuFeS2 + 2H2SO4  CuSO4 + FeSO4 + 2H2S (1)
H2S + H2SO4  S
0 + SO2 + 2H2O (2)
and
2FeSO4 + 2H2SO4  Fe2(SO4)3SO2 + 2H2O (3)
These reactions are very slow at atmospheric pressure
because of the formation of a sulphide passivation layer
on the chalcopyrite surface during the leaching process.16
There have been many efforts at the lab scale to over-
come the difficulty of the concentrate being oxidized in
the presence of several species (ferric ions, cupric ions,
bacteria, pyrite oxygen, etc.).17–22 Additionally, some
researches have been done to increase the leach rate by
favouring anodic reaction between the passivation layer
and pyrite using the following redox reactions (4, 5, 6):16
CuFeS2 + 4Fe
3+  Cu2+ + 5Fe2+ + 2S0 (4)
Anodic half-cell reaction: chalcopyrite oxidation
CuFeS2  Cu
2+ + Fe2+ + 2S0 + 4e- (5)
Cathodic half-cell reaction: reduction of ferric ions
4Fe3+ + 4e–  4Fe2+ (6)
Another process for increasing the leach rate can be
roasting before leaching due to the fact that an oxidized
and/or sulphatized concentrate has a higher dissolution
rate because of the altered sulphide passivation layer.23–28
So, the pre-heating of chalcopyrite concentrate before
the leaching in a different ambient would be an impor-
tant step for the extraction of copper. As the passivation
layer is composed of sulphur, it requires a sulphatization
or oxidation process at elevated temperatures from
500–1000 K in an oxidizing ambient. Probable reactions
during this treatment are extensively analysed in the
literature.1,29 Among them, the most important ones for
low temperatures are briefly given, in light of the rele-
vant literature. L. Meunier et al.29 studied the behaviour
of chalcopyrite in a stream of air in a lower-temperature
regime of 573–8230 K. The following reactions were
mainly formed:
2CuFeS2  Cu2S + 2FeS + S (7)
FeS  Fe3O4 + Fe2O3 (8)
Cu2S  Cu2O  CuSO4  CuO·CuSO4 
 CuO  CuFe2O (9)
The first reaction (7) describes direct oxidation of
chalcopyrite with released sulphur dioxide, occurring
preferentially at temperatures higher than 900 K. At
lower temperatures, sulphides are transformed into sul-
phates, as shown with reactions (8), (9) and (10). In the
first step, Cu and Fe are separated form chalcopyrite in
the form of sulphide and elemental sulphur. And then,
these sulphides are oxidized as iron sulphide can be
gradually transformed to magnetite and then hematite, as
given in (9); copper sulphide can be easily converted into
copper sulphate at low temperatures, and sequentially
into tenorite and copper iron oxide at higher tempera-
tures, as given in (10). Here, the produced copper iron
oxide is reported to have a higher intrinsic resistance to
an acidic attack than tenorite.30 As a result, chalcopyrite
can be transformed into sulphates and oxides, which can
be dissolved in sulphuric acid in line with the following
reaction:
CuO + H2SO4  CuSO4 + H2O (10)
The main aim of this work is to improve the leaching
rate of chalcopyrite concentrate in a sulphuric-acid solu-
tion under atmospheric pressure in order to have a high
Cu recovery at a low process cost. So, only sulphuric-
acid solution without any catalysis was used for the
leaching experiments since it can be produced from the
same concentrate. In the first stage, direct leaching expe-
riments were performed to disclose the low dissolution
rate of chalcopyrite. This rate resulted in a limited
extraction of Cu due to the existence of a sulphide layer
on chalcopyrite and was, therefore, too far from estab-
lishing an industrial process. In the second stage, the
concentrate was pre-heated at 500 °C to increase the
extraction of Cu by converting sulphides into sulphates
in the concentrate. In this case, the dissolution rate was
found to increase enormously and a Cu extraction higher
than 99.8 % was obtained in a leaching time of 180 min.
Finally, a compromise between Cu and Fe extractions
was made to have the pregnant solution free of Fe as
much as possible.
2 EXPERIMENTAL WORK
2.1 Materials, analysis and characterization of samples
The agitated acid leaching method was used at atmo-
spheric pressure for the Cu extraction from the chalco-
pyrite concentrate obtained from a newly discovered
copper mine in the Kastamonu-Hanönü region of Turkey.
The concentrate was produced by means of beneficiation
of the sulphide copper ore via the flotation technique in a
pilot-scale plant. In the first step, mineralogical/elemen-
tal analyses of the concentrate were obtained and
summarized in Table 1. We can briefly say that the
concentrate consists of 65 % chalcopyrite, 30 % pyrite
and 5 % sphalerite.
In this work, mineralogical analyses of the samples
were investigated with an X-ray diffractometer (Rigaku
Primus IV) and elemental analyses were carried out with
an inductively coupled plasma-atomic adsorption spec-
trometer (ICP-OES, Perkin Elmer Optima 2100DV) and
an atomic absorption spectrometer (AAS, Perkin Elmer
AAnalyst 400). Additionally, a particle-distribution anal-
ysis was performed by means of an optical analysis
(Nikon EPIPHOT 200); thus, the sizes of the chalco-
pyrite particles were found to be in the range of 1–60 μm
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
396 Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401
and 80 % of the particles were below 45 μm. During the
leaching operations, sulphuric acid with a purity of 96 %
and a density of 1.84 g/cm3, obtained from the MERCK
Company, and distilled water were used. The pre-heating
and drying of the samples was done using a furnace
(Protherm PL442T) and a stove (drying oven, MAS
DT104), respectively. Finally, Cu and Fe extraction per-
centages were determined with the weighting method,
using an ordinary lab scale with a 1 mg resolution.
Table 1: Mineralogical/elemental analysis of chalcopyrite concentrate
Tabela 1: Mineralo{ka/elementna analiza koncentrata halkopirita
Mineral/
Element SiO2 S Cu Zn
Au
(g/t) Fe
% 1.12 36.44 21.10 21.26 0.39 33.34
2.2 Experimental methods
Leaching experiments were performed in an 800 mL
pyrex reactor, using the batch method. The reactor was
placed in a bath, whose temperature was controlled by a
temperature controller via a thermocouple within an
error of ± 1 °C. The agitation of the leach solution was
continuously maintained by an external propeller, rotated
at a constant speed of 350 min–1 throughout the pro-
cesses. The complete system is shown in Figure 1a.
Throughout this work, a constant volume of 600 cc of
the leaching solution was used and its temperature was
increased to the preselected value prior to the addition of
60 ± 0.5 g of the concentrate, which maintained a
constant solid-content weight-to-volume ratio (w/v) of
1/10 in the solution. With the addition of the concentrate,
the leaching process was commenced. After the process
was finished, or a sample was taken out of the reactor,
the residue was obtained using white-band filter paper.
Upon determining its weight and carrying out elemental
analyses, the amounts of Cu and Fe were determined and
used for the calculation of the extraction percentage.
The process flowchart of this work is given in Fig-
ures 1b and 1c, for direct and pre-heated leaching pro-
cesses, respectively. Initially, direct-leaching experi-
ments without any pre-treatment of the concentrate were
performed and very low extraction percentages were
obtained. Secondly, the leaching process halted at the
end of every 60-min period and the pregnant solution
was replaced with a fresh solution in order to see the
effect of acid consumption. Finally, the concentrate was
pre-heated in air ambient at 773 K for 2 h in order to
achieve an alteration of sulphide minerals.
3 RESULTS AND DISCUSSION
3.1 Direct atmospheric acid leaching
Experiments of direct leaching were performed with
the parameters of temperature and sulphuric-acid con-
centration under atmospheric pressure for 360 min. The
parameters and corresponding Cu % extraction results
are reported in Table 2. Although the results of the Cu %
extraction given in Table 2 are somehow scattered, a net
increase by the temperature is observed. Thus, the effect
of the acid concentration seems to be lower than that of
the temperature as the temperature increases. More
importantly, the Cu % extraction stays below 17 %,
which is in agreement with the results reported in the
literature.31 This behaviour contradicts the one explained
above, in which galvanic reactions are reported to
dominate the process due to the addition of pyrite, for
the redox potential to be between 400 and 500 mV.15 In
our case, as the concentrate has 30 % pyrite, galvanic
effects may also be expected for the reactions without
any electric field. However, a probable built-in redox
potential seems to stay below 400 mV, since pyrite stays
undissolved. Otherwise, pyrite should be dissolved at the
redox potential of 500 mV.32–33 Another evidence of this
is the XRD analysis of the leach residue, given in Figure
2, where the pyrite and chalcopyrite peaks are clearly
seen. This confirms that the dissolution processes of
chalcopyrite and pyrite in sulphuric acid are very slow
when there is no control of the redox potential of the
leach solution.
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401 397
Figure 2: XRD spectra of: a) chalcopyrite concentrate and b) residue
of direct leaching process with 105 g of acid content at 85 °C for 360
min (chalcopyrite: ; pyrite: ; sphalerite: ; jarosite: )
Slika 2: Rentgenograma: a) koncentrat halkopirita in b) ostanek
neposrednega 360 min izlu`evanja pri 85 °C in vsebnosti kisline 105 g
(halkopirit: ; pirit: ; sfalerit: ; jarozit: )
Figure 1: Experimental setup: a) flowchart of direct, b) pre-heated
and c) leaching processes
Slika 1: Eksperimentalni sestav: a) potek neposrednega, b) toplotno
predobdelanega in c) izlu`evalnega procesa
Table 2: Results of direct-leaching experiments lasting for 360 min
Tabela 2: Rezultati preizkusa 360 min neposrednega izlu`evanja
Temperature
(C)
Acid concentration
(g/L) % Cu extraction
Room
Temperature
(25)
15
45
75
8.99
15.83
12.97
45
15
45
75
14.46
12.10
13.55
85
15
45
75
15.01
16.42
14.06
On the other hand, an increase in the acid concentra-
tion from 45 g/L to 75 g/L at room temperature and 85
°C seems to cause a decrease in Cu recoveries of 2.86 %
and 2.36 %, respectively. This might be explained with
the rapid formation of the sulphur layer, reported in the
literature.6
3.2 Effect of acid consuming
The proposed reactions for the chalcopyrite leaching,
given with Equations (1) to (3), are acid-consuming pro-
cesses. The most proton-consuming (acid-consuming)
reaction is given below:2,34
CuFeS2(s) + 4H
+
(aq) = Cu
+2
(aq) + Fe
+2
(aq) + 2H2S(aq) (11)
It is hypothesized that it is governed by two steps: (i)
a rapid dissolution to establish the equilibrium (11) be-
tween the soluble species at the chalcopyrite surface and
the bulk solid and (ii) a rate-determining diffusion of the
soluble species away from the surface.
In order to test the effect of the above reaction, the
influence of the acid consumption on the copper extrac-
tion was investigated with a solid-liquid ratio of 1/20 g/L
in a H2SO4 concentration of 105 g/L at 60 °C. Before the
process was started, a fresh acidic solution was prepared
and held at a process temperature of 60 °C. At the end of
every 60-minute period of leaching, agitation was
stopped for the solid species to precipitate in the leach-
ing bath. Five minutes after the completed precipitation,
560 mL of the solid-free leach liquor was drained out of
the reactor. At this step, a 1-mg residue was taken out of
the reactor to analyse the copper content via AAS. Then,
the process was continued with an addition of 560 ml of
the fresh solution of 105 g/L H2SO4 concentration into
the reactor at 60 °C. This procedure was repeated every
60 min throughout the process.
The effect of the acid consumption was tested by de-
termining the Cu content of the residue for each
sampling. The results are given in Table 3, where a
smooth decrease in the copper grade of the leach residue
from 17.70 % to 17.10 % is clearly seen at 60 min and
300 min., respectively. However, the dissolution rate of
Cu decreases in time and may tend to stop at around
17 %. This blockage would be caused by the passivation
layer grown on the surface of chalcopyrite minerals to
resist the leaching kinetics, as reported in the
literature.15,35 It may be concluded that the acid
consumption has little effect on direct leaching of the
chalcopyrite concentration since the results are not
scattered as in the previous case.
3.3 Atmospheric acid leaching of pre-heated chalcopy-
rite concentrate
The experiments above showed that the leaching of a
chalcopyrite concentrate in a sulphuric-acid solution is
hardly possible for industrial applications without an al-
teration of the passivation layer formed around chalcopy-
rite minerals.
Table 3: Copper concentration of the residue for each sampling when
560 mL of leach liquor was replaced with a new one
Tabela 3: Vsebnost bakra v ostanku za vsako vzor~enje, ko je bilo 560
mL teko~ine za izlu`enje zamenjane z novo
Leaching time
(min)
Added fresh
solution (mL)
Cu % of leach
residue
60 560 17.70
120 560 17.40
180 560 17.23
240 560 17.10
300 560 17.10
Therefore, the concentrate was subjected to pre-heat-
ing at 773 K for 2 h in an atmospheric ambient before
the leaching process. The change in the mineralogical
property of the concentrate was determined with an XRD
analysis whose spectra before and after the treatment are
given in Figure 3. Dominant peaks of chalcopyrite and
pyrite of the unprocessed concentrate are seen in Figure
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
398 Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401
Figure 3: XRD spectra of: a) chalcopyrite concentrate, b) heated
concentrate and c) residue of leaching process with 150 g acid content
at 85 °C for 360 min (chalcopyrite: ; pyrite: ; sphalerite: ; iron
sulphate: ; hematite: ; chalcocyanide: ; tenorite: )
Slika 3: Rentgenogrami: a) koncentrat halkopirita, b) toplotno obde-
lan in c) ostanka neposrednega 360 min izlu`evanja pri 85 °C, z vseb-
nostjo kisline 150 g (halkopirit:; pirit: ; sfalerit: ; `elezov sulfat:
; hematit: ; halkocianit: ; tenorit: )
3a. But, in Figure 3b, the chalcopyrite peaks are disap-
peared whereas the pyrite peaks are still seen for the
heated concentrate; here, hematite, iron sulphate and
chalcocyanide peaks appear as dominant peaks.36,37
Moreover, a trace of tenorite is found to have little peaks
in Figure 3b. It can be concluded from the results that
the chalcopyrite in the concentrate is almost converted to
CuSO4 and CuO, which are soluble in a sulphide-acid
solution, and to those compounds (hematite and iron
sulphide) resistant to an acid solution.1,35
The Cu and Fe extraction percentages in the leaching
experiments of the pre-heated chalcopyrite concentrate
were determined with the weighting method, via an ele-
mental AAS analysis of the residue. The Cu extraction
results for three different acid concentrations and tem-
peratures as a function of the leach duration are given in
Figure 4. The first thing to be mentioned is the fact that
the leaching process seems to be nearly finished within
60 min as the lowest Cu extraction obtained is almost
over 90 % for the lowest acid content and the lowest
temperature. Additionally, the time dependence on the
Cu extraction becomes almost flat at 85 °C, depicted in
Figure 4, so that the reactions are nearly finished and
independent of the acid concentration. Moreover, only
hematite peaks are clearly observed and any trace of Cu
bearing species is hardly found on the XRD graph given
in Figure 3c, obtained for the residue of the experiment
with the highest Cu recovery of 99.8 %. This enormous
increase in comparison with the former experiments,
obtained with the so-called direct leaching method, is
certainly caused by the alteration of the sulphides in the
concentrate of the latter case.
The pre-heating of the concentrate in air ambient
makes it soluble in sulphuric acid, i.e., the solubility of
CuSO4 is higher than for Cu2S, CuS and CuFeS2. In addi-
tion, an increase in the solubility of hematite in a strong
acid is also reported in 38.
Thus, a pregnant solution may also have FeSO4 spe-
cies making it hard to produce pure Cu directly from the
solution (liquor).
Therefore, an optimization between purity and Cu re-
covery is necessary for the copper production, since Fe
incorporated into the Cu structure especially deteriorates
the electrical characteristics of copper. In other words,
iron incorporation into copper should be as low as
possible.
The optimization of the copper production was car-
ried out by selecting the leaching process between the
Cu and Fe recoveries, which were determined from the
residue with the AAS analysis. The selectivity was
calculated by subtracting Fe % from the Cu % recovery
as a measure of the Cu purity in the solution, depicted in
Figure 5, where Cu recoveries of over 97 % are reported
for 65 °C and 85 °C. For example, the highest Cu extrac-
tion of 99.82 % was obtained from the process with the
parameters of the temperature, 85 °C, acid concentration,
150 g/L, and duration, 180 min, as seen in Figure 4c; the
highest Fe extraction of 37.31 % was also obtained, and
hence, the selectivity of the leaching process was re-
duced to 62.51 %. In the figure, the selectivity goes
through a peak with the acid concentration at 65 °C, but
it decreases and seems to reach saturation at 85 °C. This
behaviour is the evidence of an increased solubility of
hematite at high acid concentrations and temperatures.
As a result, the optimum leaching condition can be
chosen as the peak on Figure 5, for the parameters of
65 °C and the 120 g/L acid concentration. Weaker acid
concentrations and lower temperatures seemingly result
in a low Cu extraction, in which some Cu is disposed in
the residue.
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401 399
Figure 5: Selectivity of Cu in the solution with respect to the acid
concentration for leaching processes of 180 min
Slika 5: Selektivnost Cu v raztopini, glede na koncentracijo kisline pri
trajanju izlu`evalnega postopka 180 min
Figure 4: Copper extraction versus leaching time at: a) 90 g/L H2SO4,
b) 120 g/L H2SO4, c) 150 g/L H2SO4 acid concentrations ( 85 °C;
 65 °C;  45 °C)
Slika 4: Ekstrakcija bakra v odvisnosti od ~asa izlu`evanja pri kon-
centraciji kisline: a) 90 g/L H2SO4, b) 120 g/L H2SO4, c) ( 85 °C;
 65 °C;  45 °C)
4 CONCLUSION
In the course of developing a hydrometallurgical pro-
cess as an alternative to the conventional smelting and
refining method, the production of Cu directly from the
leach solution of the chalcopyrite concentrate would be a
cornerstone. The most important step in this process is to
have a Fe-free solution since Fe can be incorporated into
the Cu structure during the subsequent electrowinning
step. An observation of the selectivity between Cu and
Fe can be taken as a measure of the purity of the pro-
duced copper. For achieving this aim, the main aspects of
the kinetics of chalcopyrite leaching in the sulphuric-
acid solution is briefly reviewed in the light of the
relevant literature. Additionally, transformation reactions
of chalcopyrite at high temperatures are studied because
direct leaching of chalcopyrite is a very slow process.
Direct-leaching experiments were performed at three
different temperatures of (25, 45 and 85) °C, and acid
concentrations of (15, 45 and 75) g/L. They resulted in
very low Cu recoveries, below 15 %, due to the existence
of a sulphide layer around the chalcopyrite species. In
addition, a slight decrease in the acid concentration was
observed at 85 °C, indicating acid consumption during
the process.
Acid consumption was checked with a solid-liquid
ratio of 1/20 g/L of a 105 g/L H2SO4 concentration at
60 °C. Its effect was measured with the change in the Cu
concentration in the leach residue over time. The Cu
concentration was found to decrease initially and have a
saturation tendency, confirming that the leaching process
was limited by the motion of the species through the
passivation layer around chalcopyrite.
Finally, the pre-heating of the concentrate before the
leaching at 500 °C for 120 min was found to have con-
verted sulphides into sulphates. The subsequent leaching
experiments were performed as a function of the acid
concentration in g/L, the temperature and the time, using
a solid-to-liquid ratio of 1/10 w/v at a stirring speed of
350 min–1 under atmospheric pressure. As a result, a high
dissolution rate was obtained because of the Cu recovery
of over 90 % in the leaching process of 60 min. Thus,
very high Cu recoveries of up to 99.82 % were achieved.
Besides, over 36 % of Fe was found to have dissolved in
the concentrate during the reactions because of the
increased dissolution rate of hematite due to the increase
in the temperature and acid concentration of the solution.
Finally, the optimum selectivity of Cu and Fe was
determined for the process with the acid concentration of
120 g/L, the temperature of 65 °C and the leach duration
of 180 min. These can be taken as the proposed process
parameters for the Cu production, optimized for the
balance between the purity and Cu recovery.
Acknowledgment
Authors wish to thank ARGETEST Co. for experi-
mental facilities. This work was supported by a grant
from the Karabuk University BAP Project and, in part,
by ASYA MADEN Co.
5 REFERENCES
1 S. Prasad, B. D. Pandey, Alternative processes for treatment of chal-
copyrite – A review, Minerals Engineering, 11 (1998) 8, 763–781,
doi:10.1016/S0892-6875(98)00061-2
2 H. R. Watling, Chalcopyrite hydrometallurgy at atmospheric pres-
sure: 1. Review of acidic sulfate, sulfate-chloride and sulfate-nitrate
process options, Hydrometallurgy, 140 (2013), 163–180,
doi:10.1016/j.hydromet.2013.09.013
3 Y. N. Li, A. P. Kawashima, A. R. Chandra, A. Gerson, Review of the
structure and fundamental mechanism and kinetics of the leaching of
chalcopyrite, Adv. Coll. Int. Sci., 197–198 (2013), 1–32,
doi:10.1016/j.cis.2013.03.004
4 M. Loncnar, M. Zupancic, P. Bukovec, A. Toklic, The effect of water
cooling on the leaching behavior of EAF slag from stainless steel
production, Mater. Tehnol., 42 (2009) 6, 315–321
5 A. K. Biswas, W. G. Davenpart, Extractive metallurgy of copper, 2nd
Edition, Pergamon Press, New York 1974, 64–69
6 Ö. Gok, G. C. Anderson, Dissolution of low-grade chalcopyrite in
acidified nitride electrolyte, Hydrometallurgy, 134–135 (2013),
40–46, doi:10.1016/j.hydromet.2013.01.021
7 www.indexmundi.com/en/commodities/minerals/copper/cop-
per-t20.html
8 P. B. Munoz, J. D. Miller, Wodsworth, Reaction mechanism for the
acid ferric sulfate leaching of chalcopyrite, Metall. Trans. Process
Metall., 103 (1979), 149–158, doi:10.1007/BF02652458
9 R. Romero, A. Mazueios, I. Palencia, F. Cerranza, Copper recovery
from chalcopyrite concentrates by BRISA process, Hydrometallurgy,
70 (2003), 205–215, doi:10.1016/S0304-386X(03)00081-1
10 D. Dreisinger, N. Abed, A fundamental study of the reductive leach-
ing of chalcopyrite using metallic iron, part I: kinetic analysis,
Hydrometallurgy, 66 (2002), 37, doi:10.1016/S0304-386X(02)
00079-8
11 F. Habashi, Action of nitric acid on chalcopyrite, Soc. Min. Eng.
AIME, 254 (1973b), 224–228
12 T. Hiroto, M. Kinoshito, Y. Awokira, H. Majima, The leaching of
chalcopyrite with ferric chloride, Metall. Trans. B – Process Metall.,
17 (1973), 19–28, doi:10.1007/BF02670815
13 N. Hiroyoshi, H. Miki, T. Hirajina, M. Tsunekawa, A model for fer-
rous promoted chalcopyrite leaching, Hydrometallurgy, 57 (2000),
31–38, doi:10.1016/S0304-386X(00)00089-X
14 N. Hiroyoshi, H. Kitagawa, M. Tsunekawa, Effect of solution com-
position on the optimum redox potential for chalcopyrite leaching in
sulphuric acid solutions, Hydrometallurgy, 91 (2008), 144–149,
doi:10.1016/j.hydromet.2007.12.005.
15 J. M. S. Koleini, V. Aghazadeh, A. Sandström, Acidic sulphate
leaching of chalcopyrite concantrates in presence of pyrite, Minerals
Engineering, 24 (2011), 381–386, doi:10.1016/j.mineng.2010.11.008
16 T. R. Ingraham, R. Kerby, Roasting in extractive metallurgy – a ther-
modynamic and kinetic review, Can. Metall., 6 (1967) 2, 89–119,
doi:10.1179/cmq.1967.6.2.89
17 H. C. Linge, A study of chalcopyrite dissolution in acidic ferric ni-
trate by potantiometric titration, Hydrometallurgy, 2 (1976), 51–64,
doi:10.1016/0304-386X(76)90013-X
18 E. M. Cordoba, J. A. Munoz, M. L. Blazquaz, F. Gonzalez, A.
Ballester, Passivation of chalcopyrite during its chemical leaching
with ferric ion at 68 °C, Minerals Engineering, 22 (2009), 229–235,
doi:10.1016/j.mineng.2008.07.004
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
400 Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401
19 D. G. Dixon, D. D. Mayne, K. G. Boxter, Galvanox – a novel galvan-
ically assisted atmospheric leaching technology for copper concen-
trate, Can. Metall. Q., 47 (2008), 327–336, doi:10.1179/cmq.2008.
47.3.327
20 J. E. Dutrizac, R. J. C. McDonald, T. R. Ingraham, The kinetics of
dissolution of synthetic chalcopyrite in aqueous acidic ferric sulphate
solutions, AIME, 245 (1969), 955–959
21 R. P. Hackl, D. B. Dreisinger, E. Peters, J. A. King, Passivation of
chalcopyrite during oxidative leaching in sulfate media, Hydro-
metallurgy, 39 (1995), 25–48, doi:10.1016/0304-386X(95)00023-A
22 J. D. Miller, H. Q. Portilla, Silver catalysis in ferric sulfate leaching
of chalcopyrite, In: J. Laskowski (Ed), XII International Min. Proc.
Cont., Elsevier, Amsterdam 1979, 851–901
23 G. Debernardi, J. C. Gentina, P. Albistur, G. Slanzi, Evaluation of
processing options to avoid the passivation of chalcopyrite, Interna-
tional Journal of Mineral Processing, 125 (2013), 1–4, doi:10.1016/
j.minpro.2013.09.001
24 C. Li, J. Li, J. Lu, X. Lu, R. Wang, X. Zhao, Bioleaching of chalco-
pyrite by Acidithiobacillus ferrooxidans, Minerals Engineering, 53
(2013), 184–192, doi:10.1016/j.mineng.2013.08.008
25 M. Nicol, H. Miki, L. Velasquez, The dissolution of chalcopyrite in
chloride solutions: Part 3, Mechanisms, Hydrometallurgy, 103
(2010), 86–95, doi:10.1016/j.hydromet.2010.03.003
26 K. Kaplun, N. Li, A. Kawashima, R. Gerson, Cu and Fe chalcopyrite
leach activation energies and the effect of added Fe3+, Geochimica
et Cosmochimica Acta, 75 (2011), 5865–5878, doi:10.1016/j.gca.
2011.07.003
27 D. Dreisinger, Copper leaching from primary sulfides: Options for
biological and chemical extraction of copper, Hydrometallurgy, 83
(2006), 10–20, doi:10.1016/j.hydromet.2006.03.032
28 E. Ghahremaninezhad, D. Asselin, G. Dixon, Electrochemical evalu-
ation of the surface of chalcopyrite during dissolution in sulfuric acid
solution, Electrochimica Acta, 55 (2010), 5041–5056, doi:10.1016/
j.electacta.2010.03.052
29 L. Meunier, H. Vanderpoorten, Roasting reactions of chalcopyrite,
ATB Met., 1 (1957), 31–35
30 J. K. Gerlack, E. D. Gock, S. K. Ghosh, Activation and leaching of
chalcopyrite concentrates with dilute sulfuric acid, International
Symposium on Hydrometallurgy, Chicago, USA 1973, 403–416
31 A. F. Tshilombo, Mechanism and kinetics of chalcopyrite passivation
and depasivation during ferric and microbial leaching, PhD thesis,
UBC, 2004
32 L. M. Ou, R. A. He, Q. M. Feng, Influencing factors of pyrite leach-
ing in germ free system, J. Cent. South Univ. Technol., 14 (2007) 1,
28–31, doi:10.1007/s11771-007-0006-3
33 M. J. Nicol, I. Lazaro, The role of non-oxidative processes in the
leaching of chalcopyrite, In: P. A. Riveros, D. Dixon, D. B. Dreisin-
ger, J. Menacho, (Eds), Copper 2003-Cobre 2003, vol. VI, Canadian
Institute of Mining, Metallurgy and Petroleum, Montreal 2003,
367–381
34 M. Gericke, Y. Govender, A. Pinches, Tank bioleaching of low-grade
chalcopyrite concentrates using redox control, Hydrometallurgy, 104
(2010), 414–419, doi:10.1016/j.hydromet.2010.02.024
35 N. Li, S. Lo, C. Hu, C. Hsieh, C. Chen, Stabilization and phase trans-
formation of CuFeO4 sintered from simulated copper-laden sludge,
Journal of Hazardous Materials, 190 (2011), 597–603, doi:10.1016/
j.jhazmat.2011.03.089
36 M. Krogovic, I. Boskovic, M. Vukcevic, R. Zejak, M. Knezevic, R.
Mitrovic, B. Zlaticanin, N. Jacimovic, The influence of the mineral
content of clay from the white bauxite mine on the properties of the
sintered product, Mater. Tehnol., 45 (2011) 6, 609–612
37 M. Vukcevic, D. Turovic, M. Krgovic, I. Boskovic, M. Ivanovic, R.
Zejak, Utilization of geopolymerization for obtaining construction
materials based on red mud, Mater. Tehnol., 47 (2013) 1, 99–104
38 J. MacCarthy, A. Nusrati, W. Skinner, J. Addai, Dissolution and
Rhedogical Behaviour of Hematite and Quartz Particles in Aqueous
Media at pH 1, Chemical Engineering Journal, 92 (2014),
2509–2522, doi:10.1016/j.cej.2009.05.010
E. UZUN et al.: IMPROVEMENT OF SELECTIVE COPPER EXTRACTION FROM A HEAT-TREATED CHALCOPYRITE ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 395–401 401

E. ROMHANJI et al.: HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING FAILURE
403–407
HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING
FAILURE: ALLOY DUCTILITY AND FRACTURE
HOMOGENIZACIJA Al-Mg ZLITINE IN KROKODILJENJE:
DUKTILNOST ZLITINE IN PRELOM
Endre Romhanji, Tamara Radeti}, Miljana Popovi}
University of Belgrade, Faculty of Technology and Metallurgy, Department of Metallurgical Engineering,
Karnegijeva 4, POB 35-03, 11 120 Belgrade, Serbia
endre@tmf.bg.ac.rs
Prejem rokopisa – received: 2015-06-01; sprejem za objavo – accepted for publication: 2015-06-12
doi:10.17222/mit.2015.110
High-strength Al-Mg alloys have a propensity toward hot fracture and failure by alligatoring during hot rolling. In this study the
effect of homogenization conditions on the susceptibility of an Al-5.1Mg-0.7Mn alloy toward alligatoring was investigated. It
was found that the plates homogenized in a temperature range of 460–520 °C were prone toward alligatoring, but once
homogenized at 550 °C they were not prone toward it any longer. The characterization of the fracture showed a predominance of
the intergranular ductile fracture, but the type of the constitutive particles filling the voids changed with the homogenization
regime. Grain decohesion and grain-boundary embrittlement show that certain thermal treatments resulted in a microstructure
that promoted slip localization.
Keywords: Al-Mg alloy, thermo-mechanical processing, hot working, ductility, embrittlement
Pri Al-Mg zlitinah z visoko trdnostjo se med vro~im valjanjem pogosto pojavijo razpoke in krokodiljenje. V {tudiji je bil
preiskovan vpliv pogojev pri homogenizaciji na ob~utljivost zlitine Al-5,1Mg-0,7Mn na krokodiljenje. Ugotovljeno je, da so
plo{~e, ki so bile homogenizirane v temperaturnem obmo~ju 460-520 °C, ob~utljive na pojav krokodiljenja, medtem ko plo{~e
homogenizirane pri 550 °C, niso bile ob~utljive na krokodiljenje. Pregled prelomov je pokazal, da prevladuje interkristalni `ilav
prelom, vendar pa se je z re`imom homogenizacije spreminjala tudi vrsta delcev v jamicah. Dekohezija med zrni in krhkost po
mejah zrn ka`eta, da se dolo~ena toplotna obdelava odra`a na mikrostrukturi, ki lokalizira drsenje.
Klju~ne besede: Al-Mg zlitina, termomehanska predelava, vro~a predelava, duktilnost, krhkost
1 INTRODUCTION
One of the main concerns in the fabrication of high-
strength Al-Mg alloy sheets is their proclivity toward hot
fracture and formation of defects such as edge cracking,
central bursts and alligatoring during hot rolling.1,2 The
alligatoring defect is characterized by an opening of
rolled slab ends due to a crack formation along the
central horizontal plane of the slab. In addition to the
inefficiencies associated with the material loss, the
failure caused by it can introduce serious damage to the
equipment.
It is considered that alligatoring arises due to an
inhomogeneous deformation and a variation in the resid-
ual-stress states across the width of a rolled material.3
The modeling based on the upper-bound approach
allowed the development of the criteria for its occurrence
in terms of the roll-gap shape factor,  = h/l (h is the
average sheet thickness of the rolling gap, l is the arc of
contact).4 The prediction of a failure due to the
alligatoring taking place at  > 1.35 was verified by cold
rolling an Al6061-T6 alloy. However, other reports show
alligatoring taking place at  of 0.5–1.5.1,2 Some
studies5,6 consider the development of a sharp notch at
the concave front of a slab and the resulting triaxial state
of the stress at the notch tip to be responsible for the
failure. Similarly, the presence of a complex state of the
stress and the concave shape of the front end of a slab
play important roles in the alligatoring during the cold
rolling of spheroidized steel.7
However, the effects of other factors, such as homog-
enization conditions and the microstructures of the
Al-Mg alloys, on the alligatoring are far less understood.
This work reports about the effect of the homogenization
conditions on the occurrence of the alligatoring in an
Al-5.1Mg-0.7Mn alloy during hot-rolling experiments.
2 EXPERIMENTAL WORK
The material used in this study was an Al-Mg alloy
with higher Mg and Zn contents than a standard AA
5083 alloy; its composition was within the lower limits
of an AA 5059 alloy (Table 1). The alloy, supplied by
Impol-Seval Rolling Mill–Serbia, was industrially DC
cast.
Prior to hot rolling, plates 25 mm × 30 mm × 55 mm
were homogenized, following one of the regimes given
in Figure 1. Hot rolling was performed at a two-high
rolling mill (a roll diameter of 200 mm). The applied
schedule of hot-rolling passes (ni, i = 1 – 8) was designed
by gradually increasing the reduction from 1.5 % in the
Materiali in tehnologije / Materials and technology 50 (2016) 3, 403–407 403
UDK 666.1.031.16:669.715:669.721.5:620.17 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)403(2016)
first pass to 35 % at the end (Figure 1). The temperature
of the plates at the exit of each pass was in a 380–400 °C
range. After each rolling pass, the plate was reheated for
10 min.
SEM characterization of the fractured surfaces was
conducted with JEOL JSM-6610LV at 20 kV, equipped
with an EDS detector and Tescan Mira 3XMU at 10 kV.
3 RESULTS
3.1 Hot rolling
The effects of different homogenization treatments
on the hot-rolling outcome are shown in Table 2. The
hot rolling of the plates homogenized according to
Regimes I and II failed due to the alligatoring. The
critical passes corresponded to the partial reductions of
 20 %. The alligator crack (Figure 2a) opened at the
front side in all the cases.
The total reduction for the plates homogenized fol-
lowing Regime I was  30 %, while  was 0.85–0.90 at
the failure pass.
The hot rolling of three out of four plates homoge-
nized according to Regime II failed. In the passes lead-
ing to the alligatoring,  was in a range of 0.76–0.89.
However, the plates homogenized according to Regime
II showed higher hot ductility than in the case of Regime
I since the total reduction at failure was  45 %.
The plates homogenized at the highest temperature
(Regime III) did not alligator and all of them were suc-
E. ROMHANJI et al.: HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING FAILURE
404 Materiali in tehnologije / Materials and technology 50 (2016) 3, 403–407
Table 1: Chemical compositions of the studied alloy and standard AA 5083 and AA 5059 alloys, in mass fractions (w/%)
Tabela 1: Kemijska sestava uporabljene zlitine in zlitini AA 5083 in AA 5059 po standardu, v masnih dele`ih (w/%)
Mg Mn Cu Fe Si Zn Cr Ti Sr Zr
Alloy 5.13 .72 .013 .34 .11 .51 .008 .025 .003 -
AA5083 4.0-4.9 0.4-1.0 <.1 <.4 <.4 <.25 0.05-0.25 <.15 <.005 <0.05
AA5059 5.0-6.0 0.6-1.2 <.25 <.5 <.45 .4-.9 <0.25 <.2 <0.05 0.05-0.25
Table 2: Hot-rolling outcome and characteristic parameters after the homogenization treatments
Tabela 2: Stanje po vro~em valjanju in zna~ilni parametri po homogenizaciji
Homogenization procedure Outcome
h0
(mm)
hf
(mm)
Total
reduction
(%)
"Alligatoring" pass
Partial reduction
(%) h/l
Regime I
460 °C/ 16 h  HR
Alligatoring 20.4 14.6 28.4 18 0.90
Alligatoring 20.4 13.7 32.8 19 0.86
Regime II
430 °C/12 h + 520 °C/16h
Cooling  500 °C/1 h  HR
Alligatoring 25 13.5 46.0 20 0.82
Alligatoring 25 14.0 44.0 18 0.89
Success 25 6.5 74.0 / /
Alligatoring 21 11.8 43.0 20 0.76
Regime III
430 °C/12 h + 550 °C/16 h
Cooling  500 °C/1 h  HR
Success 25 6.0 76.0 / /
Success 25 6.1 75.6 / /
Success 25 6.0 76.0 / /
Success 21 7.0 67.6 / /
Figure 2: a) Image of an alligatoring defect after hot rolling with the
total reduction of 44 %, b) concave front profile of the plate that did
not alligator (total reduction of 74 %)
Slika 2: a) Posnetek napake krokodiljenja po vro~em valjanju s celot-
no redukcijo 44 %, b) konkavni sprednji del plo{~e, kjer ni bilo
krokodiljenja (celotna redukcija 74 %)
Figure 1: Schematic representation of the thermo-mechanical treat-
ment: homogenization and hot-rolling schedules
Slika 1: Shematski prikaz termo-mehanske obdelave: potek homo-
genizacije in vro~ega valjanja
cessfully hot rolled with a total reduction of up to
 70–75 %.
The fronts of the plates developed a concave profile
during the hot rolling, with the formation of a groove at
the centerline as the deformation progressed (Figure
2b). The shape of the side profile varied with the posi-
tion along a slab. Some lateral spreading was observed at
high total reductions.
3.2 Fractography
A macroscopic examination of the fracture surfaces
revealed two distinct regions, similarly to the report on
spheroidized steel.7 The narrow region at the front edge
of a fractured plate, 1–2 mm in width, was characterized
E. ROMHANJI et al.: HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING FAILURE
Materiali in tehnologije / Materials and technology 50 (2016) 3, 403–407 405
Figure 5: Secondary-electron SEM micrographs of a brittle intergranular fracture (Regime I): a) smooth grain surfaces, b) slip-line traces and
shearing of dispersoids
Slika 5: SEM-posnetek s sekundarnimi elektroni krhkega, interkristalnega preloma (Re`im I) : a) gladka povr{ina zrn, b) sledovi drsnih linij in
stri`enje disperzoidov
Figure 3: Secondary-electron SEM micrographs of the front edge of a
plate: a) Regime I, b) Regime II
Slika 3: SEM-posnetek prednjega roba plo{~e s sekundarnimi elek-
troni: a) Re`im I, b) Re`im II
Figure 4: Secondary-electron SEM micrographs of ductile inter-
granular fractures and EDS of the broken constitutive particles:
a) Regime I, b) Regime II
Slika 4: SEM-posnetek s sekundarnimi elektroni duktilnega, inter-
kristalnega preloma in EDS-analiza polomljenega delca: a) Re`im I,
b) Re`im II
with a number of ridges consisting of high elevations and
depressions. It was more prominent on the plates homog-
enized according to Regime II than Regime I. The topog-
raphy of the rest of the fracture was more leveled. A dull
surface typical of a ductile fracture was sprinkled with
tiny sparkles indicating the presence of cleavage facets.
A SEM characterization of the plates homogenized
according to Regime I showed that the fractured surface
extended to the very end of the front edge (Figure 3a). A
hem, with markings from the front side of the plate, was
observed only at a few points.
On the plates homogenized according to Regime II,
the hem with the front-side marks extended over the en-
tire front edge, being 100–200 μm wide (Figure 3b).
Adjacent to it was a region of shallow, elongated dimples
that is characteristic for a shear fracture and void sheet-
ing. Away from the edge, the predominant fracture mode
was ductile intergranular fracture for all the alligatored
plates (Figure 4).
An intergranular fracture proceeded due to a coales-
cence of voids created by a break-up of constitutive
particles. However, there was a distinction between the
broken constituent particles filling the voids, depending
on the homogenization regime. Most of the particles of
the plates homogenized according to Regime I had a thin
plate-like morphology (Figure 4a). The EDS analysis
showed that those particles were a Mg-Si-Sr rich phase.
On the plates homogenized according to Regime II, a
mixture of a plate like Mg-Si-Sr and more irregularly
shaped Al6(Fe,Mn) was observed (Figure 4b).
Another difference was that a grain decohesion and a
brittle intergranular fracture were observed only on the
plates homogenized according to Regime I. Smooth
grain surfaces of the brittle intergranular fracture par-
tially covered with fine dimples nucleated at grain-
boundary dispersoids are shown in Figure 5.
A cleavage, with a typical river pattern (Figure 6),
was observed on the plates homogenized according to
both Regime I and II.
4 DISCUSSION
A failure by alligatoring is frequently ascribed to a
deformation inhomogeneity across the plate cross-sec-
tion due to a low reduction and an induced distribution of
stresses.1–4 The results of this study, i.e., the alligatoring
occurring at high reductions of 20 % and  = 0.75–0.9,
show that it may not be critical. Rather, the state of the
stress leading to the alligatoring might be related to the
metal flow in the roll gap. The lateral spread along the
centerline was greater than in the surface layer due to the
friction at the roll surfaces leading to a groove formation
at the front.5 The groove could have acted as a notch and
provided a stress concentration for the crack initiation.5,6
However, the results show that the grooving is unlikely
the sole cause of the alligatoring. On the plates homo-
genized according to Regime II, the fracture by the void
sheeting adjacent to the groove (Figure 3a) is a sign of
shear stresses and a low stress triaxiality.8
Furthermore, the deepest grooves were observed on
the plates that did not fail during the hot rolling. On the
other hand, on the least ductile plates, homogenized
according to Regime I, incipient grooving was observed
only at a few points along the front profile. Since defor-
mation conditions were identical for all the plates, but
the material response varied with the homogenization
treatment (Table 2), it is clear that different ductilities
and predispositions toward the alligatoring were related
to the microstructures developed during the thermal
treatment. Fracture features such as the change in the
type of the constitutive particles filling the voids point in
that direction. The observed grain decohesion and
grain-boundary embrittlement can be related to the slip
localization and dispersoid distribution that will be
described in Part II of this study.
5 CONCLUSION
It was demonstrated that ductility and predisposition
toward alligatoring during hot rolling depend on the ho-
mogenization treatment of the Al-5.1Mg-0.7Mn alloy.
An increase in the homogenization temperature im-
proved the ductility and the plates homogenized at 550 °C
(Regime III) did not alligator. Ductile intergranular
fracture is the dominant fracture mechanism, but on the
plates homogenized at the lowest temperature (Regime I),
grain decohesion and brittle intergranular fracture were
also observed. Changes in the fracture mode and the type
of the constitutive particles broken with the homo-
genization treatment pointed out that the propensity
E. ROMHANJI et al.: HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING FAILURE
406 Materiali in tehnologije / Materials and technology 50 (2016) 3, 403–407
Figure 6: Secondary-electron SEM micrograph of a cleavage (Regime I)
Slika 6: SEM-posnetek s sekundarnimi elektroni cepilnega loma
(Re`im I)
toward alligatoring is controlled by microstructural
changes.
Acknowledgment
This research was supported by the Ministry of Edu-
cation, Science and Technological Development, Repub-
lic of Serbia, and Impol-Seval Aluminium Rolling Mill,
Sevojno, under contract grant TR 34018.
6 REFERENCES
1 M. M. Al-Mousawi, A. M. Daragheh, S. K. Ghosh, D. K. Harrison,
Some physical defects in metal forming processes and creation of a
data base, Journal of Materials Processing Technology, 32 (1992)
1–2, 461–70, doi:10.1016/0924-0136(92)90202-4
2 J. G. Lenard, Workability and Process Design in Rolling, In: G. Di-
eter, H. A. Kuhn, S. L. Semiatin (eds.), Handbook of Workability and
Process Design, ASM International, Materials Park, Ohio, USA
2003, 258–277
3 J. A. Schey, Fracture in rolling processes, Journal of Applied Metal-
working, 1 (1980) 2, 48–59, doi:10.1007/BF02833609
4 S. Turczyn, The effect of the roll-gap shape factor on internal defects
in rolling, Journal of Materials Processing Technology, 60 (1996)
1–4, 275–282, doi:10.1016/0924-0136(96)02342-4
5 A. J. Meadows, W. J. Pearson, Discussion: Edge Cracking, Lami-
nation, and Surface Cracking in Hot and Cold Rolling, Journal of the
Institute of Metals, 92 (1964) 8, 254–255
6 R. Couch, R. Becker, M. Rhee, M. Lee, Development of a rolling
process design tool for use in improving hot roll slab recovery, Re-
port UCRL-TR-206843 (2004), Lawrence Livermore National Labo-
ratory, USA 2004; available at https://e-reports-ext.llnl.gov/pdf/
312127.pdf
7 L. Xu, G. S. Daehn, Alligatoring and damage in the cold rolling of
spheroidized steels, Metallurgical and Materials Transactions A, 25
(1994) 3, 589–598, doi:10.1007/BF02651600
8 F. Bron, J. Besson, A. Pineau, Ductile rupture in thin sheets of two
grades of 2024 aluminium alloy, Materials Science and Engineering,
380A (2004) 1–2, 356–364, doi:10.1016/j.msea.2004.04.008
E. ROMHANJI et al.: HOMOGENIZATION OF AN Al-Mg ALLOY AND ALLIGATORING FAILURE
Materiali in tehnologije / Materials and technology 50 (2016) 3, 403–407 407

F. VAJKAY et al.: ASSESSMENT OF TUBULAR LIGHT GUIDES WITH RESPECT TO BUILDING PHYSICS
409–412
ASSESSMENT OF TUBULAR LIGHT GUIDES WITH RESPECT TO
BUILDING PHYSICS
OCENA CEVASTIH VODNIKOV SVETLOBE GLEDE NA
GRADBENO FIZIKO
Franti{ek Vajkay, David Be~kovský, Vladimír Tichomirov
Brno University of Technology, Faculty of Civil Engineering, Institute of Building Structures, Veveøí 95, 602 00 Brno, Czech Republic
vajkay.f@fce.vutbr.cz
Prejem rokopisa – received: 2013-10-01; sprejem za objavo – accepted for publication: 2015-06-08
doi:10.17222/mit.2013.204
Architecture and the building industry consist of several minor fields, which come together when a building is designed and
erected. One of them is the field of building physics. It primarily focuses on the evaluation of constructions, structures and
spaces with respect to thermo-technical conditions, daylighting and many more. It is believed that among the above-mentioned
sub-fields of building physics, daylighting is the most important because it influences the health of every human being.
Daylighting enables people to see the colours and objects surrounding them. Therefore, buildings have to be equipped with a
kind of daylighting system. In the past, spaces located either in the centres of buildings or in the underground areas were
illuminated solely on the basis of luminaries. Nowadays, indirect daylighting systems like optical fibres or tubular light guides
may also be utilized.
In Central Europe, the application of tubular light guides in buildings increases every year. The manufacturers say that this is an
important and maintenance-free system. It is a system that brings light into any building, thus helping to save money. On the
other hand, the designing of tubular light guides is complicated as it has to deal with daylighting, thermo-technical aspects and
moisture. A huge amount of light guides is problematic due to the condensation of water inside the pipes, which is a side effect
of a non-air-tight solution, or just due to an incorrect thermal analysis. The paper focuses on different aspects of designing
light-guiding systems through computer simulations.
Keywords: tubular light guide, illuminance, luminance, building physics, computer simulations
Arhitektura in gradbeni{tvo sestojita iz ve~ glavnih podro~ij, ki se zdru`ijo, ko je objekt na~rtovan in zgrajen. Eno od takih
podro~ij je gradbena fizika. To je podro~je, ki je usmerjeno predvsem na oceno konstrukcije, zgradbe in prostora glede na
termo-tehni~ne pogoje, dnevno svetlobo in podobno. Med omenjenimi podpodro~ji gradbene fizike je najbolj pomembna
dnevna svetloba, ker vpliva na zdravje vsakega ~loveka. Dnevna svetloba omogo~a, da vidimo barve in predmete okrog sebe.
Zato morajo biti zgradbe opremljene s sistemom za dnevno svetlobo. V~asih so bili prostori, locirani v sredini zgradbe ali v
podzemlju, osvetljeni samo na osnovi svetlobnih teles. Danes pa lahko uporabljamo tudi posredne sisteme za dnevno svetlobo z
opti~nimi vlakni ali s cevastimi vodniki svetlobe.
V centralni Evropi uporaba cevastih vodnikov svetlobe v zgradbah iz leta v leto nara{~a. Proizvajalci trdijo, da je to pomemben
sistem, ki ne potrebuje vzdr`evanja. Sistem omogo~a svetlobo v zgradbi, kar pomaga pri zmanj{evanju stro{kov. Vendar pa je
na~rtovanje cevastih vodnikov svetlobe zapleteno, glede na to ali gre za dnevno svetlobo ali termo-tehni~no na~rtovanje in
vlago. Velik dele` vodnikov svetlobe ima te`ave zaradi kondenzirane vlage znotraj cevi, kar je stranski u~inek nevodotesne
izvedbe ali pa~ samo nepravilne termi~ne analize. ^lanek je usmerjen na razli~ne vidike ra~unalni{ke simulacije pri na~rtovanju
sistema za prevajanje svetlobe.
Klju~ne besede: cevast vodnik svetlobe, osvetljenost, svetilnost, gradbena fizika, ra~unalni{ka simulacija
1 INTRODUCTION
Lighting design of indoor spaces is a rarely discussed
discipline of the building industry. More often then not,
it is in the shadows of the fields related to the thermo-
technical processes organised in a building,1 although the
light makes it possible for humans to see their
surroundings, since the visible part of optical radiation
(i.e., the visible light) causes certain photochemical
reactions in the eye balls. The information perceived by
the eye balls is then processed in the brain.2 When more
light enters the eyes through the corneas we can
distinguish more details of the surroundings. This
property should be reflected on the daylighting design of
indoor spaces as well. Human beings spend about 80 %
of their lives indoors and it is necessary to think about
how natural light can enter a building. Usually architects
use one of the available direct daylighting systems, like:
• windows and their variants,
• roof lights.
However, as we need to make buildings more
compact because of land prices or we use the under-
ground areas, architects are required to use indirect day-
lighting systems as well.
Representatives of these are optic-fibre systems
(widely applied in the USA, but not in the EU) and
tubular light guides.
Tubular light guides are a type of indirect daylighting
system. Their history can be traced back to ancient
Egypt, when the predecessors of light guides were used
to illuminate sanctuaries.3 Only after their reinvention in
the 19th century and their further development that
Materiali in tehnologije / Materials and technology 50 (2016) 3, 409–412 409
UDK 628.981:628.9.021:004.94 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)409(2016)
started at the end of the 20th century, they become what
they are today.4 Each light guide consists of at least three
elements:
• the copula (the external cover made from a material
with high light transmission),
• the tube (the pipe with a special surface finish that
has a high light-reflectance value),
• the diffuser (the internal finish of light guides
distributing the light).
Tubular light guides can differ in every aspect. They
can have different lengths and diameters of the tube, and
also different copulas and diffusers. Howbeit, once the
elements ordered are assembled on a construction site
and later built into the building, some issues might arise
because of the unexpected physical properties of the
building. One of these may be associated with the day-
lighting of the building since light guides are difficult to
design. The other difficulty might be a result of the
thermal design, as water vapour often condenses inside
the pipe.
The paper is focused on the available computer tools,
which can be used to monitor the resulting properties of
the light guides within the selected test case.
2 EXPERIMENTAL SECTION
Indoor spaces in the centres of buildings or under-
ground areas can be illuminated with indirect daylighting
systems using natural light, or with luminaries. Tubular
light guides already proved to be efficient in this respect,
at least during daytime. The first half of this section is
focused on this particular topic.
The second half of the experimental section deals
with the thermo-technical simulations of the chosen
tubular-light-guide system, since it passes through the
envelope, especially through the roof.
The dimensions of the room assessed were 7 m ×
2.5 m and the room represented a corridor. The tubular
light guides were located along the longitudinal axis of
the space. The plan of the room including the light-guide
positions is shown in Figure 1.
2.1 Daylighting design
One of the most useful computer simulation tools is
HOLIGILM.5 It was developed to assess the amount of
the visible light transmitted by straight tubular light
guides. It is simple to use, fast and it provides a graphical
output. Nonetheless, it cannot evaluate the internally
reflected component of daylight in the same way as
Radiance or Velux Daylight Visualizer.
Although all 15 sky-type definitions described by
CIE6 are implemented in HOLIGILM, considering the
local codes of European countries including the Czech
Republic, the simulations were done only for the CIE
overcast-sky conditions.7
The properties of the tubular light guides input to
HOLIGILM were as follows:
• transparency of the copula: tc = 95 %,
• surface reflectance of the pipe (Spectralight Infinity
compound): rp = 99.7 %,
• transparency of the thermally insulating element
dividing the pipe into two segments: tte = 95 %,
• transparency of the diffuser (to be on the safe side of
the calculations): td = 75 %.
Seeing that HOLIGILM does not have any input
fields for additional light-transmitting parts like the
thermally insulating element, the total light transmittance
of the diffuser had to include that value as well; hence,
this value was set to 71 % instead of 75 %.
The total length l of the tubular light-guide pipe is
4.5 m and its diameter d is 0.530 mm.
Since the space is used as a corridor two distances
(l1) were set between the diffuser and the working plane:
• 2.15 m corresponding to the distance to the base
location of the working plane at 0.85 m above the
floor since in each room there may be some furniture;
• 3.0 m at the floor level so that the corridor is suitable
for movements.
2.2 Thermo-technical design
Thermo-technical calculations were carried out in
Agros2D. Agros2D is a free multi-platform alternative to
ANSYS Workbench and COMSOL Multiphysics.
The aims were:
• to find out whether water vapour can condense within
the pipe,
• to determine the point-wise thermal conductivity of
the structure.8
The thermal-conductivity coefficients of the structu-
ral elements were set as follows:
• acryl: la = 0.2 W m–2 K,
• the thermal insulation of mineral wool around the
pipe: lti1 = 0.040 W m–2 K,
• the thermal insulation of mineral wool as part of the
roof: lti2 = 0.040 W m–2 K,
• the pipe: lp = 0.220 W m–2 K,
• steel: ls = 50.0 W m–2 K,
• the lambda value of air depends on its volume.
F. VAJKAY et al.: ASSESSMENT OF TUBULAR LIGHT GUIDES WITH RESPECT TO BUILDING PHYSICS
410 Materiali in tehnologije / Materials and technology 50 (2016) 3, 409–412
Figure 1: Plan of the corridor including the positions of designed
tubular light guides
Slika 1: Na~rt hodnika, ki vklju~uje polo`aj na~rtovanih cevastih vod-
nikov svetlobe
3 RESULTS
3.1 Daylighting design
It was assumed that the luminous efficacy of a tubular
light guide of this length under the CIE Overcast Sky
conditions would prove to be insufficient to illuminate
the working plane at 0.85 m above the floor level. The
resulting illuminance levels were rather low. The values
varied from 0 lx to 100 lx (equivalent to 0 % and 1 % of
daylight factor, Figure 2). With an additional light guide,
it was possible to shift these values. For example, the
minimum value grew by 40 lx (Figure 3).
Seeing that the design only included the sky compo-
nent of the daylight factor and that the light-trans-
mittance value of the light guide was reduced by the
thermally insulating element, it can be concluded that the
values were extraordinarily high. The peaks beneath the
light guides, at a distance of 2.15 m, increased by 10 lx,
from 100 lx to 110 lx.
Since illuminance decreases with the increasing dist-
ance, when considering the results described previously,
the evaluation at the floor level was done under the
assumption that two tubular light guides should be used
to illuminate the designed space.
As anticipated, the resulting light levels dropped. The
peak values beneath the light guides receded to 58 lx
(Figure 4).
3.2 Thermo-technical simulations
The results presented on Figure 5 were obtained with
the simulations provided under common boundary con-
ditions. External and indoor air temperatures were set to
–15 °C and 20 °C and the corresponding relative air
humidity inputs were 84 % for the exterior and 50 % for
the interior.
As can be seen on Figure 5, the contact surface
temperature on the interface of the light guide and the
roof varies between 16 °C and 20 °C.
Other results can be connected to the overall thermal
conductivity of the light pipe, which is evaluated just like
in the case of windows.
The simulations pointed out that the U point-wise
thermal conductivity of the structure is 0.37 W K–1. This
value is smaller than the one required by the Czech
legislation; therefore, the light guide meets the necessary
conditions.
F. VAJKAY et al.: ASSESSMENT OF TUBULAR LIGHT GUIDES WITH RESPECT TO BUILDING PHYSICS
Materiali in tehnologije / Materials and technology 50 (2016) 3, 409–412 411
Figure 4: Illuminance-contour plot at 3.0 m beneath the diffuser for
two tubular light guides
Slika 4: Diagram obrisa osetljenosti pri 3 m pod difuzorjem pri dveh
cevastih vodnikih svetlobe
Figure 5: Temperature profile of the light guide
Slika 5: Temperaturni profil vodnika svetlobe
Figure 3: Illuminance-contour plot at 2.15 m beneath the diffuser for
two tubular light guides
Slika 3: Diagram obrisa osvetljenosti pri 2,15 m pod difuzorjem pri
dveh cevastih vodnikih svetlobe
Figure 2: Illuminance-contour plot at 2.15 m beneath the diffuser for
one tubular light guide
Slika 2: Diagram obrisa osvetljenosti pri 2,15 m pod difuzorjem pri
enem cevastem vodniku svetlobe
4 DISCUSSION
Tubular light guides are a type of daylighting system,
used to illuminate a room or a bigger space inside a
building or construction, in which natural light is hardly
available. However, an emphasis should be put onto the
ways a light guide can be designed with respect to
daylighting, thermo-technical requirements and other
factors.
The above results for daylighting were determined
solely with HOLIGILM. The software can draw lumi-
nance maps for the diffuser and determine the total
luminous flux beneath the optical interface. It is useful
for simulations when a tubular light guide is regarded as
a luminary, practically artificial light source. For the
artificial-lighting design, luminaries are represented as
IES data files. Light guides can also be designed in this
way instead of using the approach based on a natural-
light source. This approach is useful because HOLI-
GILM cannot determine the internally reflected compo-
nent of the daylight factor, but only the sky component.
Another disadvantage of HOLIGILM is that the room
used can only have a rectangular shape. On the contrary,
when the internally reflected component of the daylight
factor is neglected, the results may be obtained for rooms
of any shape. However, simulations for different sky
types created by the software tools using one of the
global illumination models can be important in future
since the use of photorealistic rendering slowly but
steadily rises year by year.
As for the thermo-technical design of light guides, it
can be said that they lack one crucial parameter. That is
the motion of the air inside and outside the pipe. The
pipe is commonly manufactured from plain metal sheets,
thus, when put together, the joints are not sealed by
sealants. Therefore, at some places, there can be leakage
of air, with a possibility of water vapour entering the
light-guiding system. This water can condense inside the
pipe at any time, especially in winter. The condensed
water may damage the surrounding structures. Also, the
fact that the thermal conductivity of the resulting struc-
ture can be evaluated just because it goes through the
envelope of a building leaves some space for improve-
ments. Another approach similar to that of the determi-
nation of the thermal conductivity may or may not be
correct and should be researched in detail via long-term
monitoring, not just by the means of computer simu-
lations.
5 CONCLUSION
The design and application of tubular light-guiding
systems have a flaw: most of their parameters regarding
building physics are determined solely by computer
simulations, which are not required to correspond to
reality.
This could be changed by the means of experimental
activities, which are planned within the Faculty of Civil
Engineering. Through a comprehensive long-term moni-
toring, several characteristics could be observed, such as
the movement of the air in the pipe, temperature rises
and the effects of air humidity in the light guides.
Acknowledgement
This paper was prepared within project
No. LO1408 "AdMaS UP – Advanced Materials, Struc-
tures and Technologies", supported by the Ministry of
Education, Youth and Sports under National Sustain-
ability Programme I.
6 REFERENCES
1 P. Charvat, T. Mauder, M. Ostry, Simulation of Latent-Heat Thermal
Storage Integrated with Room Structures, Mater. Tehnol., 46 (2012),
239–242
2 P. Boyce, P. Raynham, SLL Lighting Handbook, CIBSE, London
2009
3 F. Moore, Concepts and Practice for Architectural Daylighting, Van
Nostrand Reinhold, New York 1985, 290
4 J. Mohelnikova, F. Vajkay, Study of Tubular Light Guides Illumi-
nance Simulations, LEUKOS, 5 (2009), 267–277, doi:10.1582/
LEUKOS.2008.05.04.001
5 M. Kocifaj, S. Darula, R. Kittler, HOLIGILM: Hollow light guide
interior illumination method – An analytic calculation approach for
cylindrical light-tubes, Solar Energy, 82 (2008) 3, 247–259,
doi:10.1016/j.solener.2007.07.003
6 Commission Internationale de l’Eclairage, CIE DS 011.2/E: 2002
Spatial distribution of daylight – CIE standard general sky, Vienna:
Commission Internationale de l’Eclairage, 2002
7 Czech Standardization Institute, ^SN 73 0580 Daylighting in
buildings – Part 1: Basic Requirements, Prague, Czech Standardi-
zation Institute, 2007
8 Czech Standardization Institute, ^SN 73 0540 Thermal protection of
buildings – Parts 1,2,3,4 Prague, Czech Standardization Institute,
2005 and 2011
F. VAJKAY et al.: ASSESSMENT OF TUBULAR LIGHT GUIDES WITH RESPECT TO BUILDING PHYSICS
412 Materiali in tehnologije / Materials and technology 50 (2016) 3, 409–412
T. FÍLA et al.: CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE
413–417
CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE
VEDENJE KRATKIH VLAKEN C/PPS KOMPOZITOV PRI LEZENJU
Tomá{ Fíla1, Petr Koudelka2, Daniel Kytýø2, Jiøí Hos1, Jan [leichrt1
1Czech Technical University in Prague, Faculty of Transportation Sciences, Konviktská 20, 110 00 Prague 1, Czech Republic
2Institute of Theoretical and Applied Mechanics, Academy of Sciences of the Czech Republic, Prosecká 76, 190 00 Prague, Czech Republic
xfila@fd.cvut.cz
Prejem rokopisa – received: 2014-08-22; sprejem za objavo – accepted for publication: 2015-04-24
doi:10.17222/mit.2014.208
Composite materials with a polymeric matrix reinforced by carbon fibres are nowadays widely used as high-tech structural
materials with excellent mechanical properties (particularly their stiffness and strength). The application of this type of
composite to structural parts exposed to thermal loading has recently been proposed. Such an application requires an
investigation and analysis of the mechanical behaviour under long-term exposure to simultaneous thermal and mechanical
loading. In this paper the measurements and results of the creep behaviour of a composite with a polyphenylene sulphide matrix
reinforced with chopped poly-acrylonitrile carbon fibres (C/PPS) are presented. The measured compound is proposed for use as
a structural material for a jet-engine frame in the aerospace industry and the internal parts of aircraft with possible thermal
loading. A custom experimental device designed for the creep measurements of composite materials was used for measurements
of the developing strain at a constant tensile stress and temperature. Short-term creep tests with continuous strain monitoring
were performed at a constant stress level at several elevated temperatures below and above the glass-transition temperature of
the matrix. The strain was measured using the digital image correlation (DIC) method. The measured data were processed to
find the strain-to-time dependency and the creep-compliance-to-time dependency. The creep-compliance-to-time data were also
fitted using Findley’s creep law for polymers to evaluate the model parameters and to analyse the applicability of the model for
a PPS polymer reinforced with chopped carbon fibres.
Keywords: creep, short fibre composite, C/PPS, Findley’s model, DIC
Kompozitni materiali s polimerno osnovo, okrepljeno z ogljikovimi vlakni, se dandanes pogosto uporabljajo kot
visokotehnolo{ki konstrukcijski materiali z izjemnimi mehanskimi lastnostmi, {e posebej to velja za njihovo togost in trdnost.
Pred kratkim je bila predlagana uporaba kompozita te vrste za konstrukcijske dele, ki so izpostavljeni toplotni obremenitvi.
Tak{na uporaba zahteva raziskavo in analizo mehanskega vedenja pri dolgotrajni izpostavljenosti isto~asni toplotni in mehanski
obremenitvi. Prispevek predstavlja meritve in rezultate deformacijskega vedenja kompozitov z matrico iz polifenilen sulfida,
okrepljenega z razcepljenimi poliakrilonitrilnimi ogljikovimi vlakni (C/PPS), ki naj bi se uporabljali kot konstrukcijski materiali
za ogrodje reaktivnih motorjev v letalski in vesoljski industriji ter za notranje dele letal, ki so pod potencialno toplotno
obremenitvijo. Za merjenje napredovane deformacije pri konstantni natezni obremenitvi in temperaturni obremenitvi, je bila
uporabljena posebna preizkusna naprava, izdelana za merjenje deformacije kompozitov. C/PPS vzorci so bili postavljeni v
toplotno komoro in segreti. Vzorci so bili nato s pomo~jo stiskalnice izpostavljeni konstantni natezni sili. Kratkotrajni preizkusi
lezenja, z nadzorovanjem neprekinjene natezne sile, so bili opravljeni pri konstantnih stopnjah obremenitve pri razli~nih
temperaturah, ki so bile vi{je ali ni`je od temperature prehoda v steklasto stanje osnove. Deformacija vzorca je bila izmerjena z
uporabo metode korelacije digitalne slike (DIC). Namen izmerjenih podatkov je bil poiskati odvisnosti deformacije od ~asovne
komponente in lezenja. Podatki o ~asu in sili lezenja so bili usklajeni s Findleyevim zakonom lezenja za polimere za oceno
parametrov modela in analizo veljavnosti modela za PPS polimere, ki so okrepljeni razceplenimi ogljikovimi vlakni.
Klju~ne besede: lezenje, kompozit s kratkimi vlakni, C/PPS, Findleyev model, DIC
1 INTRODUCTION
Composite materials with various matrices reinforced
by short or continuous fibres are nowadays widely used
as general structural materials with good mechanical
properties (particularly their stiffness and strength). One
of the very popular composite compounds that are
widely used in the aviation industry consists of a poly-
phenylene sulphide (PPS) matrix and carbon-fibre rein-
forcement. This material, in continuous fibre form, is
typically used in the production of external aircraft com-
ponents because of its excellent mechanical properties,
chemical resistance to aerospace fluids and low density.1
Composite compounds with carbon fibres and PPS
matrices (C/PPS) also have a good resistance to heat and
flames. Therefore, they are able to meet the smoke and
toxicity requirements of aviation legislation1 and recently
they have been used for the construction of aircraft’s
interior parts, particularly seats.1,2
Attempts to simplify the fabrication process for air-
craft parts and to reach a higher cost efficiency in terms
of production have led to the introduction of C/PPS in
the form of a material reinforced by chopped or short
carbon fibres.3 The parts used in the aviation industry,
e.g., seats, manufactured using a composite with a short
reinforcement can be easily fabricated using injection
moulding and their shape can be very complex and
thick.3 However, such composites consist of unit cells
with discrete boundaries, various fibre arrangements and
distributions. Thus, these materials exhibit significantly
lower strengths and are in comparison with continuous-
fibre reinforcements susceptible to the occurrence of
creep.4 For this reason the creep behaviour of short-fibre
composites has to be properly described and taken into
Materiali in tehnologije / Materials and technology 50 (2016) 3, 413–417 413
UDK 620.1:539.376:669.018.25 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)413(2016)
account as one of the most important parameters for an
estimation of a part’s lifetime.
In this study a series of isothermal creep measure-
ments was performed with the composite compound
consisting of a PPS matrix reinforced with short carbon
fibres fabricated using an untested method involving
heating cables (details of the manufacturing method are
classified, and thus only the main features of the method
will be described in the text). The tested material is
planned to be used for the construction of monocoque
shell parts of the fuselage and for structural parts with
large volumes (such as engine suspension, etc.) and thin
reinforcements with possible thermal loading.
2 EXPERIMENTAL PART
Material samples with a dog-bone shape were tested
in a custom-designed device for uniaxial creep tests at a
constant stress level. The creep was measured on a set of
samples at various temperatures above and below the
glass-transition temperature of the matrix. The strain was
observed using a digital single-lens-reflex (DSLR) came-
ra and evaluated using the digital image correlation tech-
nique (DIC). As a result, the strain-time and the creep-
compliance-time dependencies were evaluated and the
creep behaviour of the tested compound was described.
2.1 Material
The material samples were manufactured as a new
type of cost-efficient material. The composite compound
(Carbon AS4/PPS, TenCate Advanced Composites) con-
sists of a polymeric polyphenylene sulphide (PPS)
matrix and chopped carbon fibres with a length of appro-
ximately 10 mm. The material was fabricated as a
2.5-mm-thick sheet using the technique of mixing
chopped fibres with PPS particles and baking in a mould
with heating cables at 310–340 °C under 4 MPa
pressure. The outer parts of the sheet were manufactured
with an 8 mm thickness for the clamping of the samples
into sample holders. Part of the sheet is depicted in
Figure 1.
2.2 Specimen preparation
The dog-bone-shaped specimens were cut from a
sheet of material using a water-jet cutting machine. The
middle part of the specimen was prepared with a width
of 10 mm. The overall length of the specimen was 160
mm, the maximum width was 25 mm and the thickness
of the middle part was 2.5 mm. The maximum distortion
caused by the water-jet cutting was experimentally deter-
mined to be better than 0.15 mm (in all dimensions on
both surfaces of the specimen). Such a manufacturing
tolerance was considered to be sufficient for the experi-
ment and the surfaces were not additionally modified.
All the specimens were measured before the test and
particular dimensions were used for the evaluation of the
experiment. The dimensions were selected to be similar
to the dimensions of small longitudinal reinforcements
used in, e.g., aircraft seats. The specimen surface was
sprayed using an airbrush with a granite effect for better
pixel identification in the DIC technique. A specimen
with a granite coating is displayed in Figure 2.
2.3 Experimental setup
The experimental setup consists of a custom-de-
signed experimental device for the creep measurements
and a DSLR camera with accessories.
The creep experimental device was designed for uni-
axial creep testing at a constant stress level. The device
consists of a heating chamber SFL 3119 (Instron, USA)
with a temperature range of –70 °C to 350 °C, a rigid
structural steel frame, two independent aluminium alloy
lever arms with ratio of 1:10 and set of dead weights.
The device is equipped with two load cells (VTS Zlín,
Czech Republic) with a loading capacity up to 10 kN.
The strain can be measured in two possible ways: exten-
someters or DIC. Two custom-designed extensometers
(VTS Zlín, Czech Republic) with a maximum measur-
414 Materiali in tehnologije / Materials and technology 50 (2016) 3, 413–417
T. FÍLA et al.: CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE
Figure 2: Image of a specimen with sprayed granite coating on its
surface
Slika 2: Prikaz vzorca s prevleko iz granita na povr{ini
Figure 1: Image of a C/PPS sheet used to prepare a specimen
Slika 1: Posnetek C/PPS plo~evine, ki je bila uporabljena za pripravo
vzorca
able extension of 10 mm and a heat resistance up to
150 °C allow the accommodation of two samples in the
device’s heating chamber and their independent loading.
In contrast the DIC allows the measurement of only a
single specimen, but it is not limited by the maximum
temperature because all of the measuring equipment is
situated outside the heating chamber. The experimental
device is shown in Figure 3.
The DIC was used as a tool for an evaluation of the
strain held in the experiments. The specimen in the
heating chamber was observed with a DSLR camera
EOS 550D (Canon, Japan) with a macro-objective
EF100mm/1:2.8L Macro (Canon, Japan) situated on a
tripod in front of the heating chamber. The sample
surface was illuminated with a laboratory LED light
source KL 2500 (Shott, Germany). The overall view of
the experimental setup is shown in Figure 4.
2.4 Experiment
Prior to the experiment the specimen was put in the
heating chamber and the positions of the illuminators
and the camera stand were adjusted. The reference image
of the sample surface in the heating chamber was taken
to verify the focalization, the field of view and to detect
and eliminate possible reflections of illuminating light
on the chamber window. The unloaded specimen was
then heated to a given temperature and for a 30-min
period was held to reach a uniform temperature distri-
bution throughout all the heated parts. After that the
image-capturing sequence was started and the specimen
was loaded with a constant tensile force of 1.5 kN (one
measurement with a loading force of 2.5 kN was also
performed). The loading force was selected according to
data obtained during quasi-static measurements in
tension of specimens with an identical shape. A value of
1.5 kN was situated in a linear elastic region of the
material and represented approximately 70 % of the
material’s yield strength. The creep test was continued
until specimen rupture or for at least 20 h. The glass-
transition temperature of the matrix material was
according to the manufacturer’s datasheet approximately
85 °C and therefore the creep tests were performed at
temperatures of 60 °C (certainly below the glass-transi-
tion temperature), 90 °C (slightly above the glass-transi-
tion temperature), 110 °C, 130 °C and 140 °C (signifi-
cantly above the glass-transition temperature). Images
were taken at equidistant time intervals and were
labelled with a unique timestamp for precise synchroni-
zation of the DIC and the experiment time. The strain
was evaluated from the image sequence using custom
DIC software5 based on the Lucas-Kanade algorithm.6
3 RESULTS AND DISCUSSION
The evaluated strain-time dependencies are shown in
Figure 5. The creep-compliance-time dependencies were
calculated from the data using the following Equation
(1):7,8
J t
t
c
c
c
( )
( )
=



(1)
where Jc(t) represents the creep compliance in time t,

c(t) is the actual creep strain in time t (excluding the
initial strain) and c is the applied constant tensile
stress. A graph of the creep-compliance-time depen-
dency is shown in Figure 6.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 413–417 415
T. FÍLA et al.: CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE
Figure 4: Image of the experimental setup and a specimen mounted in
the thermal chamber
Slika 4: Prikaz preizkusnega sklopa in vzorca, ki je bil name{~en v
toplotni komori
Figure 3: Visualization of the custom-designed device for the creep
measurements
Slika 3: Prikaz posebne naprave za merjenje deformacije
The creep-compliance data were also fitted using
Findley’s creep law for polymers, assuming steady-state
creep behaviour according to the Equation (2):8
J t b tc
b( ) = 0 1 (2)
where J tc ( ) represents the predicted creep compliance in
time t (in seconds) and b0, b1 are material parameters
evaluated by a regression of the experimental data. The
evaluated regression curves are shown in Figure 7 and
the calculated material parameters are summarized in
Table 1.
Table 1: Evaluated parameters of Findley’s creep model
Tabela 1: Ocenjeni parametri Findleyevega modela lezenja
Temperature Loading
b0 b1(°C) (kN)
60 1.5 4.432×10–10 0.9386
90 1.5 2.286×10–9 0.6818
110 1.5 2.419×10–7 0.5711
130 1.5 1.271×10–7 0.6712
140 1.5 1.078×10–6 0.3475
140 2.5 1.832×10–7 0.7044
Based on the experimental data it can be stated that
the creep behaviour of the tested material was brittle
above the glass-transition temperature. The test of each
specimen above this temperature ended with a sudden
specimen rupture after a short time period (max. approx.
6 h).
Creep under (test at 60 °C) and slightly above (test at
90 °C) the glass-transition temperature can be considered
immeasurable because the measured maximum displace-
ments of approximately 10 μm correspond only to
double the pixel-size of the DIC image (approx. 5 μm)
and thus subject to the noise and reliability of the
method.
The majority of the measured dependencies agreed
with the standard polymer creep theory and their creep
strain-rate increased with increasing temperature or
loading. A regression analysis showed a good correlation
of the experimental data and Findley’s creep model for
polymers (Figure 7). The fitted model’s parameters have
different values than usual for clear polymers8–10 because
of the brittle behaviour of the material and the short
times to specimen failure. Still, Findley’s creep model
can be considered as suitable for a description of the
creep behaviour for the used C/PPS compound.
However, disagreement with the basic theory can be
seen in the test with a loading of 1.5 kN at a temperature
of 140 °C that exhibited a lower strain rate than ex-
pected. The strain rate at the beginning of the test was
appropriate for the temperature, but at a strain of appro-
ximately 1×10–3 a sudden decrease was observed and the
strain remained low until specimen rupture (Figure 5).
Moreover, several specimens ruptured during the initial
loading at forces even 3 times lower than the testing
force of 1.5 kN and are not presented in the results. This
discrepancy was caused by significant imperfections
detected in the material’s microstructure (missing matrix
binder, etc.) caused by an imperfect fabrication process
and size effect due to the combination of a random
material structure and the small width of the specimen
(10 mm). Hence, the delivered material was considered
to be unsuitable for parts with small dimensions in terms
of structural and thermal loading.
416 Materiali in tehnologije / Materials and technology 50 (2016) 3, 413–417
T. FÍLA et al.: CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE
Figure 7: Graph of fits of Findley’s creep model on the experimental
data
Slika 7: Prikaz prilagoditev Findleyevega modela lezenja na podatke
pri preizkusu
Figure 6: Graph of measured creep-compliance-time dependencies
Slika 6: Diagram odvisnosti izmerjenega lezenja od ~asa
Figure 5: Graph of measured strain-time dependencies
Slika 5: Diagram odvisnosti izmerjenih obremenitev in ~asa
4 CONCLUSIONS
A set of isothermal creep tests with constant loading
of the C/PPS composite with chopped fibres was per-
formed. The C/PPS material was fabricated with an
innovative technique of baking in a mould using heating
cables. The experiments were performed at several diffe-
rent temperatures below and above the glass-transition
temperature and the strain was evaluated using the DIC
technique. The experimental results showed a good
correlation with the standard creep theory for polymers
and with Findley’s creep law. However, the used
fabrication technology brought several imperfections to
the material’s structure, which together with the size
effect caused by small dimensions of the specimen, led
to a discrepancy in the results. Thus, the tested material
was considered unsuitable for small structural parts
exposed to structural and thermal loading and so further
improvements to the fabrication process are required.
Acknowledgements
The research was supported by The Technology
Agency of Czech Republic (grant No. TA03010209), by
the Grant Agency of the Czech Technical University in
Prague (grant No. SGS15/225/OHK2/3T/16) and by
RVO: 68378297. All the support is gratefully acknow-
ledged.
5 REFERENCES
1 M. Favarolo, Evaluation of a low cost thermoplastic composite for
aircraft interior applications, Proceedings of SAMPE 2010
Conference, Seattle, USA 2010
2 Composites becoming the standard in aircraft seats, JEC Composite
Magazine, 43 (2006) 23, 34
3 R. Bockstedt, J. Sajna, Low cost composites based on long carbon
fiber thermoplastics, Proceedings of SAMPE 1993 Conference,
Anaheim, USA, 1993
4 V. Monfared, M. Mondali, A. Abedian, Steady state creep behaviour
of short fiber composites by mapping logarithmic functions (MF)
and dimensionless parameter (DP) techniques, Archives of Civil and
Mechanical Engineering, 12 (2012), 455–463, doi:10.1016/j.acme.
2012.08.001
5 I. Jandejsek, J. Valach, D. Vavøik, Optimization and Calibration of
Digital Image Correlation Method, Proceedings EAN 2010, Czech
Republic, 2010, 121–126, doi:http://hdl.handle.net/11104/0190791
6 B. Lucas, T. Kanade, An iterative image registration technique with
an application to stereo vision, Proceedings of IUW, 1981, 121–130
7 W. Callister, Fundamentals of Material Science, John Wiley, 2001
8 S. Vaitkus, I. Gnip, V. Kersulis, S. Vejelis, Prediction of Creep Strain
of the Expanded Polystyrene (EPS) in Long-term Compression,
Materials Science, 13 (2007) 4, 293–296
9 Y. Guo, R. Bradshaw, Isothermal physical aging characterization of
Polyether-ether-ketone (PEEK) and Polyphenylene sulfide (PPS)
films by creep and stress relaxation, Mech Time-Depend Mater, 11
(2007) 1, 61–89, doi:10.1007/s11043-007-9032-7
10 Y. Guo, R. Bradshaw, Long-term creep of polyphenylene sulfide
(PPS) subjected to complex thermal histories: The effect of noniso-
thermal physical aging, Polymer, 50 (2009) 16, 4048–4055,
doi:10.1016/j.polymer.2009.06.046
Materiali in tehnologije / Materials and technology 50 (2016) 3, 413–417 417
T. FÍLA et al.: CREEP BEHAVIOUR OF A SHORT-FIBRE C/PPS COMPOSITE

V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
419–426
INCREASING MICRO-PURITY AND DETERMINING THE
EFFECTS OF THE PRODUCTION WITH AND WITHOUT
VACUUM REFINING ON THE QUALITATIVE PARAMETERS OF
FORGED-STEEL PIECES WITH A HIGH ALUMINIUM CONTENT
POVE^ANJE MIKRO^ISTO^E IN DOLO^ITEV U^INKA
PROIZVODNJE, Z VAKUUMSKIM RAFINIRANJEM ALI BREZ, NA
KVALITATIVNE PARAMETRE KOVANEGA JEKLA Z VISOKO
VSEBNOSTJO ALUMINIJA
Vladislav Kurka1, Jaroslav Pindor1, Jana Kosòovská1, Zdenìk Adolf2
1Material and metallurgical research, Ltd, Pohranicni 693/31, 703 00 Ostrava-Vítkovice, Czech Republic
2V[B – Technical University of Ostrava, 17. listopadu 15/2172, 708 33 Ostrava-Poruba, Czech Republic
vladislav.kurka@mmvyzkum.cz
Prejem rokopisa – received: 2014-10-13; sprejem za objavo – accepted for publication: 2015-03-24
doi:10.17222/mit.2014.258
The quality production technology for the WNr. 1.8504 steel was developed. The aim of the work was to achieve the required
internal micro-purity and determine the effects of different production technologies on the qualitative parameters of forged-steel
pieces. Firstly, polygonal ingots weighing 1600 kg were produced, using the metallurgical units, in a controlled-atmosphere
induction-melting furnace (IF) without vacuum treatment, and in a vacuum and pressurized induction-melting furnace (VPIM)
with vacuum treatment. The ingots were subsequently reshaped by open-die forging into bars with a rectangular cross-section.
The effect of the ingot-production technology was evaluated by comparing the forged-steel pieces in terms of their purity,
macrostructure and microstructure.
Keywords: vacuum, inclusion, aluminium, steel
Izvr{en je bil razvoj kakovostne proizvodne tehnologije jekla W.Nr. 1.8504. Namen je bil dose~i `eleno notranjo mikro~isto~o in
ugotoviti vpliv razli~nih tehnologij proizvodnje na kvalitativne parametre odkovkov. Najprej so bili izdelani poligonalni kovani
ingoti z maso 1600 kg, z uporabo naslednjih metalur{kih agregatov: v indukcijski talilni pe~i s kontrolirano atmosfero (IF) brez
vakuumskega rafiniranja in v vakuumski ter indukcijski talilni pe~i (VPIM) s povi{anim tlakom in z vakuumskim rafiniranjem.
Nato so bili ingoti s prostim kovanjem preoblikovani v palice s pravokotnim prerezom. Vpliv tehnologije proizvodnje ingotov je
bil ocenjen s primerjavo odkovkov z vidika ~istosti, makro in mikrostrukture.
Klju~ne besede: vakuum, vklju~ki, aluminij, jeklo
1 INTRODUCTION
Aluminium is primarily used in steel as a deoxidising
agent as well as an alloying element. In a melt, alumi-
nium occurs in the dissolved form, in a solid solution as
aluminium metal, aluminium oxide Al2O3 and, in an
interaction with nitrogen, also as aluminium nitride
AlN.1 An increased aluminium content in steel reduces
its formability due to the mechanical effect of its pre-
cipitates, and alternatively also due to its local ferrite
affecting the structural state. An increased concentration
of strongly ferrite generating aluminium may occur in
the vicinity of dissolved AlN particles. AlN is separated
in steel in the form of acicular crystals, usually on the
grain borders. The quality of WNr1.8504 (hereinafter
referred to as the "steel") is, with regard to its chemical
composition, intended for surface nitration. Nitration is a
saturation of the steel surface with nitrogen that creates
hard nitrides with the alloying elements Al, Cr, Ti and V.
Figure 1 shows an example of a nitrated steel layer. The
process of nitration takes place at temperatures of 500 °C
– 540 °C for about 50 h,2 when the nitrated steel layer
achieves a 0.3 mm thickness in 30 h, and a 0.5 mm thick-
ness in 50 h. The process depends on the temperature,
pressure, chemical composition of the steel and atmo-
spheric composition.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426 419
UDK 669.18:669.178.52:621.77 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)419(2016)
Figure 1: Example of a nitrated steel layer2
Slika 1: Primer nitrirane plasti na jeklu2
Nitrogen is thus an element that creates AlN nitrides
with Al in steel, which are very hard and non-malleable.
On the steel surface, it creates a hard, abrasion-resistant
area. AlN nitrades inside the steel are non-malleable
inclusions that impair the micro-purity of the steel. This
work is concerned with the elimination of AlN inclu-
sions from the WNr1.8504 quality steel with 0.8 % – 1.1
% of mass fractions of high Al.
An important step in the elimination of the AlN-type
inclusions with a high aluminium content from steel is
the reduction of the nitrogen content to the minimum
level.
1.1 Nitrogen in steel
Nitrogen in steel not only reacts with iron but also
with other dissolved elements, forming a wide variety of
compounds. These are dominated by nitrides, but carbo-
nitrides, oxynitrides, cyanonitrides, complex binary
nitrides and other phases of variable compositions can
occur as well. Their existence depends on a number of
factors, such as the composition of the steel, the melting
method, the temperature, the pressure, the thermal treat-
ment, etc.4 In our case, aluminium nitrides form under
the liquidus temperature and re-dissolve in the steel at
temperatures of 900 °C – 1100 °C, where the dissolution
rate is a function of the material temperature and struc-
ture.5 Increased frequency and size of, for example, the
AlN particles may lead to the generation of inter-
crystalline steel fractures. The effects of the elements on
the solubility of nitrogen in molten iron at a temperature
of 1600 °C and a nitrogen pressure of 100 kPa are pre-
sented in Figure 2.4
Surface-active elements such as sulphur and oxygen
hinder the removal of nitrogen from the molten steel
during vacuuming (nitrogen diffusion in argon bubbles
during bubbling) and from the vacuum above the steel
surface. If reducing the nitrogen content, e.g., to a
required value below 0,004 % of mass fractions, very
low values of sulphur and oxygen need to be ensured at
the same time. See Figures 3 and 4 for a graphic illustra-
tion of this dependence.6
1.2 Solubility of nitrogen in steel
The solubility of nitrogen in steel and the effects of
individual elements are described in detail in a previous
paper.7 The transition of nitrogen in steel is governed by
Sievert’s law, which presupposes its atomic dissolution.
The dependence of the nitrogen content in an iron melt at
pressure is described with relationship (1):
[ ] { }%N Fe N
N
N
rel.
= ⋅
K
f
p
2
(1)
where:
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
420 Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426
Figure 4: Nitrogen removal by tank degassing as a function of sulphur
content6
Slika 4: Zmanj{anje vsebnosti du{ika pri odstranjevanju plina v ko-
mori, kot funkcija skupne vsebnosti `vepla6
Figure 2: Effects of elements on the solubility of nitrogen in molten
iron at a temperature of 1600 °C and a pressure of 100 kPa4
Slika 2: Vpliv elementov na topnost du{ika v staljenem `elezu pri
1600 °C in tlaku 100 kPa4
Figure 3: Nitrogen removal by tank degassing as a function of the
total oxygen content6
Slika 3: Zmanj{anje vsebnosti du{ika pri odstranjevanju plina v ko-
mori, kot funkcija skupne vsebnosti kisika6
KN is the equilibrium constant of the dissolution process
(in mass fractions, (w/%)) maximum,
fN is the nitrogen activity coefficient in iron melt,1
{ }pN
rel.2
is the relative partial nitrogen pressure above
iron melt.1
The equilibrium constant KN expresses the nitrogen
solubility in iron under standard conditions, i.e., its
maximum content under a pressure of 0.1 MPa{ }pN
rel.2
= 1 a fN = 1. The temperature dependence of nitrogen
solubility is expressed with Equation (2):
[ ]lg . lg %K
TN Fe
N= − − =
188
1246 (2)
and the adequate dependence of the reaction-free en-
thalpy on the temperature is expressed with relationship,
Equation (3):
ΔG T0 3600 2386= + . (J) (3)
As it is clear from Equation (2), the solubility of ni-
trogen at a temperature of 1600 °C is 0.045 % of mass
fractions, but it drops significantly when the melt
solidifies. It slightly increases in iron 
, and then drops in
iron  to approximately 0.0015 % of mass fractions at
600 °C.
The nitrogen solubility in steel is significantly
affected by the presence of alloying elements, parti-
cularly in highly alloyed corrosion-resistant steels. The
effect of alloying elements is manifested in the value of
active coefficient fN:
[ ]lg %f eN N
X N= ∑ (4)
This effect can be expressed by means of interaction
coefficients eN(1873K)
X . The temperature dependence of the
interaction coefficients expressing the effects of elements
on the nitrogen activity was described by Chipman and,
according to the author8, it is expressed with Equation
(5):
e
T
eTN( , K)
X
N(1873K)
X= −⎛⎝
⎜ ⎞
⎠
⎟ ⋅
3280
0 75. (5)
Therefore, the dependence of the nitrogen solubility
on the temperature can be expressed with Equations (6)
and (7):
[ ] { }lg % lg lg lgN steel N N N
rel.
= − +K f p
1
2 2
(6)
[ ]lg % .
.
N steel
N(
= − −⎛⎝
⎜ ⎞
⎠
⎟ −
− −⎛⎝
⎜ ⎞
⎠
⎟ ⋅∑
188
1246
3280
0 75
T
T
e { }1873K)X N
rel.
+
1
2 2
lg p
(7)
An improvement of the calculation is, especially for
highly alloyed steels (e.g. CrNi steels) conditioned not
only by the knowledge of the first values of the inte-
raction coefficients but also of the second values and the
cross-interaction coefficients, Equation (8):
[ ] [ ] [ ]
[ ] [ ] [ ]
lg % lg % %
% % %
N N X
X X Y
steel Fe N
X
N
X 2
N
X, Y
= − ⋅ −
− ⋅ −− ⋅ ⋅
∑
∑ ∑
e
r r
2
(8)
For significantly corrosion-resistant steel alloys,
these values are quoted by, e.g., Z. Buzek8 in Table 1.
Table 1: Values of interaction coefficients 1, 2, and cross-interaction
coefficients8
Tabela 1: Vrednosti interakcijskih koeficientov 1, 2 in navzkri`nih
koeficientov8
X (in mass
fractions, (w/%))
eN(1873K)
X
rN(1873K)
X rN(1873K)
X, Y
Cr –0.0468 +0.00034 –
Nb –0.0667 +0.00019 +0.00136 (Cr-Nb)
Mo –0.0106 – +0.00002 (Cr-Mo)
Ni +0.0107 – –0.00041 (Cr-Ni)
Si +0.047 – –0.00149 (Cr-Si)
1.3 Elimination of AlN from WNr1.8504 quality steel
Elimination of the AlN inclusions from steel
commenced with the evaluation of standardly produced
steel with production-technology adjustments. The ob-
jective of the work was a reduction of the nitrogen con-
tent and thus the occurrence of AlN in the final product.
In MMR, ingots were produced in an atmospheric
induction-melting furnace (hereinafter referred to as the
IF) with a nominal batch weight of 1750 kg, and in a
vacuum and pressurized induction-melting furnace
(hereinafter referred to as the VPIM), in which vacuum
degassing (VD) at a minimum pressure of 40 Pa (a), or
vacuum oxygen decarburisation (VOD) can be carried
out by using of an oxygen-argon nozzle.
One polygonal ingot for forging, V2A, was produced
from each melt, weighing approximately 1650 kg. With
every melt, the ingot was filled from the bottom through
the casting system. The ingots were forged by open-die
forging into bars of the following dimensions: 140–160
mm × 90–110 mm.
The melts were produced and found as follows:
Melt 1 – Production of the melt in the IF with casting on
an atmospheric casting bed under a protective argon
atmosphere.
Melt 2 – Production of the melt in the VPIM, vacuum
refined with VD, with casting on an atmospheric
casting bed under a protective argon atmosphere.
Melt 3 – Production of the melt in the VPIM, vacuum
refined with VD, with casting under a protective
argon atmosphere in a cofferdam.
The chemical composition of the steel according to
the standard3 and the chemical compositions of the
monitored and evaluated melts and forged pieces are
shown in Table 2. All the melts and forged pieces
featured the required standardised chemical composition.
The content of nitrogen in the forged piece from Melt 1
was 0,0132 % of mass fractions. This amount was
reduced to the value of 0.0108 % of mass fractions,
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426 421
through the VD process, in the pieces forged from Melts
2 and 3.
Elements of C, S, N were determined with the ther-
mochemical method using equipment LECO CS 230 and
LECO TCH 600. Metal samples were melted in an
induction (for C and S) or resistor (for N) furnace in a
gas stream. The gas was analysed for the absorption of
infrared radiation (for SO2 and CO2) and the change in
the thermal conductivity was measured (for N2). Ele-
ments Si, Mn, P, S, Cr were determined with an X-ray
spectrometry apparatus, ARL ADVANT’X IntelliPower
THERMOFISHER SCIENTIFIC. The method of seque-
ntial X-ray fluorescence spectrometry is based on the
excitation of characteristic X-rays of the elements pre-
sent in a sample using an X-ray lamp. The Al element
was determined on Optima 3000SC PERKIN ELMER.
The analysed sample was dissolved in acids and
transferred into the solution, and then it was measured
using optical emission spectrometry with inductively
bounded plasma.
The pieces forged from all the melts were subjected
to non-destructive ultrasound testing according to SEP
1921/84 Group 3, Class C/c. All the forged pieces fully
conformed to the evaluation.
The work further presents an evaluation of the
micro-purity of the pieces forged from Melts 1 to 3,
according to ASTM E45-10, method A. This method
classifies the inclusions by their shape and light reflec-
tivity only, so their chemical composition plays no role.
For the above reason, a spectral microanalysis was also
performed with a scanning electron microscope JEOL
JSM-5510, equipped with an energy-dispersive analyser
from Oxford Instruments, with which the chemical
compositions of the inclusions were determined. Last but
not least, the work presents an evaluation of the forged
pieces’ macrostructures.
1.4 Micro-purity of the pieces forged from Melt 1 in
the IF
First of all, micro-purity was evaluated on the pieces
forged from Melt 1 in the IF. Very coarse inclusions, spot
D (oxidic inclusions), were observed in the specimens
that often exceeded the allowed limit of 12 μm, specified
in the classification of these inclusions. The biggest
inclusion achieved the size of 49 μm; the spot-D (oxidic)
inclusions were not quite standard, i.e., globular. They
featured a rather sharp-edged shape with s variable size.
Then there was a smaller quantity of specimens with the
inclusions arranged in lines, often in combination with
sulphides that were, using the relevant standard etalon,
evaluated as the B type – line Al2O3. A very low number
of slightly shaped A-type inclusions were then observed
in some places, exceeding the thickness of 6 μm that is
specified for the coarse A-type inclusions. Examples of
non-metallic inclusions are shown in Figures 5 and 6;
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
422 Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426
Table 2: Chemical compositions of standardised3 WNr1.8504, the melts and forged pieces (in mass fractions (w/%))
Tabela 2: Kemijska sestava normiranega3 jekla WNr1.8504, taline in odkovkov (v masnih dele`ih (w/%))
1.8504 C Si Mn P S Cr Aldiss. Albound Altotal N
Standard
min 0.30 0.15 0.60 – – 1.20 – – 0.800 –
max 0.37 0.35 0.90 0.035 0.035 1.50 – – 1.100 –
Melt 1
melt 0.33 0.30 0.81 0.017 0.006 1.39 – – 1.20 –
forged piece 0.33 0.30 0.82 0.012 0.007 1.42 1.08 0.02 1.10 0.0132
Melt 2
melt 0.33 0.27 0.70 0.015 0.007 1.44 – – 1.11 –
forged piece 0.34 0.26 0.73 0.011 0.008 1.45 1.07 0.02 1.09 0.0108
Melt 3
melt 0.35 0.24 0.76 0.018 0.007 1.47 – – 0.97 –
forged piece 0.35 0.23 0.77 0.010 0.007 1.48 0.94 0.02 0.96 0.0108
Figure 6: Non-metallic inclusions in the piece forged from Melt 1.
Magnified 330×.
Slika 6: Nekovinski vklju~ki v odkovku iz taline1. Pove~ava 65×.
Figure 5: Non-metallic inclusions in the piece forged from Melt 1.
Magnified 65×.
Slika 5: Nekovinski vklju~ki v odkovku iz taline 1. Pove~ava 65×.
see Figures 7 and 8 for the chemical compositions of the
most frequent inclusions.
Table 3 shows the results of the micro-purity eva-
luation; the table also includes the largest inclusion
found in the tested metal specimens.
1.5 Micro-purity of the pieces forged from Melt 2 in
the VPIM through the VD process and casting
under Ar atmosphere
During the examination of the polished state, most
often non-metallic inclusions of D- and A-type com-
plexes were observed in the tested chains, as shown in
Figures 9 and 10. Oval inclusion particles were often
locally dispersed in the metallic matrix in the forged
pieces, sometimes achieving a diameter of 48 μm. The
results of the micro-purity evaluation of Melt 2 are
shown in Table 3.
Locally occurring complex non-metallic particles on
the tested specimen surfaces were classified into groups
with the closest shape similarity. The majority of the
tested inclusions were observed to be globular particles
(D type) or elongated sulphides (A type). Small-scale
tiny lines of B-type inclusions were observed in the
forged-piece matrix as well.
The microanalysis detected the chemical composi-
tions of the most frequent inclusions of the AlN type
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426 423
Figure 8: EDX spectrum of non-metallic Al2O3 particles in the piece
forged from Melt 1
Slika 8: EDX-spekter nekovinskih delcev Al2O3, v odkovku iz 1.
taljenja
Figure 7: EDX spectrum of non-metallic AlN particles in the piece
forged from Melt 1
Slika 7: EDX-spekter nekovinskih delcev AlN, v odkovku iz 1.
taljenja
Figure 11: EDX spectrum of non-metallic AlN particles in the piece
forged from Melt 2
Slika 11: EDX spekter nekovinskih delcev AlN, v odkovku iz taline 2
Figure 10: Non-metallic inclusions in the piece forged from Melt 2.
Magnified 330×.
Slika 10: Nekovinski vklju~ki v odkovku iz taline 2. Pove~ava 330×.
Figure 9: Non-metallic inclusions in the piece forged from Melt 2.
Magnified 65×.
Slika 9: Nekovinski vklju~ki v odkovku iz taline 2. Pove~ava 65×.
(Figure 11), the MnS type (Figure 12), or the complex
AlN-MnS type inclusions (Figure 13).
1.6 Micro-purity of the piece forged from Melt 3 in
VPIM through the VD process and casting in the
cofferdam under Ar
As in the previous cases, the presence of a high
amount of coarse inclusions was detected in this forged
piece; due to their shape, these inclusions were classified
as D-type inclusions (oxidic inclusions). Their size
significantly exceeded the admissible diameter of up to
12 μm specified for the D-type inclusions. The occur-
rence of these inclusions was frequent and they achieved
the size of up to 50 μm; however, their shape was not
typically globular but rather angular, as shown in
Figures 14 and 15. The occurrence of oxidic inclusions
in a line arrangement was less frequent.
Besides the oxidic inclusions, A-type inclusions were
observed in the specimen, or complexes of these inclu-
sions, the occurrence of which was relatively frequent.
The results of the non-metallic inclusion evaluation are
listed in Table 3. The microanalysis revealed that, unlike
in Melt 2, AlN-type inclusions were the most frequent in
Melt 3, as shown in Figures 16, 17 and 18.
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
424 Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426
Figure 14: Non-metallic inclusions in the piece forged from Melt 3.
Magnified 65×.
Slika 14: Nekovinski vklju~ki v odkovku iz taline 3. Pove~ava 65×.
Figure 15: Non-metallic inclusions in the forged piece in Melt 3.
Magnified 330×.
Slika 15: Nekovinski vklju~ki v odkovku iz taline 3. Pove~ava 330×.
Figure 16: EDX spectrum of non-metallic AlN particles in the piece
forged from Melt 3
Slika 16: EDX-spekter nekovinskih delcev AIN, v odkovku iz taline 3
Figure 13: EDX spectrum of non-metallic AlN-MnS particles in the
piece forged from Melt 2
Slika 13: EDX-spekter nekovinskih delcev AlN-MnS, v odkovku iz
taline 2
Figure 12: EDX spectrum of non-metallic MnS particles in the piece
forged from Melt 2
Slika 12: EDX-spekter nekovinskih delcev MnS, v odkovku iz taline 2
1.7 Macrostructure evaluation
The macrostructures of the forged pieces were
revealed by etching in 10 % HNO3. Unequally distributed
insignificant segregations of a darker contrast were
revealed in the specimen of the piece forged from Melt
1, as shown in Figure 19. More or less uniform macro-
structures of the surfaces were observed for the spe-
cimens forged from Melts 2 and 3, as shown in Figures
20 and 21.
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426 425
Table 3: Micro-purity evaluation according to the ASTM E45-10 standard, method A, and the largest D-type inclusions found in the pieces
forged from Melts 1 to 3
Tabela 3: Vrednotenje mikro~isto~e po standardu ASTM E45-10, metoda A in najve~ji najdeni vklju~ki vrste D v vzorcih iz odkovkov iz taline 1
do 3
Specimen
Maximum contamination, method A Maximum
dimension of
D-type inclusionType A sulphides Type B aluminates Type C silicates Type D oxides
Fine Coarse Fine Coarse Fine Coarse Fine Coarse μm
Melt 1 1 1 2 2 2 2 49
Melt 2 2 2 2 1 – – 2 2 48
Melt 3 2 1 2 – – – 2 2 50
Figure 18: EDX spectrum of non-metallic AlN particles in the piece
forged from Melt 3
Slika 18: EDX-spekter nekovinskih delcev AlN, v odkovku iz taline 3
Figure 17: EDX spectrum of non-metallic AlN particles in the piece
forged from Melt 3
Slika 17: EDX-spekter nekovinskih delcev AIN, v odkovku iz taline 3
Figure 21: Macrostructure of the piece forged from Melt 3
Slika 21: Makrostruktura odkovkov iz taline 3
Figure 20: Macrostructure of the piece forged from Melt 2
Slika 20: Makrostruktura odkovkov iz taline 2
Figure 19: Macrostructure of the piece forged from Melt 1
Slika 19: Makrostruktura odkovkov iz taline 1
2 CONCLUSION
The objective of the presented work was to increase
the inner purity of a WNr1.8504 high-quality forged
piece. For this reason, three production technologies
(melts) were evaluated in this work.
As the performed analysis of the chemical com-
positions and non-destructive ultrasound tests indicate,
the atmospheric induction furnace and the vacuum and
pressurized induction-melting furnace with casting out-
side and inside the cofferdam are suitable for the produc-
tion of this material.
However, from the macrostructural point of view, the
production of melt in the atmospheric induction-melting
furnace proved to be unsuitable.
With respect to the micro-purity determined with the
microanalysis of the detected particles and nitrogen
content in the forged pieces, none of the three techno-
logies can be applied to achieve a reduced content of
mostly AlN inclusions. The production technology for
Melt 3 in the VPIM, with the VD process and the casting
in the cofferdam under a protective argon atmosphere,
eliminated the portion of oxidic and complex inclusions
but not the AlN inclusions.
The experiments showed that vacuum degassing
(VD) helps to reduce the Al content. The content of
nitrogen was reduced by 0.0024 % of mass fractions,
from 0.0132 % of mass fractions (the steel made in the
IF) to 0.0108 % of mass fractions (the steel made in the
VPIM). Based on this fact, the authors are preparing
another experiment that will eliminate the nitrogen
content using vacuum oxygen decarburization (VOD)
because the melt is mixed better with VOD than with
VD. VOD includes a more efficient degassing process
because, generally, a high oxygen content in a melt
decreases the solubility of nitrogen. For this process, a
newly manufactured oxygen-argon nozzle will be used.
Acknowledgement
This paper was created within Project No. LO1203
"Regional Materials Science and Technology Centre –
Feasibility Program" funded by the Ministry of Edu-
cation, Youth and Sports of the Czech Republic.
3 REFERENCES
1 J. [enberger, Z. Bù`ek, A. Zádìra, K. Stránský, V. Kafka, Metalurgie
oceli na odlitky, VUTIUM, Brno 2008, 311
2 Introduction to Metallography, PACE Technologies,
http://www.metallographic.com/Technical/Metallography-Intro.html,
24.09.2014
3 Stahlschluessel 2007, Verlag Stahlschluessel Wegst GmbH., ver.
5.01.0000, Marbach 2007
4 T. Myslivec. Fyzikálnì chemické základy oceláøství, SNTL – Nakla-
datelství technické webliteratury, Prague 1971, 448
5 J. Vi{ek, Hodnocení rùstu zrna uhlíkových a nízkolegovaných
nástrojových ocelí v závislosti na pøítomnosti AlN,
http://stc.fs.cvut.cz/pdf/VisekJaroslav- 338820.pdf, 24.9.2014
6 G. Stolte, Secondary Metallurgy, Verlag Stahleisen GmbH, Germany
2002, 217
7 V. Kurka, Z. Adolf, J. Pindor, Producing Steel with High Nitrogen
Content in Induction Melting Furnaces, European Oxygen Steel-
making Conference, Tøinec, 2014
8 Z. Buzek, Basic thermodynamic calculations in the steel industry,
Hutnické aktuality, 29 (1988) 7
V. KURKA et al.: INCREASING MICRO-PURITY AND DETERMINING THE EFFECTS OF THE PRODUCTION ...
426 Materiali in tehnologije / Materials and technology 50 (2016) 3, 419–426
O. TRUDONOSHYN ET al.: USE OF THE ABI TECHNIQUE TO MEASURE THE MECHANICAL PROPERTIES ...
427–431
USE OF THE ABI TECHNIQUE TO MEASURE THE
MECHANICAL PROPERTIES OF ALUMINIUM ALLOYS: EFFECT
OF HEAT-TREATMENT CONDITIONS ON THE MECHANICAL
PROPERTIES OF ALLOYS
UPORABA ABI TEHNIKE ZA MERJENJE MEHANSKIH
LASTNOSTI ALUMINIJEVIH ZLITIN: VPLIV POGOJEV
TOPLOTNE OBDELAVE NA MEHANSKE LASTNOSTI ZLITIN
Oleksandr Trudonoshyn1,2, Maxim Puchnin1, Olena Prach1,3
1Czech Technical University in Prague, Karlovo námìstí 13, 12135, Prague 2, Czech Republic
2Friedrich Alexander Universität Erlangen Nürnberg, Martenstraße 5, 91058 Erlangen, Germany
3Technische Universität Darmstadt, Karolinenplatz 5, 64289 Darmstadt, Germany
trudonoshyn@fhotm.kpi.ua
Prejem rokopisa – received: 2014-12-08; sprejem za objavo – accepted for publication: 2015-05-04
doi:10.17222/mit.2014.295
Effects of chemical composition and heat treatment on the microstructures and mechanical properties were investigated with
automated ball-indentation tests, scanning and transmission electron microscopy, and energy-dispersive X-ray analysis. In this
work, the automated ball-indentation (ABI) technique was compared with the standard mechanical tests. The ABI method is
based on load-controlled multiple indentations into a polished surface by a spherical indenter. The indentation depth is
progressively increased to the specified maximum limit with intermediate partial unloading. This technique allows us to
measure the yield strength, the stress-strain curve, the strength coefficient and the strain-hardening exponent. For all the test
materials and conditions, the ABI-derived results were in very good agreement with those obtained with conventional standard
test methods. We analyzed the effect of heat treatment on the alloys with different chemical compositions. Heat treatment leads
to changes in the mechanical properties of the alloys, which are the results of several processes.
Keywords: Al-alloys, heat treatment, tensile strength, yield strength, ABI hardness test
Vpliv kemijske sestave in toplotne obdelave na mikrostrukturo in mehanske lastnosti je bil preiskovan z avtomatiziranim
preizkusom trdote z vtiskovanjem kroglice, z vrsti~no in presevno elektronsko mikroskopijo in energijsko disperzijsko
rentgensko spektroskopijo. V tem delu je bila tehnika avtomatskega vtiskovanja kroglice (ABI) primerjana s standardnimi
mehanskimi preizkusi. ABI metoda temelji na kontrolirani obte`itvi pri ve~kratnem vtiskovanju krogli~nega telesa v polirano
povr{ino. Globina vtiskovanja postopoma nara{~a do maksimalne dolo~ene globine z vmesnimi razbremenitvami. Ta tehnika
omogo~a merjenje meje te~enja, krivulje raztezek-obremenitev, koeficienta trdnosti in eksponenta napetostnega utrjevanja. Za
vse preizku{ane materiale in pogoje, so se dobljeni ABI rezultati dobro ujemali z rezultati dobljenimi iz obi~ajnih metod
preizku{anja. Preu~evan je bil vpliv toplotne obdelave na zlitine z razli~no kemijsko sestavo. Toplotna obdelava povzro~i
spremembo mehanskih lastnosti zlitin, kar je posledica ve~ih procesov.
Klju~ne besede: Al-zlitine, toplotna obdelava, natezna trdnost, meja te~enja, trdota pri ABI preizkusih
1 INTRODUCTION
A large number of aluminum alloys have been deve-
loped for casting, but most of them are varieties of six
basic types (according to the AA Al alloy designation
system): Al-Cu (2XX.X), Al-Cu-Si (or Mg) (3XX.X),
Al-Si (4XX.X), Al-Mg (5XX.X), Al-Zn-Mg (7XX.X)
and Al-St (8XX.X).1 In this context, the alloys of the
Al-Mg-Si (6XXX) system have been widely used in
producing sheets, extruded parts and thin-wall castings.
It was reported by H. Sternau et al.2 that an
AlMg5Si2Mn alloy (HPDC) shows high mechanical
properties such as ductility (up to 18 %), yield strength
(up to 220 MPa) and tensile strength (up to 350 MPa),
compared with the other casting alloys.
When comparing the AlSi7Mg and AlMg5Si2Mn
casting alloys, the most evident difference is in the fact
that the highest strength of AlSi7Mg is achieved only
after the heat treatment, whereas AlMg5Si2Mn exhibits
its highest properties in the as-cast condition.
Heat-treatable Al-Mg-Si alloys are important for the
investigation because of the possible hardening process,
which leads to specific properties. The hardening effects
are shown when the dislocations interact with the preci-
pitates, which act as obstacles to the dislocation mo-
tion.1,3 It is well known that the ductility of such alloys
decreases with the increasing Si content. Brittle coarse Si
particles usually make further deformation difficult.3
In this regard, it is important to apply not only the
optimum chemical composition of the material but also
the optimum heat-treatment conditions. The main task of
this paper is to investigate the influence of heat treatment
on the mechanical properties and structure of Al-Mg-Si-
Mn alloys.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 427–431 427
UDK 669.715:620.172:67.017 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)427(2016)
The ABI test is used to measure material properties
when a conventional technique cannot be applied
(welded parts, brittle materials, samples with a high po-
rosity and the parts currently used).4,5 As shown previ-
ously6, the results obtained with this method achieve
good compatibility with those of the conventional
methods.
2 MATERIALS AND METHODS
The chemical compositions of the evaluated alloys
are shown in Table 1.6
Two types of heat treatment were applied. The first
type was the solution treatment, carried out in an electric
resistance furnace. After the solution treatment, the
specimens were quenched at water at room temperature.
The second type of heat treatment was T6, combining
solution treatment at 570 °C (60 min), quenching in
water at room temperature and artificial aging at 175 °C
during various periods.
Indentation tests were carried out using a special
device (patent CZ 304637 B1), which, due to its design,
is capable of continuously recording the load and inden-
tation depth of the used indenter. The system includes: a
recording device, an analog-to-digital converter, a PC
with software and an Instron 5582 tensile-testing ma-
chine as the force-producing mechanism. The maximum
indentation load was 2.5 kN and the indenter diameter
was 5 mm. Plane-parallel samples were used for ABI
(automatic ball indentation) testing.
Series of measurements were carried out for all the
samples. The obtained HB hardness was compared with
the hardness, measured with a standard testing machine.
In the case of a good reproducibility of the results (an
error of the order of 10 %) from the obtained curve (Fig-
ure 1), values of the yield strength and tensile strength
were calculated.
The hardness was calculated with Equation (1):
HB
P
Dh
=
π
(1)
where HB is the Brinell hardness, P is the load (kN), D
is the diameter of the indenter (mm), h is the indentation
depth (mm) (Figure 1).
For the determination of the tensile strength (Rm), we
used Equation (2):7
R c HBm = ⋅ (2)
where c is the coefficient of uncertainty. For the pre-
sented series of alloys, we used the following value of
this coefficient – 2.8.6
For the determination of the yield strength (Rp0.2), the
methodology proposed in 7, together with Equation (3),
was used. The Meyer hardness was calculated with
Equation (4):
R c HMp0.2 = ⋅ (3)
HM
P
a
=
π 2
(4)
where c is the coefficient of uncertainty (2.8), HM is the
Meyer hardness and a is the contact radius.
Using the values of the indentation depth (h),
according to Equations (5) and (6), strain values () and
the contact radius (a) were determined. The value of
deformation for the yield stress was 0.2 %, by analogy
with the tensile tests.
O. TRUDONOSHYN ET al.: USE OF THE ABI TECHNIQUE TO MEASURE THE MECHANICAL PROPERTIES ...
428 Materiali in tehnologije / Materials and technology 50 (2016) 3, 427–431
Table 1: Nominal compositions of the alloys (Al – bal.), in mass fractions (w/% )
Tabela 1: Nominalna sestava zlitin (Al – ostalo), v masnih dele`ih (w/% )
Alloys Mg Si Mn Fe Ti Cu Zn Comment
AlMg6Mn (M3) 6.0 0.4 0.6 0.3 0.1 0.1 0.1 Al-1Mg2Si-5Mg
AlMg7SiMn (MS1) 7.0 1.0 0.6 0.02 0.1 0.05 0.05 Al-3Mg2Si-5Mg
AlMg7Si2Mn (MS2) 7.0 2.0 0.6 0.02 0.1 0.05 0.05 Al-6Mg2Si-3Mg
AlMg5Si2Mn (M59) 5.0 2.0 0.6 0.02 0.1 0.05 0.05 Al-6Mg2Si-1Mg
AlMg7Si3Mn (MS3) 7.0 3.0 0.6 0.02 0.1 0.05 0.05 Al-9Mg2Si-1Mg
AlMg7Si4Mn (MS4) 7.0 4.0 0.6 0.02 0.1 0.05 0.05 Al-10.5Mg2Si-0.5Si
AlMg7Si5Mn (MS5) 7.0 5.0 0.6 0.02 0.1 0.05 0.05 Al-10.5Mg2Si-1.5Si
AlSi7Mg (S1) 0.3 6.9 0.02 0.2 - 0.05 0.05 Al-7Si
Figure 1: Scheme for determining: a) contact radius, b) indentation
curve for the AlMg6Mn alloy
Slika 1: Shematski prikaz za dolo~anje: a) kontaktnega premera, b)
krivulja vtiskovanja pri zlitini AlMg6Mn
 =
h
D
(5)
a Dh h= − 2 (6)
3 RESULTS AND DISSCUSION
3.1 Mechanical properties
The values of the HB hardness and ABI tests are
summarized in Figure 2. Figure 4 represents the
dependencies of the load-indentation depth, which were
recorded in the ABI study. The differences between two
curves given in the diagrams (Figure 4) may be
connected with the changes in the value of the load (P),
leading to a diversity in the indentation depth and diffe-
rences in the hardness of the materials (HB).6
Figure 4 shows indentation curves of the alloys in
the as-cast state and after the heat treatment (T6).
As the result of the solution treatment, both HB and
tensile-strength values are significantly decreased
(except for the LP5 alloy). In addition, artificial aging
leads to an increase in all the mechanical properties of
the alloys.
The change during the heat treatment is the result of
several processes, which occur during the heating.
3.2 Processes which occur during homogenization
The first process is the eutectic spheroidisation (Fig-
ure 3a). A higher solution-treatment temperature leads
to a faster eutectic-lamella decomposition into smaller
segments and to the spheroidising effect. In 8,9 is pre-
O. TRUDONOSHYN ET al.: USE OF THE ABI TECHNIQUE TO MEASURE THE MECHANICAL PROPERTIES ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 427–431 429
Figure 2: Mechanical properties.*Classical method
Slika 2: Mehanske lastnosti.*Klasi~na metoda
Figure 3: Processes which occur during homogenization: a) spheroidisation of eutectic lamellas, b) formation of -(Al15(Mn,Fe)3Si2)
dispersoids, c) chemical composition (EDX) of Al15(Mn,Fe)3Si2 dispersoids
Slika 3: Procesi med homogenizacijo: a) sferoidizacija lamel evtektika, b) nastajanje disperzioidov -(Al15(Mn,Fe)3Si2), c) kemijska sestava
(EDX) disperzioidov Al15(Mn,Fe)3Si2
sented a model of the spheroidisation of eutectic lamellas
in the alloys of the Al-Mg-Si system. This process leads
to a decrease in the hardness of the alloys.
The second process is the dissolution of the primary
Mn-containing phases, forming dispersoids, which
include Mn, Si and Fe (Figure 3).10 Particle morphology
is shown in Figures 3b and 5b. These particles can be
identified as the -(Al15(Mn,Fe)3Si2) phase (Figure 3c).
A lack of coherence of phase -(Al15(Mn,Fe)3Si2) with
-Al probably affects (along with a disintegration of
eutectic cells) the decrease in the hardness of the alloys.
A dissolution of the ’-Mg9Si5 particles also occurs
during the homogenization.
The last process occurs in alloys MS4 and MS5. This
is a transformation of metastable acicular-shaped -pha-
ses to a more stable state due to diffusion processes.11–13
After the solution treatment, the excess silicon from the
-phase dissolves in the -aluminum solid solution.6 As
it can be seen from Table 2 and Figure 5f, this process
improves the hardness even after the homogenization.
3.3 Processes which occur during aging
The remaining processes occur in the solid solution
and consist of the formation of nanoscale precipitates via
a decomposition of a supersaturated solid solution
O. TRUDONOSHYN ET al.: USE OF THE ABI TECHNIQUE TO MEASURE THE MECHANICAL PROPERTIES ...
430 Materiali in tehnologije / Materials and technology 50 (2016) 3, 427–431
Table 2: Average compositions of Mn-containing phases in the MS4 and MS5 alloys, measured with EDX
Tabela 2: Povpre~na sestava faz, ki vsebujejo Mn v zlitinah MS4 in MS5, izmerjenih z EDX
Phase stoichiometry Condition
Chemical composition, in mass fractions (w/%)
O Mg Al Si Mn Fe Cu
-Al4(Mn,Fe)Si2 (acicular-shaped) AC 1.5 1.1 60.4 26.8 7.6 2.2 0.4
-Al5(Mn,Fe)Si (blocky-shaped, stable) ST 0.5 0.2 59.2 11.9 25.1 2.4 0.7
Figure 4: ABI indentation curves
Slika 4: ABI krivulje pri vtiskovanju
Figure 5: TEM bright-field images of the precipitates in the AlMgSiMn casting alloy: a) as-cast state, b) after homogenization, c) after artificial
ageing
Slika 5: TEM-posnetek (svetla polja) izlo~kov v AlMgSiMn livni zlitini: a) lito stanje, b) po homogenizaciji, c) po umetnem staranju
(SSSS) during aging (Figure 5). It is established that in
the Al-Mg-Si alloys, the decomposition of the super-
saturated solid solution takes place during aging and the
precipitation sequence is SSSS  GP-I  ''  ' 
-Mg2Si where GP-I is the Guiner-Preston zone.12,13
Solid-solution grains contain plate-like particles. One
of their sides is connected with curved lines, which
might be identified as dislocations.
Authors in 14,15 reported that these particles are
formed after natural aging as a result of a heterogeneous
nucleation of dislocations. They must be particles of the
’-Mg9Si5 phase. A direct relationship between the
dislocation density and the number of particles it is
shown in 14,15.
As can be seen from the graphs (Figures 4a and 4b),
heat treatment does not have a significant effect on the
mechanical properties (both the hardness and tensile
strength) of the alloys with extra magnesium.
According to the test results (Figure 4), heat treat-
ment of the alloys with extra silicon improves the
mechanical properties. With the increasing time of artifi-
cial aging (at 175 °C), the hardness of the alloys with
extra silicon grows. This is due to a sufficient amount of
silicon in solid solution needed for forming a larger
number of strengthening particles.
4 CONCLUSIONS
A higher solution-treatment temperature leads to a
faster eutectic-lamella decomposition into smaller seg-
ments and to the spheroidising effect, which causes a
sharp reduction in the mechanical properties.
Homogenization leads to a dissolution of the primary
Mn-containing phases and a formation of dispersoids
- (Al15(Mn,Fe)3Si2).
Homogenization leads to a transformation of the
metastable acicular-shaped -phase into a more stable
state ( or ) due to diffusion processes.
Artificial aging at 175 °C leads to a formation of
strengthening particles in -Al, causing an increase in
the hardness and tensile strength of the alloys. The
difference between the best results for the mechanical
properties of the alloys with excess Mg after the heat
treatment and the properties of the as-cast alloys, is not
significant.
Comparing the AlSi7Mg and AlMg5Si2Mn casting
alloys, the most evident difference is the fact that the
highest strength for AlSi7Mg is achieved only after the
heat treatment, whereas AlMg5Si2Mn exhibits its
highest properties in the as-cast condition.
Acknowledgments
The authors gratefully thank the Visegrad Fund and
DAAD for their support of the research. Trudonoshyn O.
and Prach O. would like to thank Mykhalenkov K. for
stimulating the investigations and helping us formulate
the discussion.
This work was supported by the Ministry of
Education, Youth and Sport of the Czech Republic, pro-
gram NPU1, project No LO1207 and SGS13/186/
OHK2/3T/12 – Research on the influence of surface
treatment on the improvement of service life and
reliability of exposed water-turbine components.
5 REFERENCES
1 I. N. Fridlyander, V. G. Sister, O. E. Grushko, V. V. Berstenev, L. M.
Sheveleva, L. A. Ivanova, Aluminium alloys: promising materials in
the automotive industry, Metal Science and Heat Treatment, 44
(2002) 9–10, 365–370, doi:10.1023/A:1021901715578
2 H. Sternau, H. Koch, A. J. Franke, Magsimal-59, an AlMgMnSi-type
squeeze-casting alloy designed for temper F, Aluminium Rheinfelden
GmbH, P. O. Box 11 40, D-79601 Rheinfelden, Germany
3 S. Zajac, B. Bengtsson, C. Jonsson, Influence of cooling after homo-
genization and reheating to extrusion on extrudability and final
properties of AA6063 and AA6082 alloys, Materials Science Forum,
396–402 and 675–680, doi:10.4028/www.scientific.net/MSF.396-
402.399
4 H. Lee, J. H. Lee, G. M. Pharr, A numerical approach to spherical
indentation techniques for material property evaluation, Journal of
the Mechanics and Physics of Solids, 53 (2005), 2037–2069,
doi:10.1016/j.jmps.2005.04.007
5 K. Sharma, P. K. Singh, V. Bhasin, K. K. Vaze, Application of
Automated Ball Indentation for Property Measurement of Degraded
Zr2.5Nb, Journal of Minerals & Materials Characterization &
Engineering, 10 (2011) 7, 661–669
6 M. Puchnin, O. Trudonoshyn, O. Prach, Use of ABI technique to
measure mechanical properties in aluminium alloys, Part 1: Effect of
chemical composition on the mechanical properties of the alloys,
Mater. Tehnol., 50 (2016) 2, 247–252, doi:10.17222/mit.2014.294
7 P. I. Stoev, V. I. Moschenok, Definition of Mechanical Properties of
Metals and Alloys on Hardness, Bulletin of V. N. Karazin Kharkiv
National University, 601 (2003) 2(22), 106–112
8 C. Phongphisutthinan, H. Tezuka, T. Sato, Semi-Solid Microstructure
Control of Wrought Al-Mg-Si Based Alloys with Fe and Mn
Additions in Deformation-Semi-Solid-Forming Process, Materials
Transactions, 52 (2011) 5, 834–841, doi:10.2320/matertrans.
L-MZ201119
9 S. Otarawanna, C. M. Gourlay, H. I. Laukli, A. K. Dahle, Micro-
structure Formation in AlSi4MgMn and AlMg5Si2Mn High-Pressure
Die Castings, Metallurgical and Materials Transactions A, 40A
(2009), 1645–1659, doi:10.1007/s11661-009-9841-1
10 L. Lodgaard, N. Ryum, Precipitation of dispersoids containing Mn
and/or Cr in Al–Mg–Si alloys, Materials Science and Engineering,
283A (2000), 144–152, doi:10.1016/S0921-5093(00)00734-6
11 O. Trudonoshyn, O. Prach, V. Boyko, K. Mykhalenkov, Selection
and Optimization of Heat Treatment Process to Improve the Mecha-
nical Properties of Casting Alloys of the Al-Mg-Si System, Casting
Processes, 106 (2014) 4, 12–21
12 G. A. Edwards, K. Stiller, G. L. Dunlop, M. J. Couper, The Preci-
pitation Sequence in Al-Mg-Si Alloys, Acta Mater., 46 (1998) 11,
3893–3904, doi:10.1016/S1359-6454(98)00059-7
13 C. Ravi, C. Wolverton, First-principles study of crystal structure and
stability of Al–Mg–Si–(Cu) precipitates, Acta Materialia, 52 (2004),
4213–4227, doi: 10.1016/j.actamat.2004.05.037
14 V. Boyko, T. Link, N. Korzhova, K. Mykhalenkov, Microstructure
characterization of AlMg5Si2Mn casting alloy, Materials Science
and Technology (MS&T) 2013, Montreal, Quebec, Canada 2013,
1331–1338
15 V. Boyko, O. Prach, O. Trudonoshyn, K. Mykhalenkov, Microstruc-
ture and Natural Hardening of AlMg5Si2Mn Casting Alloy, Bulletin
of NTUU "KPI", (2014), 47–54
O. TRUDONOSHYN ET al.: USE OF THE ABI TECHNIQUE TO MEASURE THE MECHANICAL PROPERTIES ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 427–431 431

N. AKAR et al.: INVESTIGATION OF THE EFFECT OF HOLDING TIME AND MELT STIRRING ...
433–437
INVESTIGATION OF THE EFFECT OF HOLDING TIME AND
MELT STIRRING ON THE GRAIN REFINEMENT OF AN A206
ALLOY
PREISKAVA VPLIVA ^ASA ZADR@EVANJA IN ME[ANJA TALINE
NA ZMANJ[ANJE VELIKOSTI ZRN ZLITINE A206
Neºet Akar1, Ziya Tanyel3, Kadir Kocatepe1, Ramazan Kayikci2
1Gazi University, Department of Metallurgical and Materials Engineering, Ankara, Turkey
2Sakarya University, Department of Metallurgical and Materials Engineering, Sakarya, Turkey
3Gazi University, Graduate School of Natural and Applied Sciences, Ankara, Turkey
rkayikci@sakarya.edu.tr
Prejem rokopisa – received: 2014-12-12; sprejem za objavo – accepted for publication: 2015-06-17
doi:10.17222/mit.2014.302
Effects of melt stirring and holding time were studied with an Al-4.5 % of mass fractions of Cu (A206) alloy. The optimum
level of a grain refiner was determined in conjunction with both continuously stirred and non-stirred melts during the holding
time. Results showed that a mass fraction of a Ti addition of as low as 0.03 % was sufficient to obtain the 82 μm average grain
size, while a Ti addition exceeding the mass fraction of 0.03 % showed no significant effect on the grain size of castings. The
grain refinement tends to fade with a long holding time in a non-stirred liquid. The results also showed that an effective grain
refinement of the A206 alloy can be achieved with a long holding time as long as the liquid alloy is continually stirred.
Keywords: grain refinement, melt stirring, holding time, Al-4.5Cu alloy, A206 alloy
Raziskan je bil vpliv me{anja in ~asa zadr`evanja taline zlitine Al-4,5 % masnega dele`a Cu (A206). Dolo~ena je bila optimalna
koli~ina udrobnjevalca zrn, v povezavi s stalnim me{anjem ali brez me{anja taline med zadr`evanjem. Rezultati so pokazali, da
je bilo `e 0,03 % masnega dele`a dodatka Ti, dovolj za doseganje povpre~ne velikosti zrn 82 μm, medtem ko dodatek Ti ve~ji od
0,03 % masnega dele`a, ni pokazal vpliva na velikost zrn ulitkov. U~inkovitost drobnjenja zrn se zmanj{a pri dolgih ~asih
zadr`anja in brez me{anja taline. Rezultati so pokazali {e, da je mogo~e dose~i u~inkovito drobnjenje zrn tudi pri dolgih ~asih
zadr`evanja, dokler se talina stalno me{a.
Klju~ne besede: drobnjenje zrn, me{anje taline, ~as zadr`evanja, zlitina Al-4.5Cu, zlitina A206
1 INTRODUCTION
Al-Cu alloys are one of the most important Al-based
alloys because they provide good castability and
excellent mechanical properties.1–6 Due to their superior
mechanical properties, Al–Cu alloys can be used in
many areas such as aircraft construction, military field
and automobile manufacturing.4 Grain refinement of
Al–Cu alloys significantly improves the microstructure
and mechanical properties.7–9 Along with many advant-
ages, hot tearing, which frequently occurs during solidi-
fication due to a long freezing range, is a severe problem
in producing cast components with these alloys. Previous
works on casting and solidification of Al–Cu alloys
consistently indicated that hot tearing can be eliminated
with a good grain refinement.1–4,7–9 Grain refinement was
also found to be effective for reducing the amount of the
porosity and size of the pores, and improving the feeding
of cast Al–Cu alloys.9 Titanium and boron are added in
the form of Al–Ti–B master-alloy rods to cast aluminium
alloys for the grain refinement. Rod-type additions were
found to be more effective for providing, controlling and
optimizing TiB2 particles than the salt form.10,11 It was
reported in recent years that Al–Ti–C master alloys are
also effective grain refiners.4,12 Due to its high refining
potential, an Al5Ti1B alloy in the rod form is one of the
most commonly used grain refiner and has been widely
accepted in controlling the grain size and microstructures
of aluminium alloys in industrial applications.13
The A206 alloy is a well-known Al-Cu casting alloy
and a research on grain refinement with this alloy indi-
cated that a proper grain refinement can be achieved with
a 0.15–0.30 % titanium mass fraction of the final cast
part.14 On the contrary, in recent years other resear-
chers3,5,7–9 showed that modern grain refiners containing
Al–Ti and B are more suitable for an acceptable grain
refinement of A206 alloys if the Ti content is lower than
0.15 %.
Despite a number of studies conducted in the past to
investigate the grain refinement of Al–Cu
alloys,3,5–9,12,15–16 studies on the effects of the holding time
and melt stirring on the grain-refinement process have
not been published. Therefore, in this work, the optimum
amount of the grain refiner for the A206 alloy, using an
Al5Ti1B rod-shape grain refiner, in combination with the
holding time and melt stirring, was studied.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 433–437 433
UDK 620.181:622.795:669.3:669.715 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)433(2016)
2 EXPERIMENTAL PROCEDURES
A206 alloys with two different chemical compo-
sitions were prepared using commercial-purity Al ingots
and commercially pure Cu wires in a new SiC crucible.
The final chemical compositions of these alloys before
the grain-refiner additions are given in Table 1. A Spec-
tro-type optical spectrometer was employed to perform a
chemical analysis of the alloys throughout this study. An
industrial electrical-resistance furnace with a 600 kg
capacity and a SiC crucible were used for the melting.
An Al5Ti1B master alloy was introduced into the
liquid A206 alloy at 730 °C followed by rotary degassing
for 10 min with dry argon. The temperature of the liquid
metal in the melting furnace was continuously controlled
with a K-type thermocouple connected to the control unit
of the furnace to ensure the holding of the liquid at
730 °C.
Grain-refining experiments were carried out using the
Alcan standard grain-refining test10 with two sets of
samples. In the first set of experiments, samples were
taken from the crucible at 730 °C, before the grain-refi-
ner addition. Samples were also taken after each addition
of the Al5Ti1B rod-type master alloy introducing (0.01,
0.02, 0.03, 0.05, 0.1, 0.2, 0.3) % Ti contents into the
liquid alloy. After each addition, the liquid alloy was re-
degassed for 10 min at 730 °C. This resulted in 10-min
intervals between the sample-taking processes.
The second set of experiments was carried out to
determine the effects of the holding time coupled with
melt stirring on the grain refinement. To implement this,
a new heat of the A206 alloy was melted in a new cru-
cible and heated up to 730 °C. The Al5Ti1B master alloy
was introduced into the melt to obtain a 0.05 % Ti con-
tent within the alloy. After a 10-min degassing treatment,
the purging argon was turned off and the molten alloy
was continuously stirred using the graphite lance of the
rotary degassing unit with a rotating speed of 150 min–1
for 90 min. Samples for the Alcan grain-refining tests
were taken during the whole stirring period, in 30-min
intervals from the beginning to the end.
Finally, the melt was re-degassed for 10 min and held
for another 90 min without any stirring actions. During
this second 90-min period, Alcan test samples were also
taken in 30-min intervals, in the same manner as ex-
plained above. At the end of this period, the melt was
re-stirred for only 1 min and the final sample was taken.
The specimens for the metallographic examinations
were cut as shown in Figure 1. The surface of each
specimen was electro-polished using 5 mL of HClO4,
15 mL of 2-Butoxyethanol, 60 mL of ethanol and 20 mL
of distilled water. The average grain size was determined
with the linear intercept method according to the ASTM
E112 standard, at different regions of each sample.
3 RESULTS AND DISCUSSION
3.1 Grain-size measurements
The mean values of the measured grain sizes with va-
rious Ti contents are shown in Table 2, which indicates
that the titanium contents in the alloy were determined
within a narrow variance. Table 2 also shows that the
grain size of the samples decreased dramatically with the
addition of the Al5Ti1B master alloy regardless of the
titanium content. The average grain size versus the Ti
content is also shown in Figure 2. Figure 2 indicates
that the addition of the grain refiner, even with a Ti con-
tent as low as 0.01 %, resulted in a remarkable reduction
in the grain size of the alloy. Figure 2 also indicates that
the lowest grain size of the alloy is about 80 μm, which
was obtained with a 0.03 % Ti content. A higher Ti con-
N. AKAR et al.: INVESTIGATION OF THE EFFECT OF HOLDING TIME AND MELT STIRRING ...
434 Materiali in tehnologije / Materials and technology 50 (2016) 3, 433–437
Table 1: Chemical compositions of A206 alloys before the grain refinement, in mass fractions (w/%)
Tabela 1: Kemijska sestava zlitin A206, pred udrobnjenjem zrn, v masnih dele`ih (w/%)
Experiment
number
in mass fractions (w/%)
Si Fe Cu Mn Mg Zn B Ti Al
1 0.041 0.105 4.510 0.281 0.214 0.098 0.001 0.000 bal.
2 0.039 0.107 4.610 0.353 0.199 0.084 0.002 0.007 bal.
Table 2: Average grain size of the samples with different Ti contents
Tabela 2: Povpre~na velikost zrn pri vzorcih z razli~no vsebnostjo Ti
Ti (w/%)
Target 0 0.01 0.02 0.03 0.05 0.1 0.2 0.3
Realized 0 0.013 0.019 0.032 0.044 0.098 0.194 0.292
Average grain size (μm) 970 112 92 82 81 79 81 79
Figure 1: Alcan test dimensions and grain-size measuring surface
Slika 1: Dimenzije Alcan preizku{anca in povr{ina za merjenje
velikosti zrn
tent within the A206 alloy did not further reduce the
grain size of the samples. These results are in good
agreement with the previous work carried out by Sig-
worth and co-workers.3,7–8
The relationship between the titanium recovery and
the microstructure is given in Figure 3. Figure 3a shows
the microstructure of the sample obtained with no grain-
refiner addition. The microstructure consists of coarse
dendrites, heterogeneously distributed in equiaxed
grains. Figure 3b shows a small addition of Ti, as low as
a 0.01 % mass fraction, which caused the grain refine-
ment of the A206 alloy. However, a Ti content of up to
0.05 % was found to be more effective for further
reducing the average grain size of the alloy, as seen in
Figure 3c. The microstructure obtained with a 0.3 %
mass fraction of Ti in the melt is shown in Figure 3d. It
indicates that the increased amount of Ti no longer
affected the reduction of the grain size of the A206 alloy
used in this study.
The Al5Ti1B master alloy is an effective grain refiner
of Al-Cu alloys as it increases the number of hetero-
geneous nucleation sites for achieving a finer equiaxed
grain structure.10 Figure 3 clearly shows that a Ti content
exceeding the mass fraction of 0.01 % is adequate to turn
the solidification morphology of the alloy from a fully
dendritic to a globular or near-dendritic structure. Thus,
the grain refinement of the A206 alloy by adding the
Al5Ti1B master alloy can be expected to increase the
mechanical properties of A206 castings via reducing the
amount of solidification defects such as hot tearing,
micro-shrinkage and micro-segregation. These results
are in good agreement with the report from H. Kamali et
al.5 who reported that Ti additions of 0.05–0.3 % showed
no significant effect on the grain size, although the mini-
mum Ti mass fractions of 0.05 % was necessary to
eliminate the hot-tearing defects.
3.2 Effects of the holding time and melt stirring on the
grain size
In the present study, possible effects of melt stirring
during the holding period after the master-alloy addition
were also investigated. Experiments were carried out
with the castings from two different heats of stirred and
non-stirred melts. Samples were cast in 30-min intervals
throughout the 90-min holding time, during which the
amount of Ti was fixed at around 0.05 % of mass
fractions.
N. AKAR et al.: INVESTIGATION OF THE EFFECT OF HOLDING TIME AND MELT STIRRING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 433–437 435
Figure 3: Microstructure of the A206 alloy with different Ti
additions: a) 0 % of mass fractions, b) 0.01 % of mass fractions, c)
0.05 % of mass fractions and d) 0.3 % of mass fractions
Slika 3: Mikrostruktura zlitine A206 z razli~nimi dodatki Ti: a) 0 %
masnega dele`a, b) 0,01 % masnega dele`a, c) 0,05 % masnega dele`a
in d) 0,3 % masnega dele`a
Table 3: Effect of the holding time on the average grain size
Tabela 3: Vpliv ~asa zadr`evanja na povpre~no velikost zrn
Liquid-metal
condition
Holding time
(min)
Ti recovery
(w/%)
B recovery
(w/%) Ti:B ratio
Average grain size
(μm)
Heat 1
Stirred melt
0 0.049 0.009 5.4 84
30 0.046 0.008 5.8 84
60 0.047 0.008 5.9 85
90 0.048 0.009 5.3 84
Heat 2
Non-stirred melt
0 0.046 0.009 5.1 84
30 0.036 0.005 7.2 88
60 0.035 0.004 8.8 102
90 0.031 0.003 10.3 111
90* 0.048 0.009 5.3 85
(90*) Stirring after 90 min.
Figure 2: Effect of the Ti content on the grain size of the A206 alloy
Slika 2: Vpliv vsebnosti Ti na velikost zrn zlitine A206
A summary of the quantitative results obtained from
the test samples throughout these experiments are shown
in Table 3. Titanium and boron measured in the samples,
cast with the stirred melt, show that both elements
remained at almost their initial values during the 90-min
holding time. Accordingly, the initial Ti:B ratio also
remained constant, which was around 5:5. The average
grain size of the stirred melt was about 85 μm regardless
of the holding time. These results emphasise the fact that
the refining efficiency of the Al5Ti1B type master alloy
is consisted, at least during the 90-min holding time,
even with the Ti recovery as small as 0.05 % of mass
fractions of Ti in the A206 alloy. On the other hand, the
results obtained for the non-stirred melt reveal that both
titanium and boron recovery decreased with the
increasing holding time. Interestingly, the diminution in
boron was faster compared to titanium, especially in the
first 30 min of the holding time, for the non-stirred melt.
Therefore, the measured Ti:B ratio shows a continuous
increase as the period of the holding time increases.
In Table 3, the average measured grain size for the
samples of the non-stirred melt also shows a consisted
increase with the holding time. This is associated with
the diminishing of the Ti and B recovery and the ever
increasing Ti:B ratio with the increasing holding time for
the non-stirred heat. It is also obvious from Table 3 that
an immediate stirring action followed by a 90-min hold-
ing time facilitated the Ti and B levels as well as the Ti:B
ratio to remain almost at their initial values. This also
caused a similar consequence in the measured grain size.
These results are in good agreement with a number of
previous studies on Al-Cu alloys. Grain-refining studies
on non-stirred melts were concluded so that the grain
size increased continuously with the increasing holding
time;11–12,16 however, it began to decrease when the
stirring action was resumed.16
The phenomenon of decreasing Ti and B with the in-
creasing holding time in the non-stirred A206 alloy can
be attributed to the settling of TiB2. This compound has
been widely accepted as one of the potential nucleation
sites during the solidification of aluminium.19–22 Since
the density of a solid TiB2 compound is higher (4.48
g/cm3)17 than that of the liquid A206 alloy (2.78 g/cm3),
it is quite probable that some potential TiB2 nuclei are
disqualified because they sink to the bottom of the
crucible as the holding time increases.
The microstructures representing the Alcan test sam-
ples obtained from the heats after the 90-min holding
time are shown in Figures 4a and 4b. A comparison of
the two microstructures provides an explanation of the
difference between the stirred and the non-stirred melts.
The larger grain size for the non-stirred melt in Figure 4b
can be associated with a weaker grain-refining action of
the master alloy during the 90-min holding. This can also
be related to the fading of the potential nuclei probably
due to the gravity action of the TiB2 compound.18
4 CONCLUSIONS
The grain-refining effects of the Al5Ti1B rod-type
master alloy on a commercial A206 alloy with different
addition levels were studied. The effects of the holding
time under stirring and no-stirring conditions were also
studied. From the experimental results, the following
conclusions can be drawn:
1) Different amounts of titanium recovery ranging from
0.01 to 0.3 were formed in the A206 alloys. Measure-
ments showed that the grain size of the Alcan test
samples decreased dramatically with an addition of
the Al5Ti1B master alloy regardless of the titanium
recovery.
2) The smallest average grain size of the A206 alloy
was 82 μm, achieved with a 0.03 % of mass fraction
of Ti recovery in the melt. Increasing the Ti recovery
up to a 0.3 % of mass fractions did not result in a
further decrease in the grain size of the A206 alloy.
3) Grain-size measurements for two different heats
showed that the average initial grain size of the
samples increased from 84 μm to 111 μm at the end
of the 90-min holding time for the non-stirred liquid.
However, a constant grain size was achieved for the
stirred melt throughout the holding period.
4) The results obtained for the non-stirred melt revealed
that both titanium and boron recovery decreased with
N. AKAR et al.: INVESTIGATION OF THE EFFECT OF HOLDING TIME AND MELT STIRRING ...
436 Materiali in tehnologije / Materials and technology 50 (2016) 3, 433–437
Figure 4: Microstructures of Alcan test samples after a 90 min
holding time and an Al5Ti1B master-alloy addition. Samples were
cast from: a) continuously stirred melt and b) non-stirred melt.
Slika 4: Mikrostruktura Alcan preizku{ancev po 90 min zadr`evanju
po dodatku predzlitine Al5Ti1B. Vzorci so bili uliti iz: a) kontinuirno
me{ane taline in b) taline brez me{anja.
the increasing holding time. During the holding pe-
riod, the diminution in boron was larger compared to
titanium, as measured on the cast samples. This was
attributed to the formation of a TiB2 compound in the
melt, which is widely accepted as a heterogeneous
nucleation site for aluminium. Since the density of
TiB2 is higher than that of the liquid A206 alloy, the
compound tends to settle in the non-stirred melt,
which may be the reason for the fading of the grain-
refining efficiency as the holding time increases.
5) The refining efficiency of the Al5Ti1B-type master
alloy can be stimulated via re-stirring the melt. The
results showed that an immediate stirring action
followed by a 90-min holding time reverted Ti and B
and also the Ti:B ratio almost to their initial con-
ditions. The Alcan test samples of the re-stirred melt
also showed a well-refined grain structure.
Acknowledgements
Authors thank Mr. Ahmet Cevdet Altun at Altun
Döküm A.ª., Konya, Turkey, for the casting experiments
carried out at the premises of the company.
5 REFERENCES
1 H. K. Kamga, D. Larouche, M. Bournane, A. Rahem, Mechanical
properties of aluminium-copper B206 alloys with iron and silicon
additions, International Journal of Cast Metals Research, 25 (2012),
15–25, doi:10.1179/1743133610Y.0000000012
2 H. K. Kamga, D. Larouche, M. Bournane, A. Rahem, Hot tearing of
aluminum–copper B206 alloys with iron and silicon additions, Mate-
rials Science and Engineering, 527A (2010), 7413–7423,
doi:10.1016/j.msea.2010.08.025
3 G. K. Sigworth, F. DeHart, Recent developments in the high strength
aluminum-copper casting alloy A206, AFS Transactions, 111 (2003),
341–354
4 T. Wang, T. Gao, P. Zhang, J. Nie, X. Liu, Influence of a new kind of
Al–Ti–C master alloy on the microstructure and mechanical proper-
ties of Al–5Cu alloy, Journal of Alloys and Compounds, 589 (2014),
19–24, doi:10.1016/j.jallcom.2013.11.187
5 H. Kamali, M. Emamy, A. Razaghian, The influence of Ti on the
microstructure and tensile properties of cast Al–4.5 Cu–0.3 Mg alloy,
Materials Science and Engineering A, 590 (2014), 161–167,
doi:10.1016/j.msea.2013.10.032
6 Y. D. Kwon, Z. H. Lee, The effect of grain refining and oxide inclu-
sion on the fluidity of Al–4.5 Cu–0.6 Mn and A356 alloys, Materials
Science and Engineering A, 360 (2003), 372–376, doi:10.1016/
S0921-5093(03)00504-5
7 G. K. Sigworth, Improved hot crack resistance, U.S. Patent US6,
645, 321, 11. 11. 2003
8 G. K. Sigworth, Providing melt of aluminum base alloy, maintaining
dissolved titanium in range of 0.005 to 0.05 wt.% to improve resis-
tance of alloy to hot cracking, adding nucleating agent selected from
metal carbides, aluminides, borides, solidifying, U.S. Patent US6,
368, 427, 9. 4. 2002
9 G. K. Sigworth, T. A. Kuhn, Grain refinement of aluminum casting
alloys, AFS Transactions, 115 (2007), 1–12
10 G. P. Jones, J. Pearson, Factors affecting the grain-refinement of alu-
minum using titanium and boron additives, Metallurgical Transac-
tions B. 7 (1976), 223–234, doi:10.1007/BF02654921
11 K. Sato, M. C. Flemnigs, Grain refining of Al-4.5 Cu alloy by adding
an Al-30TiC master alloy, Metallurgical and Materials Transactions
A, 29 (1998), 1707–1710, doi:10.1007/s11661-998-0093-2
12 P. Cooper, A. Barber, Review of the latest developments and best use
of grain refiners, 2th International Melt Quality Workshop, Prague,
Czech Republic 2003
13 Alcan AA206 Primary Foundry Alloys 206.2, A206.2 and B206.2,
Alcan Primary Products Corporation Inc., Robards, KY, USA 2004
14 S. Li, K. Sadayappan, D. Apelian, Role of Grain Refinement in the
Hot Tearing of Cast Al-Cu Alloy, Metallurgical and Materials Tran-
sactions B, 44 (2013), 614–623, doi:10.1007/s11663-013-9801-4
15 M. A. Kearns, S. R. Thistlewaite, P. S. Cooper, In W. Hale (Ed.),
Light Metals TMS, Warrendale 1996
16 H. Li, T. Sritharan, Y. M. Lam, N. Y.Leng, Effects of processing
parameters on the performance of Al grain refinement master alloys
Al-Ti and Al-B in small ingots, Journal of Materials Processing
Technology, 66 (1997), 253–257, doi:10.1016/S0924-0136(96)
02536-8
17 L. Zhou, Z. Fan, Effect of Free Ti on Grain Refinement of Alumi-
nium Inoculated with Potent TiB2 Particles, Materials Science
Forum, 790–791 (2014), 155–160, doi:10.4028/www.scientific.net/
MSF.790-791.155
18 P. S. Mohanty, J. E. Gruzleski, Mechanism of grain refinement in
aluminium, Acta Metallurgica et Materialia, 43 (1995) 5,
2001–2012, doi:10.1016/0956-7151(94)00405-7
19 P. S. Mohanty, J. E. Gruzleski, Grain refinement mechanisms of
hypoeutectic Al-Si alloys, Acta Materialia, 44 (1996) 9, 3749–3760,
doi:10.1016/1359-6454(96)00021-3
20 C. Wang, M. Wang, B. Yu, D. Chen, P. Qin, M. Feng, Q. Dai, The
grain refinement behaviour of TiB2 particles prepared with in situ
technology, Materials Science and Engineering A, 459 (2007),
238–243, doi:10.1016/j.msea.2007.01.013
21 Z. Fan, Y. Wang, Y. Zhang, T. Qin, X. R. Zhou, G. E. Thompson, T.
Pennycook, T. Hashimoto, Grain refining mechanism in the
Al/Al–Ti–B system, Acta Materialia, 84 (2015), 292–304,
doi:10.1016/j.actamat.2014.10.055
22 X. Wang, Z. Liu, W. Dai, Q. Han, On the Understanding of Alumi-
num Grain Refinement by Al-Ti-B Type Master Alloys, Metallur-
gical and Materials Transactions B, (2014), doi:10.1007/s11663-
014-0252-3
N. AKAR et al.: INVESTIGATION OF THE EFFECT OF HOLDING TIME AND MELT STIRRING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 433–437 437

R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
439–445
INVESTIGATING THE INFLUENCE OF CUTTING SPEED ON THE
TOOL LIFE OF A CUTTING INSERT WHILE CUTTING DIN 1.4301
STEEL
PREISKAVA VPLIVA HITROSTI REZANJA NA ZDR@LJIVOST
VLO@KA ZA REZANJE PRI REZANJU JEKLA DIN 1.4301
Rozmarína Dubovská1, Jozef Majerík2, Robert ^ep3, Karel Kouøil4
1University of Hradec Kralove, Faculty of Education, Department of Technical Subjects, Rokitanského 62,
500 03 Hradec Králové, Czech Republic
2Alexander Dubcek University of Trencin, Faculty of Special technology, Department of Engineering, Pri Parku 19, 911 05 Tren~ín, Slovakia
3Technical University of Ostrava V[B, Faculty of Mechanical Engineering, Department of Machining and Assembly, 17.listopadu 15,
708 33 Ostrava, Czech Republic
4Brno University of Technology, Faculty of Mechanical Engineering, Institute of Manufacturing Technology, Technická 2896/2,
616 69 Brno, Czech Republic
robert.cep@vsb.cz
Prejem rokopisa – received: 2015-02-10; sprejem za objavo – accepted for publication: 2015-05-20
doi:10.17222/mit.2015.036
The main aim of this paper is to assess the tool life T = f(vc) during the dry turning of 1.4301 austenitic stainless steel with a
CNMG 120408 coated carbide cutting insert. Experimental tests of the selected material were realized in an Aero Turn BT-380
CNC machine tool with a Fanuc 21i TB control system. The effect of the applied cutting parameters on the surface finish, tool
wear, tool life and surface roughness were investigated during the realized experiments. The aim of the present paper is to focus
scientific research on the impact of the various cutting speeds during the outer longitudinal turning. The presented approach and
results will be helpful for understanding the machinability of 1.4301 austenitic stainless steel during dry turning. This paper,
together with the achieved results, is a basis to optimize the performance of the machining (i.e., turning) of austenitic stainless
steel 1.4301 used for special industrial applications with their dominant functional areas.
Keywords: austenitic stainless steel, CNC turning, cutting speed, tool life, surface finish
Glavni namen ~lanka je oceniti preiskovano zdr`ljivost orodja T = f(vc) pri stru`enju, brez mazanja avstenitnega nerjavnega jekla
1.4301, s karbidnim rezalnim vlo`kom CNMG 120408 s prevleko. Preizkusi izbranega materiala so bili izvr{eni na CNC stroju
Aero Turn BT-380 s Fanuc 21i TB kontrolnim sistemom. Med preizkusi je bil preiskovan vpliv uporabljenih parametrov pri
rezanju na kvaliteto povr{ine, obrabo orodja, zdr`ljivost orodja in hrapavost. Namen ~lanka je usmeriti raziskavo na vpliv
razli~nih uporabljenih vrednosti hitrosti rezanja pri zunanjem vzdol`nem stru`enju. Vsi predstavljeni pribli`ki in rezultati bodo
pomagali pri razumevanju obdelovalnosti avstenitnega nerjavnega jekla 1.4301 pri stru`enju brez mazanja. Dobljeni rezultati so
osnova za optimiranje stru`enja avstenitnega nerjavnega jekla 1.4301, ki se ga, na podlagi posebnih lastnosti, uporablja pri
posebnih industrijskih namenih.
Klju~ne besede: avstenitno nerjavno jeklo, CNC stru`enje, hitrost rezanja, zdr`ljivost orodja, kvaliteta povr{ine
1 INTRODUCTION
High productivity and reliability are necessary in
today’s very highly competitive world of production. In
this context, the appropriate selection of cutting tool
geometry and tool material is crucial to be competitive,
especially in the field of difficult-to-machine materials,
such as stainless steels.1,2 Problems such as poor surface
finish and high tool wear are common in the machining
of austenitic stainless steel. The authors3 carried out turn-
ing tests on the 1.4301 austenitic stainless steel to deter-
mine the optimum machining parameters. Austenitic
stainless steel is among the difficult-to-cut material and
difficulties such as poor surface finish and rapid tool
wear are common.4–8 Stainless steels are widely used in
several industrial sectors, such as engine production, the
medical and chemical industries. Their high strength,
low thermal conductivity, high ductility and high ten-
dency towards work hardening are the main factors for
their poor machinability.9 The turning of parts made of
austenitic stainless represents nearly 24 % of all ma-
chined parts made of steel. Various special chemical
compositions of stainless steels are a challenge for all
machining technologies. High-speed machining (HSM)
is applied with significantly higher cutting speeds vc with
relatively small cross-sections being cut. HSM techno-
logy is realized with extremely hard and heat-resistant
cutting tools.10 With the trend in technology develop-
ment, stainless steel has been broadly adopted because it
has the characteristics of high toughness, low thermal
conductivity, and a high strain hardening coefficient.
This has a negative effect on the surface finish of
a machined product and results in a reduced tool life.11
Such is the case for austenitic stainless steels, which in
spite of being materials of high economic and techno-
logical value, their behaviour with respect to machining
is still not well understood in some aspects. There are
Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445 439
UDK 620.179.5:621.9:669.018.26 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)439(2016)
not reliable and updated technological data about
austenitic stainless steels in industry.12 The austenitic
stainless steel AISI 304 (according to DIN 1.4301) is the
second most widely used anti-corrosive material with
excellent corrosion resistance, cold formability and
weldability. The steel 1.4301 is resistant to water, steam,
humidity, edible acids, weak organic and inorganic
acids.13 Austenitic stainless steel is not hardenable. How-
ever, its strength can be increased by cold forming. It is
characterized by the need for a solution treatment to
ensure corrosion resistance in contact with a wide range
of substances. The AISI 304 is approved for a thermal
stress of 300 °C. When machining the 1.4301 it is
necessary to work with sharp cutting tools of high-speed
alloyed steel or cemented carbides because of the
tendency to harden. The steel 1.4301 (AISI 304) is used
in the engineering and nuclear industries, in architecture,
in transport facilities, the food industry, the pharma-
ceutical and cosmetic industries, the construction of
chemical apparatus and vehicles, the manufacture of
surgical instruments, sanitary installations, objects and
appliances and works of art. The shape of the individual
components for the automotive industry and subse-
quently the time and relative complexity of the conver-
sion work and tool paths for CNC (Computer Numerical
Control) program preparation led to the establishment of
internal and external graphics support for the creation of
individual programs.14 CNC machines are commonly
used in automated factories for producing machined
parts. In this study, the AISI 304 austenitic stainless steel
was used to help the manufacturers. In this work, the
values for the flank wear were investigated and in this
way the best cutting parameters were determined. Apart
from classic methods, it was also investigated that the
process sound generated during machining could be used
to assess machinability.15 Surface integrity is an import-
ant factor in evaluating the machinability of the steels.16
Numerous experimental investigations have been carried
out over the years to study the effect of the cutting para-
meters and tool geometries on the workpieces’ surface
integrity using several types of workpiece materials.17
Currently, companies prefer to order material according
to DIN or AISI.
2 MATERIALS AND METHODS
The basic factor that causes the flank wear of carbide
tools is the high temperature of the cutting edge. In order
to increase tool life, we have to reduce this temperature.
For machining is necessary to use a stable and solid
CNC machine tool with the appropriate cooling. The
workpiece material and cutting tools must be firmly
clamped in the CNC machine with a small overhang. It is
important to create the conditions for cutting, to prevent
the formation of vibrations. Progressive solutions in
terms of cutting tools seem to be new cutting materials
for machining stainless, especially austenitic, steels.
There are cutting tools with changeable cutting inserts
with a fine-grained or ultrafine-grained substrate. The
carbide grain size is from 0.3 μm to 0.5 μm. On the ce-
mented carbide is deposited a multilayer with the coat-
ings type TiC+Al2O3+TiN on the surface, the substrate is
WC+Co. The austenitic stainless steels are generally
annealed for austenitizing, so that they are heated to
1000–1150 °C. Subsequently, they are quickly cooled in
water or air, to prevent the precipitation of the carbides at
the grain boundaries. This resulted in a homogeneous
austenitic structure. The structure increases with the
resistance of these steels to intergranular corrosion and
the metallurgical point of view is correct. The disadvant-
age is a significant increase in the ductility and plasticity
of these steels, which is highly undesirable during ope-
ration. From the metallurgical point of view, the disting-
uishing feature of poor machinability is the kinematic
coarse austenite, almost carbides. The sign of good
machinability is a fine-grained austenite with plenty of
finely distributed carbides. Machinability is related to the
economy of production. The aim is to produce the maxi-
mum performance with the available resources. Machin-
ability influences and even determines the cutting forces,
heat and cutting temperature, chip formation, wear and
tool life, but also the surface integrity. Cutting parame-
ters such as the cutting speed vc and the feed rate f play
critical roles in the cutting temperature and the surface
roughness in the turning processes. The surface rough-
ness, which is used for the evaluation of the product
quality, is an important performance characteristic in
turning processes.18 That is, cutting speeds out of the
range recommended by tool manufacturers (cutting
speed in the range of vc = 180–250 m min–1) were tested.
The objective was to analyse the effect of cutting speed
over the work material–toolpair.19 I. Korkut et al.19 and
I. Ciftci et al.20 reported that during the turning of AISI
304 austenitic stainless steel using a multilayer (CVD)
coated tool, the tool flank wear decreases with an
increasing cutting speed up to 180 m min–1 and the sur-
face roughness values decrease with the increasing the
cutting speed. The poor performance of the tool at lower
cutting speeds can be explained by the influence of the
heat on the cutting tool. That is because, metal cutting
involves the generation of a large amount of heat and in
the machining of AISI 304 stainless steel it is not
dissipated rapidly due to the low thermal conductivity of
this material. The heat generation principally occurs in
three areas: the shear zone, the rake face and on the
clearance side of the cutting edge.20 W. Grzesik et al.21,
studied the machinability of AISI 304 and C45 steel
using CVD TiC, TiN/ TiC and TiN/Al2O3/TiC coated and
uncoated (P20) cemented carbide tools. They found that
in the case of TiC and TiN/Al2O3/TiC coating the spe-
cific cutting pressure decreases and for the TiC/TiN
coating it increases. In another study they found a low
value of the surface roughness for the coating
TiC/Al2O3/TiN. In addition, they found that as the
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
440 Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445
cutting speed increases the cutting force and the contact
length decreases. A. Hosokawa et al.22 reported that
during its own realization of turning tests of stainless
steel (AISI 304) were carried out in order to examine the
tool-wear characteristics. W. I. H. Liew23 investigated the
wear characteristics of PCBN (polycrystalline cubical
boron nitride) tools in the ultra-precision machining of
stainless steel.22 During lathe turning, the machined
surface is work hardened. This work-hardened surface is
machined on the next lathe turning step, which accele-
rates the tool wear.23 This results in the degradation of
the surface quality and the acceleration of the adhesive
wear of the tools used.24 All of these experiments deter-
mine that the purpose and the machinability significantly
affect the cutting process. The machining of high-
strength materials can cause brittle fracture of the cutting
edge parts. This is due to the high cutting forces during
machining. In a practical assessment of the size of the
wear it is most common to use parameters such as the
width of the wear pads on the back of the VBk and the
depth of the groove at the forefront of the KT (according
to ISO 3685). The intensity of the wear may affect the
cutting conditions. The biggest influence is the cutting
speed, then the feed rate and a smaller minimum is the
depth of cut. When turning austenitic stainless steels,
they are a generally chosen criterion of the flank wear
VBk, given the increase in cutting forces with increasing
wear. There is also an increase in the temperature and the
intensity of the wear of the cutting tool. The duration of
the work of the cutting tool with a target of VBk wear, the
tool life T can be determined within minutes of the
machining time, the number of machined parts or the
cutting tool path. The cutting tool’s ability to restore its
sharpness or setting a new cutting edge with a cutting
tool exchange. In practice we try to choose the criterion
of wear, so that we have the maximum tool life.
Selection criteria and the process limits the wear and
machined surface roughness achieved, increased cutting
forces, and the emergence of oscillations in the system’s
working. We are talking about the technological criterion
of wear. The tool life depends on the cutting conditions,
the geometry of the cutting wedge-shaped tools, the
cutting material, the fluency of the cutting process, the
method and type of operation and the workpiece
material.
3 EXPERIMENTAL DETAILS
The main aim of the paper was the measurement of
the tool life T = f (vc) for various cutting speeds. The
experiments were performed in the tool AERO TURN
BT-380 CNC machine (Figure 1) with a maximum
spindle speed nmax = 4500 min–1 and performance Pc =
11.5 kW with a turret for clamping the cutting tools
(Figure 2) and a CNC control system FANUC Series 0i
– TC. The workpiece material was austenitic stainless
steel 1.4301(microstructure can be seen in Figure 3),
Ø 60 mm × 200 mm, hardness HB 190. The clamping
was in a three-jaw chuck with the turned inside diameter
Ø 60 mm in length l = 15 mm, clamping the workpiece
by the tail stock. The cutting tool was a side cutting tool
holder r = 95°, with geometry PCLNR 2525 (PRAMET
Tools). The carbide cutting insert was CNMG
120408E-NM, carbide type GC 2025 (PRAMET Tools),
used for rough machining of the austenitic stainless
steels (SEM microstructure of cutting edge’s appearance
can see in Figure 4).
Cutting conditions:
• Depth of cut ap0 = 1.0 mm,
• Feed rate f0 = 0.15 mm/rev,
• Cutting speed vc1 = 250 m min–1, vc2 = 200 m min–1,
vc3 = 150 m min–1 with usage of the coolant E5%.
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445 441
Figure 2: a) Cutting area of CNC machine tool with clamped work-
piece, b) control panel of the AERO TURN BT-380 CNC machine
tool with system FANUC Series 0i – TC, c) view of the workpiece
(tested part) used for the turning experiments
Slika 2: a) Podro~je CNC stru`enja z vpetim obdelovancem, b) kon-
trolna plo{~a CNC stru`nice AERO TURN BT-380 s sistemom
FANUC Serije 0i-TC, c) izgled obdelovanca (vzorca) na katerem so
bili izvr{eni preizkusi stru`enja
Figure 1: Overall view of the AERO TURN BT-380 CNC machine
tool
Slika 1: Izgled CNC stru`nice AERO TURN BT-380
4 RESULTS AND DISCUSSION
To determine the dependence of T = f (vc) we must
satisfy the condition vcmax = 2.5 vcmin and the cutting-tool
wear criteria of VBk = 0.2 mm. The scheme of the
gradual removal of material during the turning in the
virtual interface of the CATIA V5 Lathe Machining
system is shown in Figure 5. The turning tests were
carried out for the following values of the cutting speeds
vc1 = 250 m min–1, vc2 = 200 m min–1, vc3 = 150 m min–1,
and the cutting speed vc4 = 100 m min–1 was determined
by the calculation method. The CNC machine tool cal-
culates vc = const. directly from the turning diameter. For
the average diameter Ø58 the cutting speed vc = 250
m min–1 and the spindle speed n = 1348 min–1. For the
maximum removed diameter the cutting speed vc = 254
m min–1. With each cut the spindle speed of the CNC
machine tool also changes. With a change of the dia-
meter from Ø60 mm to Ø30 mm there is also a change of
the feed velocity vf. Even the machining time for one cut
also changes as follows.
Different criteria can be used for the measurement of
tool life such as the average of the maximum flank wear,
the surface roughness and the number of components per
tool. In this experimental investigation the criterion of an
average flank wear VB = 0.2 mm was considered for the
tool-life measurement. Fernández-Abbia et al.1,2 reported
that the cutting speed in the range of 200 to 300 m min–1
is favourable for the machining of 1.4301 stainless steel.
After each pass, the cutting insert was used for a
measurement of the tool wear and thus wear progress
was obtained. Figure 6 shows the tool-life curves at
(150, 200, 250) m min–1 cutting speed for a constant feed
rate and depth of cut. The tool life demonstrates three
wear stages. Flank wear VB of carbide insert at vc2 = 200
m min–1 can be seen in Figure 7.
Table 1: The calculation table to determine the tool life T (min)
Tabela 1: Tabela za izra~un ~asa T (min) zdr`ljivosti orodja
N vci Ti log vci log Ti log vci .log Ti
log vci2
1 150 48.8 2.17609 1.68842 3.67416 4.73537
2 200 28.9 2.30103 1.46090 3.36157 5.29474
3 250 19.5 2.39794 1.29003 3.09343 5.75012
 – – 6.87506 4.43935 10.12915 15.77998
Auxiliary calculation ( log vci)2 = 6.875062 = 47.2664
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
442 Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445
Figure 4: Surface morphology of the CNMG 120408E-FM carbide
cutting insert (edge) surface appearance, SEM image
Slika 4: SEM-posnetek morfologije povr{ine CNMG 120408E-FM
karbidnega rezalnega vlo`ka (rob)
Figure 3: The microstructure of DIN 1.4301 (AISI 304) with an auste-
nitic structure. The microstructure consists of large grains of austenite
(grey) and small grains of carbides (black).13
Slika 3: Avstenitna mikrostruktura jekla DIN 1.4301 (AISI 304).
Mikrostruktura je sestavljena iz velikih zrn avstenita (sive barve) in
majhnih karbidnih zrn (~rne barve).13
Figure 6: The graphical dependence of the tool flank wear on the
machining time during the turning of DIN 1.4301
Slika 6: Grafi~na odvisnost obrabe boka orodja od ~asa stru`enja jekla
DIN 1.4301
Figure 5: 3D simulation of the longitudinal turning process in CATIA
Slika 5: CATIA tridimenzionalna simulacija vzdol`nega procesa
stru`enja
In the process of investigating the quality indicators
in terms of the surface integrity during the turning of
1.4301 austenitic stainless steel, the authors of the article
also dealt with the dependence of the arithmetic mean
surface roughness Ra = f (vc) in changing the fillet radius
of the used cutting tool (the difference can see in Figure
8). The surface roughness measurement was carried out
on the machined surfaces using a Taylor Hobson measur-
ing device. The value reported represents the average of
the surface roughness value obtained from at least three
measurements. The surface quality of the machined
surface is mainly dependent on the used cutting condi-
tions and it plays a significant role in the functionality
and fatigue life of the component.
The measured spindle speeds ni, the feed velocity
rates vfi and the machining times for the individual cuts
were determined from a calculation. Sample no. 4 was
machined with the same method and with the same
cutting parameters as the sample no. 3. Flank wear con-
trol was carried out after 23.3 min, 28.8 min, and 33.06
min, and then the sample no. 5 was machined with a
flank wear measurement after 39.83 min, 45.33 min, and
50.83 min. The measured values of the flank wear VBmax
for vc1 = 250 m min–1, vc2 = 200 m min–1, vc3 = 150
m min–1, with ap = 1.0 mm and f = 0.15 mm/rev are
determined in this experiment. Then follows an exchange
(rotation) of the cutting insert again. Then continued the
completion of the turning of sample no. 5 from the
diameter Ø40 mm to the diameter Ø30 mm on the L =
180 mm with vc2 = 200 m min–1 = const. with the same
cutting parameters ap = 1.0 mm, f = 0.15 mm/rev, with
usage of coolant. The measured spindle speeds ni, the
feed velocity rates vfi and the machining times for the
individual cuts were determined using a calculation.
During this phase of the experiment we turned samples
no. 5, 6, 7, and 8 from the overall number of 10 pieces.
The graphical dependence of the cutting tool wear on the
machining time for the cutting speeds vc1, vc2, vc3 from the
turning of austenitic stainless steel 1.4301 with a cutting
parameter depth of cut ap = 1.0 mm and feed rate f = 0.15
mm/rev, with coolant is shown in Figure 6. Three points
of the measurement in the dependence T = f (vc) accord-
ing to the relevant equation and determines the shape of
the curve as linear in the logarithmic coordinates. For the
calculation we used values directly from Table 1. The
PRAMET Tools (Sandvik Group Sweden) is recom-
mended for the tool life of the cemented cutting inserts
with the coating type GC 2025 at the cutting speed vc =
250 m min–1 with the value of tool life T = 18 min, esta-
blishing the criterion of wear VBk = 0.2 mm. Since the
cemented carbide insert type GC 2025 has a multilayer
coating TiN+Al2O3+TiC on the fine-grained substrate
WC+Co it achieves an even higher durability.
There is an analytical description for determining the
dependence of the tool life T = f (vc) with a value of vc =
100 m min–1 in the following section. The linear regres-
sion of the single parameter is:
y = b0 · x0 + b1 · x1 (1)
Then the x0 is a fictitious value, which has a value of
1 for the integer scale. For the logarithmic scale of log 10
= 1, the x1 is an independent variable, the b0 is an additive
constant, which shows the growth on the axis "y", and
the b1 indicates the slope of the regression function. The
values b0, b1 are then calculated using the following
Equation (2) (x) to (y):
b
N T v T v
N v
i i
i
N
i
N
i
N
i i
1
1 11
2
=
⋅ ⋅ − ⋅
⋅
= ==
∑ ∑∑ (lg lg ) lg lg
(lg
c c
c ci i
v
i
N
i
N
) lg−
⎛
⎝
⎜ ⎞
⎠
⎟
==
∑∑
1
2
1
(2)
Substituting b1 into the Equation (2) we obtain the
constant b0, Equation (3):
b
T b v
N
i
i
N
i
N
i
0
1
11=
−
==
∑∑ lg lg c
(3)
where i to N is the number of measurements
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445 443
Figure 7: Flank wear VB of cemented carbide insert at cutting speed
vc2 = 200 m min–1, established criterion of the flank wear VB = 0.2 mm
Slika 7: Obraba roba VB karbidnega vlo`ka pri hitrosti rezanja vc2 =
200 m min–1, uveljavljeno merilo za obrabo roba VB = 0,2 mm
Figure 8: The graphical dependence Ra = f (vc) in the turning of DIN
1.4301 steel with CNMG 12048E with different r

Slika 8: Grafi~na odvisnost Ra = f (vc) pri stru`enju jekla DIN 1.4301
z CNMG 12048E, z razli~nimi r

b1
3 1012915 4 43935 6 87506
3 15 77998 47 266
=
⋅ − ⋅
⋅ −
( . ) ( . , )
. . 4
= –1.81327
b0
4 43935 181327 687506
3
=
− − ⋅. ( . ) .
= 5.63523
y = b0 · x0 + b1 · x1 = 5.63523 · x0 – 1.81327 · x1
The introduction of the substitution for b0 = log CT
and for y = log T is then CT = 105,63523 = 431,748.103.
Then the value is tg  = 1.81327, thereof  = arctg 1.81327
and consequently the size of the angle is  = 61° 21’.
The inclination angle of the line in the logarithmic
coordinates is –b1 = m = tg , and from which we obtain
the value of the angle .
The shape of the linear regression for the tool life has
the following form:
log T = log CT – m log vc (4)
The equation according to Taylor (2) for the material
1.4301 (AISI 304) will be in the following form for the
cutting conditions ap = 1 mm and f = 0.15 mm/rev.
T
v c
=
⋅431748 10 3
1 81327
.
.
Then for the tool-life calculation of cutting edge for
the cutting speed vc4 = 100 m min–1 the following for-
mula is used:
T
T
v
v
c
c
m
4
1
1
4
=
⎛
⎝
⎜
⎞
⎠
⎟ then T4
1 81327
195
250
100
102 7= ⎛⎝
⎜ ⎞
⎠
⎟ =. .
.
min
5 CONCLUSIONS
The main aim of the presented paper is an experi-
mental determination of the tool life depending on the
cutting speed according to Taylor in turning the auste-
nitic stainless steel 1.4301 (AISI 304). The cutting speed
and the feed rate have a significant effect on the flank
wear. The tool life is significantly influenced by the
cutting parameters, the surface roughness and the flank
wear. The obtained results are statistically processed
using a linear regression analysis with the method of
least squares. The results and values are shown in Table 1,
and the graphical dependence of the flank wear VBk on
the time is shown in Figure 6. The flank wear (Figure 7)
and the tool life of the cemented carbide insert were
monitored so that there is no reduction in the quality of
the surface finish. These areas were defined by the
technical documentation. For the cemented carbide insert
wear always occurs at one point when there is a variable
depth of cut (can see in Figure 7). For example, this can
be avoided by using of the CNC program preparation in
the three-dimensional CATIA interface (Figure 5). The
CNC program divides the allowance for machining so
that the next depth of cut was slightly smaller than the
previous one. The advantage of this cycle is the fact that
the cutting tool is not still loaded in the same area, but
over the range of applied depth of cuts. When machining
austenitic stainless steels we should follow these rules,
which help to increase the durability of the cutting edge
of the carbide cutting tool and thus the quality of the
machined surface.
• The first rule is basically to use cutting inserts coated
with CVD+PVD (for example, TiC+Al2O3+TiN on
the cutting tool surface).
• The second is to use a washer of cemented carbide
directly below the cemented carbide inserts.
• The third rule is to visually diagnose and timely eli-
minate the causes of premature damage to the cutting
edge (notch).
• The fourth rule is to use the best applications for the
changeable cutting inserts for dimensionally demand-
ing workpieces.
In terms of the defined cutting parameters the
greatest impact comes from the cutting speed vc, a lower
feed rate f and the least depth of cut ap. We did not study
more tool life depending T = f(f) and T = f(ap) for this
reason in CNC machine tools, and the size of the ob-
served flank wear. This creates space for the realization
of further research in this area. The applied, discovered
knowledge from the literature sources and the experi-
ments conducted here can be used in the future for the
manufacture of specific parts on CNC machine tools
with new, progressive cutting tools.
Acknowledgement
This paper was supported in the frame of the project
"Alexander Dub~ek University of Tren~ín – Faculty of
special technology wants to offer high-quality and
modern science/research and education", ITMS code
26110230099, based on the Operational Programme
Education. Modern education for knowledge society /
The project is co-funded by European Social Fund and
also includes the results of the grant VEGA no.
1/9428/02 titled "The technological heredity of the
machined surfaces – surface integrity".
6 REFERENCES
1 A. I. Fernández-Abia, J. Barreiro, J. Fernández-Larrinoa, L. N.
López de Lacalle, A. Fernández-Valdivielso, O.M. Perreira, Beha-
viour of PVD coatings in the turning of austenitic stainless steels,
Procedia Engineering, 63 (2013), 133–141, doi:10.1016/j.proeng.
2013.08.241
2 A. I. Fernández-Abia, J. Barreiro, L. N. López de Lacalle, S. Martí-
nez, Effect of very high cutting speeds on shearing, cutting forces
and roughness in dry turning of austenitic stainless steels, Inter-
national Journal of Advanced Manufacturing Technology, 57 (2011)
1–4, 61–71, doi:10.1007/s00170-011-3267-9
3 M. A. Xavior, M. Adithan, Determining the influence of cutting
fluids on tool wear and surface roughness during turning of AISI 304
austenitic stainless steel, Journal of Material Processing Technology,
209 (2009) 2, 900–909, doi:10.1016/j.jmatprotec.2008.02.068
4 A. P. Kulkarni, G. G. Joshi, V. G. Sargade, Dry turning of AISI 304
austenitic stainless steel using AlTiCrN coated insert produced by
HPPMS technique, Procedia Engineering, 64 (2013), 737–746,
doi:10.1016/j.proeng.2013.09.149
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
444 Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445
5 S. S. Wagh, A. P. Kulkarni, V. G. Sargade, Machinability studies of
austenitic stainless steel (AISI 304) using PVD cathodic arc evapo-
ration (CAE) system deposited AlCrN/TiAlN coated carbide inserts,
Procedia Engineering, 64) (2013), 907–914, doi:10.1016/j.proeng.
2013.09.167
6 S. Emura, M. Kawajiri, X. Min, S. Yamamoto, K. Sakuraya, K. Tsu-
zaki, Machinability improvement and its mechanism in SUS304
austenitic stainless steel by precipitated hexagonal boron nitride, ISIJ
International, 53 (2013) 10, 1841–1849, doi:10.2355/isijinterna-
tional.53.1841
7 J. Paro, H. Hänninen, V. Kauppinen, Tool wear and machinability of
X5CrMn18 stainless steel, Journal of Materials Processing Tech-
nology, 119 (2001) 1–3, 14–20, doi:10.1016/S0924-0136(01)
00877-9
8 F. Koné, C. Czarnota, B. Haddag, M. Nouari, Modeling of velo-
city-dependent chip flow angle and experimental analysis when
machining 304L austenitic stainless steel with groove coated-carbide
tools, Journal of Materials Processing Technology, 213 (2013) 7,
1166–1178, doi:10.1016/j.jmat-protec.2013.01.015
9 S. E. Cordes, Thermal stability of 
-alumina PVD coatings and
analysis of their performance in machining of austenitic stainless
steels, CIRP Journal of Manufacturing Science and Technology, 5
(2012) 1, 20–25, doi:10.1016/j.cirpj.2011.11. 003
10 R. ^ep, A. Janásek, A. Slíva, M. Neslu{an, M. Binder, Experimental
tool life tests of indexable inserts for stainless steel machining,
Tehnicki Vjesnik, 20 (2013) 6, 933–940
11 V. T. G. Naves, M. B. Da Silva, F. J. Da Silva, Evaluation of the
effect of application of cutting fluid at high pressure on tool wear
during turning operation of AISI 316 austenitic stainless steel, Wear,
302 (2013) 1–2, 1201–1208, doi:10.1016/j.wear.2013.03.016
12 A. I. Fernández-Abia, J. Barreiro, L. N. López de Lacalle, S.
Martínez-Pellitero, Behavior of austenitic stainless steels at high
speed turning using specific force coefficients, International Journal
of Advanced Manufacturing Technology, 62 (2012) 5–8, 505–515,
doi:10.1007/s00170-011-3846-9
13 R. K. C. Nkhoma, C. W. Siyasiya, W. E. Stumpf, Hot workability of
AISI 321 and AISI 304 austenitic stainless steels, Journal of Alloys
and Compounds, 595 (2014), 103–112, doi:10.1016/j.jallcom.2014.
01.157
14 J. Jambor, Improving the quality of the machining process through
new technologies, Annals of DAAAM for 2012 & Proceedings –
World Symposium Intelligent Manufacturing & Automation, 23
2012, 623–626
15 Z. Tekiner, S. Yesilyurt, Investigation of the cutting parameters
depending on process sound during turning of AISI 304 austenitic
stainless steel, Materials and Design, 25 (2004) 6, 507–513,
doi:10.1016/j.matdes.2003.12.011
16 T. Akasawa, H. Sakurai, M. Nakamura, T. Tanaka, K. Takano,
Effects of free-cutting additives on the machinability of austenitic
stainless steels, Journal of Materials Processing Technology,
143–144 (2003) 1, 66–71, doi:10.1016/S0924-0136(03)00321-2
17 M. Kaladhar, K. Venkata Subbaiah, C. Srinivasa Rao, Parametric
optimization during machining of AISI 304 austenitic stainless steel
using CVD coated Duratomic™ cutting insert, International Journal
of Industrial Engineering Computations, 3 (2012) 4, 577–586,
doi:10.5267/j.ijiec. 2012.04.002
18 M. Aydin, C. Karakuzu, M. Uçar, A. Cengiz, M. A. Cavuºlu, Predic-
tion of surface roughness and cutting zone temperature in dry turning
processes of AISI304 stainless steel using ANFIS with PSO learning,
Int. Journal of Advanced Manufacturing Technology, 67 (2013) 1–4,
957–967, doi:10.1007/s00170-012-4540-2
19 I. Korkut, M. Kasap, I. Ciftci, U. Seker, Determination of optimum
cutting parameters during machining of AISI 304 austenitic stainless
steel, Materials and Design, 25 (2004) 4, 303–305, doi:10.1016/
j.matdes.2003.10.011
20 I. Ciftci, Machining of austenitic stainless steels using CVD multi-
layer coated cemented carbide tools, Tribology International, 39
(2006) 6, 565–569, doi:10.1016/j.triboint. 2005.05.005
21 W. Grzesik, The role of coatings in controlling the cutting process
when turning with coated indexable inserts, Journal of Material Pro-
cessing Technology, 79 (1998) 1–3, 133–143, doi:10.1016/S0924-
0136(97)00491-3
22 A. Hosokawa, T. Ueda, R. Onishi, R. Tanaka, T. Furumoto, Turning
of difficult-to-machine materials with actively driven rotary tool,
CIRP Annals – Manufacturing Technology, 59 (2010) 1, 89–92,
doi:10.1016/j.cirp.2010.03.053
23 W. I. H. Liew, B. K. A. Ngoi, Y. G. Lu, Wear characteristics of
PCBN tools in the ultra-precision machining of stainless steel at low
speeds, Wear, 254 (2003) 3–4, 265–277, doi:10.1016/S0043-1648
(03)00002-4
24 M. Hua, W. Xicheng, L. Jian, Friction and wear behaviour of SUS
304 austenitic stainless steel against Al2O3 ceramic ball under rela-
tive high load, Wear, 265 (2008) 5–6, 799–810, doi:10.1016/j.wear.
2008.01.017
R. DUBOVSKÁ et al.: INVESTIGATING THE INFLUENCE OF CUTTING SPEED ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 439–445 445

A. MICHALCOVÁ et al.: NiAl INTERMETALLIC PREPARED WITH REACTIVE SINTERING ...
447–450
NiAl INTERMETALLIC PREPARED WITH REACTIVE SINTERING
AND SUBSEQUENT POWDER-METALLURGICAL
PLASMA-SINTERING COMPACTION
REAKCIJSKO SINTRANJE IN ZGO[^EVANJE S PLAZEMSKIM
SINTRANJEM NiAl INTERMETALNE ZLITINE
Alena Michalcová1, Dalibor Vojtìch1, Tomá{ Franti{ek Kubatík2, Pavel Novák1,
Petr Dvoøák1, Petra Svobodová1, Ivo Marek1
1University of Chemistry and Technology, Department of Metals and Corrosion Engineering, Prague, Technická 5, 166 28 Prague 6,
Czech Republic
2Institute of Plasma Physics AS CR, v. v. i., Za Slovankou 1782/3, 182 00 Prague 8, Czech Republic
michalca@vscht.cz
Prejem rokopisa – received: 2015-04-30; sprejem za objavo – accepted for publication: 2015-06-02
doi:10.17222/mit.2015.089
This paper proposes a novel method for powder-metallurgy preparation of compact NiAl intermetallics. In the first step, the
NiAl powder is prepared with the reactive-sintering procedure. The porous NiAl product of the SHS reaction is milled to a fine
powder and consequently compacted by SPS processing. The compaction of powder metals and alloys is a very difficult field
due to the need of preserving the unique properties of the initial materials. One of the few possible methods of a successful
compaction is plasma sintering. To describe detailed structures of powder-metallurgy materials, it is necessary to use advanced
microscopy methods such as SEM and TEM. In this study, the structure of a NiAl intermetallic compound is described. The
material was first produced, with reactive sintering, from pure elements. Subsequently, the NiAl porous master alloy was milled
and compacted with the spark-plasma-sintering (SPS) technique. The particle size of the NiAl powder was comparable to the
grain size of the compacted material, which exhibited a low porosity. It was proved that the interconnection of the NiAl particles
is made by a thin layer of nanocrystalline oxides.
Keywords: SPS, intermetallics, powder metallurgy
^lanek predlaga novo metodo za pripravo kompaktne NiAl intermetalne zlitine s pomo~jo metalurgije prahov. V prvem koraku
je bil NiAl prah pripravljen s postopkom reakcijskega sintranja. Z SHS reakcijo proizvedeni porozni NiAl je bil zmlet v droben
prah in nato kompaktiran z SPS postopkom. Kompaktiranje kovinskega prahu je te`avno zaradi potrebe po zadr`anju enkratnih
lastnosti za~etnih materialov. Ena od redkih uspe{nih metod kompaktiranja je sintranje s plazmo. Za podroben opis mikro-
strukture materialov v metalurgiji prahov je potrebno uporabiti napredne mikroskopske metode, kot sta SEM in TEM. V {tudiji
je opisana struktura intermetalne zlitine NiAl. Material je bil najprej izdelan z reakcijskim sintranjem iz ~istih elementov. Nato
je bila porozna zlitina NiAl zmleta in kompaktirana s tehniko iskrilnega plazma sintranja (SPS). Velikost delcev prahu NiAl je
bila primerljiva z velikostjo zrn v kompaktiranem materialu, ki je imel tudi majhno poroznost. Dokazano je bilo, da se povezava
delcev NiAl izvede s tanko plastjo nanokristalini~nih oksidov.
Klju~ne besede: SPS, intermetalne zlitine, metalurgija prahov
1 INTRODUCTION
Like many other transition metal aluminides, nickel
aluminide exhibits properties that are very interesting for
industrial utilization. These are a high melting point
(1638 °C), a low density (5.95 g/cm3), a high thermal
conductivity (70 W m–1K–1), an excellent corrosion resis-
tence1,2 and a very good wear resistance.3 These
properties allow intermetallics to be used in the applica-
tions where metallic and ceramic materials fail. In
addition, nickel aluminide is easily produced in atmo-
spheric air with a self-propagating high-temperature
synthesis (SHS)2,4–5 even when pre-pressed into a green
body.4 This makes the Ni-Al system to be an ideal model
for the study of a possible powder preparation using
metallurgical methods based on SHS.
The intermetallic materials usually exhibit good me-
chanical properties at elevated temperatures, but
unfortunately, they seem to be quite brittle at room
temperature. When decreasing the grain size of a ma-
terial, this factor limiting its utilization can be solved.
One of the promising ways is to produce fine-grained
intermetallics with a two-step powder-metallurgy
method: in the first step, an intermetallic is formed with
the SHS procedure; then it is milled to a very fine pow-
der and compacted with the spark-plasma-sintering
(SPS) procedure. The advantage of the SPS process lies
in extremely short sintering times, due to which there is
almost no grain coarsening.6–9 The SPS method is well
described for ceramics, but for metals and especially for
intermetallics, the description of the process is still being
formed.6–9
The spark-plasma sintering method has been very
popular in the last two decades, mainly in the field of
compaction of ceramics. It is an ideal tool for obtaining
homogenous nanocrystalline bulk materials with a high
density, i.e., fine-grained ceramics, thermo-electric semi-
Materiali in tehnologije / Materials and technology 50 (2016) 3, 447–450 447
UDK 669.245:669.715:621.762.5 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)447(2016)
conductors and biomaterials.6 Compared to the other
compaction methods (cold and hot isostatic pressing),
the SPS is distinguished by a low overall sintering tem-
perature, short sintering times and better properties of
the prepared bulk materials.7
Using the SPS method, many successes were
achieved in the fields of increasing the superplasticity of
ceramic materials, improvement of magnetic properties,
reduction of the amount of impurities segregated at the
grain boundaries, improvement in the binding quality
and many others.7 From the historical point of view, the
first machine comparable to SPS was built in Germany,
as reported in reference6. In 1933, in the USA, F. Taylor
was awarded a patent for the first resistance-sintering
method used for sheets.10
Basically, the SPS method for sintering materials can
be divided into four generations: the first SPS was built
in Japan (in 1962) and called spark sintering (SS).11 The
next generation can be described as plasma-activated sin-
tering (PAS), followed by spark-plasma sintering (SPS),
while the fourth and currently the last generation is the
one described in 12.
The study of a NiAl alloy prepared with SPS can be
used, in future, as a milestone for the preparation of
NiAl-based composites.2 Preparation conditions can be
easily changed by adding reinforcements to the reaction
system before the SHS reaction or by adding them to the
powder before the SPS compaction.
2 EXPERIMENTAL WORK
The NiAl intermetallic compound was prepared with
an SHS synthesis. A high-purity nickel powder with a
particle size <100 μm and an aluminium powder with a
purity of 99.99 % and a particle size of 200–400 μm
were mixed and pressed at room temperature with a
pressure of 260 MPa using a LabTest 5.250SP1-VM
universal testing machine. Reactive sintering of the
pressed powder mixtures was carried out at 900 °C for
15 min in the usual furnace (air) atmosphere. The sin-
tered particles with an approximately cylindrical shape
and a size of 1 cm in diameter and 1 cm in height were
milled with a laboratory vibration mill VM4. The ob-
tained NiAl powder was leached in a 20 % NaOH solu-
tion to dilute any residual Al. The NiAl powder was
compacted with the SPS procedure (model SPS 10-4
thermal technology) at a temperature of 1100 °C, for a
compaction-process time of 5 min and at a pressure of 80
MPa. The SPS die is made of carbon and its internal dia-
meter is 19.3 mm. To separate the sintered material from
the die, a carbon foil with a thickness of 0.15 mm was
used. The amount of compacted material was approxi-
mately 5 g for each experiment.
The structures of the SHS material, the NiAl powder
and the SPS-compacted material were observed with an
Olympus PME3 light microscope and a TESCAN VEGA
3 LMU scanning electron microscope equipped with
EDS and EBSD detectors (Oxford Instruments). The
phase compositions of the materials were determined us-
ing X-ray diffraction (PAN analytical X’Pert PRO +
High Score Plus, Cu anode). TEM samples were pre-
pared by ion polishing using Gatan PIPS Model 691 and
consequently observed with a Jeol JEM 3010 trans-
mission electron microscope. SAED patterns were
integrated and phases were identified using Process
Diffraction software. The hardness of the materials was
measured with a FUTURE-TECH FM700 hardness
tester with loads of 10 g and 1 kg.
3 RESULTS AND DISCUSSION
The samples prepared with the SHS procedure had
approximately cylindrical shapes. They were mainly
composed of the NiAl phase with a low amount of resid-
ual Al in the surroundings of the pores. As illustrated in
Figure 1, the porosity of the SHS samples is extremely
high.
The NiAl particles were milled into a powder, whose
structure is shown in Figure 2. The particles have irregu-
lar shapes, as expected after milling a brittle material.
A. MICHALCOVÁ et al.: NiAl INTERMETALLIC PREPARED WITH REACTIVE SINTERING ...
448 Materiali in tehnologije / Materials and technology 50 (2016) 3, 447–450
Figure 2: Structure of NiAl powder (LM)
Slika 2: Struktura NiAl prahu (LM)
Figure 1: Structure of a NiAl particle prepared with SHS procedure
(SEM/BSE)
Slika 1: Struktura delca NiAl, izdelanega z SHS postopkom
(SEM/BSE)
The size of the majority (96 %) of the particles is less
than 140 μm. The phase composition of the powder is
given in Figure 3. Peaks of residual aluminium are also
visible in the milled powder. Although the powder was
leached with a 20 % NaOH solution, areas of residual Al
are still shown in Figure 1.
Subsequently, the powder was compacted with the
SPS method at 1100 °C for 5 min. The structure of the
SPS-prepared material is given in Figure 4. The particles
of the initial powder are clearly distinguishable. The dark
parts in the structure are pores. The porosity of the
SPS-prepared material is 1.9 ± 0.9 %, which is satisfac-
tory for the material prepared by powder-metallurgy
processing.
The grains of the compacted material are formed by
the particles of the initial powder and no grain coarsen-
ing is observed. It can be supposed that the grain size of
the compacted material depends only on the particle size
of the initial powder. The plot in Figure 5 shows the
particle-size distribution of the initial NiAl powder and
the grain-size distribution of the SPS-compacted mater-
ial. It seems that the powder contains more particles with
a size of up to 20 μm. This slight disagreement can be
caused by a measurement error. Small particles located
at the grain boundaries cannot be distinguished as easily
as the separate particles in the mounting material.
The EBSD analysis (Figure 6) of the SPS-compacted
material proved that the area of the initial-powder parti-
cles is monocrystalline. Between the large, clearly seen
particles (grains of the compacted material), there are
areas where the crystallographic orientation is not very
clear. These areas at the grain boundaries can exhibit a
large misorientation or can be oxidised.
The amount of residual Al is lower than a 2–5 % of
mass fraction because it is not detectable with XRD, as
shown in Figure 3. The same is true of the oxide content
in the initial powder and also in the SPS-compacted
A. MICHALCOVÁ et al.: NiAl INTERMETALLIC PREPARED WITH REACTIVE SINTERING ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 447–450 449
Figure 3: XRD pattern of NiAl powder and compacted material
(1 = NiAl, 2 = graphite)
Slika 3: Rentgenogram prahu NiAl in kompaktiranega materiala
(1 = NiAl, 2 = grafit)
Figure 6: SEM micrograph of NiAl material compacted from powder
with SPS method (1100 °C/5min) and EBSD scan of the area
Slika 6: SEM-posnetek NiAl materiala, kompaktiranega iz prahu po
SPS metodi (1100 °C /5 min) in EBSD posnetek podro~ja
Figure 4: Structure of NiAl material compacted from powder with
SPS method (1100 °C/5 min) (LM)
Slika 4: Struktura NiAl materiala, kompaktiranega iz prahu po SPS
metodi (1100 °C/5 min) (LM)
Figure 5: Grain (particle) size distribution of NiAl powder and com-
pacted material
Slika 5: Razporeditev velikosti delcev prahu NiAl in kompaktiranega
materiala
Figure 7: TEM micrograph of NiAl material compacted from powder
with SPS method (1100 °C/5min)
Slika 7: TEM-posnetek NiAl kompaktiranega materiala iz prahu po
metodi SPS (1100 °C /5 min)
product. The only excess peak in the XRD pattern of the
SPS-compacted material relates to the graphite from the
protection graphite foil used in the SPS process.
A detailed material observation made using TEM is
given in Figure 7 and it shows the structure of a grain
boundary. In the left bottom part, a dark NiAl particle is
located. It can be seen that the particles are connected by
a nanocrystalline oxide interlayer. The amount, thickness
and crystallinity of the oxide layer are not sufficient to
be detected with XRD or EBSD analysis, but they can be
distinguished with selected area electron diffraction
(SAED), as shown in Figure 8.
The fact that the weak parts of the material are the
grain boundaries is also proved with the hardness mea-
surement. While the microhardness (inside individual
particles) is the same for the SHS material and for the
SPS-compacted material, the macrohardness (measured
with a load of 1 kg) varies significantly (Figure 9).
These results indicate that micro-properties stay the
same after a consolidation, while macro-properties
change significantly due to the formation of an oxide
interlayer during a compaction. The question is what
would happen if the SHS process was performed in an
inert atmosphere.
4 CONCLUSION
The powder-metallurgy preparation of NiAl consist-
ing of the SHS NiAl preparation, the milling and the SPS
compaction is a promising method for obtaining bulk
intermetallic materials. The grain size of an SPS-com-
pacted material is mainly determined by the grain-size
distribution of the initial powder. The grain size was
estimated to be less than 40 μm. It was proved that the
particles of the initial powder are interconnected by a
thin oxide layer, which decreases the macroscopic and
also microscopic properties of the material.
Acknowledgement
This research was financially supported by the Czech
Science Foundation, project No. P108/12/G043.
5 REFERENCES
1 D. Tingaud, L. Stuppfler, S. Paris, D. Vrel, F. Bernard, C. Penot, F.
Nardou, Time-Resolved X-ray Diffraction Study of SHS-produced
NiAl and NiAl–ZrO2 Composites, International Journal of Self-
Propagating High-Temperature Synthesis, 16 (2007) 1, 12–17,
doi:10.3103/S1061386207010025
2 P. Novák, D. [otka, M. Novák, A. Michalcová, J. [erák, D. Vojtìch,
Production of NiAl–matrix composites by reactive sintering, Powder
Metallurgy, 54 (2011) 3, 308–313, doi:10.1179/003258909X
12518163
3 J. Guo, Z. Wang, L. Sheng, L. Zhou, C. Youya, Z. Chen, L. Song,
Wear properties of NiAl based materials, Progress in Natural Sci-
ence: Materials International, 22 (2012) 5, 414–425, doi:10.1016/
j.pnsc.2012.10.008
4 O. Ozdemir, S. Zeytin, C. Bindal, A study on NiAl produced by pres-
sure-assisted combustion synthesis, Vacuum, 84 (2010), 430–437,
doi:10.1016/j.vacuum.2009.09.006
5 A. Biswas, S. K. Roy, Comparison between the microstructural evo-
lutions of two modes of SHS of NiAl: key to a common reaction
mechanism, Acta Materialia, 52 (2004), 257–270, doi:10.1016/
j.actamat.2003.08.018
6 M. Tokita, Spark Plasma Sintering (SPS) Method, Systems, and Ap-
plications, Handbook of Advanced Ceramics, (2013), 1149–1177,
doi:10.1016/B978-0-12-385469-8.00060-5
7 Z. A. Munir, U. Anselmi-Tamburini, M. Ohyanagi, The effect of
electric field and pressure on the synthesis and consolidation of ma-
terials: A review of the spark plasma sintering method, Journal of
Materials Science, 41 (2006) 3, 763–777, doi:10.1007/s10853-
006-6555-2
8 D. M. Hulbert, D. Jiang, D. V. Dudina, A. K. Mukherjee, The synthe-
sis and consolidation of hard materials by spark plasma sintering, In-
ternational Journal of Refractory Metals and Hard Materials, 27
(2009), 367–375, doi:10.1016/j.ijrmhm.2008.09.011
9 L. Wang, J. Zhang, J. Wan, Recent development in reactive synthesis
of nanostructured bulk materials by spark plasma sintering, Interna-
tional Journal of Refractory Metals and Hard Materials, 39 (2013),
103–112, doi:10.1016/j.ijrmhm.2013.01.017
10 G. F. Taylor, Apparatus for making hard metal compositions, US Pat-
ent 1, 896, 854, 1993
11 K. Inoue, Electric discharge sintering, US Patent 3, 241, 956, 1966
12 M. Tokita, Trends in advanced SPS spark plasma sintering systems
and technology, Journal of the Society of Powder Technology, 30
(1993) 11, 790–804, doi:10.4164/sptj.30.11_790
A. MICHALCOVÁ et al.: NiAl INTERMETALLIC PREPARED WITH REACTIVE SINTERING ...
450 Materiali in tehnologije / Materials and technology 50 (2016) 3, 447–450
Figure 9: Microhardness and macrohardness of samples after SHS
preparation and powder-metallurgicy preparation with SPS
Slika 9: Mikrotrdota in trdota vzorcev po SHS pripravi in po meta-
lur{ki obdelavi z SPS metodo
Figure 8: SAED pattern of the grain boundary and diffraction pattern
obtained by integrating it in Process Diffraction software. The grain
boundary is composed of NiAl and Al2O3 phases.
Slika 8: SAED-posnetek meje zrna in posnetek uklona, dobljen z vsta-
vitvijo v programsko opremo Difrakcija procesa. Meja zrna je sestav-
ljena iz faz NiAl in Al2O3.
A. KRA^UN et al.: MICROSCOPIC CHARACTERIZATION AND PARTICLE DISTRIBUTION ...
451–454
MICROSCOPIC CHARACTERIZATION AND PARTICLE
DISTRIBUTION IN A CAST STEEL MATRIX COMPOSITE
MIKROSKOPSKA KARAKTERIZACIJA IN RAZPOREDITEV
DELCEV V KOMPOZITU Z MATRICO LITEGA JEKLA
Ana Kra~un1,2, Matja` Torkar1, Jaka Burja1, Bojan Podgornik1
1Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia
2International postgraduate school Jo`ef Stefan, 1000 Ljubljana, Slovenia
ana.kracun@imt.si
Prejem rokopisa – received: 2015-10-01; sprejem za objavo – accepted for publication: 2015-12-24
doi:10.17222/mit.2015.310
The purpose of this investigation was to identify the distribution of ultrafine particles in a steel matrix introduced through a
conventional melting and casting method, and above all to determine the methodology and analysing techniques suitable for the
analysis and identification of ultrafine particles incorporated into the steel matrix. In the frame of this work, steels dispersed
with Al2O3 ultrafine particles were produced by a conventional casting method and their microstructure investigated with light
microscopy (LM), scanning electron microscopy (SEM) and auger electron spectroscopy (AES). Microstructural analyses show
that the distribution of the Al2O3 ultrafine particles is non-uniform and has a high degree of agglomeration. Furthermore, for a
detailed analysis of the nanoparticles a specific preparation and characterization using advanced microscopic techniques is
required.
Keywords: particle distribution, microscopic characterization, steel matrix
Namen raziskave je bil ugotoviti porazdelitev ultrafinih delcev v jekleni matrici, ki je bila proizvedena s konvencionalnim
postopkom litja, predvsem pa dolo~iti metodologijo in analizne tehnike, primerne za analizo in identifikacijo ultrafinih delcev,
ki so bili vklju~eni v jekleno matrico. Delci Al2O3 so bili dodani med procesom konvencionalnega litja in so bili analizirani s
pomo~jo razli~nih analiznih tehnik, in sicer: z uporabo opti~nega mikroskopa (LM), vrsti~nega elektronskega mikroskopa
(SEM) in spekroskopije Augerjevih elektronov (AES). Analiza mikrostrukture je pokazala neenakomerno porazdelitev in
aglomeracijo Al2O3 delcev. Za podrobno analizo je potrebna karakterizacija mikrostrukture s pomo~jo naprednih mikroskopskih
tehnik.
Klju~ne besede: porazdelitev delcev, mikroskopska karakterizacija, jeklena matrica
1 INTRODUCTION
The insertion of ceramic reinforcements into metal
matrices to produce composite materials with improved
properties has been a subject of intensive research during
the past three decades.1–3 Ceramic particulates such as
borides, carbides, oxides and nitrides are added to metal
matrix composites (MMCs) to improve their elastic mo-
dulus, wear resistance, creep and strength.4–5
The ductility of MMCs, however, deteriorates at high
ceramic particle concentrations5. The metal matrix, the
so-called metal-matrix nano-composite (MMnCs) is
strengthened by nano-sized ceramic particles.6 These
nanoparticle reinforcements can significantly increase
the mechanical strength of the metal matrix, as they pro-
mote particle hardening more effectively than micro
particles. Moreover, MMnCs improve the performance
significantly at elevated temperatures, because the cera-
mic nanoparticles can maintain their properties at high
temperatures.6
Steel matrix composites commonly have a combi-
nation of hard ceramic (e.g., TiC, TiB2, WC and Al2O3)
reinforcements and a ductile metallic matrix, which
makes them promising candidates for high-strength and
wear-resistance applications. There are several methods
for fabricating particulate-reinforced steel matrix compo-
sites, such as powder metallurgy, conventional melting
and casting, reactive sintering and self-propagating
high-temperature synthesis (SHS). The casting process is
simple and more economical than the other available
routes for integrating nanoparticles into the microstruc-
ture of steel. However, it is extremely difficult to obtain a
uniform dispersion of ceramic nanoparticles in liquid
metals due to the poor wettability and the difference in
the specific gravity between the ceramic particles and the
metal matrix.7
The microstructure of metals is generally charac-
terized by advanced microscopic techniques (e.g., LM,
SEM and TEM) which probe and map the surface and
sub-surface structure of a material. These techniques can
use photons, electrons, ions or physical cantilever probes
to gather data about a sample’s structure on a wide range
of length scales.8 Auger electron spectroscopy (AES)
also provides quantitative elemental information from
the surfaces of solid materials.9
The current work aims at contributing to the knowl-
edge and understanding of the conventional casting route
for ultrafine particle inoculation in a steel matrix. This
Materiali in tehnologije / Materials and technology 50 (2016) 3, 451–454 451
UDK 621.74:620.18:669.13 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)451(2016)
production route seems to show potential and offers
more cost efficiency in achieving the dispersion of
second-phase ultrafine particles compared to the powder
and metallurgical techniques used until now. The aim of
the work was therefore to study the influence of Al2O3
ultrafine particles on the microstructure of a steel matrix
using a conventional casting method. The additional aim
is to determine the methodology and analysing tech-
niques suitable for analysing and necessary to identify
the ultrafine particles incorporated in the steel matrix.
2 EXPERIMENTAL WORK
2.1 Material
Austenitic stainless steel was used for the work,
mainly due to the distinctive two-phase microstructure of
austenite and ferrite. The chemical composition of this
alloy is given in Table 1. These are the most used group
of stainless steels. They are paramagnetic, have a face-
centred cubic lattice and excel with a good combination
of hot and cold workability, mechanical properties and
corrosion resistance.
Table 1: Chemical composition of austenitic stainless steel in mass
fractions, (w/%)
Tabela 1: Kemijska sestava avstenitnega nerjavnega jekla v masnih
dele`ih, (w/%)
Elements w/%
C 0.02
Si 0.33
Mn 1.24
Cr 17.4
Ni 10.1
Cu 0.36
Mo 1.29
V 0.08
As the reinforcement particles, commercial ultrafine
Al2O3 powder with a mean particle size of 500 nm was
used, as shown in Figure 1. The Al2O3 ultrafine particles
were selected due to their high chemical stability to Fe
and high specific gravity. In particular, it was reported
that the wetting angle  between Al2O3 and molten iron
alloy is less than 50°, even at high temperatures and in
many different types of atmospheres.10
2.2 Specimens preparation
A weighed quantity (10 kg) of the austenitic stainless
steel was melted in an induction furnace. In the first
experiment 20 g of the ultrafine Al2O3 particles were
wrapped in an Al foil and put into the ingot and the
molten metal was poured over it into the same ingot. In
the second experiment a mixture of 24 g of Al2O3 and
2.4 g of dry glue was prepared. The mixture was then
filled in the steel tube and flooded with paraffin. The
tube was inserted into the molten metal and when
melted, the molten metal was poured into the ingot.
2.3 Characterization
The microstructural changes and the dispersion of the
ceramic particles in the steel matrix were observed and
analysed using light microscopy (LM), scanning electron
microscopy (SEM) and auger electron spectroscopy
(AES). Samples for the microstructure analysis were
taken from the bottom, middle and top portions of the
cast piece. Metallographic samples were prepared by
grinding, polishing, followed by chemical etching and
analysed to reveal the particle distribution. Samples for
Auger electron spectroscopy were prepared by grinding
and polishing the surface. These samples were attached
to the bracket, placed in an experimental container-air-
lock, pumped to UHV and transferred into an analytical
container. The surface of the sample was ion etched and
analysed to determine the elemental composition in the
surface region of the sample.
3 RESULTS AND DISCUSSION
Figure 2 shows a LM micrograph of the microstruc-
ture of pure austenitic stainless steel with a distinctive
A. KRA^UN et al.: MICROSCOPIC CHARACTERIZATION AND PARTICLE DISTRIBUTION ...
452 Materiali in tehnologije / Materials and technology 50 (2016) 3, 451–454
Figure 1: SEM image of Al2O3 ultrafine particles at various magni-
fications
Slika 1: SEM-posnetek ultrafinih delcev Al2O3 pri razli~nih pove-
~avah
two-phase microstructure of austenite and -ferrites. A
LM micrograph of the microstructure and ultrafine
particles’ distribution of the sample produced by the
casting process of the austenitic stainless steel poured
over the Al2O3 ultrafine particles is shown in Figure 3.
As shown in Figure 3, the microstructure of the
austenitic stainless steel is modified after the addition of
Al2O3 ultrafine particles, being incorporated into the me-
tal matrix. However, the distribution of Al2O3 particles is
non-homogeneous and concentrated in a certain area.
In Figure 4 the particle distribution of the sample
taken from the second experiment, where the steel tube
filled with Al2O3 particles was inserted into the melt is
shown. As in the case of the first experiment, with the
molten steel being poured over the Al2O3 ultrafine parti-
cles, the distribution of the particles is non-uniform and
has a high degree of agglomeration (Figure 4). However,
the degree of particles is lower when inserting the parti-
cles-filled steel tube into the molten metal.
From the SEM elemental analysis, shown in Figure
5, it was confirmed that the bright, small, spot-like feat-
ures represent the Al2O3 ultrafine particles that are non-
uniformly distributed in the steel matrix.
In Figure 6 the AES spectrum of the Al2O3 ultrafine
particles in the cast microstructure of austenitic stainless
steel is shown. The spectra of particles (P1 and P2)
showing only O and Al peaks confirm the successful
introduction of Al2O3 ultrafine particles into the steel
matrix (P3) without any intermetallic reaction taking
place.
4 CONCLUSIONS
Steel matrix composites with non-uniformly dis-
persed Al2O3 ultrafine particles were produced by a con-
ventional melting and casting method. The purpose of
this investigation was to determine the methodology and
analysing techniques suitable for the analysis and identi-
A. KRA^UN et al.: MICROSCOPIC CHARACTERIZATION AND PARTICLE DISTRIBUTION ...
Materiali in tehnologije / Materials and technology 50 (2016) 3, 451–454 453
Figure 5: SEM elemental analysis of Al2O3 ultrafine particles in the
cast microstructure of austenitic stainless steel
Slika 5: SEM-posnetek elementne analize Al2O3 ultrafinih delcev v
liti mikrostrukturi avstenitnega nerjavnega jekla
Figure 3: Cast microstructure of austenitic stainless steel with 6 % of
-ferrite and Al2O3 ultrafine particles
Slika 3: Lita mikrostruktura avstenitnega nerjavnega jekla s 6 % -fe-
rita in Al2O3 ultrafinimi delci
Figure 2: Cast microstructure of austenitic stainless steel with 6 % of
-ferrite
Slika 2: Lita mikrostruktura avstenitnega nerjavnega jekla s 6 % -fe-
rita
Figure 4: Cast microstructure of austenitic stainless steel with Al2O3
ultrafine particles inserted into the melt (experiment 2)
Slika 4: Lita mikrostruktura avstenitnega nerjavnega jekla z Al2O3
ultrafinimi delci, vstavljenimi v talino (preizkus 2)
fication of ultrafine particles incorporated in the steel
matrix.
The microstructural changes and the dispersion of the
Al2O3 ultrafine particles in the steel matrix were ob-
served and analysed by light microscopy (LM), scanning
electron microscopy (SEM) and auger electron spectro-
scopy (AES). This work clearly shows that for a proper
analysis and identification of the successful nano-parti-
cles’ incorporation, different analysing techniques need
to be used and combined.
Based on the experimental results the dispersion of
the Al2O3 ultrafine particles in the steel matrix is non-
homogeneous and concentrated in certain areas.
In order to be able to obtain a homogeneous distri-
bution of reinforcements in the metal matrices the
following factors need to be understood and taken into
consideration for future work:
• particle density, size, shape and volume fraction will
influence the reinforcement settling rate,
• surface properties of the particles will affect the
wetting with molten metal,
• rheological behaviour is influenced by the reaction of
the particles with the melt and each other,
• in general, the reinforcement particles occupy inter-
dendritic or between secondary dendrite arm spa-
cings, while the particle distribution is also metal-
matrix dependent.
5 REFERENCES
1 Y. Q. Liu, H. T. Cong, W. Wang, C. H. Sun, H. M. Cheng, AlN nano-
particle-reinforced nanocrystalline Al matrix composites: Fabrication
and mechanical properties. Met.Sic.Eng.A, 505 (2009), 151–156,
doi:10.1016/j.msea.2008.12.045
2 C. S. Goh, J. Wei, L. C. Lee, M. Gupta, Ductility improvement and
fatigue studies in Mg-CNT nanocomposites. Comps. Sci. Tech. 68
(2008), 1432–1439, doi:10.1016/j.compscitech.2007.10.057
3 Z. Razavi Hesabi, A. Simchi, S. M. Seyed Reihani, Structural evolu-
tion during mechanical milling of nanometric and micrometric Al2O3
reinforced Al matrix composites. Mater. Sci. Eng. A, 428, (2006),
159–168, doi: 10.1016/j.msea.2006.04.116
4 J. Llorca, Fatigue of particle-and whisker reinforced metal-matrix
composites. Prog.Mater.Sci., 47 (2002), 283–353, doi:10.1016/
S0079-6425(00)00006-2
5 B. N. Chawla, Y. Shen, Mechanical Behavior of Particle Reinforced
Metal Matrix Composites, Adv.Eng.Mater., 3 (2001) 6, 357–370,
doi:10.1002/1527-2648(200106)3:6<357::AID-ADEM357>3.3.CO;2-9
6 R. Casati, M. Vedani, Metal Matrix Composites Reinforced by
Nano-Particles – A Review, Metals (Basel), 4 (2014) 1, 65–83,
doi:10.3390/met4010065
7 S. H. Lee, J. J. Park, S. M. Hong, B. S. Han, M. K. Lee, C. K. Rhee,
Fabrication of cast carbon steel with ultrafine TiC particles. Trans
Nonferrous Met. Soc. China (English Ed., 21 (2011), 54–57,
doi:10.1016/S1003-6326(11)61060-1
8 R. Konwar, A. B. Ahmed, Nanoparticle: an Overview of Preparation,
Characterization and Application. Int. Res. J. Pharm., 4 (2013) 4,
47–57, doi:10.7897/2230-8407.04408
9 C. Linsmeier, Auger electron spectroscopy. Vacuum, 45 (1994) 6–7,
673–690, doi:10.1016/0042-207X(94)90108-2
10 S. Y. Cho, J. H. Lee, Anisotropy of wetting of molten Fe on Al2O3
single crystal. Korean J. Mater. Res., 18 (2008) 1, 18–21,
doi:10.3740/ MRSK.2008.18.1.018
A. KRA^UN et al.: MICROSCOPIC CHARACTERIZATION AND PARTICLE DISTRIBUTION ...
454 Materiali in tehnologije / Materials and technology 50 (2016) 3, 451–454
Figure 6: AES spectrum of the Al2O3 ultrafine particles in the cast
microstructure of austenitic stainless steel
Slika 6: AES-spekter analize Al2O3 ultrafinih delcev v liti mikrostruk-
turi avstenitnega nerjavnega jekla
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
455–460
A COMPARISON OF AS-WELDED AND SIMULATED HEAT
AFFECTED ZONE (HAZ) MICROSTRUCTURES
PRIMERJAVA MIKROSTRUKTURE TOPLOTNO VPLIVANEGA
PODRO^JA VARJENEGA IN SIMULIRANIH VZORCEV
Roman Celin1, Jaka Burja1, Gorazd Kosec2
1Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia
2Acroni d.o.o., Cesta Borisa Kidri~a 44, 4270 Jesenice, Slovenia
roman.celin@imt.si
Prejem rokopisa – received: 2016-01-05; sprejem za objavo – accepted for publication: 2016-01-29
doi:10.17222/mit.2016.006
The high-strength steel grade S690QL and a filler welding wire Mn3Ni1CrMo were the materials chosen for welding a
V-shaped butt weld. In order to prevent the weld’s cold cracking, a multi-pass welding technique was applied. A metallographic
investigation revealed microstructure variations in different areas of the weld’s heat-affected zone. A reverse-engineering
approach was used to test a dilatometer’s capabilities to simulate different HAZ microstructures. Hollow steel-cylinder
specimens were subjected to several weld thermal cycles in order to generate similar microstructures as in the real weld’s HAZ.
The microstructures of the as-welded and simulated heat-affected zone specimens were investigated. Good agreement was found
between the dilatometer-simulated HAZ microstructures and those in a real HAZ weld.
Keywords: welded joint, microstructure, high-strength low-alloy steel, simulation, dilatometer, heat-affected zone
Visokotrdno jeklo S690QL in varilna `ica Mn3Ni1CrMo kot dodatni material, sta bila uporabljena pri varjenju so~elnega
V-spoja. Za prepre~itev pokanja v hladnem je bila uporabljena tehnika ve~varkovnega varjenja. Metalografska preiskava je
odkrila razli~ne mikrostrukture v razli~nih predelih toplotno vplivanega podro~ja zvara. Za preizkus delovanja dilatometra je bil,
za simulacijo mikrostruktur razli~nih podro~ij toplotno vplivanega podro~ja, uporabljen princip povratnega in`enirstva.
Razli~na mikrostruktura toplotno vplivanega podro~ja je bila simulirana z izpostavitvijo votlih cilindri~nih vzorcev razli~nim
toplotnim ciklom. Opravljena je bila metalografska preiskava realnega zvarjenega spoja in simuliranih vzorcev. Primerjava
rezultatov je pokazala dobro ujemanje simuliranih in realnih mikrostruktur.
Klju~ne besede: zavarjen spoj, mikrostruktura, visokotrdno jeklo, simulacija, dilatometer, toplotno vplivano podro~je
1 INTRODUCTION
The decision to use high-strength steel depends on a
number of application requirements, such as thickness
reduction, corrosion resistance, formability and weld-
ability. The quenched and tempered low-alloy (QTLA)
steels, usually containing less than 0.25 % carbon and
less than 5 % alloying elements, are strengthened prima-
rily by quenching and tempering to produce micro-
structures containing martensite and bainite.1 S690QL is
such a steel grade with high strength and toughness.2
Any common welding procedure can be used to join
QTLA steels.3 For any given steel, the welded joint’s
microstructures and the mechanical properties in the
weld metal (WM) and the heat-affected zone (HAZ) are
influenced mainly by the welding thermal cycle.
Due to the welding thermal cycle and the associated
peak temperature, a change of the parent metal’s micro-
structure and the mechanical properties happens in the
HAZ. With increasing distance from the fusion line the
peak temperature decreases, thus forming different
microstructures. The width of the HAZ depends on the
welding procedure, the thermal conditions and the
physical properties of the parent metal.
Figure 1 shows a simplified presentation of different
HAZ regions of a multi-pass welded joint. Weld pass 2
has a portion of HAZ that can be treated as a single weld
pass and can be divided into four regions. These regions
are defined by the peak temperature to which the region
was exposed during the weld thermal cycle:4,5
• Coarse-grain region (CG HAZ) – material adjacent to
the fusion line that reaches a peak temperature TP
above 1300 °C,
• Fine-grain region (FG HAZ) – TP is lower, but still
above AC3,
• Inter-critical region (IC HAZ) – TP is between AC1 <
TP < AC3,
• Sub-critical region (SC HAZ) – TP is lower than AC1.
However, the welding of thick steel components
usually requires the application of a multi-pass welding
procedure. In this case, the first pass (weld pass 1) HAZ
regions are reheated to different peak temperatures
during the second weld pass thermal cycle (weld pass 2).
Figure 1 also shows a simplified presentation of the
different reheated first weld pass HAZ regions:4,5
• region 1: Super-critically reheated coarse-grain HAZ
– SCR CG HAZ,
Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460 455
UDK 67:017:669.14.018.298:621.791.05 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)455(2016)
• region 2: Inter-critically reheated coarse-grain HAZ –
IR CG HAZ,
• region 3: Sub-critically reheated coarse-grain HAZ –
SC CG HAZ,
• region 4: Super-critically reheated fine-grain HAZ –
SCR FG HAZ,
• region 5: Inter-critically reheated fine-grain HAZ –
IR FG HAZ,
• region 6: Sub-critically reheated fine-grain HAZ –
SC FG HAZ.
In a real weld the HAZ regions are narrow in relation
to the weld. To achieve regions of uniform micro-
structure suitable for investigations, weld simulators are
used.6-12 Usually, these weld thermal simulations are
made in conjunction with weldability investigations to
determine the proper welding parameters.
In the presented investigation a dilatometer with a
controlled heating and cooling fixture was used to simu-
late the weld thermal cycles. The goal of the investi-
gation was to test the dilatometer’s capability to simulate
the microstructures that correspond to the real weld’s
HAZ microstructures using a reverse-engineering
approach.
2 EXPERIMENTAL PART
The chemical analysis of the Micral 690 sample with
an ICP spectrometer was made prior to any further inve-
stigations. The determination of the welding parameters
for the Micral 690 welding procedure was carried out by
considering the EN 1011-2 recommendations, the steel
manufacturer’s specifications and the data from several
papers.13–16
The two plates, with a thickness of 15 mm, a length
of 500 mm, a width of 150 mm and a V-groove joint
geometry, were welded with 8 passes using the sequence
shown in Figure 2. A manual MAG welding procedure
in a flat (PA) position was used. The filler material was
grade Mn3Ni1CrMo welding wire with a diameter of
 1.2 mm, according to SIST EN ISO 1683417. The
shielding gas used was a mixture M21 (82 % Ar + 18 %
CO2). The welding was carried by a skilled, certified
welder and without preheating.
The welding current, voltage and time were recorded
during the procedure. The weld pass 5 thermal cycle was
recorded by dipping a type-S thermocouple (Pt Rh–Pt)
directly into the weld’s molten bead. The approximate
position of the thermocouple is marked with a dot.
Standard18 non-destructive examinations and mechanical
tests of the weld joint were also carried out and are
described in a previous paper.19
Seven hollow, 10-mm-long cylinders with a 4-mm
outer diameter and 1.5-mm inner diameter were ma-
chined from the weld’s parent material (base metal)
plate. Holes were bored in order to obtain a uniform
heating and cooling of the sample during the thermal
simulation.
The simulations were carried out with a TA instru-
ments DIL805A/D quenching dilatometer in a vacuum
atmosphere, with argon used as a coolant. Control of the
thermal cycle was maintained via a type-S thermocouple
that was spot welded directly onto the hollow cylinder
sample.
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
456 Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460
Figure 2: Sketch of the welded joint
Slika 2: Skica zavarjenega spojaFigure 1: Simplified representation of the different regions of the
HAZ
Slika 1: Poenostavljena predstavitev razli~nih obmo~ij toplotno vpli-
vanega podro~ja
Figure 3: Schematic diagrams for the thermal cycle simulations
Slika 3: Diagrami za potek simulacij termi~nih ciklov
Figure 3 shows schematic diagrams with the parame-
ters for seven thermal cycle simulations used to generate
the corresponding microstructure:
• single weld pass CG, FG, IC and SC regions of the
HAZ,
• reheated weld pass SCR CG, IC CG, SC CG regions
of the HAZ, as in Figure 1.
For the reheated simulated thermal cycles the same
parameters were used as for the single-cycle simulation.
The time–temperature dependence (Figure 4) recorded
during weld pass 5 was used to determine the simulation
heating rate, the peak temperatures, the holding times
and the cooling rates.
During a single coarse-grain (CG) thermal cycle
simulation the specimens’ phase-transition temperatures
AC1 and AC3 were recorded as well as the dilatation and
temperature data from each simulation. The specimen
was prepared by grinding and polishing with diamond
paste, followed by a chemical etching with 5 % Nital.
Macro- and microscopic examinations20–22 were per-
formed using a light microscope (LM) to characterize the
microstructures of the deposited weld metal, the heat-
affected zone, the parent metal and the dilatometer-simu-
lated hollow cylinder specimens of the parent metal.
A comparison between the as-welded and the dila-
tometer-simulated microstructures was carried out.
3 RESULTS AND DISCUSSION
The results of the parent metal’s quantitative chemi-
cal analysis are presented in Table 1. The steel contains
strong carbide-forming elements such as Nb, Mo, Cr and
Ti, which ensure high strengths, even at elevated tempe-
ratures.
Table 1: Chemical composition of parent metal in mass fractions, w/%
Tabela 1: Kemi~na sestava jekla v masnih dele`ih, w/%
C Si Mn P S Cr Ni
0.164 0.29 0.72 0.006 0.001 0.54 0.20
Cu Mo Ti Nb Al B N
0.25 0.263 0.02 0.028 0.028 0.001 0.011
The applied welding parameters and the calculated
welding speed and heat input were:
• weld length, 500 mm
• welding time, 132–422 s
• voltage, 21–22 V
• current, 100–165 A
• weld inter-pass temperature, 130 °C
• welding speed, 1.8–2.89 mm/s
• heat input, 1.03–1.63 kJ/mm
Lower-value welding parameters were applied for the
root pass welding.
After the temperature–time data acquisition from the
temperature-measurement instrument’s memory card and
data analysis for the welding pass 5, a cooling time t8/5 of
7.6 s was determined and the peak temperature TP was
1370 °C (Figure 4).
The shape of the cooling curve (Figure 4) is affected
by the inter-pass temperature. Also, the weld’s cooling
time is prolonged due to the reduced temperature diffe-
rence between the weld material and the surrounding
parent metal.
Figure 5 shows the ICR CG HAZ recorded thermal
cycle simulation with the temperature and the dilatation
curve. From the dilatation data analysis of the phase
transformation during the CG thermal cycle simulation
the transition temperatures AC1 and AC3 were found to be
830 °C and 885 °C, respectively. During the rapid heat-
ing of steel, as in the case of welding, the phase-trans-
formation temperatures are increased, the observed
transformation temperatures AC1 and AC3 are well above
equilibrium, i.e., about 710 °C and 785 °C respectively.
The increase of the transformation temperatures is
attributed to the diffusion process of the transformation
from ferrite to austenite, which is time and temperature
dependent. The dilatation curve in Figure 5 clearly
indicates that a transformation took place below 500 °C,
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460 457
Figure 5: ICR CG HAZ thermal simulation with temperature-time and
dilatation-time dependence
Slika 5: Potek temperature in raztezka v odvisnosti od ~asa pri simu-
laciji interktiti~nega grobozrnatega TVP
Figure 4: Weld pass 5 – recorded temperature – time dependence
Slika 4: Polnilni varek 5 – zabele`ena odvisnost temperatura – ~as
during the first cooling, and that another transformation
took place during the second cooling, but it was not as
intense as the first one, thus proving that the inter-critical
temperature was indeed reached.
A good fit was found between the theoretical (Figure
3) and dilatometer-simulated thermal IR CG HAZ cycle
(Figure 5). The heating rates, cooling rates, and holding
times were very close to those programmed, with minor
deviations due to the response of the dilatometer control
and the regulation system. The first thermal cycle peak
temperature was 1360 °C, with an inter-critical peak
temperature of 866 °C.
Figure 6 shows the welded joint’s macro section with
each weld pass and the corresponding HAZ. Between
subsequent weld passes an unaffected region of the
coarse-grain HAZ microstructure can be distinguished.
The reheated pockets of the CG HAZ regions are small
and discontinuous, which makes their microstructure
difficult to identify and investigate. During metallo-
graphic investigations the dimensions of the IR CG HAZ
microstructure region were estimated to be 0.8 mm long
and 0.3 mm wide.
Figure 7 shows the weld pass 5 HAZ area not
affected by a subsequent weld pass thermal cycle. It is an
area with a longitudinal microstructure transition from
the weld metal (WM) through the heat-affected zone
(HAZ) microstructures to the parent metal (PM).
The weld metal (WM) consists of columnar dendrites
with a bainite microstructure. Some individual marten-
sitic grains are also present in the middle of the deposi-
ted weld metal. The microstructure of the HAZ consists
of martensite and bainite with a transition to the
unaffected tempered martensitic microstructure of the
parent metal (PM) (Figures 8 and 9).
Figure 8.1 shows the coarse grains (CGs) that are
present in the HAZ adjacent to the fusion boundary of
the weld. The fine-grain (FG, Figure 8.3) region follows
due to the peak temperature above AC3 but lower than in
the CG region, the temperature and time were not
sufficient to cause severe grain growth. With increasing
distance from the fusion boundary there are inter-critical
(IC, Figure 8.5) and subcritical (SC, Figure 8.7) regions
of the HAZ.
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
458 Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460
Figure 8: Real and simulated HAZ microstructures
Slika 8: Realne in simulirane mikrostrukture TVP
Figure 6: Welded joint macro section
Slika 6: Makroposnetek zavarjenega spoja
Figure 7: Transition from weld metal (WM) to parent metal (PM)
Slika 7: Prehod iz vara (WM) v osnovni material (PM)
The reheated inter-critical region (Figure 9.11 and
9.12) is very susceptible to failure, due to the fact that
the phase transformation into austenite began on the
grain boundaries; these small areas were then quickly
cooled and are in turn hard and brittle. The subcritical
regions are mainly tempered bainite and martensite with
precipitated carbides and therefore represent no danger
to the structural integrity of the weld. The simulated
microstructures in Figure 8 and Figure 9 are labelled
with even numbers. Although the grain size of the simu-
lated specimen is slightly larger than that of the real
welded joint when comparing the identical thermal
cycle, we can distinguish a similarity between the
welded HAZ and the dilatometer-simulated microstruc-
tures. The reason for the larger grains is that the thermal
pinning is not considered in the thermal cycle simulation
process.6
4 CONCLUSIONS
Normally, a dilatometer is used to observe a speci-
men’s dimensional changes under a controlled heating or
cooling rate. It can also be used to construct a conti-
nuous-cooling- transformation (CCT) diagram or an
isothermal time-temperature-transformation (TTT) dia-
gram. The TA DIL805A/D dilatometer with controlled
heating and cooling fixtures was tested to simulate a real
weld’s HAZ microstructure. Based on the investigation
the following can be concluded:
• The use of a simulated HAZ microstructure is a con-
venient way to study the weldability of a given steel.
• The presented investigation was limited to micro-
structure, due to the specimen’s size and geometry.
• The simulation of a weld’s HAZ microstructure is
possible within the limits of the dilatometer’s
capabilities and the specimen’s size.
• Hollow cylinder samples had a better response to the
heating and cooling rate change than a solid cylinder.
• The inter-critical temperature during welding of
S690QL is between 830 °C and 885 °C.
Acknowledgment
The authors are thankful to the Acroni d.o.o. and
Elektrode Jesenice for the technical support related to
work presented in this paper.
5 REFERENCES
1 F. J. Winsor, ASM Handbook, Vol. 6: Welding, Brazing and Solder-
ing, ASM, International, Materials Park, OH 1993, 662
2 SIST EN 10025-6:2005+A1:2009, Hot rolled products of structural
steels – Part 6: Technical delivery conditions for flat products of high
yield strength structural steels in the quenched and tempered condi-
tion
3 S. Kou, Welding metallurgy, 2nd ed., Wiley Interscience, Hooboken
2003, 406
4 M. Hamada, Control of strength and toughness at the heat affected
zone, Welding International, 17 (2003) 4, 265-270
5 T. Lolla, S. S. Babu, S. Lalam, M. Manohar: Understanding the Role
of Initial Microstructure on Intercritically Reheated Heat- Affected
Zone Microstructures and Properties of Microalloyed Steels, Proc. of
the 9th Inter. Conf., June 4–8, 2012 Chicago, 34-42
6 Y. Shi, Z. Han, Effect of weld thermal cycle on microstructure and
fracture toughness of simulated heat-affected zone for a 800 MPa
grade high strenght low alloy steel, Journal of Materials Processing
Technology, 207 (2008), 30–39, doi: 10.1016/j.jmatprotec.2007.
12.049
7 G. R. Goodall, J. Gianetto, J. Bowker, M. Brochu, Thermal simu-
lation of HAZ regions in modern high strength steel, Canadian
Metallurgical Quarterly, 51 (2012), 1, 58–66, doi:10.1179/
1879139511Y.0000000023
8 I. Samard`i}, A. Stoi}, D. Kozak, I. Kladaric, M. Dun|er, Applica-
tion of Weld Thermal Cycle Simulator in Manufacturing Engineer-
ing, Journal of Manufacturing and Industrial Engineering, Journal of
Manufacturing and Industrial Engineering, 12 (2013), 1–2, 7–11,
doi: http://dx.doi.org/10.12776/mie.v12i1-2.177
9 9G. L. Liang, S.W. Yang, H. B. Wu, X. L. Liu, Microstructure and
mechanical performances of CGHAZ for oil tank steel during high
heat input welding, Rare Metals 32 (2013) 2, 129–133, doi:10.1007/
s12598-013-0036-y
10 Y. Adonyi, Heat-Affected Zone Characterization by Physical
Simulations, Welding Journal, 10 (2006), 42-47
11 J. Wang, S. Lu, L. Rong, D. Li, Thermal cycling, microstructure and
mechanical properties of 9Cr2WVTa steel welds, Journal of Mate-
rials Processing Technology, 222 (2015) 8, 434–443, doi:10.1016/
j.jmatprotec.2015.03.017
12 M. Dun|er, I. Samard`i}, T. Vuherer, Dependence of hardness and
impact energy on cooling time t8/5 and temperature for S960QL,
Metalurgija, 54 (2015) 3, 539–542, doi:10.12776/mie.v12i1-2.177
13 Welding – Recommendations for welding of metallic materials – EN
1011-2 Part 2: Arc welding of ferritic steels
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460 459
Figure 9: Real and simulated HAZ microstructures; reheated regions
Slika 9: Realne in simulirane mikrostrukture TVP; ponovno pregreta
podro~ja
14 I. Hajro, O. Pa{i}, Z. Burzi}, Karakterizacija zavarenih spojeva na
visoko~vrstom konstrukcionom ~eliku S690QL, Zavarivanje i zava-
rene konstrukcije, 58 (2010) 4, 123–129
15 M. Shome, Effect of heat-input on austenite grain size in the heat
affected zone of HSLA-100 steel, Materials Science and Engineering
A, 39 (2007) 13, 454–460, doi:10.1016/j.msea.2006.09.085
16 Acroni Micral 690 Datasheet, https://www.metalandsteel.com/docu-
mentserver/store/17606/2b66fa1b-adaa-4b42-9700- 1ab07d115e60.pdf
17 SIST EN ISO 16834: – Welding consumables – Wire electrodes,
wires, rods and deposits for gas shielded metal arc welding of high
strength steels – Classification
18 SIST EN ISO 15614-1: Specification and Qualification of Welding
Procedures for Metallic Materials – Welding Procedure Test – Part 1:
Arc and Gas Welding of Steels and Arc Welding of Nickel and
Nickel Alloys
19 R. Celin, J. Berneti~, D. A. Skobir Balanti~, Welding of the steel
gradeS890QL, Mater. Tehnol., 48 (2014) 6, 931–935
20 S. Ko`uh, M. Goji}, I. Ivani}, B. Kosec, Microstructure of welded
austenitic stainless steel after annealing at 900 °C, Zavarivanje i
zavarene konstrukcije, 61 (2013) 4, 149–156
21 F. Tehovnik, B. Arzen{ek, B. Arh, D. Skobir, B. Pirnar, B. @u`ek,
Microstructure evolution in SAF 2507 super duplex stainless steel,
Mater. Tehnol., 45 (2011) 4, 339–345
22 M. Pournavari, S. P. H. Marashi, H. L. Jaber, DP780 dual-phase-steel
spot welds: critical fusion-zone size ensuring the pull-out failure
mode, Mater. Tehnol., 49 (2015) 4, 579–585, doi:10.17222/
mit.2014.184
R. CELIN et al.: A COMPARISON OF AS-WELDED AND SIMULATED HAZ MICROSTRUCTURES
460 Materiali in tehnologije / Materials and technology 50 (2016) 3, 455–460
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
461–466
DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
DEGRADACIJA REZERVOARJA IZ AISI 304 NERJAVNEGA JEKLA
Matja` Torkar, Irena Paulin, Bojan Podgornik
Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia
matjaz.torkar@imt.si
Prejem rokopisa – received: 2016-01-28; sprejem za objavo – accepted for publication: 2016-02-10
doi:10.17222/mit.2016.027
Some austenitic stainless steels are sensitive to the stress corrosion cracking (SCC) that appears only localized and with the
specific combination of a relatively high internal tensile stress and the presence of corrosion media with chloride ions. The
presented results are from an investigation of the leakage from a tank containing medical disinfection liquids. Due to a crevice
below the insulation and the presence of condensed moisture, the corrosion pits started to grow on the surface of the tank. These
crevice corrosion pits led to the appearance of SCC, which was responsible for the leakage from the AISI 304 stainless-steel
tank.
Keywords: tank, AISI 304, leakage, crevice corrosion, stress corrosion cracking, brittle fracture
Nekatera avstenitna nerjavna jekla so ob~utljiva na napetostno korozijsko pokanje (SCC), ki se pojavi samo lokalno pri
specifi~ni kombinaciji relativno velikih notranjih nateznih napetosti in prisotnosti korozijskega medija s kloridnimi ioni.
Predstavljeni so rezultati raziskave pu{~anja rezervoarja z medicinskimi dezinfekcijskimi teko~inami. Zaradi {pranje pod
izolacijo in prisotnosti kondenzirane vlage, so pri~ele rasti korozijske jamice na povr{ini rezervoarja. Jamice, nastale pri
{pranjski koroziji, so omogo~ile pojav SCC, ki je odgovoren za pu{~anje rezervoarja iz AISI 304 nerjavnega jekla.
Klju~ne besede: rezervoar, AISI 304, pu{~anje, {pranjska korozija, napetostno korozijsko pokanje, krhek prelom
1 INTRODUCTION
Stress corrosion cracking (SCC) is a localized form
of corrosion that occurs under the simultaneous action of
a tensile stress and a corrosive environment such as a
chloride. SCC is characterized by fine cracks that can
propagate extremely rapidly, leading to failure of the
component and, potentially, of the associated structure.1,2
Typical for SCC is the conjoint action of stress and a
corrosive environment, which leads to the formation of a
crack that would not have developed from the action of
the stress or environment alone.3
Extensive research studies indicate that SCC appears
only localized and under a specific combination of three
conditions:
• the use of susceptible grades of material,
• a relatively high tensile stress relative to the yield
strength (0.2 % proof strength), either from structural
loading or present as residual stresses from forming
or welding operations during manufacture and
installation,
• the presence of a specific aggressive environment
with chlorine-containing compounds (for instance
by-products of disinfection) that can produce a highly
corrosive film, which can lead to SCC.
Some grades of stainless steel, including 1.4301(AISI
304) and 1.4401 (AISI 316), have long been recognised
as susceptible to SCC, but generally only above 55 °C.
However, failures in swimming pools in recent years
occurred at around 30 °C, in highly stressed components
that had not been washed by pool water or frequently
cleaned.3 Suitable steels for safety-critical and load-bear-
ing components in a pool hall atmosphere3 are the grades
1.4547, 1.4529 and 1.4565. All three types of austenitic
stainless steels have a high content of molybdenum and
nickel, which provide good resistance to chloride SCC.
Several studies were made on the influence of the
surface condition of austenitic stainless steel on SCC.4–6
It was also revealed how highly cold-worked material
showed a higher crack propagation rate.4,5
A model was proposed for the crack propagation
based on brittle fracture, localized oxidation and shear-
ing near the crack tip.6 In all cases the cracks were
initiated at the pitting sites.7
The mechanical fracture model of SCC assumes that
the crack essentially propagates by dissolution, and then
the remaining ligaments fail as a result of mechanical
fracture (ductile or brittle). There are several proposed
models described by the mechanical fracture model: the
film-induced cleavage model, the tarnish rupture model,
the tunnel model, the adsorption model and the hydrogen
models.
The dissolution mechanism of SCC assumes that the
crack propagation is due to active dissolution at the crack
tip. The different models under this mechanism are the
slip-dissolution model or the film-rupture model and
intergranular SCC.8
SCC causes a rapid, brittle failure of the steel without
any prior indication, and for this reason it is considered
to be catastrophic. Several major disasters have been
Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466 461
UDK 620.19:669.14.018.8:52-14 ISSN 1580-2949
Professional article/Strokovni ~lanek MTAEC9, 50(3)461(2016)
attributed to the SCC of steel equipment, including the
rupture of high-pressure gas-transmission pipes, boiler
explosions and severe damage to power stations and oil
refineries.8–10
The SCC propagation mechanisms can be divided
into two groups: those which involve embrittlement of
the metal due to corrosive reactions (mechanical fracture
models) and those in which the cracks grow due to a
localized dissolution process.8
The disinfection liquid storage tank was part of an
industrial washer-disinfector assembly used for cleaning
medical equipment and was made of 1.8-mm-thick
cold-rolled AISI 304 stainless steel. The whole assembly
was in operation for 2 years. Two kinds of liquids were
stored in the tank: 1 % of volume fractions of solution of
cleaning liquid in demineralized water with pH = 11.3,
and 1 % of volume fractions of solution of neutralizing
liquid in demineralized water with pH = 2.6.
The aim of this paper is to present a corrosion degra-
dation investigation and the reasons for the leakage from
the storage tank.
2 EXPERIMENTAL PART
The samples for the investigation were cut from the
wall of the tank, in the corrosion-damaged area where
the wall leaks. Metallographic samples of the cross-sec-
tion of the wall were prepared by a standard metallo-
graphic procedure. The samples were observed with a
Nikon Microphot FXA light microscope with a video ca-
mera and analySIS software. The surface and the cracks
were observed using a JSM-6500F FE SEM scanning
electron microscope and analysed by EDS (Energy-Dis-
persive Spectroscopy). The EDS analyses were per-
formed on the corroded internal surface and on the frac-
tures. The base material was analysed using the X-ray
Fluorescence (XRF) method.
3 RESULTS AND DISCUSION
3.1 Visual examination
The corroded areas inside and outside the tank under
the insulation are shown in Figure 1 and Figure 2.
Inside the tank the corrosion damage is limited to the
area shown in Figure 2, while the corroded area on the
outside surface below the insulation is spread much
wider (Figure 1) and looks more uniform. In addition,
the place where the sample material was cut from the
tank wall is presented in Figure 1.
Inner surface of the tank, shown in Figure 3, looks
like a general pitting-corrosion attack with some pits
joining together to form interconnected pits.
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
462 Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466
Figure 3: Corrosion damage on the inner surface of the tank
Slika 3: Korozijske po{kodbe na notranji povr{ini rezervoarja
Figure 1: Corroded outer surface below the insulation
Slika 1: Korodirana zunanja povr{ina rezervoarja pod izolacijo
Figure 2: Local corrosion damage inside the tank
Slika 2: Lokalne korozijske po{kodbe znotraj rezervoarja
Figure 4 depicts the sample material cut from the
tank wall. Corrosion pits due to crevice corrosion and
cracks due to stress corrosion are present on the outer
surface of the tank.
3.2 Chemical analysis
An XRF analysis was performed on the sample ma-
terial from the tank. The results of the analysis are shown
in Table 1.
The sample chemical composition analysis results are
in accordance with the AISI 304 grade stainless steel’s
specification requirements.
3.3 Metallographic examination
During the cutting of the samples for metallography
it was evident that the material is brittle in the region of
the corrosion and did not resist even a small bending
force. In contrast, out of the region of corrosion the
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466 463
Figure 6: Cross-section of the tank wall
Slika 6: Presek stene rezervoarja
Figure 7: Crevice-corrosion pit
Slika 7: [pranjska korozija
Figure 5: Microstructure of the cross-section of the wall
Slika 5: Mikrostruktura preseka stene
Table 1: Chemical composition of the tank sample in mass fractions
(w/%)
Tabela 1: Kemijske sestave vzorca rezervoarja v masnih dele`ih
(w/%)
C Si Mn Cr Ni Mo Fe
tank
sample 0.07 0.68 1.2 19.2 9.1 – 69.75
AISI
304
Max
0.08
Max.
1.00
Max.
2.0 18–20 8–10.5 –
Figure 4: Corroded external surface of the tank with corrosion pits
and cracks
Slika 4: Zunanja povr{ina rezervoarja s korozijskimi jamicami in
razpokami
material behaved normally and was bent at an angle of
180° without any damage.
The microstructure of the cross-section in the sound
part of the material is presented in Figure 5a. In the cen-
tral part (Figure 5b) are the elongated grains. The
elongated inclusions of MnS are shown in Figures 5c
and 5d, both oriented in the cold-rolling direction.
The cross-section of the degradation area, below the
tank insulation, is presented in Figure 6. The crack
originates from the crevice corrosion on the external
surface of the tank.
The presence of condensed moisture in the gap
between the insulation and the tank’s external surface
caused the development of a crevice-corrosion pit shown
in Figure 7.
The cracks (Figures 6 and 8) are typical for SCC and
spread through the wall, mostly across the grains (trans-
crystalline). Through the wall the cracks caused the
leakage of the liquid stored in the tank. The cracks are
connected with the pits and the pits act as initiation
places for the cracks’ formation. We can conclude that
the corrosion processes started as crevice corrosion on
the external surface below the insulation.
The internal surface of the tank is shown in Figure 9.
Wide corrosion pits can be seen (Figure 9). These pits
can be classified as pitting corrosion. Branched stress
corrosion cracks are also present.
3.4 Scanning electron microscopy
The brittleness of the material was observed only in
the areas damaged by corrosion. The fracture of the wall
is brittle (Figure 10) and typical for the SCC of stainless
steel. The region of the EDS analysis (SEM) is marked
as Spectrum 1.
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
464 Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466
Figure 10: Brittle fracture of the wall
Slika 10: Krhek prelom stene
Figure 8: Internal surface pitting corrosion stress corrosion cracks
Slika 8: Jami~asta korozija na notranji povr{ini rezervoarja in nape-
tostne korozijske razpoke
Figure 9: Trans-crystalline crack formed during SCC
Slika 9: Transkristalna razpoka, nastala pri napetostnem korozijskem
pokanju
Figure 11: Surface with deep-etched grain boundaries and corrosion
products with marked areas of EDS analysis (SEM)
Slika 11: Povr{ina z globoko jedkanimi mejami med zrni in korozijski
produkti z ozna~enimi podro~ji EDS-analize (SEM)
The EDS analysis of the fracture surface (Table 2)
detected the presence of chlorine, which is a regular
companion in the corrosion of stainless steels and acts as
the main accelerator of the pitting corrosion.
The deep-etched grain boundaries presented in
Figure 11 are due to the pickling of the sheet’s surface.
The surface is partially covered with corrosion products.
The EDS analysis (Table 3) of the grain (Spectrum 1)
revealed an increased content of oxygen, silicon,
chlorine, manganese and nickel. On the other two areas
analysed (Spectrum 2 and 3) traces of rust prevailed with
contents of iron, chromium, nickel and oxygen. The
origin of the chlorine is not known. This chlorine was
detected on the corroded external surface as well as on
the surfaces of the brittle cracks.
Table 2: EDS analysis of a brittle fracture surface, in mass fractions
(w/%)
Tabela 2: EDS-analiza povr{ine krhkega preloma, v masnih odstotkih
(w/%)
Spectrum Cr Ni Si Mn Fe Cl
1 19.24 8.99 0.68 1.14 69.56 0.40
Table 3: EDS analysis of corroded surface (w/%)
Tabela 3: EDS analiza korodirane povr{ine (w/%)
Spectrum O Si Cl Mn Cr Ni Fe
1 35.62 0.61 5.78 4.37 5.76 23.14 24.71
2 14.51 0.49 1.08 13.83 6.41 63.69
3 12.92 0.54 1.14 14.46 6.78 64.15
The cracks were formed due to the internal or
external stresses or a combination of both in the presence
of the corrosion media. It is typical for the cracks to form
at relatively low stresses. The crack spreads through the
material, either in a trans-crystalline or inter-crystalline
direction, depending on the material, the stress and the
corrosion environment.
The real mechanism of SSC is not quite understood,
despite there being several explanations for it. In general,
the mechanism of SSC can be divided into two main
parts: the mechanism of anodic dissolving and the ca-
thode mechanism that causes hydrogen embrittlement of
the material.2
For some types of austenitic stainless steels, like
AISI 304 and AISI 316, it has been known for a long
time that they are sensitive to SCC, but mostly at tem-
peratures above 55 °C.
The performed investigations revealed that the local
damage and leaks are a result of SCC phenomena. The
SCC in austenitic stainless steel appears locally in the
form of thin branched cracks that can grow very quickly
and can cause failure of the structure.
One possible reason for SCC of the tank is the
longitudinal orientation of the crystal grains due to cold
rolling of the sheet (Figure 5). The directed microstruc-
ture is evidence that the material was not properly
recrystallization annealed after the cold rolling, and thus
the internal stresses remained in the material. This was
also confirmed with cracks in the longitudinal direction
following the longitudinally deformed crystal grains.
Additional possible sources of stresses are the cold
forming and the welding of the tank.
The elimination of internal stresses and the stabi-
lization of austenite is possible in austenitic stainless
steel with annealing to a temperature of 1050 °C,
followed by rapid cooling in water. Such measures also
prevent the formation of brittle phases in the steel,
typical for a slow cooling process.
In general, SCC can be prevented by the:
• selection of a more resistant material,
• elimination of internal stresses with material anneal-
ing,
• elimination of chloride ions in the storage liquid.
The corrosion pits were observed on the external
surface of the tank due to crevice corrosion in the gap
between the insulation and the tank surface. Crevice
corrosion is typical for narrow crevices with a lack of
oxygen. At such places the formation of a new protective
layer of chromium oxide (re-passivation) on the steel
surface is not possible and crevice corrosion proceeds.
The corrosion process is also accelerated by the presence
of chloride ions. Crevice corrosion led to the formation
of pits where the initiated cracks and SCC started. The
brittle cracks due to SCC that spread from the surface
through the tank wall are probably a consequence of
hydrogen embrittlement.
Both stored media in the tank have a pH from 2.6 to
11.3. The content of the tank was either acid or alkaline,
both of which accelerate the corrosion processes on the
internal surface of the tank.
For the investigated tank the corrosion process could
be prevented by good contact between the insulation and
the wall. This would prevent the formation of condensed
water on the surface of the tank from the trapped
moisture in the gap.
4 CONCLUSIONS
Our investigations confirmed that the leakage of the
tank, made of AISI 304 austenitic stainless steel, is a
consequence of the interaction of several localized corro-
sion processes and the presence of internal stresses in the
material. The SCC originates in the crevice corrosion
pits on the external surface of the tank. The moisture
captured in the gap below the insulation condensed
during the cooling of the tank and together with the
presence of chloride ions enabled the start of crevice
corrosion.
All the necessary factors were met for the appearance
of SCC: sensitive material, internal stresses in the mate-
rial and corrosion media. With the proper combination of
all three, the crevice corrosion pits start to grow and later
the SCC developed. Cracks, as a result of SCC, are
responsible for the leakage of the tank.
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK
Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466 465
5 REFERENCES
1 C. L. Page, R. D. Anchor, Stress corrosion cracking of stainless
steels in swimming pools, The Structural Engineer, 66 (1988) 24,
418
2 S.P. Lynch, Mechanisitc and fractographic aspects of stress-corrosion
cracking (SCC), Stress Corrosion Cracking, Theory and Practice , A
volume in Woodhead Publishing Series in Metals and Surface
Engineering, 2011, 3–89, doi:10.1533/9780857093769.1.3
3 N. Baddoo, P. Cutler, Stainless steel in indoor swimming pool
buildings, The Structural Engineer, 82 (2004) 9, 24–26
4 M. Tsubota, Y. Katayama, Y. Saito, Stress Corrosion Crack Growth
Behavior of Cold Worked Austenitic Stainless Steel in High Tem-
perature Water, Environmental Degradation of Materials in Nuclear
Systems XII, Proc. 12th International Conference on Environmental
Degradation of Materials in Nuclear Power Systems – Water
Reactors, ed. by T.R. Allen, P.J. King and L. Nelson, TMS, Warren-
dale, PA 2005, 109–115
5 A. Turnbull, K. Mingard, J.D. Lord, B. Roebuck, D. Tice, K. Motter-
shead, N. Fairweather, A. Bradbury, Sensitivity of stress corrosion
cracking of stainless steel to surface machining and grinding pro-
cedure, Corros. Sci., 53 (2011) 10, 3398–3415, DOI:
10.1016/j.corsci. 2011.06.020
6 S. Lozano-Perez, T. Yamada, T. Terachi, M. Schroder, C.A. English,
G.D.W. Smith, C.R.M. Grovenor, B.L. Eyre, Multi-scale characteri-
zation of stress corrosion cracking of cold-worked stainless steels
and the influence of Cr content, Acta Mat., 57 (2009) 18,
5361–5381, doi:10.1016/j.actamat.2009.07.040
7 G. Hinds , L. Wickström, K. Mingard, A. Turnbull, Impact of surface
condition on sulphide stress corrosion cracking of 316L stainless
steel, Corrosion Science, 71 (2013) 43–52, doi:10.1016/j.corsci.
2013.02.002
8 A. Bhattacharya, Stress Corrosion Cracking of Duplex Stainless
Steels in Caustic Solutions, Georgia Institute of Technology, Atlanta,
GA 2008, 8
9 J. Orlikowski, S. Krakowiak, Pitting corrosion and stress corrosion
cracking of buffer tanks in a brewery, Engineering Failure Analysis,
29 (2013) 75–82, doi:10.1016/j.engfailanal.2012.10.013
10 J. Woodtli, R. Kieselbach, Damage due to hydrogen embrittlement
and stress corrosion cracking, Engineering Failure Analysis, 7 (2000)
6, 427–450
466 Materiali in tehnologije / Materials and technology 50 (2016) 3, 461–466
M. TORKAR et al.: DEGRADATION OF AN AISI 304 STAINLESS-STEEL TANK