UDK 621.791.05:539.55 Ocena lomne žilavosti popravljenih zvarnih spojev na vrtalnih ploščadih Fracture Toughness Evaluation of Repair Welded Joints for Offshore Application INOSLAV RAK - MUSTAFA KOCAK - BLAGOJ PETROVSKI Ocena lomne žilavosti popravljenih zvarnih spojev z odprtjem vrha razpoke CTOD (dva strjena zvara, mehanske neenakomernosti in velike zaostale napetosti) terjajo nekatere spremembe geometrijske oblike običajno uporabljanih preizkušancev kakor tudi postopka preizkušanja. Prispevek podaja prve rezultate na enostransko zarezanih upogibnih preizkušancih s plitvimi (a/w = 0,16) in globokimi (a/w = 0,3 do 0,5) zarezami. Preizkušanci so bili odvzeti iz 30 mm debelih večvarkovnih EPP (elektro pod žlindro) zavarjenih spojev, ki so bili zavarjeni s hiperbarskim (16 bar) varilnim postopkom, in sicer na temenu prejšnjega zvara EPP. Rezultati dokazujejo, da je mogoče doseči zadovoljivo žilavostno raven tudi za popravljene zvare, izvedene pod tlakom 16 bar (160 m pod gladino vode). Enostransko izvedeni zvarni spoj plošče z debelino 30 mm, kakovosti jekla StE 445.7 TM, ima majhno žilavost v predelu toplotno vplivnega področja (lokalno krhko področje). Ocena lomne žilavosti CTOD (glede na standard B S5762 in CTOD (S5) merilno metodo) je bila izvedena s številnimi površinsko in globoko zarezanimi (skozi debelino in s površine zvara) upogibnimi preizkušanci, na različnih mestih popravljenega zvara. Neposredno merjene CTOD (S5) vrednosti se ujemajo z izračunanimi vrednostmi CTOD po standardu B S5762. Vpliv strjenega zvara z višjo mejo plastičnosti glede na osnovni material je bil opažen v upogibnih preizkušancih s plitvo zarezo, ki je potekala skozi debelino. V teh preizkušancih se je pojavila majhna žilavost CTOD kljub vpetju na konici razpoke. The crack tip opening displacement (CTOD) toughness evaluation of repaired weld joints (presence of two weld deposits, complex mechanical heterogeneity and high residual stresses) requires some modifications to commonly used specimen geometry and testing procedures. This paper presents the results of single edge bend (SENB) specimens with shallow (a/w = = 0.16) and deep notches (a/w = 0.3 to 0.5) extracted from 30 mm thick multipass submerged arc welded (SAW) joints which had been repaired with a hyperbaric (16 bar) welding process at the toe region of the original SAW weld deposits. The results show that an acceptable toughness level can be achieved for hyperbaric repair weld metals deposited at 16 bar (160 m water depth). The repaired single bevel-joint on 30 mm thick StE 445.7 TM steel contained low toughness heat affected zone (local brittle zone). The CTOD toughness evaluation (according to BS 5762 standard and CTOD (S5) techniques) of both plates contained a series of shallow and deep cracked SENB specimens notched (through thickness and surface) at various locations of the repaired weld joints. The directly measured CTOD (S5) values are consistent with the calculated CTOD BS 5762 values. An effect of weld metal overmatching on through thickness shallow cracked SENB specimen result was observed, since these specimens can produce low CTOD values despite their lower crack tip constraint. 0 UVOD Površinske napake ali razpoke na temenu ali v korenu zvara je treba često popravljati z varjenjem v suhi ali mokri atmosferi v zahtevanih visokotlačnih razmerah. V načelu sta prostornina in globina popravljanega zvara manjši od prejšnjega zvara z napako. Ocena lomne žilavosti z odprtjem vrha razpoke CTOD tako nastalega zvara (dva različna dela strjenega zvara in večje zaostale napetosti) nedvomno terja nekatere spremembe oblik običajno uporabljanih preizkušancev kakor tudi spremembo postopka preizkušanja. 0 INTRODUCTION The shallow defects or cracks at the toe or root regions of the weld joints of offshore structures sometimes have to be repaired under harsh service conditions with dry or wet hyperbaric welding processes. Generaly, the volume and depth of the deposited repair welds are smaller compared to the defected original weld. The CTOD toughness evaluation of these complex weld joints (presence of two weld deposits and higher residual stresses) certainly requires some modifications to commonly used specimen geometry and testing procedures.
Standardi CTOD BS5762:1979 til, ASTME 1290-91 121 in EGF Pl 91 131 priporočajo uporabo upogibnega preizkušanca z globoko razpoko (a/ W = =0,5). Lomna žilavost, določena na takem preizkušancu, je konservativna glede izbire materiala, verifikacije varilnega postopka in postopkov, ocenjevanja vpliva napak. V popravljenih zvarih se navadno pojavljajo napake v obliki plitvih, površinskih razpok na temenu ali v korenu. Vpliv takšnih napak lahko glede na omenjene standarde ocenimo napačno, posebej pri zelo majhnih vrednostih lomne žilavosti. Na drugi strani pa imamo često opraviti z dvema strjenima deloma zvara s povsem različnimi trdnostnimi lastnostmi v smeri debeline, kar še dodatno zaplete preizkuse lomne žilavosti in analizo rezultatov. Neposredna uporaba običajnih metod za določevanje lomne žilavosti ni mogoča, ker se v popravljenih zvarih lahko pojavijo dodatna lokalna krhka področja - LKP, ki vplivajo na rezultate. Tako je pri oceni kakovosti popravljenega zvara uporaba preizkušancev s plitvo razpoko nujno dopolnilo k preizkušancem z globoko razpoko. Dosedanje preiskave so pokazale, da so vrednosti elastoplasto lomne žilavosti pri začetku razpoke (<5, in J{) pri preizkušancih s plitvo razpoko v primerjavi s preizkušanci z globoko razpoko 141 do 1101 lahko večje- Znano je, da formula CTOD v standardih pri upogibnem preizkušancu temelji na modelu plastičnega členka. Vrednost plastičnega faktorja rotacije rp pomembno vpliva na izračunano vrednost CTOD. Znatno zmanjšanje globine razpoke (a/\v) lahko premakne vrtišče plastičnega členka ( W-a) rp v nenačetem delu pred konico razpoke. Zaradi tega je pri določevanju CTOD žilavosti ključnega pomena določitev vrednosti rp 141 do 1121 za preizkušance s plitvo razpoko. Upoštevati je treba tudi heterogenosti snovi (neskladnost). Odvisnost lomne žilavosti od velikosti in oblike preizkušanca je povzročila velik odziv pri elastoplastičnih metodah merjenja lomne žilavosti. Določevanje CTOD ali integrala J na popravljenih zvarih in njim pripadajočih toplotno vplivnih področij (TVP) se še dodatno zaplete zaradi pojavljajočih se mikro- in makroheterogenosti. Sedanje delo je nadaljevanje prejšnjih preiskav [131 in je namenjeno raziskovanju vpliva oblike zvarnega žleba, mikro in makroheterogenosti, ter lokacije in globine razpoke na lomno žilavost CTOD. 1 METODA RAZISKAVE Večvarkovni EPP (elektro pod žlindro) zvarni spoji so bili izvedeni na mikrolegiranem trdnem jeklu StE 52, debeline 30 mm v navadnih atmosferskih razmerah. Žleb za popravni zvar je bil However, the CTOD test standards BS5762: 1979 111, ASTME 1290-91 [21 and EGF Pl-91 [31 recommend the use of deep cracked (a/w= 0.5) SENB specimens. Fracture toughness data determined on such specimens are bound to lead to the use of conservative toughness data on material selection, welding qualification and defect assessment procedures. Yet, particularly in repaired weld joints, defects are often found to be in the form of shallow toe or root cracks. Obviously, the significance of such defects may thus be assessed in an unduly conservative manner, especially if very low toughness values are used. Also, having two weld deposits, which can often cause remarkable mismatched properties in the thickness direction in the repaired joints, futher complicates fracture mechanics testing and analysis of the results. A direct application of the fracture toughness testing practice of unrepaired original welds to repaired joints is not a straightforward task, since repair welding can produce additional local brittle zones (LBZ) to be tested. Hence the use of a shallow cracked specimen, in addition to standard specimen geometry, is essential to characterize the repaired weld joints. Various studies have already shown that the elastic-plastic fracture toughness values at crack initiation (5, and J{) can be higher on shallow cracked specimens than on deep cracked ones 141 to [101. It is known that the CTOD formula used in testing standards is based on the plastic hinge model of bend specimens. The value of the plastic rotation factor rp significantly affects the calculated CTOD values. A considerable decrease of the notch depth (a/\v) can cause a shift of the plastic hinge point location ( W-a) rp in the unnotched ligament ahead of the crack tip. Hence, the determination of rp values for shallow cracked specimens plays an important role in CTOD testing [41 to 1121, particularly of shallow notched weld specimens. Material heterogeneity (mismatching) should particularly be taken into account. In this sense, the specimen size or geometry dependent fracture toughness data have caused increasing interest in elastic-plastic toughness testing procedures. The CTOD or J integral testing of the repaired weld joints and their head affected zones (HAZ) further complicates the issue, due to their micro- and macro-heterogeneities. The present study is the extension of previous research [131 focusing of the effects of the original weld groove preparation, weld joint micro- and macro-heterogeneities and crack depth location on the CTOD testing procedure. 1 EXPERIMENTAL PROCEDURE Multipass submerged arc welds (SAW) were prepared on 30 mm thick StE 52 and low carbon HSLA pipeline steel in half-k grooves under atmospheric conditions. The repair weld grooves
Repair weld groove SA weld Sl. 1. Priprava žleba za popravno varjenje Fig. 1. Repair weld groove preparation pripravljen tako kakor prikazuje slika 1. Zvari so bili izdelani v varilni komori v vodoravni legi po suhem postopku MAG (Metal Active Gas) pod tlakom 16 bar (160 m pod vodno gladino). Vnesena energija pri varjenju je znašala 2,7 oziroma 1,8 kj/mm za zvar EPP in MAG. Makroprerez celotnega zvara prikazuje slika 2. Za prvotni EPP in popravni zvar MAG sta bila dodajna materiala izbrana tako, da je v obeh primerih trdnost zvara presegala trdnost osnovnega materiala (pregi. 1). were machined as shown in fig. 1. The repair welds were deposited in a flat position by using the GMA welding procedure under 16 bar pressure (160 m water depth) in the welding chamber. The heat inputs were 2.7 kj/mm and 1.8 kj/mm for SAW and GMA welds respectively. The macro-section of the repair weld joint shown in fig. 2. For the original weld (SAW) and hyperbaric repair weld deposits, consumables were selected to obtain overmatching weld metals with respect to their respective base metal yield strength as shown in table 1. SI. 2. Makro prerez zvara po popravljenem varjenju Fig. 2. Macro-section of the repaired weld joint Preglednica 1: Mehanske lastnosti osnovnega materiala in strjenega zvara Table 1: Base material and weld metals mechanical properties Material Yield Stress, Meja plastičnosti (MPa) Tensile Strength, Natezna trdnost (MPa) Elongation, Raztezek Mis-matching factor M (%) Base Material, 30 mm 388 550 29.7 - Osnovni material SAW Weld Metal,(1 bar) 701 23.4 1.50 EPP strieni zvar Repair Weld Metal, 630 704 24.7 1.62 Reparatumi strjeni zvar M 6 bars-drv ) M=Weld metal yield stress/base metal yield stress, Meja plastičnosti zvara/meja plastičnosti os. materiala Tritočkovni upogibni preizkušanci dimenzije 28 X 28 mm so bili izrezani iz osnovnih EPP in popravljenih zvarov. Zareze so bile nameščene na različnih mestih z različnimi dolžinami, kar prikazuje preglednica 2, tako da so razmerja a /W bila 0,16; 0,3; 0,4; in 0,5. Preizkušance smo pred preizkusom utrujali tako, da je bil vrh razpoke nameščen v zaželenem delu zvarnega spoja. Posamezna področja so bila Three point bend SENB specimens (28 x 28 mm) were extracted from the original SAW and repair welded 30 mm thick plates with various notch locations and notch length as shown in table 2. The specimens were prepared with a /W ratio of 0.16; 0.3; 0.4; 0.5, and then fatigue precracked. These extensive notch locations and cofigurations were selected in order to screen potentially
Preglednica 2: Lege razpok na preizkušancih iz 30 mm debelih popravljenih zvarnih spojev in uporabljena»ekvivalentna Ov«za uporabo v formuli CTOD(BS) Table 2: Various notch position for repair welded joint - 30 mm thick plates and»ekvivalent oy «used in CTOD(BS) formula izbrana tako, da bi razkrila potencialna lokalna krhka področja celotnega zvara po popravilu zvara in da bi bila mogoča primerjava med lomno žila vostjo, izmerjeno na preizkušancih z globoko in s plitvo razpoko. Vsi preizkusi so bili izvedeni pri -10 C. Vrednost CTOD je bila določena po standardu BS 5762 in z neposrednim merjenjem po metodi GKSS (<55), ki omogoča meritev razmikanja dveh točk ob konici utrujenostne razpoke na razdalji 5 mm. Zapišemo ga na obeh stranskih ploskvah preizkušanca 1141, 1151. Ker je odnosnost med obema navedenima meritvama zelo dobra 1161, je mogoče izmerjeno vrednost č>5 vstaviti za preizkušance z globoko in plitvo razpoko v formulo BS in tako določiti dejanski plastični faktor rotacije rp na podlagi preizkusa: embrittled zones of these complex weld joints and to compare the shallow and deep notched specimen toughness results. All CTOD tests were conducted at -10 C. The CTOD values were calculated in accordance with BS 5762 (S5) and also directly measured with the GKSS developed S5 clip gauges (c>5) on the specimen s side surface at the fatigue crack tip over a gauge length of 5 mm 114, 151. Since there is a very good agreement between the two CTOD measurements 1161, it is thought that the c$5 values can be substituted into the BS formula to determine the plastic rotation factor rp, experimentally for deep and shallow notched bend specimens: CTODBS - K 2( 1-v2) r p (W - a)v p 2 öye rp( W - a) + a +Z = ( 1). Iz razmerja (1) določimo plastični rotacijski faktor r p: From eq. (1), the plastic rotation factor rp can be obtained as: a + Z p ~ W-a 2oyES,-K2(\-v2) 2 öye(vp-sb) +K3 (]-v2) (2). Dobljena vrednost rp pri različno globokih razpokah {a/ W =0,1; 0,3 do 0,5) se giblje v območju od 0,2 do 0,45 1161. To smo ugotovili v enakih razmerah za vrednosti CTODBS in CTOD^ ). Faktor rotacije za plitve in globoke razpoke, določen v drugih laboratorijih 171, 1101, znaša za preizkušance s plitvimi razpokami (a/ W ~ 0,1) prav tako okoli The rotation factors obtained in this manner for various shallow (a/w =0.1) and deep notched {a/ IV' =0.3 to 0.5) specimens produced values of about 0.2 and 0.45 respectively 1161. It is clear that these rp values were found under equal conditions of CT(j Dbs and CTOD(<55) values. The plastic rotation factor determined for shallow and deep cracked specimens in other studies 171, 1101 also
0,2. Na podlagi tega je bila za izračun CTODBS pri upogibnih preizkušancih s plitvo razpoko (a/w = = 0,16) z rp pri 28 mm debelih preizkušancih izbrana vrednosti 0,25. Za opazovanje stabilne rasti razpoke je bila med preizkušanjem uporabljena tehnika merjenja padca potenciala enosmernega toka. Merili smo tudi pomikanje točke delovanja obremenitve (LLD-Load Line Displacement) z referenčnim merilnikom 1171, 1181. Utrujanje preizkušancev je bilo izvedeno po postopku SHR 1191, 1201 (step-wise high R-ratio), pri kateri se uporablja razmerje napetosti R = 0,1 za začetek ter večanje razpoke do okoli 1 mm, potem pa R =0,7 z največjo dopustno obremenitvijo, dokler ni dosežena zaželena vrednost a/w. Slika 3 prikazuje površino loma preizkušanca s popravljenim profilom konice razpoke. Po preizkusih je sledil razrez preizkušancev in metalografska preiskava za določitev mikrostrukture v okolici konice razpoke. showed a value of about 0.2 for shallow cracked (a/w ~ 0.1) specimens. By following up this approach and various other results developed at the GKSS, the rp value of 0.25 was used for 28 mm thick specimens to calculate the <5BS for shallow cracked specimens (a/w = 0.16). During the test, the DC potential drop technique was applied for monitoring stable crack growth. The load-line-displacement (LLD) was also measured with reference bar 1171, 1181. Fatigue precracking was carried out with a»step-wise high R-ratio«(SHR) procedure 1191, 1201 for all specimens. The SHR technique uses two stress ratia, R = 0.1 for crack initiation and growth of about 1 mm, then R = 0.7 with allowable maximum load, until the required a/w ratio is obtained. Fig. 3 shows the fracture surface of a specimen with an improved prefatigue crack front profile. After CTOD tests, post-test sectioning and microstructural examinations were conducted for all HAZ specimens to identify the fatigue crack tip microstructure. SI. 3. Prelomna površina preizkušanca (mesto 4a, a /W =0,5) z utrujenostno razpoko, narejeno z dvema stopnjama utrujanja, prikazuje izboljšanje profila razpoke, izdelane po tem postopku Fig. 3. Fracture surface of specimen (position 4a, a /W =0.5) fatigue precracked by two»step-wise high R-ratio«showing an improvement of the fatigue crack front profile by this technique 2 REZULTATI IN OBRAVNAVA 2.1 Ocena izm erjenih m lkrotrdot Slika 4 prikazuje krivulje merjenja trdot na 30 mm debelem popravljenem zvaru. Namen je bil ugotoviti razporeditev mikrostrukturo kritičnih področij, ki so bila pred tem določena z metalografsko preiskavo. Interkritično segreti grobozrnati del TVP (ICCGHAZ) zvara EPP se je pokazal kot najbolj krhek in je na sliki 4 tudi označen. To področje je kot najbolj kritično navedeno tudi v drugih raziskavah 1211, 1221. Slika 5 prikazuje porazdelitev trdot po krivuljah a in b. Opazno je povečanje trdote v korenskih varih in TVP popravljenega zvara v primerjavi s trdoto, izmerjeno v temenskih varih. Karakteristične največje trdote, izmerjene v smeri debeline (krivulje d, e in f) in iz tega preračunane ustrezne vrednosti za mejo plastičnosti (231 so prikazane v preglednici 3. 2 RESULTS AND DISCUSSION 2.1. M icro-hardness survey Figure 4 shows the micro-hardness paths for 30 mm thick weld. This hardness survey was made to determine the hardness distribution across the microstructurally critical zones recognized by metallographic examination. The inter critically coarse grained heat affected zone (ICCGHAZ) region of the SAW weld is found to be the most brittle zone of this joint as shown in figure 4. This zone is also depicted as the most critical microstructure in other studies 1211, 1221. Figure 5 presents the hardness distribution obtained from paths a and b which indicates the hardness increase at the root pass of the repair weld (RW b) and root HAZ compared to its cup passes (RW a). The characteristic maximum hardess values obtained in the thickness direction (paths d,e and f) and corresponding calculated yield stresses 1231 for these peak hardness values are given in table 3.
Sl. 4. Krivulje Izmerjenih trdot in meja plastičnosti osnovnega materiala ter obeh strjenih delov zvara Fig. 4. Hardness measurement directions, and the yield strengths of the base and weld metal 0 5 10 15 20 25 30 35 40 Distance, mm Razdalja SI. 5. Porazdelitev trdote Fig. 5. Hardness distribution Preglednica 3: Meje plastičnosti delov zvara v smeri debeline, izračunane iz naj večji h trdot Table 3: Weld metal yield stresses in thickness direction calculated from max.hardness values Plate-30 mm Ploiča HV Micro-hardness measurement, direction -f, Smer merjenja mikro trdote -f Yield stress-calculated*, Izračunana meja plastičnosti (M Pa) Mismatching faktor M RW-cup pass, Krovni varek 204 474 1.28 RW-root pass, Korenski varek 238 582 1.58 RW-OW HAZ, TVP 250 620 1.69 OW 214 506 1.37 BM. OM 170 367.5 - CTyw=3.15HV-168 Opaziti je veliko razliko v faktorju heterogenosti M, če ga primerjamo s celotnimi vrednostmi v preglednici 1. Očitno je, da bodo tako kompleksne mehanske heterogenosti zajete vzdolž črte konice utrujenostne razpoke gotovo vplivale na značilnosti nastajanja in širjenja razpoke pri obremenjevanju. 2.2 Preizkusi CTOD na popravi jenem zvarnem spoju V preglednici 2 so glede na lokacijo razpoke podane vrednosti meje plastičnosti za izračun CTOD po standardu BS 5762. Tako imenovana ekvivalentna meja plastičnosti je bila dobljena iz povprečja lastnosti sodelujočega materiala v najožji okolici konice utrujenostne razpoke. Za primer lokacije razpoke na mestu 4a in 4b je to povprečje določeno z: Obviously, the mismatching factor M, shows difference compared to the global M values of table 1. Hence this complex mechanical heterogeneity at the vicinity of the crack tip will certainly influence the characteristics of crack initiation and growth. 2.2 CTOD testing of repair welded joint Various yield strength values used (in relation to the notch position) in the calculation of CTOD values according to the BS 5762 formula are given in table 2. The so-called»equivalent yield strength«value was obtained on the average material combination at the vicinity of the crack tip, for example for notch positions 4a and 4b: {a. RW y + a. ow + o B M )/3,
kjer pomenijo: r w popravljeni zvar, o w prvotni zvar, BM osnovni material. Namen tega je bil upoštevati vsaj približen vpliv trdnostne heterogenosti v bližini konice utrujenostne razpoke pri določitvi lomne žilavosti CTOD. Pri tem sta upoštevana dva različna dela strjenega zvara in osnovni material, ustrezno legi zareze. Vpliv izbire različnih napetosti meje plastičnosti na vrednosti CTOD (BS) za obe vrsti preizkušancev je razviden iz preglednice 4. where: RW repaired weld, o w original weld, BM base material. The purpose of this exercise was to include the approximate effect of strength mismatching of the crack tip vicinity into the CTOD toughness determination. Two weld metal and a base metal mechanical properties in varying combinations are used with respect to the notch positions. The effect of yield stress selection on CTOD(BS) values can be seen in table 4, in which two specimen results are presented. Preglednica 4: Primerjava vrednosti CTOD (BS)-CTOD (S5) za preizkušance z a /W = 0,16 in 0,5 Table 4: Comparison of values CTOD (BS)-CTOOD (S5) for specimens with a /W = 0.16 and 0.5 CTOD(BS)-Pos 4b a/w=0.16 CTOD(BS)-Pos 4a a/w =0.5 Specimen 10.12 Specimen 11.2 Specimen 10.2 Specimen 10.7 Preizkušanec Preizkušanec Preizkušanec Preizkušanec 1. ay =388, BM 0.096 0.018 0.139 0.054 2. oy=582, OW 0.081 0.013 0.122 0.041 3. cry =630, RW 0.079 0.012 0.120 0.039 4. ay =606 av.* 0.080 0.012 0.121 0.040 5. ay=533 av.** 0.082 0.013 0.123 0.041 CTOD(55) 0.092 0.025 0.159 0.061 * 1/2(RW+OW) ** 1/3(RW+OW+BM) Za izračun vrednosti CTOD(BS) je bilo uporabljenih pet različnih ravni meje plastičnosti. Za obe lokaciji konice razpok lahko pričakujemo znaten vpliv mehkega (zelo žilavega) osnovnga materiala na oblikovanje plastične cone ob konici razpoke. Očitno je približevanje vrednosti CTOD(BS) vrednosti CTOD(55), ko uporabimo pri izračunu mejo plastičnosti osnovnega materiala. Navedeni rezultati kažejo prednost neposredne meritve CT0D (i55), saj te vrednosti ni treba izračunavati na podlagi neke druge mehanske lastnosti (meje plastičnosti). Druga prednost te metode je, da jo lahko uporabimo za preizkušance s plitvo ali globoko razpoko, brez prilagajanja faktorja vrtenja. Primerjava vrednosti CTOD, določenih po obeh metodah, je podana na sliki 9. Slika 6 prikazuje izmerjene vrednosti CTOD (BS) glede na dejansko lego razpok. Glede na različno vpetost materiala v neposredni okolici konice razpoke je opazna velika razlika v žilavosti med preizkušanci z globoko in plitvo razpoko, ko ta poteka po celotni debelini (la in lb v preglednici 2). Česa podobnega ni mogoče jasno videti na preizkušancih, ki imajo površinske razpoke v strjenem zvaru (2a in 2b v preglednici 2), razmerje a/wpa 0,4 in 0,16. Preizkušanci s plitvo razpoko (2b) izkazujejo le malce večjo žilavost od preizkušancev z globoko razpoko (2a), pri katerih se plastična cona lahko razširi v mehkejši osnovni material. Pri preizkušancu s plitvo razpoko se plastična cona razvije in ostane v področju zvara z večjo trdnostjo. Five yield stress levels were used to calculate the CTOD(BS) values. For these notch positions, a significant contribution of the soft (high toughness) base metal side on the plastic zone development at the crack tip can be expected. Apparently, if the yield stress of the base metal is used, the CTOD(BS) results approach closer to CTOD (<$5) values. These results indicate the potential advantage of directly measured CTOD (<$5) values, since they do not need to be calculated by using specific yield stress values. The direct application of the CTOD (<J5) technique on shallow and deep notched specimens without any plastic rotation factor correction offers another advantage of this technique. The comparison of two CTOD values obtained according to the BS 5762 standard and CTOD (<S5) procedure for deep and shallow notched specimens extracted from 30 mm thick plate is shown in figure 9. The CTOD(BS) results are plotted in Fig. 6 with respective notch positions. Large differences in toughness results between shallow and deep through thickness notched weld metal specimens (see notch positions la nad lb in table 2) can be observed due to the differences in constraint. However, a similar observation could not be made so clearly on surface cracked repair weld metal specimens (positions 2a and 2b) with a/w ratios of 0.4 and 0.16 respectively. Shallow cracked specimens (posit.2b) showed only slightly higher toughness values than the test pieces of position 2a in which crack tip plastic zone developments can easily extend into the softer base metal, compared to the shallow cracked specimen in which crack tip plastic zone development predominantly remains within the high strength hyperbaric weld metal.
a/w 0.16 0.5 0.4 0.16 0.5 0.5 0.16 0.4 a/w Notch position (see Table 2), Mesto zareze (glej tabelo 2) Sl. 6. Vrednosti CTOD(BS) za vse lege razpok na debelini 30 mm (T = -10 C) kažejo vplive lege in globine razpoke Fig. 6. The CTOD(BS) values for all notch positions on 30 mm thick plate (T = -10 C) showing the effects of notch position and a /W ratio Najmanjše vrednosti CTOD so bile dobljene na preizkušancih z razpoko na mestu 4a. Konica utrujenostne razpoke je potekala skozi popravljeni strjeni del zvara in skozi TVP prvotnega zvara EPP (pregi. 2). Prav tako majhne vrednosti so dobljene tudi s prelzkušanci z razpoko na mestu 4b. Očitno je na teh mestih utrujenostna razpoka potekala prek najbolj kritičnih mikrostruktur kompleksnega popravljenega zvarnega spoja. Proti pričakovanju prikazujejo rezultati preizkusov z razpoko na mestu 3, kjer je TVP popravljenega zvara v osnovnem zvaru EPP, večjo žilavost kakor na mestu 4a( pregi.2 in sl. 6). Zaradi tega so bile izvedene obširne mikrostrukturne preiskave za določitev mesta začetka krhkega loma na mestih 4a in 4b. Značilne prelomne površine pri preizkušancih z globokimi in plitvimi razpokami so razvidne s slik 7a in 7b. V obeh primerih se je krhki lom pojavil v grobozrnatem področju TVP osnovnega zvara EPP, ki je bilo pri popravilu ponovno segreto na temperaturo med AC3 in AC1 (shematsko je to področje prikazano na sliki 4). Pojav duktilne razpoke in njeno razširjanje je bil opazen v popravljenem zvaru. Metalografska preiskava preizkušanca z razpoko na mestu, ki je bilo v grobozrnatem delu TVP ponovno segreto na temperaturo med AC3 in AC1, je odkrila krhko interkristalno mikrostrukturo, ki je prikazana na sliki 8a. Slika 8b daje slutiti navzočnost blokovne mikrostrukture na kristalni meji z očitno velikim deležem cementitnih izločkov, ki so vzrok medfazne dekohezije in pojav cepilne razpoke v skladu z mehanizmom kopičenja dislokacij predlaganega z lokalnim kriterijem lomarkr 1241, 1251. Clearly, the lowest toughness data (lower than the permissible CTOD value of 0.12 mm) was obtained for specimens having notches at position 4a. In this notch position, the fatigue crack was sectioning the repair weld metal and HAZ of the SAW deposit, table 2. The shallow cracked specimens with this notch position (4b) also produced low pop-in toughess values comparable to the deep notched results. This notch position is clearly sampling the most critical regions of the repaired weld joint. Contrary to expectations, test results obtained from notch position 3 indicate that the repair HAZ region developed within the original SAW weld metal has higher toughness properties that the regions sampled by position 4a, see table 2 in figure 6. Extensive microstructral examination was therefore conducted to determine the location of cleavage crack initiation for these two notch positions. Typical fracture surfaces of the deep and shallow notched specimens for positions 4a and 4b are shown in figures 7a and 7b. For both cases, brittle fracture started at the ICCGHAZ region of the original SAW weld (schematically shown in figure 4). Ductile crack initiation and growth was observed at the repair weld metal side of the specimens. The metallographic examination of the ICCGHAZ on sectioned specimens revealed the embrittled grain boundary microstructure shown in figure 8a. Figure 8b suggests that the presence of a blocky grain boundary structure containing a large amount of cementite particles can cause interfacial decohesions and cleavage crack initiation in accordance with dislocation piling-up mechanisms suggested by RKR local fracture criterion (241, 1251.
Slika 7 a) Prelomna površina preizkušanca (mesto 4a, a /W = 0,16) s krhkim in duktil ni m potekom loma prek TVP osnovnega EPP in popravljenega zvara a) Fracture surface of the specimen (pos. 4a, a /W = 0.5) showing the brittle and ductile fracture path corresponding to the original weld HAZ and repair weld respectively b) Prelomna površina preizkušanca (mesto 4b, a /W = 0,16) z duktilno rastjo razpoke v popravljenem zvaru (desna stran) in krhkim lomom v grobozrnatem delu TVP osnovnega zvara EPP b) Fracture surface of the specimen (pos. 4b. a /W = 0.16) showing at the right side a ductile crack growth occurred at the repair weld metal and brittle fracture at the CCHAZ of the SAW joint a) b) Slika 8 a) Mikrostruktura TVP v točki pojava krhkega loma, optični mikroskop (povečava 500-krat) a) ICCCHAZ microstructure at the brittle fracture initation point, optical micrograph (mag.soox) b) Mikrostruktura na kristalni meji, vrstični elektronski mikroskop (povečava 4000-krat) b) High magnification showing grain boundary microstructure scan electron microscope (mag, 4000 x)
Sl. 9. Vrednosti SBS in S5 preizkušancev s plitvo in globoko razpoko na 30 mm debelih preizkušancih dokazujejo dobro ujemanje Fig. 9. Relationship between SBS and S5 for shallow and deep notched specimens for 30 mm thick plates showing close agreement Zanimiva je ugotovitev, da ima grobozrnato področje TVP zvara EPP, ki je bilo ponovno segreto med AC3 in AC1, majhno žilavost le v preizkušancih s popravljenim zvarom. Vrednosti CTOD v TVP z nepopravljenim zvarom EPP so vključene v sliko 6 na mestu 4A. Te so večje od onih na mestu 4 a, ki ustrezajo popravljenemu z varnemu spoju. To daje slutiti, da je popravljeni zvar povzročil dodatno krhkost grobozrnatega dela TVP v prvotnem zvaru EPP. Vpliv velike razlike v trdnosti vzdolž konice razpoke v smeri debeline je tudi lahko vzrok za majhno žilavost. Zelo trdno popravljeni zvar (Re = 630 MPa) ščiti del konice razpoke, nameščene v popravljenem delu zvara pred večjim plastičnim deformiranjem in tako prisili drugi del konice, ki poteka v bližini sosednjega manj trdnega osnovnega materiala (Re = 388 MPa), da se obsežneje plastično deformira. Nesimetrična trdnostna porazdelitev vzdolž konice razpoke na mestu 4a in 4b lahko v preizkušancih zveča kritično napetost za pojav cepilne razpoke v grobozrnatem delu TVP zvara EPP, ki je bil segret na temperaturo med AC3 in AC1. Negativni učinek višje meje plastičnosti popravljenega zvara se kaže tudi pri preizkušancih s plitvo razpoko, čeprav je vpetost na konici razpoke zaradi manjše globine razpoke manjša. Rezultati preiskav na mestu kjer se križata dve TVP preizkušancev s površinsko razpoko (mesto 5, pregi. 2), prikazani na sliki 6, kažejo veliko žilavost v primerjavi s preizkušanci, kjer razpoka poteka prek celotne debeline na mestu 4a in 4b. Slika 6 prikazuje tudi določeno podobnost med preizkušanci s površinskimi razpokami (2b in 5). Verjetno je večja žilavost na mestu 5 rezultat pozitivnega vpliva bližnjega osnovnega It is interesting to note that the ICCGHAZ of the SAW weld metal showed this low toughness behaviour only in specimens with repair weld deposits. The CTOD results of the unrepaired SAW HAZ specimens are included in figure 6 at notch position 4a. These specimens showed higher values compared to the position 4a results obtained from repaired joints, figure 6, notch position 4a. This may imply that the repair weld caused further embrittlement of the CGHAZ of the original SAW weld and hence low toughness values were only obtained after repair weld deposition. The effect of large strength differences (mismatching) along the crack front in the thickness direction may also play a role in low toughness results. The highly overmatched repair weld metal (Re = 630 MPa) protects the crack tip portion which samples the repair weld metal from applied deformation and hence forces the other part of the crack front neigbouring the low strength base metal (Re = 388 MPa) to accommodate the applied deformation. This asymmetric strength distribution along the crack front of the position 4a and 4b specimens can enhance the attainment of the critical stress for cleavage crack initiation at the ICCGHAZ region of the original SAW weld metal. Due to this negative effect of the overmatching repair weld metal, the shallow cracked specimens (notch position 4b) also showed low CTOD results, despite their lower crack tip constraint (if one considers the crack depth only). The intersection of two HAZs of two weld deposits (notch position 5, table 2) was tested by using surface notched specimens. The results are shown in figure 6, and indicate a high toughness level compared to the through thickness notched 4a and 4b specimens. Fig. 6 further shows a certain similarity of the surface notched specimen results, positions 2b and 5. It is believed that the higher toughness values of position 5 were again
materiala, ki zlahka dopusti sprostitev napetosti ob konici razpoke z razvitjem večje plastične cone na strani materiala z manjšo mejo plastičnosti in največjo žilavostjo (sl. 6, na mestu primerjave s 4a). Iz tega je moč sklepati, da mehanske neenakosti vzdolž konice razpoke (v tem primeru strjeni popravljeni zvar z višjo mejo plastičnosti od osnovnega materiala) lahko pomembno vplivajo na velikost izmerjenih vrednosti CTOD z možnostjo nesimetrične porazdelitve deformacij vzdolž konice razpoke. Očitno je, da bodo izmerjene vrednosti lomne žilavosti vedno kazale vpliv trdnostne neenakosti v bližini konice razpoke. Zaradi tega naj bi takšne podatke označevali raje kot»dejansko žilavost«, ne pa kot»resnično žilavost«mikrostrukture ob konici razpoke, ki je karakteristika materiala. Še vedno ostaja odprto, kako definirati vpliv trdnostne neenakosti na napetostno stanje ob konici razpoke oziroma na vpetost in s tem na»dejansko žilavost«, ker se trdnostne neenakosti močno kažejo pri nastanku in usmeritvi razpoke. 3 SKLEPI Raziskali smo lastnosti večvarkovnega zvarnega spoja EPP, ki je bil narejen z varilnim postopkom MAG pri tlaku 16 bar (160 m pod gladino vode). Izvedli smo metalografsko preiskavo in lomnomehanska preizkušanja s plitvimi in globokimi razpokami s preizkušanci CTOD. Vrednosti CTOD so bile določene v skladu s standardom BS 5762 in s postopkom CTOD (č>5). Analiza eksperimentalnih rezultatov ponuja naslednje sklepe: Ze pred odvzemom preizkušancev in izdelavo zarez je priporočljivo izvesti mikrostrukturno raziskavo popravljenega zvara za določitev glavnih kritičnih področij. Izvedba popravnih varkov z različnimi tehnikami lahko povzroči dodatna lokalna krhka področja (LKP). Meritev trdot popravljenega zvara pomaga identifikaciji LKP. Bistvena je uporaba preizkušancev z različnimi legami razpok za popolno določitev žilavosti spoja. Popravni strjeni zvar, izdelan pri visokem tlaku je pokazal dobro žilavost CTOD. To dokazuje, data tako imenovani hiperbarski varilni postopek, uporabljen v tej raziskavi, omogoča izdelavo zvarov v globini 160 m pod vodno gladino, z žilavostjo strjenega zvara, ki je enaka stopnji žilavosti poprejšnjega okvarjenega dela strjenega zvara. Najbolj kritično področje (TVP osnovnega zvara EPP, izvedenega na debelini 30 mm) je bilo odkrito na preizkušancih s plitvimi in globokimi razpokami v smeri debeline spoja na mestih (4a in 4b), ki so pod vplivom višje trdnosti popravnega zvara. Pred popravilom isto področje v TVP ni imelo majhne žilavosti. Preizkušanci s plitvimi razpokami so dali podobne vrednosti CTOD kakor tisti z globokimi. Rezultati meritev CTOD (<S5) so skladne z izračunanimi vrednostmi CTOD po standardu BS 5762 v primeru preizkušancev z globokimi in plitvimi razpokami, vendar le, če je za plitve razpoke (a/w= 0,16) upoštevana vrednost za faktor vrtenja rp = 0,25. Uporaba merilne metode CTOD obtained under the positive influence of the neighbourning soft base metal (which readily provides some crack tip relaxation due to the plastic zone development at he undermatched base metal side). Evidently, these high CTOD results are influenced by a high base metal toughness level included in Fig. 6 at position 4a. It can be concluded that the mechanical heterogeneity along the crack front (In this case the presence of overmatching hyperbaric weld metal) can significantly influence the CTOD values due to the unsymmetric distribution of the applied deformation along the crack front. It is apparent that the present toughness values obtained from such complex weld joints will always reflect the effect of the strength mismatching of the crack tip vicinity. So such data should be called»apparent toughness«rather than the material parameter»intrinsic toughness«of the microstructure at the crack tip. There is still a need to define to effect of mismatching on the crack tip stress state (constraint) and on toughness, since mismatching significantly changes the fracture initiation behaviour and crack path direction. 3 CONCLUSIONS A characterization of a multipass SAW weld joint repaired with the hyperbaric MAG welding process at 16 bar pressure (160 m water depth) was carried out by microstructural examination and testing of shallow and deep notched CTOD specimens. The CTOD was determined according to the BS 5762 standard and the CTOD (<S5) procedure. The analysis of the experimental results led to the following conclusions: Prior to the extraction of CTOD test pieces and notching, detailed microstructural examination of the repaired joint is recommended to establish the most critical zones, since repair weld deposition with a different welding process can procedure additional local brittle zones. A hardness survey of the repaired joint may help to identify such zones. Various notch configurations are essential for full toughness characterization of the joint. The hyperbaric repair weld metals showed good CTOD toughness levels. This indicates that the hyperbaric welding process used in this study for a water depth of 160 m can provide weld metals with toughness levels equal to the original defective weld deposit. The most critical zone (HAZ of the original SAW weld joint of 30 mm thick plates) was depicted by through thickness shallow and deep notched specimens (position 4a and 4b) under the influence of the overmatched repair weld deposit. Prior to the repair the same HAZ region did not produce low toughness results. Shallow cracked specimens in this case produced similar CTOD results compared to deep notched ones. The CTOD (<S5) measurements are consistent with the calculated CTOD values according to the BS 5762 standard for both deep and shallow crakked specimens (a/w= 0.16) if the rp value of 0.25 is used in the latter. The application of the CTOD
(<S5) na preizkušancih s plitvo in globoko razpoko, odvzetih iz popravljenih zvarov, obeta preprosto in hitro tehniko ocenjevanja žilavosti. Preizkušanje lomne žilavosti po metodi CTOD na popravljenih zvarih s plitvo razpoko (a/w= 0,16), kaže, da so te vrednosti v splošnem odvisne od geometrijske oblike preizkušanca. Očitno so pomembnejši dejavniki pri ocenjevanju lomne žilavosti popravljenih zvarov naslednji: oblika osnovnega oziroma poprejšnjega zvara, geometrijska oblika popravljenega zvara in lega konice razpoke glede na kritične mikrostrukture. Grobozrnati del TVP je najbolj kritičen tudi zaradi heterogenosti porazdelitve vpetja vzdolž konice razpoke. ( <55) technique on the shallow and deep notched specimens extracted from repaired weld joints offers a simple and quick toughness estimation technique. CTOD fracture toughness testing of repair welds with shallow notched specimens ia/w =0.16) gives generally geometry dependent toughness values. It seems that the most importat factors for complex repaired welded joint fracture toughness evaluation are: original welded joint design, geometry of the repair weld, crack tip sampling of the ICCGHAZ/UCGHAZ micro structures and heterogeneous crack tip constraint distribution. 4 LITERATURA 4 REFERENCES ili BS 5762:1979, Methods for Crack Opening Displacement (COD) Testing, The British Standards Institution (1997). 12) ASTME 1290-91, Standard Test Method for Crack-Tip Opening Displacement (CTOD) Fracture Toughness Measurement, 1991. (31 EGF Recommendations for Determining the Fracture Resistance of Ductile Materials, EGF Pl-90, European Group on Fracture, December 1989. 141 Li, Quing-Fen: A Study about Ji and di in Three- Point Bend Specimens with Deep and Shallow Notches, Eng. Fracture Mechanics, Voi. 22, No. 1, pp. 9-15, 1985. 151 Kocak, M. et. al.: Evaluation of HAZ Toughness by CTOD and Tensile Panels, European Symp. on EPFM: Elements of Defect Assessment, Oct. 1989, Freiburg, F.R.G. 161 Matsoukas, G.-Cotterel, B.-Mai, Y.-W.: The Effect of Shallow Cracks on Crack Tip Opening Displacements. Eng. Fracture, Mechanics, Voi. 24, No. 6, pp. 837-842. 1986. 171 Wu, S.-X.: Evaluation of CTOD and J-integral for Three-Point Bend Specimens With Shallow Cracks. Proc. of the 7th Int. Conf. on Fracture, 1CF 7, Houston. Texas, March 1989, Vol. 1, pp. 517-524. 181 Sorem, W.A., et al: An Analytical Comparison of Short Crack fand Deep Crack CTOD Fracture Specimens of an A36 Steel. To be Published. 19] Li, Q.F.: et al. The Effect of a/w Ratio on Crack Initiation Values of COD and J-integral. Eng. Fracture Mechanics, Voi. 23, No. 5, 1986, pp. 925-928. 1101 Zhang, D.Z.-Wang, H.: On the Effect of the Ratio a/w on the Values of di and Ji in a Structural Steel. Eng. Fracture Mechanics, Voi. 26, No. 2, 1987, pp. 247 250. 1111 Kolednik, O.: On the Calculation of COD from the Clip-gage Displacement in CT and Bend Specimens. Eng. Fracture Mechanics. Voi. 29, No. 2, 1988. pp. 173-188. 1121 Wu, S.X.: Plastic Rotational Factor and J-COD Relationship of Three Point Bend Specimen. Eng. Fracture Mechanics, Voi. 18, No 1, 1983, pp. 83-95. [131 Rak, l.-kocak, M.-Golesorhk, M.-Heerens, J.: CTOD Toughness Evaluation of Hyperbaric Repair Welds with Shallow and Deep Notched Specimens. IIW Document X-1242-92. 1141 Hellmann, D.-Schwalbe, K.-H.: Geometry and Size Effects on J-R and 5-R Curves under Plane Stress Conditions. ASTM STP 833, 1984, pp. 577-605. 1151 GKSS-Forschungszentrum Geesthacht: GKSS -Displacement Gauge Systems for Applications in Fracture Mechanics, 1991. 1161 Kocak, M.-Yao, S.-Schwalbe, K.H.-Walter, F.: Effects of Crack Depth (a/w) on Weld Metal Fracture Toughness. International Conference Welding 90, Geesthacht, Hamburg, pp. 255-268. 1171 Schwalbe, K.-H.-Hellman, D.: Application of the Electrical Potential Method to Crack Length Measurements Using Johnson s formula. JTEVA. Voi. 9, Nr. 3, 1981, pp. 218-221. 1181 Schwalbe, K.-H.-Hellmann, D.-Hereens, J.-Knack, J.-Miiller-Rose, J.: Measurement of Stable Crack Growth Including Detection of Initiation of Growth Using the DC Potential Drop and the Partial Unloading Methods. ASTM STP 856,E.T. Wessel, and F. Loss, Eds., American Society for Testing and Materials, 1985, pp. 3 22. 1191 Kocak, M.-Seifert, K.-Yao, S.-Lampe, H.: Comparison of Fatigue Precracking Methods for Fracture Toughness Testing of Weldments: Local Compression and Step-Wise High R-Ratio. International Conference Welding 90, Geesthacht-Hamburg, pp. 307-318. 1201 Kocak, M,-Yao. S.-Schwalbe, K.-H.: Effects of Welding Residual Stresses on Fatigue Precracking of CTOD Specimens, Metal Behaviour &Surface Engineering, IITT Technology Transfer Series, 1989, 249-254. 1211 API Spec. 2Z. First Edition 1987: Preproduction Qualification for Steel Plates for Offshore Structures. 1221 McHennry, H.I.-Denys, R.M.: Measurement of HAZ Toughness in Steel Weldments. 5th International Fracture Mechanics Summer School, Dubrovnik 1989, pp. 211-222. 1231 Pargerter, R.J.: Yield Strength from Hardness-a Reappraisal for Weld Metal. Welding Research Bulletin Nov. 1978. 124) Ritchie, R.O.-Knott, J.F.-Rice, J.R.: On the Relationship Between Critical Tensile Stress and Fracture Toughness in Mild Steel, JMPS, Voi. 21, 1973, pp. 3 9 5-4 1 0. 1251 Knott, J.F.: Macroscopic/Microscopic Aspects of Crack Initiation. Advances in Elasto-Plastic Fracture Mechanics, L.H. Larson, Ed., Commission of the European Communities. Joint Research Center, Ispra Establishment, Italy. 1979, pp. 1-41. Naslov avtorjev: prof. dr. Inoslav Rak, dipl. inž. Authors' Address: Fakulteta za strojništvo Faculty of Mechanical Engineering Smetanova 17, 62000 Maribor Mustafa Kocak, Dipl. Ing., Blagoj Petrovski. Dipl. Ing. GKSS Research Center, Geesthach Institute of Materials Research 2054 Geesthacht. Germany Prejeto: Received: 9.12.1994 Sprejeto: Accepted: 21.12.1994