ISSN 1318-0010 KZLTET 32(1-2)2(1998) WELD JOINT FRACTURE BEHAVIOUR OF HSLA STEELS DISSIMILAR IN STRENGTH POTEK LOMA V ZVARNEM SPOJU DVEH TRDNOSTNO NEENAKIH VISOKOTRDNOSTNIH JEKEL INOSLAV RAK, A. TREIBER University of Maribor, Faculty of Mechanical Engineering IKGS, Welding Laboratory, Smetanova 17, 2000 Maribor Prejem rokopisa - received: 1996-10-01; sprejem za objavo - accepted for publication: 1997-12-19 Effect of strength differences (mis-match) between weld metal and two base metals as well as local variations of strength within weld metal/HAZ zones on the toughness properties are discussed. Significance of local fracture toughness measurement technique is also discussed by comparing the CTOD results of 85 and British Standard 8bs. Some differences in two techniques are discussed in particular for CGHAZ toughness of similar and dissimilar joints. Key words: strength mis-matching, hardness, materials, fracture toughness, fracture path Obravnavan je vpliv trdnostne neenakosti med strjenim zvarom in dvema osnovnima materialoma, kakor tudi razlike v trdnosti med strjenim zvarom in TVP podro~ji, na lomne lastnosti. Poleg tega je obravnavan pomen tehnik merjenja lomne ilavosti s primerjavo CTOD vrednosti 85 z 8bs vrednostjo dobljeno po britanskem standardu. Obravnavane so bile nekatere razlike, ki nastopajo pri obeh tehnikah merjenja, posebno za GZTVP v istovrstnih in razli~nih zvarnih spojih. Klju~ne besede: trdnostna heterogenost, trdota, materiali, lomna ilavost, potek loma 1 INTRODUCTION The narrow gap welding procedure is suitable for welding of thick sections because of beneficial arc operation and smaller base material dilution range and HAZ width when compared to conventional welding processes. Furthermore the narrow welding groove preparation enables lower weld joint misalignment and residual stresses. The technique can be particularly useful for welding of two steels of different strength with welding consumable which provides the weld metal (WM) strength properties in between. In this way substantial differences in strength properties (mis-matching) of two welded base materials (BM), WM, and both heat affected zones (HAZ) can occur particularly in the welded structures of high strength steels. The aim of the present work was to establish the yielding behaviour and fracture toughness of WM and HAZ of similar and dissimilar weld joints of HSLA steels. The differences in mechanical properties among different weld regions affect the development of plastic zone and hence strain distribution around the crack tip during the fracture toughness test and consequently may influence the CTOD fracture toughness values. It is believed that both strength mis-match and toughness would control the fracture behaviour of the dissimilar weldjoint depending on the testing temperature. Extensive amount of work is currently being carried out to understand the behaviour of mis-matched weld joints and to develop methodologies for treatment of this issue with respect to testing and defect assessment procedures. Further the aim of this investigation was to determine the fracture toughness properties of similar and dissimi- KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 lar weld joints and hence establishing the effect of neighbouring base plate strength (mis-match) on the toughness of HAZ and WM. This has been achieved by comparing the experimental CTOD results obtained from measurement of local 85 and standard fracture toughness evaluation techniques. Furthermore, special treatment was given to the fatigue pre-cracking of CTOD specimens taken from as welded plate. If one uses the prescribed procedure of existing toughness testing standards (valid for uniform materials)1-3, then the fatigue crack tip front will not be straight due to bi-modal residual stress distribution in thickness direction and the heterogeneous hardness distribution along the crack tip4,5. To overcome this problem the modified version of High R-ratio so called GKSS "Step Wise High R-ratio" (SHR) method is useful6. By using this method and taking into account the highest value of WM yield stress, more straight crack tip fronts are achieved for full thickness specimens made of as-welded weld joints. 2 MATERIALS, WELD JOINT AND EXPERIMENTAL PROCEDURE Mechanical Properties and Mis-Matching Factor Determination Commercial high strength low alloyed (HSLA) steels in thickness of 50 mm were used in quenched and tempered condition (Q+T) corresponding to grade HT50 and HT80. For welding of steel coupons narrow gap welding procedure was selected producing over-matched and un- 117 I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... der-matched BM similar weld joints. All weld joints were produced with the same welding consumable. Chemical composition both BMs, real weld metal and original wire are presented in Table 1. The mechanical properties of base plates and WM regions of multipass weld joints are presented in Table 2. Due to the special feature of the dissimilar weld joints (two base materials) the strength mis-match factor M should be determined separately between WM and both sides of the weld joint (MHT50 and MHT80). Table 1: Chemical composition of base plates and weld metal. Wire composition as provided by the manufacturer is also included Tabela 1: Kemi~na sestava osnovnih plo{~ in dejanskega zvara. Navedena je tudi sestava varilne ice. Designation Composition (wt%) Pcm C Si Mn P S Cr Ni Cu Mo HT80 0.11 0.28 0.27 0.011 0.003 1.03 2.7 0.19 0.27 0.256 HT50 0.07 0.28 0.43 0.012 0.005 0.51 0.23 0.44 0.27 0.101 Weld Metal 0.084 0.34 0.32 0.026 0.009 0.13 2.15 0.02 0.205 Wire comp. 0.11 0.32 1.22 0.014 0.007 0.06 2.4 0.05 Pcm - Cold cracking parameter Table 2: Mechanical properties of base metal and dissimilar weld joint Tabela 2: Mehanske lastnosti osnovnega materiala in zvara trdnostno neenakega zvarnega spoja Designation Y.S. (MPa) U.T.S. (MPa) Elongation (%) Charpy (J) at-40°C Mis-match Factor, M Mht80 MHT50 HT80 690 765 23.0 70 HT50 380 490 33.8 300 WM cap 565 690 21 30(at 0°C) 0.82 1.49 WM middle 595 680 24 70(at 0°C) 0.86 1.57 WM root 585 665 22 0.85 1.54 Average WM 581.6 678.3 22.3 60(at 0°C) Average 0.85 1.53 M = Y.S.wm/Y.S.bm, M<1 - under-matching condition, M>1 -over-matching condition The WM mechanical properties were determined by cylindrical tensile specimens extracted from the top, Figure 1: Macro cross showing the position of round tensile specimens and notches for WM and HAZ CTOD specimens Slika 1: Prerez zvara z vrisanimi poloaji nateznih preizku{ancev in CTOD-preizku{ancev middle and root region, in the weld axis direction of the narrow gap welded joint shown in Figure 1. These data represent only the mechanical properties of regions were the specimens were taken from. The yield strength of the WM was calculated by using formula of ayw = 3.15HV-168 which is proposed for WM7. The WM HV1 hardness measurements with the distance of 1 mm between indentations in the three through thickness directions in Figure 2 have provided the average values of distribution of WM mis-matching factor M across the weld joint, as shown in Table 3. They can be compared with those values in Table 2. The only exception was the very narrow band on both sides of the fusion line in the WM, the mismatching factor was locally changed due to alloying/dilution mechanisms from base metal to molten pool and which can not be determined by standard all-weld metal cylindrical tensile specimens. These bands might affect the yielding behaviour at the vicinity of the crack tip and influence the value of locally measured CTOD. The distribution of global/local mis-matching factor M calculated by hardness values over the weld joint is shown in Figure 3. Table 3: WM mis-matching factor M distribution across dissimilar WM determined by average hardness values Tabela 3: Porazdelitev M-faktoija skozi zvar dolo~ena preko povpre~nih vrednosti trdote _Mis-matching factor M_ WM HT50 WM HT80 WM hardness -0.15 +0.2 +1 WM midd. line +1 +0.2 -0.15 at fusion line mm mm mm mm mm mm Average hard- 217 223 225 225 224 227 274 ness M 1.36 1.40 1.42 1.42 0.78 0.78 0.80 1.14 Y.S.bmht50 = 380 MPa Y.S.BMHT80 = 690 MPa Series of Charpy impact toughness and full thickness SENB specimens were extracted from the weld joints. Charpy impact toughness specimens were taken from both BMs, CGHAZ, WM cap passes, WM upper section (between cap and middle passes) and WM middle passes. The specimen notch positions are shown in Fig- Figure 2: Hardness measurement directions in WM of dissimilar narrow gap weld joint Slika 2: Podro~ja merjenja trdot v prerezu trdnostno neenakega zvarnega spoja z ozko reo 14 KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... Figure 3: Global/local mis-match factor M in the dissimilar weld joint cross section Slika 3: Globalni/lokalni M-faktor v prerezu trdnostno neenakega zvarnega spoja ure 4. The geometry of SENB specimen was B x 2B (B = 40 mm). The through thickness fatigue cracks with ratio a/W ~ 0.5 were positioned in the WM and at the vicinity of the fusion line as shown in Figure 1. The testing temperature was -10°C, however some specimens were also tested at -20°C and +10°C. During the test DC potential drop technique was applied for stable crack growth monitoring8. The load line displacement (LLD) was also measured with the reference bar to minimise the effects of possible indentations of the rollers. The CTOD values were calculated in according to BS 5762 (§BS)1 and directly measured with GKSS developed §5 clip gauge on the specimen's side surfaces at the fatigue crack tip over gage length of 5 mm9. 3 RESULTS 3.1 Charpy Impact Toughness The results of the Charpy-V specimens are shown in Figure 5 in terms of seven average transition curves which represent different notch locations. Figure 5: Charpy-V impact toughness values obtained from dissimilar narrow gap weld joint (see figure 4); two base metals, BM HT80 and BM HT50 curves are also included for comparison reason Slika 5: Rezultati preizkusov udarne ilavosti trdnostno neenakega zvarnega spoja z ozko reo (glej sliko 4); dodani so tudi rezultati osnovnih materialov BM HT80 and BM HT50 za primerjavo The lowest Charpy toughness represents the material of WM at the weld cap position. The remaining WM material shows better impact toughness than in the cap layer but lower than both HAZs. Both BMs show excellent impact toughness and in fact all data of BM HT50 represent the upper shelf toughness. Both HAZ transition curves are drastically decreased when compared to the original base plate curves, particularly at lower temperature regime. HT80 HAZ transition temperature is shown at approx. -40°C (~70 J) and for HT50 HAZ at approx. -35°C (~160 J). The upper and middle WM transition temperature is shown at approx. 0°C (~70 J) and for BM HT80 at approx. -40°C (~200 J) while BM HT50 shows in the whole tested region upper shelf values. Figure 4: Charpy-V impact toughness specimens extracted from WM and HAZ of dissimilar narrow gap weld joint Slika 4: Poloaj Charpy ilavk v prerezu trdnostno neenakega zvarnega spoja z ozko reo Figure 6: WM and BM R-curves for similar and dissimilar narrow gap weld joints Slika 6: Odpornostne krivulje osnovnih materialov in zvarov istovrstnih in trdnostno neenakih zvarnih spojev 436 KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... 1.0 0.9 0.8 0.7 ¥ 5 0.6 m (O 0.5 O O H 0.4 U 0.3 0.2 0.1 0.0 > oBM-HISO : A HAZ-HT50 n ° BM-HT80 HAZ-HT80 HAZ-dHT80 HAZ-dHT50 O ! tfa A* É1 < A ' rf o o ou ■ U H o 3 t¥ « * o» O • o ti*' §5 = 85 (W-ao)rp (W-ao)rp+ao-a§, (1) Table 4: SENB CTOD values of BM HT50 and HT80 and their weld joints Tabela 4: CTOD vrednosti osnovnih materialov in njunih zvarnih spojev 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Aa (mm) Figure 7: HAZ and BM R-curves for similar and dissimilar narrow gap weld joints Slika 7: Odpornostne krivulje osnovnih materialov in TVP istovrstnih in trdnostno neenakih zvarnih spojev 3.2 CTOD Fracture Toughness In Tables 4 and 5 the typical values of "apparent" CTOD fracture toughness and stable crack initiation toughness, CTOD at Aa = 0.2 mm is intrinsic fracture toughness for both, CTOD (BS) and CTOD(85) are presented to show typical differences. All measured §5 values are corrected due to difference between average ao (because of penny shaped fatigue crack front) and the 85 clip gauge position on both specimen sides in accordance with the proposal in reference10. According to this proposal the following formula (1) for §5 correction was used: Material/weld Aa CTOD CTOD CTOD CTOD ao a§5 joint, Testing (mm) §5 8BS §5 8BS meas. meas. temperature corr. (mm) Aa=0.2 0.2mm (mm) (mm) (mm) (mm) BM-HT80 1.02 0.540 0.620 0.222 0.267 41.6 40.4 at -10°C 0.85 0.566 0.549 0.242 0.243 43.9 41.6 " 0.38 0.334 0.376 0.233 0.265 41.1 40.5 BM-HT50 2.57 2.424 2.435 0.405 0.405 43.4 40.0 at -10°C 2.95 2.685 2.723 0.314 0.310 43.5 40.3 " 2.68 2.525 2.667 0.281 0.289 42.5 40.0 WM-HT80 0.51 0.251 0.305 0.165 0.202 42.0 42.0 at -10°C 0.13 0.154 0.190 - - 41.2 41.2 " 0.05 0.098 0.125 - - 41.4 41.4 " 0.11 0.142 0.183 - - 41.6 41.6 WM-HT50 3.19 2.076 2.218 0.199 0.247 41.0 40.6 at +10°C 2.34 1.367 1.521 0.189 0.239 41.1 40.4 at -10°C 0.22 0.070 0.123 - - 42.5 41.5 " 2.07 1.260 1.277 0.218 0.260 41.8 40.0 " 2.10 1.066 1.260 0.136 0.190 41.5 40.8 at -20°C 0.07 0.072 0.097 - - 41.9 41.1 HAZ-HT80 0.84 0.609 0.697 0.259 0.294 40.5 40.3 at -10°C 0.48 0.312 0.346 0.227 0.252 41.2 40.2 " 0.89 0.526 0.635 0.203 0.247 40.6 40.2 HAZ-HT50 at 2.28 1.863 2.429 0.273 0.349 39.5 39.5 +10°C at -10°C 2.35 1.841 2.362 0.241 0.339 40.0 40.4 " 2.62 1.994 2.603 0.263 0.376 39.4 40.4 " 2.67 2.162 2.766 0.309 0.290 40.1 41.4 " 2.67 2.280 2.712 0.286 0.340 39.5 40.0 §5 = sides surface measured CTOD at the crack tip W = specimen depth ao = average specimen half crack length rp = rotation factor a§5= half crack length measured at the specimen sides surface In Table 4, the SENB CTOD values for each weld joint made on similar BM while in the Table 5 the SENB CTOD values for weld joint made on both dissimilar BMs are presented. In Figure 6, the R-curves which represent the lowest CTOD values for both similar and dissimilar weld joints WMs compared to BMs R-curves are shown, while in Figure 7 the R-curves which represent the lowest CTOD values for both dissimilar and similar weld joints HAZs compared to BMs R-curves are shown. Figure 8 presents one of the dissimilar WM specimen failure behaviour; fracture path deviation towards lower strength HT50 base metal and occurrence of brittle fracture at the fusion line. 4 DISCUSSION It was expected that global mis-matching ahead of the crack tip would play an important role in yielding behaviour, crack initiation and fracture path development in dissimilar weld joint. These result from the complex mis-match condition which govern in the dissimilar weld Figure 8: Fracture initiation in the weld metal, slow crack growth deviation to BM HT50 and brittle fracture propagation along the fusion line on WM side Slika 8: Za~etek loma sredi zvara, stabilna rast razpoke do osnovnega materiala BM HT50 in krhki lom vzdol' linije zlitja znotraj zvara 16 KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... Table 5: SENB CTOD values of BM HT50 and HT80 dissimilar weld joints Tabela 5: CTOD vrednosti trdnostno neenakih zvarnih spojev jekel HT50 and HT80 Material/weld Aa CTOD CTOD CTOD CTOD ao a85 joint, Testing (mm) 85 8BS 85 8BS meas. meas. temperature corr. (mm) Aa=0.2 0.2mm (mm) (mm) (mm) (mm) WM 2.05 0.905* 1.160 - 0.245 40.4 40.1 at -10°C 0.06 0.083° 0.098 - - 40.1 39.8 " 0 0.088° 0.108 - - 41.1 40.4 " 4.12 2.473** 2.800 0.150 0.216 36.5 38.7 HAZ-HT80 0.74 0.350 0.640 0.175 0.319 39.8 38.9 at -10°C 0.63 0.392 0.650 0.216 0.346 38.3 38.8 " 1.12 0.676 1.110 0.231 0.368 38.2 38.2 " 0.53 0.323 0.690 0.187 0.383 37.5 38.7 HAZ-HT50 3.18 2.272 2.470 0.186 0.253 44.1 40.7 at -10°C 3.05 2.289 2.930 0.243 0.325 37.9 38.7 " 3.01 2.651 3.040 0.273 0.377 38.1 39.5 " 2.68 0.939 2.630 0.073 2.63 37.8 38.5 * Fracture initiation started at the WM middle and stable crack growth deviated towards the HT50 BM but arrested in the WM (8m) "Fracture initiation at the WM middle section and stable crack deviated towards HT50 BM, passed WM and at the weakest ICCGHAZ portion of HAZ the brittle fracture occurred following the fusion line (8u) ° Fracture initiation at the WM and crack remained in WM as LBZ (8c) joint due to different strength and toughness micro structural properties near the fusion line area. 4.1 Weld Joint Impact Toughness The lowest Charpy impact toughness was obtained WM in the cap region, however in general dissimilar weld metal joint showed lower toughness compared to both base plates, Figure 5. From this observation it can be concluded that WM produced by narrow gap welding dissimilar steels would be the most sensitive portion and potential region for LBZ occurrence. 4.1.1 Weld Metal Fracture Toughness The lowest fracture toughness of dissimilar WM is in two cases equal to some cases appearing in both similar WMs. One can compare the lowest value for both similar and dissimilar WM in Table 4 and 5. This fracture behaviour is governed by microstructural toughness and fracture initiation following the WM dendrites grain boundaries as brittle fracture. In other cases, where the fracture toughness was higher, obviously the fracture initiated and propagated through tougher WM regions by ductile crack growth mechanism. In the two remaining dissimilar weld joint cases the fracture is governed by lower HT50 strength properties of HT50 steel. Figure 8 shows the fracture initiation at the WM and crack, which deviated by slow growth toward HT50 BM. At the weakest HAZ-ICCGHAZ area, before the crack reached BM-HT50- Figure 9, the brittle fracture occurred following 17 KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 the unfavourable less tough fusion line. It seems that crack initiation and propagation behaviour in dissimilar weld joints is rather a complex event. We assume that crack propagation after crack initiation in this case was governed by HT50 lower strength properties until the vicinity of higher toughness HT50 HAZ was reached where the fracture was further governed by toughness properties which were lower in ICCGHAZ near the fusion line. So, it its clearly shown that strength mis-match in combination local material properties can affect the fracture behaviour. Similar behaviour has occurred also in another specimen in which the max. loading was attained before fracture occurred and before ductile crack growth has reached the HAZ HT50 fusion line. As mentioned earlier, fracture behaviour has shown high CTOD fracture toughness in both specimens as it can be seen from Table 5. It seems that elongated dendrites offer high resistance against fracture propagation perpendicularly to their primary axis. The R-curves, Figure 6, present the results of both similar WMs (provided on HT80 and HT50) and dissimilar WM comparing to both BMs. The lowest value of CTOD fracture toughness is for all three WMs rather equal (0.07 mm for WMHT50, 0.098 mm for WMHT80 and 0.083 mm for WM diss.) and bellow initiation at Aa = 0.2 mm. The similar WM-HT80 has shown some amount of crack extension before fracture while WM-HT80 and dissimilar WM specimen have failed immediately after crack initiation. Unstable events are designated by the arrows on Figure 6. On the other hand and as shown in Table 4 and 5 higher fracture toughness values are predominant for other specimens. Obviously, the CTOD testing in accordance with standards valid for uniform materials (a/W ~ 0.5) is too conservative for toughness assessment of welded joints provided on HSLA steels. This conclusion results also from the fact that LBZ can appear after some amount of slow crack growth also in HT80 BM despite its high impact toughness as it can be seen from Figure 6 and Figure 5. It would be better to evaluate weld joints (also produced on dissimilar steels) by testing wide plate specimens where the constraint conditions are approx. similar to weld joints operating under their full loading. 4.1.2 HAZ Fracture Toughness Average HAZ similar weld joint fracture toughness presented in Table 4 (HAZ-HT80, HAZ-HT50) is in the same order of magnitude when compared to the average HAZ dissimilar weld joint fracture toughness (HAZ-HT80, HAZ-HT50) presented in the Table 5. From both tables one can recognise that HT80 HAZ shows lower fracture toughness than HT50 HAZ. This is understandable due to the lower steel strength level but higher impact toughness of the later. Some of specimens tested did not show any instability until the end of testing. This is the consequence of excellent BM toughness and of the influence of narrow gap welding procedure where the heat input direction does not attack the BM in the way as I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... fi 3i-L pii'li; f:|liiin lln-rrajli ^ M Figure 9: Slow crack growth passing WM at ICCGHAZ (coarse grain M-A constituents) area transformes into brittle fracture mode Slika 9: Stabilna rast razpoke na prehodu iz zvara v podro~je IKGZTVP preide v krhki lom it does by welding of conventional bevelled weld joints (X,V,K grooves). CTOD fracture toughness of HAZs is much better comparing to similar and dissimilar WMs. Their lower band toughness is presented by R-curves comparing to R-curves of BMs on Figure 7. In all cases considerable crack extension was recognised before fracture occurred or before reaching the max. loading value. The events where the instability has occurred mainly at the crack extension higher than 1.0 mm as shown are designated by arrows in the figure. The behaviour of HAZ HT80 specimens taken from similar and dissimilar narrow gap weld joints was quite interesting. At testing LBZ appeared in WM after initiation and after transferring the slow crack growth to softer WM. Obviously the influence of softer WM and HT50 BM, comparing to HT80 BM on the crack tip was essential for the development of the plastic zone in WM and the measured CTOD fracture toughness does not represent the toughness property of similar and dissimilar HAZ HT80. This can be observed in Figure 7 for HAZ-HT80 in similar weld joint and HAZ-dHT80 dissimilar weld joint. The fracture behaviour of both R-curves is similar after the crack entering into WM. The HAZ-HT50 fracture toughness for similar and dissimilar weld joint does not represent the real toughness properties as well. The fracture initiates at the HAZ-HT50, but later deviates into the BM-HT50 due to WM-HT50 over-matching properties. It seems that CGHAZ fracture toughness testing should be modified/changed. The crack tip location perpendicular to weld joint direction a few mm before fusion line on straight HAZ side offers a good solution for CGHAZ fracture toughness determination. 1 5 CONCLUSIONS An experimental programme has been carried out to compare strength and toughness of mis-matched narrow gap weld joints produced on two HSLA steels in BM similar and dissimilar state by the equal weld consumable. The results of this research work can be summarised as follows: 1. Due to the beneficial arc operation by narrow gap welding procedure of two dissimilar BMs the base material dilution range as the HAZ width range are smaller than in conventional welding processes. The consequence of it is clearly expressed as global strength mismatch condition. It seems that the only exception is very narrow band along the fusion line on the WM side where weak local mis-matching can appear. 2. Charpy impact toughness in WM and HAZ is for similar and dissimilar weld joint approx. on the same level. The lowest impact toughness is revealed at WM cap layer where the hardness is the highest and no beneficial tempering heat from the subsequent passes is available. The impact toughness of WM is lower than that of both BM HAZs. 3. WM CTOD fracture toughness is in both similar weld joints in some cases good enough in some cases very low. Obviously, the fracture initiation depends on fatigue crack angle to dendrites orientation and either the crack tip is sampling particles/impurities on the dendrite grain boundary or is located in the grain matrix. The same can appear in the dissimilar weld joint. But in case of slow crack growth the crack deviates in the direction of the softer BM. When the fracture passes through WM the brittle fracture can appear if the crack tip reaches the KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2 I. RAK, A. TREIBER: WELD JOINT FRACTURE BEHAVIOUR OF HSLA ... local less tough ICCGHAZ area before BM with high toughness. It seems that the fracture initiation and propagation is complex and can be governed by toughness and strength properties around the crack tip. 4. The dissimilar weld joint HAZ CTOD fracture toughness is excellent comparing to WM CTOD one. HAZ toughness of the BM with higher strength can be overestimated due to effect of softer WM and propagation of the crack from HAZ perpendicular to WM dendrites. The same can be recognised for HAZ toughness of the BM with lower strength due to softer BM and crack deviation from CGHAZ to BM. 5. Two CTOD measurement methods used (CTOD §5 and CTOD §BS) especially on dissimilar weld joints have shown in some cases huge differences. Once again it was shown, how important the local measurement is without the need to infer from remotely measured quantities. All measured CTOD §5 values were corrected due to difference between average ao and the §5 clip gauge position on both specimen sides in accordance with the mentioned proposal. ACKNOWLEDGEMENT This work is partially funded by International Bureau DLR Bonn, Germany and Slovenian Ministry of Science and Technology Ljubljana. Authors would like to thank GKSS Research Centre, M. Koçak and his colleagues for their help in the experimental work and for his comments to the manuscript. 6 REFERENCES 1 BS 5762: 1979. Method for Crack Opening Displacement (COD) Testing, the British Standards Institution 2 ASTM E 1290-91. Standard Method for Crack-Tip Opening Displacement (COD) Fracture Toughness Measurement 3 European Structural Integrity Society, ESIS Recommendation for Determining the Fracture Resistance of Ductile Materials, ESIS P1-92 4 K.-H. Schwalbe, M. Koçak: Fracture Mechanics of Weldments: Properties and Application to Components, Keynote Lecture on the 3rd International Conference on Trends in Welding Research, June 1-5, 1992, Gatlinburg, Tennessee, USA 5 Y. Mukai, A. Nishimura: Fatigue Crack Propagation Behaviour in the Hardness Heterogeneous Field; Transactions of the Japan Welding Society, 14 (1983) 1, April 6M. Koçak, K. Seifert, S. Yao, H. Lampe: Comparison of Fatigue Pre-cracking Methods for Fracture Toughness Testing of Weldments: Local Compression and Step-Wise High R-ratio, Proc. of the Int. Conf. Welding-90, Oct. 1990, Geesthacht, FRG (ed. by M. Koçak), 307-318 7 R. J. Pargerter: Yield Strength from Hardness a Reappraisal for Weld Metal; Welding Research Bulletin, Nov. 1978 8 K.-H. Schwalbe, D. Hellmann: Application of the Electrical Potential Method to Crack Length Measurement using Johnson's Formula; JTEVA, 9 (1981) 3, 218-221 9 GKSS-Forschungszentrum Geesthacht, "GKSS-Displacement Gauge Systems for Application in Fracture Mechanics", 1991 10 A. Treiber, I. Rak: Fracture Properties of a Strength Mismatched Narrow Gap Weld Joint; The First ASM Conference on Welding and Science Technology, Madrid, Spain 10-12 march 1997 2 KOVINE, ZLITINE, TEHNOLOGIJE 32 (1998) 1-2