ZELEZARSKI ZBORNIK IZDAJAJO ŽELEZARNE JESENICE, RAVNE, ŠTORE IN METALURŠKI INŠTITUT LETO 25 LJUBLJANA DECEMBER1991 Vpliv varilne tehnologije in izbire dodajnega materiala na lomne lastnosti EPP zvarnega spoja na nizko ogljičnem finozrnatem jeklu The Influence of Welding Technology and Welding Material Selection on Fracture Properties of Submerged Are Welded, Low Carbon, Finegrained Steel Plate I. Rak1, V. Gliha2, F. Vodopivec3, M. Tavčar2 1. UVOD Razvoj nizko ogljičriih jekel, ki so izdelana na termo-mehanski način ali kaljena in popuščana in ki so uporabljana v konstrukcijah z zahtevnimi obremenitvenimi pogoji, je v veliki meri spremenil varilno tehnologijo. Zaradi nizkega ogljika in nizke vsebnosti difuzijskega vodika v zvaru izdelovalci jekel ne priporočajo več predgrevanja pred varjenjem. Pri tem ni nevarnosti, da bi v TVP nastopila razpokljivost v hladnem, kar je mogoče npr. preveriti z Durenovim in Suzuki konceptom /1, 2/. Če navedeno prenesemo na dejanske zvarne spoje (hlajenje pod 50° C po vsakem varku in začetek varjenja brez predgrevanja), ki so zvarjeni z nizko dovedeno toploto (10—15kJ/cm), da bi zajamčili dobro žilavost v grobozr-natem predelu TVP, se lahko v raztopljenem zvaru pojavi nizka žilavost kot posledica tvorbe podolgovatih M/A strukturnih faz glede na kemično sestavo jekla in dodajnega materiala. V primeru napetostnega žarjenja pri 580° C se pojavi neugodni efekt izločevanja Fe3C iz M/A faz, ki predstavljajo prenasičeno raztopino. Pri tem je žilavost lahko še nižja, če je staljeni zvar občutljiv na re-verzibilno popuščno krhkost. V tem prispevku so obravnavane samo lastnosti raztaljenega dela zvarnega spoja. Zaradi jasnega efekta vpliva znižanega dovoda toplote pri varjenju in zaradi zmanjšanja velikosti zrna in primarnega ferita po kristalnih mejah je bil izbran za raziskavo dodajni material z 0,4% Cr in 0,2% Mo ter z dodatkom Ti-B. Uporabljena je bila metoda elektro varjenja pod praškom - EPP. Žilavost zvara je bila ugotavljana z udarnim Charpijevim kladivom. Zareze v Charpy preizku-šancih so bile locirane v kovini in korenski legi X simetričnega zvara pravokotno na debelino. Razlog za to je bilo pričakovano različno razmešanje, predvsem s stališča Nb (NB = 0,01 % v krovni legi in 0,04% v korenski legi). Preizkusi so bili opravljeni na celotni debelini zvara po metodi CTOD z upogibnimi preizkušanci, zarezanimi in utrujanimi pravokotno na debelino v sredini zvara. Lomne površine so bile preiskane z vrstičnim elektronskim mikroskopom predvsem glede pojava in lokaci- 1 Tehniška fakultata Maribor, VTO-S, 2 RRI, Metalna Maribor, 3 Inštitut za kovinske materiale in tehnologije 1. INTRODUCTION Development of LC steels, produced on thermome-chanical way or by quenching and tempering and used in constructions under sophisticated load conditions, has largely changed the vvelding technology. Because of the L C and lovv diffusibie hydrogen content in the weld, the steel producers do not recommend preheating before vvelding. There is no danger of cold cracking appearance in HAZ what is possible to check up vvith Duren and Suzuki theory for exampie /1, 2/ Transferring these statements into the real vveldment (interpass temperatures under 50 degrees Celsius and no preheating before vvelding) vvelded vvith low heat in-put (10-15 KJ/cm) to ensure good toughness in coarse grain area of HAZ, can in the melted part of the weld cause the appearance of low toughness values as the result of oblongated M/A struetural phase formation de-pending on the chemical composition of steel and vvelding material, in the čase of stress reiieved heating at 580 degrees Celsius there appears the undesirable effect to Fe3C precipitation from M/A phases representing a satu-rated solution. The toughness values can be lovvered even more if the sensitivity of the melted part of the vveld on a reversible temper brittieness is present. Because of the dear effects of iow heat input at vvelding and reduced grain size and primary ferrite on crystai boundaries, the vvelding material vvith 0,4% Cr and 0,2% Mo and Ti-B additions vvas chosen in these re-searches. The SA W vvelding method vvas used. The vveld toughness vvas determined by Charpy V-notch impact test. The notehess of Charpy impact specimens vvere lo-cated in face and root location of doubie V butt type vveld, right angled on the thickness of the plate. The rea-son vvas in the expected uneven dilution, first of aH from the Cb point of vievv (Cb= 0,01% in the face of the vveld and 0,04% in root of the vveld). Further tests vvere carri-ed out on the compiete vveld thickness by the CTOD method on the bend type specimens notehed and fati-gued in the rightangled direetion to the thickness in the middle of the vveld. Fractured surfaces vvere examined vvith line EM, first of aH on the appearance of LBZ. Both, line EM and opti-cal microscope vvere used by the examinations of mic-rostruetures in the vvelded joint. The presented resuits je LKP. Mikrostrukture zvara so bile preiskane z optičnim in vrstičnim elek. mikroskopom. Rezultati dokazujejo, da staljeni del zvara kljub drobnemu zrnu in nizki vsebnosti primarnega ferita po kristalnih mejah ni vedno dovolj žilav. Torej ocena struktur z optičnim mikroskopom po priporočilu IIW /3/ ni vedno zadostna za analizo in določitev vzrokov za dobro ali slabo žilavost staljenega zvara. Dodatne metode, kot npr. uporaba vrstičnega mikroskopa, lahko pokažejo dejanski vzrok za nizko ali visoko žilavost v staljenem zvaru v izhodnem in napetostno žarjenem stanju. 2. LASTNOSTI IN MIKROSTRUKTURA NIZKO OGLJIČNEGA EPP STALJENEGA ZVARA Poznan je odnos med mikrostrukturo zvara in žila-vostjo /4/. Razmešanje raztaljenega zvara v času varjenja spreminja transformacijsko kinetiko staljenega zvara, pri čemer imajo lahko vključki znaten vpliv. Gostota, velikost in porazdelitev vključkov narekujejo razvoj velikosti avstenitnih zrn po strditvi. Oksidni in drugi vključki ter koncentracija avstenitne trdne raztopine skupaj s hitrostjo ohlajevanja vplivajo na različne feritne morfologije v času premene (y/a /5a). Puščičasti ferit se npr. pojavi v EPP zvaru, če je koncentracija 02 višja od 450 ppm; njegova rast iz primarnega ferita na kristalni meji v zrno je intergranularna. Pri nižjih vsebnostih 02 se pojavi več ugodnega acikularnega ferita, medtem ko nizke vsebnosti 02 vodijo do tvorbe bainitne strukture. Poleg vsebnosti 02 je pomembna velikost in enakomerna porazdelitev vključkov. Vključki velikosti >0,2 (im bodo povzročili pospešeno tvorbo acikularnega ferita in znižali vsebnost primarnega ferita, pri čemer se povečajo primarna den-dritna zrna s fino porazdeljeno intergranularno strukturo /5b/. V času tvorbe primarnega in puščičastega ferita se preostali avstenit znatno obogati s C ter se lahko trans-formira v faze, ki vsebujejo zaostali avstenit in martenzit ali bainit-M/A strukturne faze. Takšno strukturo često najdemo pod izrazom ferit s sekundarno fazo. Prisotni vključki so tvorci puščičastega ferita, katerega oblika je odvisna od hitrosti ohlajevanja /6/. Višje hitrosti ohlajevanja pospešujejo tvorbo Widmanstattenskega ferita, ki ga spremlja neugodna porazdelitev M/A faz. Dodatki Nb ta fenomen še pospešujejo, kar se kaže v nižji žilavosti. Na drugi strani dodatki Ti in B izboljšajo žilavost, ker pospešujejo tvorbo acikularnega ferita, ki se pojavi na drobnih intergranularnih vključkih. Večina elementov vpliva na tvorbo in hitrost rasti puščičastega ferita. Dodatki Si in Al pospešujejo tvorbo puščičastega ferita, medtem ko jo Mn in Mo zavirata. Pretvorbo iz acikularnega ferita v ferit s sekundarno fazo, kot npr. M/A fazo, pospešujeta Cr in Mo, ki tudi povišujeta mejo plastičnosti in trdnost /7/. 3. RAZISKAVA SOČELNEGA EPP STALJENEGA ZVARA 3.1. Izbira osnovnega in dodajnega materiala Raziskave so bile opravljene na EPP zvarnem spoju nizkoogljičnega poboljšanega jekla Niomol 390. Lastnosti osnovnega in dodajnega materiala so navedene v tabeli 1. Navedeni dodajni material je bil izbran, da bi poudarili razlike pri nižjem in višjem dovodu toplote zaradi njegove povišane zakaljivosti, kar je razvidno iz višjega Pcm v primerjavi z osnovnim materialom. V predhodnih raziskavah /8/, kjer je bil uporabljen komercialni dodajni material (z 1 % Ni s Pcm = 0,149 in je bilo varjenje opravljeno prav tako brez predgrevanja, so se pojavila vidna LKP v prelomu CTOD preizkušanca v show, that the melted part of the vveid is not aiways tough enough in spite of fine grained coarse and iow pri-mary ferrite on crystai boudaries. Therefore, the estima-tion of structures vvith opticai microscope, as it is rec-ommended by IIW /3/, is not aiways sufficient enough for the anaiyse and determination of good or bad toughness in the melted part of the vvelded joint. Additional methods, as line eiectron microscope for example. can shovv the real reason for lovv or high toughness values in the melted part of the vvelded joint in as-welded or stress relieved conditions. 2. THE PROPERTIES AND MICROSTRUCTURE OF THE LOVV CARBON MELTED PART OF THE SUBMERGED ARC VVELDED JOINT The relationship betvveen microstructure and toughness of the vvelded joint is knovvn /4/. Dilution of the melted part of the vvelded joint during vvelding is chang-ing the transformation kinetic of melted vveldment and the inclusions can have a significant influence on it. Den-sity, size and distribution of inclusions do dictate the size development of austenitic grains after the solidifica-tion. Oxide and other types of inclusions and concentra-tion of austenitic solid solution together vvith the cooling speed. are influencing upon different ferrite morpholo-gies during the transformation y/a /5a/. Acicuiar ferrite can appear for example in the SA vvelded joint, if the concentration of oxygen is higher than 450ppm; its grovvth is intragranular from primary ferrite on crystal boundary into the grain. A t the lovver oxygen concentra-tions more favourable acicuiar ferrite appears, vvhiie lovv oxygen concentration lead to the formation of bainite structure. Beside the lovv oxygen concentration the size and uniform distribution of inclusions is important. The inclu-sion sizes over 0,2[im will cause an accelerated formation of acicuiar ferrite and reduce the content of primary ferrite increasing at the same time the size of the den-drite grains vvith fine distributed intragranular structure /5b/. During the formation of primary and acicuiar ferrite the remained content of austenite becomes significantly rich vvith carbon and can transform in phases containing residual austenite and martensite or bainite-M/A struc-tural phases. Such structure we can often find in the ex-pression ferrite vvith the second phase. Present inclusions are authors of acicuiar ferrite vvhich shape de-pends on cooling speed /6/. Higher cooling speed ac-ceierates the formation of VVidmanstatt-ferrite accom-panied by unfavourable distribution of M/A phases. The additions of Cb accelerates this phenomena vvhat results in the iovver toughness. On the other side the additions of Ti and B improve the toughness because they accelerate the formation of acicuiar ferrite, vvhich appears on intragranular inclusions. The most of elements have their effect on the formation and grovvth speed of acicuiar ferrite. The additions of Si and Al accelerate the formation of arrovv-shaped ferrite, vvhiie Mn and Mo re-tard it. The transformation from acicuiar ferrite to ferrite vvith secondary phase, as for example M/A phase. is accelerated vvith Cr and Mo vvhich also increase the yield point and strength /7/. 3. RESEARCH OF THE BUTT — SA VVELDED MOL TEN JOINT 3.1. The choice of the base and vvelding material The researches vvere carried out on the SA vvelded, lovv carbon, guenched and tempered steel plate Niomol izhodnem in napetostno žarjenem stanju celo pri + 20°C. Preiskave na rasterskem mikroskopu so odkrile več kot 10 % M/A strukturne faze. Dodatek Ti-B je v tej preiskavi bil izbran z namenom preprečiti oz. znižati tvorbo primarnega ferita po kristalnih mejah in pospešiti tvorbo acikularnega ferita v strukturi staljenega zvara /9/. Vpliv grobega zrna in primarnega ferita na žilavost staljenega zvara je bil tako z izbiro navedenega dodajnega materiala izključen. Namen preiskave je bil ugotoviti vpliv M/A faz z ozirom na njihovo sestavo (visok C), velikost, usmerjenost in gostoto na udarno in lomno žilavost v odvisnosti od hitrosti ohlajevanja in legiranosti staljenega zvara /10, 11/. Nadaljnji namen je bil ugotoviti vpliv napetostnega žarjenja na razpad M/A faz in tako na žilavost staljenega zvara. Tabela 1: Lastnosti uporabljenih materialov Lastnosti osnovnega materiala Debelina Re a5 Žilavost CTODi, -40° C /mm/ /MPa/ /MPa/ /%/ -60OC/J/ /mm/ 30 432 528 25 198,161,149 0,51 Kemična sestava 0.08C 0,30Si 1.11Mn 0.006S 0.015P 0.049Nb 0,0 8Sn 0,18Ni 0,012As 0.047AI Pcm = 0,1- CE = 0,27- 49 0 Lastnosti čistega dodajnega materiala EPP- R„ Rm a5 Žilavost CTOD dod.mat. /MPa/ /MPa/ /%/ -60°C/J/ /mm/ OP121TT 480 520-620 24 >50, _ Fluxocord varjeno 35.22 >35, — zarjeno Kemična sestava 0.05C 0,20Si 1,2Mn 0,04Cr 0,20Mo 0.005B Tidod. Pcm = 0,1 -CE = 0,37- 75 0 Kemična sestava dejanskega EPP zvara 0,05C 0,36Si 1,67Mn 0,70Cr 0,41Mo 0.025V 0,031Ti 0.023AI 0,04Sn 0,04Sb 0.006B 0,007As 0.009S 0.020P 0,022Nb Pcm = 0.2-CE = 0,55- 02 = 227p- HD<2,6 - 46 7 pm ml /100 gr 3.2. Varjenje preizkusnih plošč in priprava preizkušancev Uporabljeno je bilo večvarkovno varjenje na simetrično pripravljenem X zvarnem žlebu. Keramični vložek je bil uporabljen z izvedbo korenskega varka. Zvarjeni sta bili dve preizkusni plošči z različnima tehnologijama, A-brez predgrevanja in B-s predgrevanjem. Podatki o varjenju: debelina plošče — 30 mm predgrevanje — brez (A), 150°C (B) vnesena toplota — 15kJ/cm At8/5 — 5,4s(A), 10,5s(B) vmesna temp. — 50°C(A), 200°C(B) pogrevanje — 200° C 390. The properties of base and vveiding material are gi-ven in Table 1. Given vveiding material has been chosen for the reason to point out the differences at lower and higher heat input because of its higher quench capabi/i-ty. what is obvious from higher Pcm value in comparison vvith base material. In previous researches /8/, vvhere commercial vveiding material (vvith 1% Ni) vvas used, vvith Pcm= 0.149 and vvhere vveiding vvas also carried out vvithout preheating, visibie LBZ have appeared on fractured CTOD specimen surfaces in as-vvelded and stress relieved conditions even at + 20 degrees Celsius. Raster microscope ex-aminations have determined more than 10% M/A struc-tural phase. Ti-B addition in this research has been chosen for the reason to prevent or to reduce primary ferrite formation on crystrai boundaries and to acceier-ate the acicular ferrite formation in the structure of molt-en weld /9/. With such vveiding material selection the influence of coarse grain in primary ferrite on molten weid toughness vvas expelled. The aim of the examination vvas to find out the influence of M/A phases concerning their composition (high C(, size. orientation and density on fracture toughness of the molten vveld depending on its cooling speed and chemical composition /10.11/. Another aim vvas to find the influence of stress relieving on the disin-tegration of M/A phases in the molten vveld. Table 1: Properties of material used Base material properties Thickness Re Rm 85 Tough- CTODI. /mm/ /MPa/ /MPa/ /%/ ness -40° C -6CPC/J/ /mm/ 30 432 528 25 198,161,149 0.51 Chemical composition 0.08C 0,30Si 1,11Mn 0.006S 0,015P 0,049Nb 0.08Sn 0,18Ni 0,012As 0.047AI Pcm= 0,1- CE= 0,27- 49 0 Weid metal properties S A l/V Re Rm 5 5 Tough- CTOD ness I/Ve Id. Mat. /MPa/ /MPa/ /%/ -6CPC/J/ /mm/ OP121TT 480 520—620 24 >50, _ Fluxocord as welded — 35.22 >35, — stress re/. Chemical composition 0.05C 0,20S i 1,2Mn 0.04Cr 0.20Mo 0.005B Ti add. Pcm= 0,1-CE= 0.37- 75 0 Chemical composition of reai weid meta/ 0.05C 0.36Si 1,67Mn 0,70Cr 0.41 Mo 0.025V 0.031 Ti 0.023AI 0,04Sn 0,04Sb 0.006B 0,007As 0,009S 0,020P 0,022Nb Pcm= 0,2-CE= 0,55-O227p- HD< 2,6- 46 7 pm ml /100 g 3.2. Welding of test plates and specimen preparation A vveiding technique vvith more passes vvas used on double-V grooved steel plate. For the root vveld accom-piishment a ceramic insert vvas used. There vvere two Po varjenju je bila polovica vsake plošče napetostno odžarjena pri 580°C v času 3 ur, nekaj izrezanih preizkušancev pa še pri 650° C ter s kombinirano termično obdelavo. Preizkušanci so bili izrezani iz zvarnega spoja v izhodnem in napetostno odžarjenem stanju. Pripravljeni so bili metalografski preizkušanci, žilavostni preizkušanci zarezani v krovni in korenski legi ter CTOD preizkušanci iz celotne debeline; vsi zarezani pravokotno na površino plošče v sredini staljenega zvara. 3.3. Udarna in lomna žilavost staljenega zvara Udarni Charpijevi preizkušanci so bili pretežno odvzeti iz krovne lege zvara in preizkušani v izhodnem in različno termično obdelanih stanjih. Namen je bil ugotoviti vpliv M/A strukturnih faz na udarno žilavost. Nekaj preizkušancev je bilo odvzetih iz korenskega dela zvara z namenom ugotoviti vpliv Nb, ki se je izcejal iz osnovnega materiala v zvar ter ugotoviti, ali je prisotna termična reverzibilna krhkost. Rezultati za udarno in lomno žilavost staljenega zvara v izhodnem stanju, zvarjenega z in brez predgrevanja, so dani v tabeli 2. Tabela 2: Udarna žilavost in vrednosti za CTOD Stanje Lokacija Tempera- Žilavost CTOD Trdota zareze tura /°C/ /J/ /mm/ /HV5/ 1. 2. 3. 4. 5. 6. Brez krovni + 20 92 pred- sloj, -20 50 — 274 grevanja. korenski + 20 88 varjeno sloj celotna + 20 0,148-au debelina 0 0,094-ac —20 0,064-oc Brez krovni + 20 72 pred- sloj, -20 24 grevanja, korenski + 20 20 zarjeno sloj -20 8 580° C celotna + 20 0,034ac debelina Brez krovni -20 23 260 pred- sloj grevanja, zarjeno 650° C Predg. in krovni -20 65 257 varjeno sloj Predg. in krovni + 20 103 žar. 580° C sloj -20 29 Predg. in krovni + 20 114 236 žar 650° C sloj -20 42 Brez predgrevanja, varjeno Ref/8/, 1 % Ni krovni sloj —20 78 celotna debelina + 20 0 -20 0,420-ou 0,062-ac 0,091-ac Brez predgrevanja, žarjeno, 580° C, Ref. /8/, 1 % Ni krovni sloj -20 65 celotna debelina + 20 0 -20 0,323-au 0,103-ac 0,062-ac test plates prepared, vvelded vvith two different technolo-gies; A-without preheating and B-vvith preheating. VVeiding data: — 30 mm - vvithout (A), 15CPC (B) — 15 KJ/cm - 5,4 s (A), 10,5 s (B) — 50oC (A), 200° C (B) - 20CPC piate thickness preheating heat input A/s/5 interpass temp. heating after vveiding After vveiding each ha/f of the plate vvas stress re-lieved at 580° C in a period of 3 hours but some speci-mens vvere heated to 65CPC vvith combinated heat treatment. Test sampies vvere cut from the vveid in as vvelded and stress relieved condition . Test specimens vvere cut out for metal/ographic examination, toughness, notched in face and root area of the vveld, and CTOD specimens form the vvhole vveid thickness, ali notched perpendicu-iary on the plate surface vvith the notch location in the middle of the molten vveid. 3.3. Notch and fracture toughness of the molten vveld metal Table 2: Notch toughness and CTOD vaiues Condition Notch Tempera- Tough- CTOD Hardness location ture/0 C/ ness /mm/ /HV5/ /J/ 1. 2. 3. 4. 5. 6. VVithout face + 20 92 preheat.. vveld -20 50 — 274 as vvelded root + 20 88 vveld -20 16 vvhole + 20 0,148-8U thickness 0 0,094-8c -20 0,064+0c VVithout face + 20 72 preheat.. vveld -20 24 heated. root + 20 20 58CPC weld -20 8 vvhole + 20 0,034-?> c thickness VVithout face -20 23 260 preheat.. weid heated. 650° C Preheatedface -20 65 257 as vvelded vveld Preheatedface + 20 103 heated weid -20 29 580° C Preheatedface + 20 114 heated vveld —20 42 236 650° C VVithout face -20 78 preheat., weld as vvelded, vvhole + 20 0,420-6 u Ref. /8/, thickness 0 0,062-8 c 1 % Ni -20 0,091-8C VVithout face —.20 65 preheat., weid heated, 58CPC, vvhole + 20 0,323-8 u Ref. /8/. thickness 2 0,103-8 c 1% Ni —20 0,062-8 c Za primerjavo so dani rezultati iz reference /8/, kjer je bil uporabljen komercialni dodajni material (1% Ni). Varjeno je bilo z isto nizko vnešeno toploto in brez pred-grevanja kakor pri tej preiskavi. Tudi v tem primeru so se pojavila vidna LKP na preloma vrednosti, dobljene pri CTOD preizkusu ob pojavu "pop-in" efekta, so uporabljene različne oznake. Za nastop "pop-in" efekta po počasni rasti razpoke je oznaka (au), za nastop "pop-in" efekta takoj za otopitvijo razpoke pa (oc). Občutljivost na popuščno reverzibilno krhkost /13/ se je določala z VVatanabe faktorjem J /14/ in z udarno žilavostjo; rezultati so dani v tabeli 3. Tabela 3: Popuščna krhkost, določana z udarno žilavostjo Tehnologi- Brez predgrevanja Predgrevanje na 150°C ja varjenja Vmesna temp. 50° C Vmesna temp. 200° C Termična Lokacija T Žilavost Lokacija T Žilavost obdelava* zareze /° C/ /J/ zareze /°C/ /J/ 550° C krovna + 20 72 krovna + 20 103 lega, lega, krovna -20 23 krovna -20 30 lega, lega, korenska -20 10 — — — lega 550° C + krovna -20 55 krovna + 20 108 750° C lega lega 550° C + krovna -20 30 krovna + 20 86 710°C + lega lega 550° C * Pri vsaki temperaturi po 4 ure; VVatanabe faktor J = (Si + Mn) + (P + S) x 104; v našem primeru je J = 512 Za J= <200 je nizka občutljivost na reverzibilno popuščno krhkost Za J= >400 je visoka občutljivost na reverzibilno popuščno krhkost 3.4. Metalografske preiskave z optičnim in vrstičnim mikroskopom Preiskave so bile opravljene na metalografskih obru-sih, odvzetih iz plošč, varjenih brez in s predgrevanjem. Sliki 1 in 2 prikazujeta stebričaste dendrite z drobno in-tragranularno strukturo in nizko vsebnostjo ferita po kristalnih mejah v staljenem zvaru, zavarjenem brez predgrevanja. Drobna struktura, posneta z vrstičnim mikro- Slika 1 Brez predgrevanja. Fig. 1 VVithout preheating. Notch toughness specimens were mostiy cut out from face area and tested in as we/ded and different heat treated conditions. The aim was to find out the influence of M/A structural phases on notch toughness. Some specimens vvere cut out from root area vvith the aim to find out the influence of Nb, segregated from the base material to the vveld and also if the reversible tem-per embrittlement is present. The result of notch and fracture toughness of the molten vveld metal in as vvelded condition and vvelded vvith and vvithout preheating are shovvn in Table 2. For comparison, the results from the reference /8/ are given, vvhere the commercia/ vvelding material (1% Ni) vvas used. Welding vvas carried out vvith the same lovv heat input energy and vvithout preheating as in this ex-amination. In this čase visible LBZ on the fractured sur-faces of the specimen also appeared at + 20 degrees Celsius Different designations are used for the CTOD vaiues at "pop-in" effect appearance. At "pop-in" effect appearance, after the slovv crack grovvth, the designa-tion (t)J is used and designation (8J at "pop-in" effect appearance immediately after the crack tip is blunted. The sensitivity on reversible temper embrittlement /13/ vvas determined vvith VVatanabe factor J /14/ and vvith notch toughness; the results are shovvn in Table 3. Table 3: Temper embrittlement determined vvith notch-tough-ness VVelding VVithout preheating Preheating 15CPC technolo-- gy Interpass temp. 50° C Interpass temp. 200° C Heat treatment' Notch location T /°C/ Toughness /J/ Notch location T /° C/ Toughness /J/ 500° C face vveld root vveld + 20 -20 —20 72 23 10 face vveld + 20 -20 103 30 550° C 750° C face vveld -20 55 face vveld + 20 108 550° C+ 75CPC+ 55CPC face vveld -20 55 face weid + 20 108 'at each temperature 4 hours; VVatanabe factor J= (Si+ Mn)+ (P+ S)x 10*; in our čase is J= 512 For J— < 200 is sensibiiity on reversible temper embrittlement lovv For J= > 400 is sensibility on reversible temper embrittlement high 3.4. Meta/lographic examinations vvith optic and line microscope Metalographic specimens vvere cut out and ex-amined from steel plates, vvelded vvithout and vvith preheating. Fig. 1 and Fig. 2 shovv columnal dendrites vvith fine intragranular structure and lovv ferrite content on crystal boundaries in the molten vveld, vvelded vvithout preheating. Fine structure taken off with a line microscope in the face vveld area is shovvn in Fig. 3; oblong M/A phases along intragranular precipitated ferrite are visible. Fig. 4 shovvs the microstructure after stress re-lieved heating; severa! cementite precipitations on the boundary betvveen M/A phase and ferrite are perceived. Microstructure of the molten vveld, vvelded vvith preheating is shovvn in Fig. 5; there are less M/A phases and or- Slika 2 Ista struktura kot slika 1. Fig. 2 The same structure as in Fig. 1. Slika 3 SEM mikrostruktura zvara, brez predgrevanj. Fig. 3 SEM microstructure of the weid, vvithout preheating. Slika 4 Ista struktura - nap. žarjeno pri 580 °C. Fig. 4 The same structure - stress - reiieved heated at 580°C. Slika 5 SEM mikrostruktura zvara, s predgrevanjem. Fig. 5 SEM microstructure of the weid. vvith preheating. Slika 6 Ista struktura - nap. žarjena pri 580°C. Fig. 6 The same structure - stress - reiieved heated at 580°C. ientation is iess distinctive. The same microstructure after stress-reiieved heating is shovvn in Fig. 6: a strong tendency of cementi te coaguiation is visibie. The examination of LBZ areas has detected a quasi brittle fracture in the fractured area of the specimen in as-vvelded condition - Fig. 7. The fracture area of stress reiieved specimen shovvs besides the quasi brittle frac-tures also the presence of intergranular brittleness along the co/umnar dendrites as it is shovvn in Fig. 8. Slika 7 LKP - kvazi krhki transkristalni prelom. Fig. 7 LBZ - guasi brittle transcrystal fracture. skopom v krovni legi. je razvidna iz slike 3; vidne so podolgovate M/A faze vzdolž intragranularno izločenega ferita. Mikrostruktura po napetostnem žarjenju je razvidna iz slike 4; zaznavni so številni cementitni izločki na meji med M/A fazo in feritom. Mikrostruktura staljenega zvara, zavarjenega s predgrevanjem, je razvidna iz slike 5; M/A faz je manj in usmerjenost je manj izrazita. Ista mikrostruktura po napetostnem žarjenju je razvidna iz slike 6; vidna je tendenca močnega skepljanja cementi-ta. Pregled površin LKP je odkril kvazi krhki lom v prelomu preizkušanca v izhodnem stanju — slika 7. Površina preloma v napetostno žarjenem preizkušancu je poleg kvazi krhkega preloma pokazala še nastop intergranular-ne krhkosti vzdolž stebričastih dendritov, kot je razvidno iz slike 8. 4. DISKUSIJA REZULTATOV Preiskave so pokazale bistveno razliko v udarni Charpijevi žilavosti med osnovnim materialom in staljenim zvarom. Kljub dodatkom Ti-B in ugodni vrednosti 02 (ca. 227 ppm) v zvaru, ki pospešuje tvorbo acikularnega ferita in preprečuje tvorbo primarnega ferita, je bila dosežena udarna žilavost nižja od pričakovane. Lomna žilavost v prisotnosti ostre utrujenostne razpoke pa je pokazala popolno krhkost že pri temperaturah pod 0°C. Vzrok za to je varjenje brez predgrevanja in z nizko dovedeno toploto. Posledica tega je tvorba drobne mikrostrukture. ki vsebuje ferit s sekundarno fazo v obliki M/A strukturne faze namesto acikularnega ferita. Dodatki Cr, Mo in Nb težijo k pospeševanju tvorbe M/A faze, v kateri vsebnost C naraste tudi nad 1% /10/. Razpotegnjene M/A faze vzdolž intrgranularno izločenega ferita so lahko potencialni izvori zgodnjega začetka loma in zato nizke žilavosti. Dodatek Nb in njegovo izcejanje (0,04% v korenu in 0,01% v temenu) imata bistveni vpliv na udarno žilavost. Kemična sestava staljenega zvara je povišala VVata-nabe faktor J >400, tako da je posledica pojava popuš-čne reverzibilne krhkosti (SI. 8). Zaradi navedenega ni iz- Slika 8 LKP - intergranularni krhki prelom. Fig. 8 LBZ - intergranuiar brittte fracture. 4. RESULTS DISCUSSION The examinations have shovvn the essential differ-ences in Charpy toughness values of the base material and the molten vveld. In spite of Ti-B additions and fa-vourable oxygen content (ca. 227ppm) in the vveld, ac-celerating the acicular ferrite formation and prevention of primary ferrite formation, the notch-toughness values achieved vvere iovveras it vvas expected. But the fracture toughness in the presence of sharp fatigue crack has shovvn the absolute brittleness also at temperatures under 0 degree Celsius. The reason is vvelding vvithout preheating and vvith lovv heat input energy. The result is formation of the fine microstructure containing ferrite vvith secondary phase in the shape of M/A structural phase in state of acicular ferrite. The addition of Cr, Mo and Nb has a tendency of M/A phase formation in vvhich the carbon content can rich the values greater than 1 % /10/. Oblong M/A phases aiong the intragranutar precipitated ferrite can be the potential sources of early fractures be-ginning and the reason for lovv toughness values. Chemical composition of the molten vveld has increased the VVatanabe factor J= > 400, so that it resuits in the reversibie temper embrittlement appearance /8/. That is the reason why there is no toughness improvement at heating above 650 degrees Celsius vvhere the causes for temper embrittlement are dissolving, on the other side an intensive precipitating of Fe3C appears on the boundary betvveen oblong ferrite and M/A phase. The size of Fe3C precipitates are increasing vvith eievat-ed temperature under the coagulation mechanism vvhat is the reason for brittle fracture at elevated temperature /15/ (Fig. 4). From the above we can conclude, that stress-re-lieved heating does not have a favourable effect on molten vveld toughness if M/A structural phases are present into it. M/A phases formation is a consequence of too high alloyment or/and too high cooling speed, so that their presence depends on the vveld material selection and vvelding technology. Using the vvelding technology vvithout preheating the content of M/A structural phase boljšanja žilavosti po segrevanju nad 650° C, kjer se povzročitelji za popuščno krhkost raztapljajo, na drugi strani pa se hkrati pojavi intenzivno izločanje Fe3C na meji med razpotegnjenim feritom in M/A fazo. Velikost izločkov Fe3C se povečuje s povišanjem temp. po mehanizmu skepljanja, kar privede pri višji temp. do krhkega loma /15/ (SI. 4). Iz zgornjega je mogoče sklepati, da napetostno žar-jenje nima ugodnega učinka na izboljšanje žilavosti staljenega zvara, če so v njem prisotne M/A strukturne faze. Tvorba M/A faz je posledica previsoke nalegiranosti ali/in previsoke hitrosti ohlajevanja, tako da je njihova prisotnost odvisna od izbire dodajnega materiala in varilne tehnologije. Pri uporabi varilne tehnologije brez predgrevanja je znašala vsebnost prisotne M/A faze okrog 38% v staljenem zvaru. Z uporabo predgrevanja se je ta vsebnost znižala na okrog 33%. Iz raziskave je mogoče sklepati, da je pri varjenju brez predgrevanja in nizki dovedeni toploti za dosego visoke žilavosti v TVP potrebno uporabiti dodajne materiale, ki ne bodo tvorili večjih količin razpotegnjenih M/A faz. Navedeno lahko dosežemo z nižjimi vsebnostmi Si, Al, Cr in Mo in z dodatki Ti-B, ki pospešujejo tvorbo aciularnega ferita. Na drugi strani je podoben efekt mogoče doseči z uporabo predgrevanja in višjo dovedeno toploto ne glede na prisotnost omenjenih legirnih elementov. Vendar je takšna varilna tehnologija uporabna le, če izberemo jeklo, ki ni občutljivo na rast zrna, ker bomo v nasprotnem primeru dobili nizke vrednosti za žilavost v TVP. Omenjena moderna jekla so izdelana na osnovi TiN in BN izločevalnih efektov s vsebnostjo raztopljenega N pod 10 ppm. Pri varjenju preizkusnih plošč se vmesna temperatura visoko dvigne in lastnosti staljenega zvara niso primerljive z lastnostmi na dejanskem zvarnem spoju, ki je narejen brez predgrevanja ali z nizkimi At8'5 časi. Namesto da bi se varilo brez predgrevanja, dejansko varimo plošče z visokim predgrevanjem (često >200°C). Takšni rezultati niso uporabni za varjenje brez predgrevanja na dejanskih zvarnih spojih, kjer se zaradi dolgih zvarov in velikih površin vmesna temperatura zniža oz. pade celo na sobno. Takšen varilni postopek lahko znatno vpliva na žilavost izhodnih zvarnih spojev in tako na varnost celotne zavarjene konstrukcije. Torej je navedeno vzrok za pazljivo in temeljito izbiro verifikacije varilne tehnologije, ki mora odgovarjati dejanskim pogojem pri izvedbi varjenja v delavnici in na montaži. 5. ZAKLJUČEK Preiskava EPP zavarjenih raztaljenih zvarov, ki so za-varjeni brez predgrevanja in z nizko dovedeno toploto z uporabo več varkov na nizko ogljičnem finozrnatem jeklu, je odkrila naslednje: — Udarna in lomna žilavost raztaljenega zvara je odvisna od izbire dodajnega materiala in varilne tehnologije. — Pri varjenju in preizkušanju zavarjenih plošč morajo biti izpolnjeni pogoji, ki bodo reprezentirali dejanske pogoje v delavnici in na montaži. — Varilna tehnologija brez predgrevanja in z nizko dovedeno toploto, da bi dosegli dobre žilavostne lastnosti v TVP, lahko povzroči v raztaljenem zvaru v zrnih s primarnim feritom po kristalnih mejah in intragranularnim feritom s sekundarno fazo tvorbo krhkih M/A faz, ki lahko znatno znižajo žilavost. Odločitev, ali uporabiti pred-grevanje ali ne, je odvisna od ekvivalenta Pcm osnovnega in staljenega zvara in pričakovanih lastnosti. in moiten weid vvas 38 %. vvhiie vvith using the preheating this content vvas decreased to ca. 33 %. From this examination we can conciude that for vvelding vvithout preheating and lovv input energy to achieve high toughness in HAZ it is necessary to use vvelding materials vvhich vvill not form higher quantities of oblongated M/A phases. We can achieve that vvith iovver contents of Si, Al, Cr and Mo and vvith additions of Ti-B, vvhich accelerate the acicular ferrite formation. On the other side we can achieve a similar effect vvith the use of preheating and higher heat input energy regardless to above mentioned al/oying elements. But such vvelding technoiogy is useful only if we choose steel vvhich is not sensible to grain grovvth because in the opposite čase we will get iow toughness values in the HAZ. The above mentioned modem steels are produced on the base of TiN and BN precipitation effects vvith the nitrogen content under 10 ppm. During the vvelding of experimental plates, the interpass temperature raise high up and the properties of the moiten vveld are not comparable vvith the properties of the actual vveld vvelded vvithout preheating or vvith lovv At8/5 time. Instead of vvelding vvithout preheating we actually vveld them vvith high preheating (often > 200 degrees Celsius). Such results are not useful for vvelding vvithout preheating on actual vvelded joints because of their long-ness and greatness the interpass temperature de-creases or even tal/s dovvn to room temperature. Such vvelding process can have a considerable effect on as vvelded vveldments and in this way it effects on security of the vvhole vvelded construction. AH that has been stat-ed is the reason for a careful and profound choose of verification of vvelding technoiogy vvhich has to respond to actual vvelding conditions in the vvorkshop and as-sembling. 5. CONCLUSION The examination of SA vvelded moiten vvelds, vvelded vvithout preheating and vvith lovv heat input energy, vvith more run technique on lovv carbon fine grained steel has discovered the follovving: — Notch and facture toughness of the moiten vveld depends on vvelding material choose and vvelding tech-nology. — During vvelding and testing of vvelded plates such conditions has to be fulfilled that can represent the actual condition in the vvorkshop and assembling. —- Welding technology vvithout preheating and vvith lovv heat input energy can in the moiten vveld in the grains vvith primary ferrite vvith secondary phase, cause the formation of brittle M/A phases, vvhich can con side r-ably decrease the toughness. Decission to use or not to use the preheating depends on Pcm equivaient of base metal and moiten vveld and on properties expected. — By CTOD method determinated fracture toughness is othervvise conservative but gives a good insight into the quality estimation of the vvelded joint. — The LBZ appearance in the moiten vveld on the fractured area of CTOD specimens has to be already estimated at room temperature vvith iarge test (l/Vide Plate Tests) vvith fatigue crack inserted on the surface area. Test has to be carried out at the lovvest operating temperature of the future construction /16/. — Stress releaving vvith heating has negative conse-quences because of the M/A structural phases pres-ence. Cementite precipitation on the boundary betvveen ferrite and M/A phase additionally decrease toughness. In the čase of stress-reteaved heating it is recom- — Lomna žilavost, določana po metodi CTOD, je sicer konservativna, vendar daje dober vpogled v oceno kvalitete zvarnega spoja. — Pojav vidnih LKP v staljenem zvaru na lomni površini CTOD preizkušancev že pri sobni temp. je potrebno oceniti z velikimi preizkusi (Wide Plate Tests) s površinsko vgrajeno utrujenostno razpoko. Preizkus je potrebno opraviti pri najnižji temp. obratovanja bodoče konstrukcije /16/. — Termično sproščanje zaostalih napetosti ima negativne posledice zaradi prisotnosti M/A strukturnih faz. Izločanje cementita na meji med feritom in M/A fazo še dodatno zniža žilavost. V primeru uporabe napetostnega žarjenja je priporočljivo preveriti možnost pojava reverzi-bilne popuščne krhkosti, ki je odvisna od količine in izcej legirnih elementov ter nečistoč v raztaljenem zvaru. mended to control the possibility of reversible temper embrittlement appearance, vvhich depends on the quan-tity and segregations of alloying elements and impure-ments in the molten vveid. LITERATURA i REFERENCES 1. Suzuki, H.: Revised Cold Cracking Parameter PHA and Its Application, IIW IX-1311-84. 2. Durren, C.: Determining the Preheating Temperature for the Field-vvelding of Large-diameter Pipe, IIW IXG-318-84. 3. Guide to the Light Microscope Examination of Ferrite Steel Weld Metals, IIW IX-1533-88. 4. Cochran, R.C.: VVeld Metal Microstructure - a State of the Art Revievv, VVelding in the World, Vol. 21, No. 1/2, p. 16-25, 1983. 5a. Abson, D.J.: Non-metallic Inclusion in Ferritic Steel VVeld Metals. A Revievv IIW IX-1486-87. 5b. Liu, S., Olson, D.L.: The Role of Inclusions in Controlling HSLA Steel VVeld Microstructures, Supplement of VVelding Journal AWS, June 1986. 6. Easterling, K.: Introduction to the Physical Metallurgy of VVelding, Buttervvorths, 1983, p. 88-103. 7. Evans, G.M.: The Effect of Chromium on the Microstructure and Properties of C-Mn All-weld Metals Deposits, Oerlikon-Schweissmiteilungen 47/89, No. 120. 8. Rak, I., Gliha, V., Sidjanin, L., Petrovski, B.: J-integral Test-ing of SA VVeldments on HSLA Steel Regarding Different Treatments After VVelding, European Symposium on Elas- tic-Plastic Fracture Mechanics, Elements of Defect Asses-ment, Freiburg, 1989. 9. Okhita, S., Homma, H., Tsushima, Mori, N.: The Effect of Oxide Inclusions on the Microstructure of Ti-B Containing VVeld Metal, IIW IX-1070-86. 10. Ikavva, H., Oshige, H., Fukada, Y.: Effects of Martensite Austenite Constituents on HAZ Toughness of a High Strength Steel, IIW IX-1156-80. 11. Komizo, Y., Furusavva, J., Fukada, Y.: Development of High Toughness in Heat Affected Zone, JOM-3 Conference, Hel-singor, 1989. 12. Suzuki, H.: Root Cracking and Maximum Hardness in High Strength Steel VVelds, IIW IX-1983. 13. Dhooge, A., Ostyn, K., Magula, V., Vinckier, A.: Temper Embrittlement in Modem Off-shore Structural Steels. 14. Komizo, Y.: A Review on Reversible Temper Embrittlement in Cr-Mo Steel VVeld Metal, IIW IX-1488-87. 15. Knott, J.K.: Microscopic Aspects of Crack Extention, Adv-ances in Elasto-Plastic Fracture Mechanics, Applied Science Publishers London, 1979. 16. Denys, R.: Incentives for Fracture Toughness Testing, IIW X-1131-86.