B. Ule*1, F. Vodopivec*2, M. Pristavec*3, F. Grešovnik*4
UDK: 620.192.49:621.785.72:669.14.018.258 ASM/SLA: Q26s, N8a, TSk
Popustna krhkost utopnega jekla za delo v vročem s 5 % kroma
Temper Embrittlement of 5 wt.-% Cr Hot Work Die Steel
Raziskali smo vpliv parametrov popuščanja na udarno žilavostin lomne značilnosti utopnega jekla za delo v vročem Č.4751 s postmartenzitno mikrostrukturo.
Prevladujoči dejavnik, ki kontrolira razvoj krhkosti, je segregiranje fosforja na mejah zrn in na drugih med-plastjih v kovini. Segregacijski učinek fosforja spremlja kosegregacija silicija skupaj s sočasno precipitacijo karbidov.
1.	UVOD
Orodno jeklo s sekundarnim utrjevanjem z 0,4 % C, 5 % Cr, 1,3 % Mo, 0,4 % V in 1 % Si (C.4751) se uporablja predvsem za orodja in utope, ki obratujejo pri povišanih temperaturah, ker združuje dobro obstojnost trdote z veliko obrabno odpornostjo, zadostno žilavost ter majhno občutljivost na pokanje v vročem. Jeklo Č.4751 kaže efekt sekundarnega utrjevanja, ako je popuščano pri temperaturah okrog 500° C, ugotovljeno pa je bilo tudi, da s pričetkom sekundarnega utrjevanja sovpada znaten padec žilavosti. Mehanizem pojava je povezan s precipitacijo karbidov MSC in MC iz martenzita, kot tudi z atermalno transformacijo zaostalega avstenita. Ugotovljeno je bilo1, da to krhkost lahko skoraj odpravimo z odstranjevanjem silicija iz jekla in zdi se, da je posledica vpliva silicija na precipitacijo karbidov med letvicami martenzita.
Ako pa jeklo Č.4751 popuščamo nekaj ur v temperaturnem območju med 550° C in 600° C, opazimo določen zastoj v pričakovani evoluciji žilavosti2. Mikrostrukturni izvor te krhkosti do sedaj še ni bil zadovoljivo pojasnjen, zato smo ga podrobno raziskali.
2.	EKSPERIMENTALNI DEL
Majhno količino jekla Č.4751, trgovske kvalitete, smo pretalili v 20-kg vakuumski peči, ga ulili v ingot ter le-tega v vročem izvaljali v palico, premera 15 mm. Kemična sestava jekla je prikazana v tabeli 1.
Tabela 1: Kemična sestava preiskanega jekla v ut. %
C	Si Mn P S Cr Mo V Al
0,39 0,93 0,30 0,023 0,014 5,86 1,12 0,28 0,038
*1 mag. Boris Ule, dipl. ing. met., SŽ — Metalurški intštitut Ljubljana, Lepi pot 11, 61000 Ljubljana *2 SŽ — Metalurški inštitut Ljubljana *3 Inštitut Jožef Štefan, Ljubljana
*4 SŽ — Železarna Ravne
The influence of tempering parameters on impact energy and fracture characteristics of Č.4751 hot work die steel with post-martensitic microstructure has been investigated.
The dominant factor controlling the evolution of brit-tteness is the segregation of phosphorus on grain boundaries as well as on other interfaces in metat. The segregation effect of phosphorus is accompanied by cosegregation of silicon together with simultaneous car-bide precipitation.
1.	INTRODUCTION
Secondary hardening tool steel with 0,4 % C, 5% Cr, 1,3 % Mo, 0,4 %V and 1 % Si (Č.4751) is used primarily for tools and dies vvhich operate at eievated temperatures because it associates good hardness retention vvith high resistance to wear. sufficient toughness and iittie susceptibility to heat checking. Č.4751 steel exhi-bits secondary hardening effect when tempered around 500° C and a sharp drop in toughness, coincident vvith the onset of secondary hardening, was observed too. The mechanism involved is associated vvith precipitation of M6C and MC type carbides from the as-quenched martensite and also due to athermal transformation of retained austenite. It has been established1 that this embrittlement can be almost etiminated by removal of silicon and it seems that the embrittlement resulted from the influence of silicon on interlath carbide precipitation.
But when Č.4751 steel is tempered at temperature range of 550° C to 600° C for a few hours some standstill in the expected evolution of toughness is observed2. The microstructure origins of this embrittlement have not been ctarified yet that is why we investigated them thoroughiy.
2.	EXPERIMENTAL
A small quantity of commerciai Č.4751 steel vvas remelted in a 20 kg vacuum-induction furnace, then it vvas čast in an ingot and hot-rolled to a 15 mm rod. The chemical composition of the steel is shovvn in Table 1
Table 1: Chemical composition of the investigated steel, wt. - %
C	Si	Mn	P S	Cr	Mo	\/	Al
0,39	0,93	0,30	0,023 0,014	5,86	1,12	0,28	0,038
Charpy V-notch specimens vvere machined from the rod, vvhich was previously normalized and soft annealed
Iz palice, ki je bila normalizirana in mehko žarjena 2 uri pri temperaturi 800° C, smo izdelali Charpyjeve pre-izkušance z V zarezo. Preizkušanci so bili v vakumski peči avstenitizirani 15 min. pri 960° C, kaljeni v toku plinastega dušika pri tlaku 0,5 MPa. nato dvakrat popuščeni po 2 uri pri 710° C, z vmesnim podhlajevanjem pri -196° C, ter nazadnje dodatno popuščeni v temperaturnem območju od 450 do 660° C 2 uri, 5 ur oziroma 24 ur, s končnim ohlajanjem v vodi.
Merjenja žilavosti pri sobni temperaturi so bila opravljena z vsaj petimi Charpyjevimi preizkušanci za vsako temperaturo popuščanja, medtem ko je bila Brinellova trdota merjena na vsakem preizkušancu.
Mikrofraktografske preiskave prelomnih površin Charpyjevih preizkušancev smo opravili z vrstičnim elektronskim mikroskopom JEOL JSM-35 (SEM), medtem ko smo mikrostrukturo preiskali s presevnim elektronskim mikroskopom JEOL FX (TEM), ki je bil opremljen z analizatorjem karakterističnih rentgenskih žarkov (EDS).
Opravili smo tudi elektrolitsko izolacijo karbidov, izo-lat pa je bil analiziran z rentgensko difrakcijsko tehniko.
3. REZULTATI
V diagramu na sliki 1 so prikazane odvisnosti med udarno žilavostjo in temperaturo dodatnega popuščanja, različno dolge čase popuščanih Charpyjevih V-preizku-šancev, ki so bili pred tem kaljeni ter dvakrat popuščeni pri 710° C, z vmesnim podhlajevanjem v tekočem dušiku.
Podhlajevanje med začetnim dvojnim popuščanjem je bilo potrebno, da bi se izognili vplivu morebitnega zaostalega avstenita.
Medtem ko trdota jekla ostaja navidezno neodvisna od parametrov popuščanja, pa Charpyjeva udarna žila-vost po 24-urnem popuščanju pri 600° C pade od začetnih 42 J na vsega okrog 20 J. Močan padec Charpyjevih vrednosti pri jeklu, ki je bilo 5 ur, zlasti pa 24 ur popuš-čano pri 600° C, je verjetno posledica razvoja reverzi-bilne popustne krhkosti, kot je bilo ugotovljeno v eni prejšnjih raziskav3.
Segregacije, ki nastopajo pri reverzibilni popustni krhkosti, so ravnotežnega tipa. Začetna segregacija
at 800" C for 2 hours. Specimens vvere then austenitized at 960°C for 15 min. in a vacuum furnace. quenched in a ftovv of gaseous nitrogen at a pressure of 0.5 MPa. tem-pered tvvice at 710° C for 2 hours vvith intermediate undercooiing at — 196° C, then additionaiiy tempered in a temperature range of 450° C to 660°C for 2 hours, 5 hours and 24 hours respectiveiy and finaiiy cooied in vvater.
At ieast five Charpy V-notch specimens vvere tested at room temperature for each tempering temperature, whiie the Brine/1 hardness was measured on every speci-men. The m/crofractographic examination on fracture surfaces of Charpy specimens vvas carried out in the JEOL JSM-35 scanning eiectron microscope (SEM), vvhiie the microstructure vvas investigated in trans-mission eiectron microscope JEOL FX (TEM) equipped vvith energy dispersive spectroscopy (EDS). The electro-/ytic isolation of carbides vvas a/so carried out and the isotate vvas anaiyzed vvith X-ray diffraction technique.
3. RESULTS
Charpy V-notch energy versus tempering tempera-tures curves as obtained for specimens additionaiiy tempered for various periods, after quenching and doubie tempering at 710° C vvith intermediate undercooiing in /iquid nitrogen, are shovvn in Fig. 1 Undercooiing in betvveen the initial doubie tempering vvas necessary to avoid the influence of eventua/ retained austenite.
The hardness of steel remains virtually independent of the tempering parameters. vvhereas the Charpy V-notch impact energy drops from initial value of 42 J to onty about 20 J after tempering 24 hours at 600° C. The drastic drop in the Charpy values of steel after tempering for 5 hours but particularly 24 hours at 600°C is probably partly due to reversibie temper embrittlement. as established in one of eariier investigations3.
The segregations involved in reversibie temper embrittlement are of equi/ibrium type. namely vvhen the ageing temperature is risen. the initial segregation rate increases follovving the temperature dependence of the buik diffusion coefficient, vvhiie the maximum i. e. steady
Kaljeno in dvakrat popuščeno pri 710°C Ouenched and tempered twice at 710 °C
Ponovno popuščano>-, Additionally tempered for ^
(5)" - IM hours,y hours and ^ hours respectively
A50	500	550
Temperatura popuščanja v °C Tempering temperature (°C)
Slika 1:
Vpliv dodatnega popuščanja na trdoto in udarno žilavost jekla Č.4751, ki je bilo pred tem kaljeno in dvakrat popuščeno pri 710° C, z vmesnim podhlajenjem v tekočem dušiku Fig. 1
Influence of additional tempering on hardness and impact ener-gies of steel Č.4751 vvhich has been previously quenched and tempered tvvice at 710° C vvith intermediate undercooiing in liquid nitrogen.
Qa = 160.3 - A.5 kJ/m	ol
	
	
	
11
1
U
,3
1.3
T'
Slika 2:
Izvrednotenje aktivacijske energije za segregiranje fosforja z uporabo Arrheniusove enačbe Fig. 2
Evaluation of the activation energy of segregation of phospho-rus according to the Arrhenius eguation.
namreč z dviganjem temperature žarjenja raste ter sledi temperaturni odvisnosti koeficienta volumske difuzije. Maksimalna, to je stacionarna interkristalna koncentracija pri tem pada, in to razlaga reverzibilnost krhkosti pri visokih temperaturah feritnega območja, kot je bilo tudi ugotovljeno v že citirani referenci3.
Udarna žilavost je sorazmerna intenziteti segregacij, enak nivo segregacij je bil zato dosežen s popuščanjem, bodisi 2 uri pri 592° C bodisi 5 ur pri 555° C bodisi 24 ur pri 505° C. Ako predpostavimo, da še celo po 24 urah popuščanja ni dosežena končna največja intenziteta segregacije, potem je levi del diagrama na sliki 1 mogoče uporabiti za določitev aktivacijske energije za volumsko difuzijo oligoelementa, ki kontrolira razvoj krhkosti. Iz naklona premice v dvojnem logaritemskem diagramu časa popuščanja proti recipročni vrednosti temperature popuščanja, prikazanem na sliki 2, je bila izračunana aktivacijska energija približno 160 kJ/mol, kar je zelo blizu aktivacijski energiji za difuzijo fosforja v feritu. Zares je že bilo potrjeno z Augerjevo spektroskopijo, da v jeklih podobne vrste segregira zlasti fosfor. Romhanyi s sodelavci4 je v orodnem jeklu s 5 % kroma, ki je bilo avstenitizirano pri 1100° C, kaljeno in popuščano 2 uri pri 600° C, našel na mejah kristalnih zrn do 6% fosforja in 1 % žvepla, pa tudi sledove dušika in antimona. Augerjev spekter, prikazan na sliki 3, dokazuje, da so segregacije fosforja in žvepla v soodvisnosti z obogatitvijo s kromom (8 do 9%), vanadijem (približno 2 %) in molibdenom (3 do 5 %). Nadalje je opazen izrazit ogljikov pik (8 do 9 %) z deloma karbidno strukturo. Augerjevi piki kažejo rahlo lokalno fluktuacijo, vendar pa bi lahko poudarili soodvisnosti med kromovim pikom ter vsoto segregiranega fosforja in žvepla.
Že omenjene spremembe v udarni žilavosti se odražajo tudi na morfologiji prelomnih površin. Po dvakratnem popuščanju 2 uri pri 710° C, z vmesnim podhlajeva-njem v tekočem dušiku, je frakturna površina Charpyje-vih preizkušancev transgranularna, drobno jamičasta, torej duktilna, kakršna je prikazana na sliki 4. Po dodatnem popuščanju 24 ur pri 600° C se pot napredovanja razpok spremeni, prelom postane intergranularen vzdolž meja primarnih avstenitnih zrn (slika 5). čeprav so opaženi tudi kvazicepilni detajli ter posamični duktilni grebeni (slika 6).
x 3
dN(E) dE
r
N 381
512 0 Sb — 462
272C
437473 I v ^ 489 5 29Cr Cr
703 Fe
E(eV) Slika 3:
Augerjev spekter intergranularne prelomne površine jekla s 5 % kroma, avstenitiziranega pri 1100° C, kaljenega ter popuščen-ega pri 600" C (Lit. 4) Fig. 3
Auger spectrum of the intergranuiar fracture surface of steel vvith 5wt.-% chromium. austenitized at 1100°C. quenched and tempered at 600°C (Ftef. 4).
state grain boundary concentration decreases. This accounts for the "reversibiiity" of embrittlement at higher temperatures of the ferritic range, as observed a/so in the aiready quoted reference3.
The impact energy is proportiona/ to the leve/ of segregation and the same leve/ of segregation is there-fore achieved after tempering for 2 hours at 592° C or 5 hours at 555" C or 24 hours at 505° C. If the assump-tion is considered that the fina/ maximal leve/ of segregation is not attained even after 24 hours of tempering, then the /eft side of the diagram shovvn in Fig. 1 could be used for the determination of the activation energy for bulk diffusion of residuals, vvhich controis the development of embrittlement. An activation energy of about 160 kJ/mol was derived from the stope of a log-log p/ot of tirne i/s. reciprocat tempering temperature in Fig. 2, which is very ciose to that for bulk diffusion of phospho-rus in ferrite. It was a/ready confirmed indeed by the Auger spectroscopy, that part/cular/y phosphorus segregates in such type of tooi steel. Romhanyi and covvorkers4 found up to 6 % of phosphorous and 1 % of sulphur as weH as traces of nitrogen and antimony on the grain boundaries in 5wt. —% chromium tooi steel, austenitized at 1100"C, quenched and tempered at 600"C for 2 hours. The Auger spectrum shovvn in Fig. 3 proved, that the segregation of phosphorus and sulphur is in correlation vvith the enrichment of chromium (8—9 %), vanadium (approx. 2 %) and molybdenum (3—5 %). Further. the strong carbon peak (8—9 %) vvith partly carbide structure is remarkab/e. The Auger peaks exhibit slight local fluctuaction, but the correlation of the chromium peak vvith the sum of phosphorus -l- sulphur segregated should be underlined.
The already mentioned changes in the impact values are a/so ref/ected in fracture surfaces morpho/ogy. After tempering twice at 710° C for 2 hours vvith intermediate
Slika 4:
Fraktografski posnetek Charpyjevega preizkušanca, kaljenega in dvakrat popuščenega pri 710° C, z vmesnim podhlajenjem v tekočem dušiku Fig. 4
Fractographs of Charpy specimen, quenched and tempered twice at 710° C vvith intermediate undercooiing in liquid nitrogen.
Slika 5:
Fraktografski posnetek Charpyjevega preizkušanca, kaljenega in dvakrat popuščenega pri 710° C, z vmesnim podhlajenjem, nato dodatno popuščenega 24 ur pri 600° C. Interkristalno krhko Fig. 5
Fractographs of Charpy specimen. quenched and tempered twice at 710° C with intermediate undercooling. then addltionally
tempered at 600° C for 24 hours. Intercrystalline brittle.
V literaturi5-8 so bili kot najvažnejši identificirani pre-cipitati v kaljenem in pri povišanih temperaturah popu-ščenem orodnem jeklu s 5 % kroma, navedeni M2C, M7C3 ter M6C. Okuno7 je ugotovil, da so v orodnem jeklu H-13, kaljenem v olju in popuščenem pri 600° C, le karbidi MC in M6C, karbide M7C3 v obliki aglomeriranih zrn najdemo po popuščanju pri višjih temperaturah, med 600 in 650° C.
Tip precipitatov je bil določen z uporabo TEM in EDS analize, kot tudi z rentgensko difrakcijsko tehniko, upoštevaje pri tem citirane podatke, nanašajoče se na termično zgodovino jekla. Rentgenska difrakcija elektrolitskega izolata odkrije v jeklu, ki je bilo kaljeno in med dvakratnim popuščanjem pri 710° C še podhlajeno, karbide M7č3 in M6C v enakem deležu. Elektronski posnetek mikrostrukture tega vzorca jekla, prikazan na sliki 7, kaže značilno substrukturo visoko popuščenega mar-tenzita, sestavljeno iz malih podzrn, nanizanih vzdolž prvotnih martenzitnih igel, vrste precipitatov v glavnem vzdolž meja podzrn in redke precipitate v letvicah mar-tenzita.
Po dodatnem popuščanju istega jekla 24 ur pri 600° C smo ugotovili majhen porast količine karbidov M6C, posledično pa je bila zmanjšana količina karbidov, vrste M7C3. Posamični precipitati M6C so bolj grobi, vendar pa število precipitatov vzdolž meja podzrn ter vzdolž letvic ostaja skoraj nespremenjeno (slika 8). Relativno majhna razlika v količini in obliki karbidov ne more biti odgovorna za povsem drugačno morfologijo preloma, povzročeno z dodatnim popuščanjem jekla 24 ur pri temperaturi 600° C. Pri EDS analizi smo pozornost usmerili zlasti na precipitate na mejah podzrn in med letvicami, ker bi lahko bili povezani s potjo razpoke. Tipični EDS spekter precipitata v jeklu, ki je bilo po kaljenju in dvakratnem popuščanju pri 710° C, z vmesnim podhlaje-
Slika 6:
Fraktografski posnetek Charpyjevega preizkušanca, kaljenega in dvakrat popuščenega pri 710° C, z vmesnim podhlajenjem, nato dodatno popuščenega 24 ur pri 600° C. Kvazicepilno s posamičnimi duktilnimi grebeni Fig. 6
Fractographs of Charpy specimen, quenched and tempered twice at 710° C with intermediate undercooling. then additionally tempered at 600° C for 24 hours. Quasi-cleavage and single ductile tearing.
undercooling in liquid nitrogen. the fracture surface is small dimpled and transgranuiar, therefore ductile (Fig. 4). After an additional tempering at 600°C for 24 hours, the cracks propagation path changed and an intergranular fracture along preaustenite grain bounda-ries (Fig. 5) quasi-cleavage fracture details and single ductile tearing are observed too (Fig. 6)
In references5~8 the main precipitates present in as quenched and at elevated temperature tempered 5wt.-% chromium tool steel vvere identified as M2C. M-C3 and M6C. Okuno7 estabiished that in oii quenched and at 600° C tempered H-13 tool steel only carbides MC and M6C are found. M7C3 carbides in shape of agglomer-ated grains are found after tempering at higher tempera-tures betvveen 600 ° C and 6500 C.
The type of precipitates was estabiished using TEM and EDS analysis as vvell as X-ray diffraction technique, also considering the quoted data relating to the thermal history of the steel. X-ray diffraction of electrolytic iso-late reveaied in steel. as quenched and tempered twice at 710" C vvith intermediate undercooling. M7C3 and M6C carbides in equal portion. Electron micrographs of this specimen in Fig. 7 shovv a characteristic substructure of high-tempered martensite, consisting of small subgrains arranged along the former martensite needles. rovvs of precipitates mainly along the subgrain boundaries and rare intralath precipitates.
After additional tempering of the same steel for 24 hours at 600° C a small increase of the quantity of M6C carbides and consequential decrease of the quan-tity of M7C3 precipitates are found. Some M6C precipitates are coarser, hovvever the number of precipitates along iath and subgrain boundaries remains nearly unchanged (Fig. 8) The relatively small difference in quantity and size of carbides couldn 't be responsible for
Slika 7:
TEM mikrografski posnetek jekla Č.4751, kaljenega in dvakrat popuščenega pri 710° C. z vmesnim podhlajenjem v tekočem dušiku. Fig. 7
TEM micrographs of steel Č.4751, quenched and tempered twice at 710° C with intermediate undercooling in !iquid nitrogen.
! 1 ij	f[ tJ |	j		
!|c K r" i ji • i' ■s : 11		1 ! i I		
Vi I / l / c j W		1 i > h	a A	>1
Slika 9:
Značilni EDS rentgenski spekter medplastja martenzita letvica/ karbid v jeklu Č.4751, kaljenem in dvakrat popuščenem pri 710° C, z vmesnim podhlajenjem, nato dodatno popuščenem 24 ur pri 600° C Fig. 9
Characteristic EDS X-ray spectrum of martensite lath/carbide interface in steel Č.4751, quenched and tempered twice at 710° C with intermediate undercooling. then additionaiiy tempered at 600° C for 24 hours.
Slika 8:
TEM mikrografski posnetek jekla Č.4751, kaljenega in dvakrat popuščenega pri 710° C, z vmesnim podhlajenjem, nato dodatno popuščenega 24 ur pri 600° Č Fig 8
TEM micrographs of steel Č.4751. quenched and tempered twice at 710° C with intermediate undercooling. then additionally tempered at 600° C for 24 hours.
the completely different fracture morphology produced by the additional tempering of steel for 24 hours at 600° C.
A t EDS X-ray analysis the attention was focused on precipitates at lath and subgrain boundaries. because they could be connected to the crack path. A typical EDS spectrum for a precipitate in steel, quenched and tempered twice at 710° C with intermediate undercooling then additionally tempered at 600° C for 24 hours, is shown in Fig. 9 It is characteristic for chromium rich carbide but it is not clear vvhether iron and moiybdenum are detected from carbide phase or from the matrix. By focusing the beam to the martensite lath/carbide interface. the strong peak of silicon and phosphorus was regular/y detected. The peak of sulphur could be covered with that of molybdenum. For this reason it is not possible to detect an eventual segregation of sulphur at grain boundaries.
4. DISCUSSION
Temper embrittlement of Č.4751 tool steel is easily provoked when the high-tempered steel with post-mart-ensitic microstructure is additionally tempered for a few hours at 600° C.
Fracture morphology of additionally tempered steel changes from transcrystalline and ductile into mixed mode i. e. intercrystalline and quasi-cleavage transcrys-
vanjem, še dodatno popuščeno 24 ur pri temperaturi 600° C, je prikazan na sliki 9. Spekter je značilen za s kromom bogat karbid, čeprav ni jasno, ali sta železo in molibden detektirana iz karbidne faze ali iz osnove. S fo-kusiranjem snopa na mejo martenzita letvica/površina karbida smo redno zabeležili močan silicijev in fosforjev pik. Pik, ki pripada žveplu, bi bil lahko prekrit z molibde-novim pikom. Iz tega razloga ni mogoče zaznati morebitnih segregacij žvepla na mejah zrn.
4.	RAZPRAVA
Popustno krhkost orodnega jekla Č.4751 lahko povzročimo tako, da pri visokih temperaturah popuščeno jeklo s postmartenzitno mikrostrukturo dodatno popuščamo nekaj ur pri temperaturi 600° C. Morfologija preloma dodatno popuščenega jekla se spremeni od trans-kristalne duktilne v mešano obliko, to je interkristalno ter kvazicepilno transkristalno. Med duktilnim in krhkim jeklom pa nismo opazili nobene pomembne razlike, nanašajoče se na morfologijo ter sestavo karbidnih pre-cipitatov.
Iz rentgenske EDS analize sledi, da vsebnost molibdena v trdni raztopini ni bila bistveno spremenjena z dodatnim popuščanjem nekaj ur pri 600° C. Po drugi strani je očividno, da prisotnost molibdena v trdni raztopini ni v celoti preprečila segregiranja fosforja na notranjih površinah. Segregiranje fosforja se pojavlja vzporedno s segregiranjem silicija, aktivacijska energija za volumsko difuzijo silicija v feritu (258 kJ/mol) pa je mnogo višja od aktivacijske energije za volumsko difuzijo fosforja v feritu (167 kJ/mol). Ako bi segregacije obeh elementov, silicija in fosforja, imele enak učinek, bi pričakovali, da bo kinetika porajanja krhkosti kontrolirana z difuzijo počasnejšega silicija na meje kristalnih zrn. Dejstvo, da je kinetika krhkosti kontrolirana z difuzijo fosforja, bi lahko razložili na dva načina, bodisi tako, da ima segregiranje silicija le manjši vpliv na energijo loma, bodisi z naravo segregiranja silicija, ki naj bi bila drugačna od fosforjeve, namreč neodvisna od volumske difuzije.
Stopnjevanje krhkosti zaradi sočasne prisotnosti fosforja in silicija v jeklu je omenjeno tudi v referencah 1 in 8. Ugotovljeno je bilo še, da bi silicij lahko vplival na tvorbo karbidov, vrste M5C (ref. 5). Segregiranje silicija ter precipitiranje karbidov M6C na mejah in v medplastjih smo opazili tudi mi. Tega pa še ne moremo imeti za dokaz, da segregiranje silicija vpliva tudi na tvorbo karbidov M6C, saj smo našli na mejah zrn in med letvicami martenzita tudi karbidne precipitate, vrste M7C3.
5.	SKLEPI
Popustna krhkost visoko popuščenega orodnega jekla 6.4751 s postmartenzitno mikrostrukturo je nastala zaradi segregiranja fosforja na primarnih mejah avstenit-nih zrn, kot tudi na različnih drugih medplastjih. Posledica tega je interkristalni oziroma krhek transkristalni prelom jekla.
Sočasno s segregiranjem fosforja je bilo opaženo tudi segregiranje silicija. Segregiranje silicija ter izločanje karbidov ima sicer lahko pomembno vlogo v procesu nastajanja krhkosti, vendar pa ugotovljena aktivacijska energija, približno 160 kJ/mol, izhajajoča iz časovno-temperaturne odvisnosti poslabšanja žilavosti zaradi dodatnega popuščanja, dokazuje, da je difuzija fosforja v feritu odločujoči dejavnik, ki kontrolira kinetiko krhkosti.
ta/line respectively. No significant distinction in morphol-ogy and composition of carbide precipitates is observed between ductiie and brittie state of steel. From EDS X-ray analysis ensues that the content of molybdenum in solid soiution was not substantially modified at the addi-tiona/ tempering for a few hours at 600° C. On the other hand, it is evident that the presence of molybdenum in solid soiution did not prevent entirely the interfacial segregation of phosphorus. The segregation of phos-phorus occurs simultaneously with that of silicon. The activation energy of bulk diffusion of silicon in ferrite (258 kJ/mol) is much greater than that for bulk diffusion of phosphorus (16/ kJ/mol). If the segregation of both elements, silicon and phosphorus. had the same effect one would expect that the kinetics of embrittlement vvould be controlled by the diffusion of the slower silicon tovvards grain boundaries. The fact that the kinetics of embrittlement is controlled by the diffusion of phosphorus could be explained in two ways. either the segregation of silicon has only a minor effect on fracture energy or that the nature of silicon segregation is different from that of phosphorus. i. e. independent of bulk diffusion. An enhanced embrittlement due to the simultaneous presence of phosphorus and silicon in steel is quoted also in ref. 1 and 8. It was also found that silicon could even affect the formation of carbides of M6C type5. The presence of a segregation of silicon and precipitates of the M6C carbide at boundaries and interfaces was also observed in this work. This could not be considered as evidence that the segregation of silicon did affect the formation of M6C carbide. since precipitates of M7C3 carbide vvere found on grain boundaries and martensite lath interfaces.
5. CONCL USIONS
Temper embrittlement of Č.4751 tool steel vvith high-tempered post-martensitic microstructure was pro-duced by segregation of phosphorus at preaustenite grain boundaries as well as at other different interfaces. This further caused intercrystalline or transcrystalline brittie fracture mode respectively.
A segregation of phosphorus was observed simultaneously vvith that of silicon. The segregation of silicon and precipitation of carbides may play an important role in embrittlement process. hovvever the activation energy of about 160 kJ/mol derived from time-temperature rela-tionship of toughness reduction because of additional tempering proves that the diffusion of phosphorus in ferrite is a dominant factor controlling the kinetics of embrittlement.
LITER ATU R A/ REFERENCES
1.	W. M. GARRISON, Jr.: Mater. Sci. Technol., 1987, 3, 256-259
2.	F. VODOPIVEC, L. KOSEC, A. RODIČ in J. RODIČ: Rud., geolog, i metal., 1984, 1, 35—42
3.	B. ULE, F. VODOPIVEC in A. RODIČ: Žel. metal, zbornik, 1985, 19, (3), 79-83
4.	K. ROMHANYI, ZS! SZASZ CSIH, G. GERGELY and M. MENYHARD: Kristali und Technik, 1980, 15, (4), 471 -477
5.	B. R. BANERJEE: J. Iron Steel Inst., 1965, 203, 166—174
6.	JIN YU and C. J. McMAHON, Jr.: Metali. Trans. A, 1980, 11 A, 277; ibid., 291
7.	T. OKUNO: Trans. ISIJ, 1987, 27, 51—59
8.	L. A. NORSTROM and N. OHRBERG: Metals Technol., 1981, 1, 22—26