Int J Fract (2014) 185:179–185 DOI 10.1007/s10704-013-9900-x
© Springer Science+Business Media Dordrecht 2013
LETTERS IN FRACTURE AND MICROMECHANICS
BRITTLE CRACK-ARREST FRACTURE TOUGHNESS IN A HIGH HEATINPUT THICK STEEL WELD Gyu Baek An1, Wanchuck Woo2*,Jeong-Ung Park3 1
Technical Research Laboratories, POSCO, Pohang, 790-300, Korea Neutron Science Division, Korea Atomic Energy Research Institute, Daejeon, 305-353, Korea 3 Department of civil Engineering, Chosun University, Gwangu 501-759, Korea 1 2 3 e-mails:
[email protected],
[email protected],
[email protected] 2
Abstract. Brittle crack-arrest fracture toughness (KIa) was determined as a function of the temperature in a large-scale (1 x 1 m) 50-mm thick steel base metal and its high heat-input (32 kJ/mm) weld. An impact initiates crack propagation from a notch at low temperature toward a higher temperature region where the crack stopped due to the improved fracture toughness. The relationship between the toughness and crack-arrest temperature provides the KIa at -10 Û&RIDERXW 03D¥PLQWKHweld specimen, which is a significant decrease compared to that of WKHEDVHPHWDO 03D¥P )XUWKHUPRUHWKHSDWKRIWKHFUDFNSURSDJDWLRQZDV discussed in terms of the grain size, hardness, and Charpy impact energy of the localized crack region. Keywords: Brittle fracture, crack arrest, toughness, thick steel, welds 1. Introduction. As a complement to the classical crack-initiation fracture toughness (KIc), the crack-arrest fracture toughness (KIa) has become an important mechanical property to approach the material fracture and failure mechanics (Masubuchi 1980; Ripling and Crosley 1982). Although a crack may start in the region of high stress and/or localized faults/embrittlement, its extension can be restricted by surrounding materials having sufficient resistance to the crack propagation resulting in prevention of a catastrophic failure (Smedley 1989, Priest 1998). A number of types of tests have been proposed and experimented to characterize the brittle crack propagation and resistance up to date. Robertson (1951) early introduced a crack-arrest test method, which examines an impactinitiated crack growing under the uniform load and the crack stop due to the high enough toughness at an elevated temperature. It can determine the crack-arresting (transition) temperature when it is above a certain point the cracks are arrested under fairly high stress. Thus, it has been widely studied to provide a relationship between the critical stress and the crack-arresting temperature by Feely et al. (1954), Yoshiki et al. (1960). Recently, it has become essential to assess the integrity of the pressure vessel components in nuclear power plant applications at various temperatures and sample geometries (Pugh et al. 1986, Naus et al. 1989, Bass et al. 2005). Several models and simulations propose the fracture mechanics
123
180
G. B. An et al.
and dynamics related to the crack toughness and propagations (Sumi 1990, Hajjaj et al. 2008). As the scale of many ships and offshore plants significantly increases in heavy industries, it is necessary to understand the brittle fracture phenomena for the fullscale components at low temperature (Inoue et al. 2007, An et al. 2009). A few previous studies elucidate the strong dependence of the KIa on the scale and localized material properties of specimens (Priest 2003, Bezensek and Hancock 2007). Furthermore, a new high heat-input (over 30 kJ/mm) welding process is extensively used to join the thick (over 50 mm) steel plates with the better efficiency. Consequently, it is of importance to determine the quantitative value of KIa in the thick steel base metal and its high heat-input weld in the perspective of the safety design and engineering. In recent, a guideline by the International Association of Classification Societies (IACS, 2013) requires that the KIa should EHKLJKHUWKDQ 03D¥P corresponding to 6000 N/mm3/2) at the lowest design temperature (- Û& RI VKLSV ZKHQ LW XVHV RYHU -mm thick steel base metal plates. Inoue et al. (2007) reported that it is not easy to arrest the running brittle cracks of the high heat-input welds when the plate thickness is over 65 mm at -10 Û&GXH WR WKH ORZ IUDFWXUH WRXJKQHVV LQVLGH WKH WKLFN ZHOGV +RZHYHU LW LV YHU\ limited in documents to show the KIa as a function of crack-arrest temperature and the crack propagation path in the high heat-input large-scale weld components. The purposes of this paper are (i) to determine the KIa as a function of crackarrest temperature in both 50-mm thick base metals and high heat-input welds and (ii) to examine the crack propagation path associated with material properties including grain size, hardness, and Charpy impact energy. It can provide a quantitative guide of the KIa to avoid the unexpected abrupt failure in the largescale high heat-input welding structures.
C
D
Z
%QQNKPI EJCODGT
.QY6
%TCEM CTTGUV
HCEG
[
EGPVGT +ORCEV PQVEJ
*KIJ6
9GNF
WPKV OO
%QPUVCPVNQCF
+ORCEV PQVEJ
OO
TQQV EGPVGTNKPG
Figure 1. (a) Schematic of the brittle fracture test specimen in welds, (b) macrostructure of the 50-mm thick high heat-input steel weld.
123
Brittle crack-arrest fracture toughness
181
2. Experimental procedure. The chemical composition of as-received commercial 50-mm thick high-strength steel (called as EH36 for the shipbuilding steel grade) was 0.07C, 0.14Si, 1.54Mn, 0.004P, 0.002S, and balance Fe in wt.%. The base metal (BM) provides the yield strength of 450 MPa, tensile strength of 540 MPa, elongation of 26%, and Charpy impact energy of 340 J at -Û&7KH two BM plates (each 1000-mm long × 500-mm wide × 50-mm thick) have been joined using 1-pass 2-pole electrogas welding technique, Fig. 1(a). The joint experienced very high heat input of 32 kJ/mm by using 400 A, 42 voltage electrode (face pole) and 360 A, 42 voltage electrode (root pole) both traveling parallel to the rolling direction of the plate at a constant speed of 0.1 mm/s. The cross-sectional macrostructure was shown in Fig. 1(b). Note that the angle of the V type groove was 25 degree. The chemical composition of the welding consumable was 0.08 C, 0.32 Si, 1.67 Mn, 0.01 P, and balance Fe in wt.%. Brittle crack-arrest toughness (KIa) was measured in the BM and weld. The specimen was prepared with the size of 600 x 600 x 50 mm (BM) and 1000 x 1000 x 50 mm (weld), Fig. 1(a). It was installed to the 3,000-ton tensile machine by attaching to the large-scale specimen in the total dimension of about 5000-mm long, 1800-mm high, and 150-mm thick. A constant tensile load in the range of 50 ~ 170 MPa was applied along the transverse direction (y) to the crack propagation in each test and the temperature gradieQWZDVFRQWUROOHGSUHFLVHO\DVÛ&PPLQ the range of - a Û& E\ OLTXLG QLWURJHQ DQG FRLO KHDWHU DORQJ WKH SDUDOOHO direction (x) to the crack propagation, Fig. 1(a). When the tensile load and temperature gradient were stabilized, a notch impact energy (2.7 kJ) was applied to the full-thickness penetrated notch, and the propagated crack length (Ca) and the temperature (Ta) at the location of the crack arrested were recorded, Fig. 2(a). Note that the KIa 03D¥P ZDV FDOFXODWHG E\ WKH IROORZLQJ HTXDWLRQ (Smedley 1989, IACS 2013): r p = σ πj
πj Yi πj Yi
(1)
where ıg (P/Bt) is the gross stress (N/mm2), P is the test load (N), B is the specimen width (mm), t is the specimen thickness (mm), Ca is the arrest-crack length (mm). The applied stresses and arrest temperatures were summarized in Table 1. Note that the KIa values at -Û&ZHUHGHWHUPLQHGXVLQJWKHWKUHHVHSDUDWH experiments performed with the different applied stresses and arrest temperatures. Microstructure was investigated at the cross-section of the weld. Vickers microhardness (Hv) was measured along the 2-mm below the top surface (face),
123
182
G. B. An et al.
mid-thickness (center), and 2-mm above the bottom surface (root), Fig. 1(b), with 1-mm horizontal spacing up to 30 mm from the weld centerline. Charpy impact energy were measured at -Û& XVLQJ WKUHH VHWV RI VSHFLPHQV WDNHQ IURP HDFK notch located at the weld centerline, fusion line (FL), and 1 and 3 mm from the FL, respectively. Table 1. Brittle crack-arrest toughness (KIa03D¥P DWDSSOLHGVWUHVVHVDQGFUDFN-arrest temperature (Ta in WKHVWHHOEDVHPHWDOV %0 DQGKLJKKHDW-LQSXWZHOGV ZHOG
Test specimens
KIa at Ta (MPa m)
Applied stress (MPa)
Arrest temperature (Ta, Ș)
BM 1
110
-25
153
BM 2
160
-2
BM 3
170
-7
Weld 1
50
-23
9GNF
120
-10
9GNF
130
24
3. Results and discussion. Figure 2(b) shows the variation of the KIa 03D¥P DV a function of the crack-arrest temperature (Ta Û& 2YHUDOO LW VKRZV D OLQHDU relationship between the KIa and Ta. The KIa decreases as the Ta decreases in both BM and welds. For example, the KIa RI 03D¥P DW - Û& GHFUHDVHG WR 03D¥P DW - Û& LQ %0 DQG VLPLODUO\ 03D¥P DW Û& GHFUHDVHG WR 03D¥PDW-Û&LQZHOGV7DEOH. +ORCEV PQVEJ
C
$/
9GNF
D
350
BM Weld
300
+ORCEV PQVEJ
250
-+C /2C O
%TCEM CTTGUV
200
150
100
%TCEM CTTGUV
(WUKQPNKPG
50
0 -30
-20
-10
0
6C Q%
10
20
30
Figure 2. (a) Brittle crack-propagation path and crack-arrest location of the base metal (BM) and heat-input weld (Weld), (b) brittle crack-arrest toughness (KIa 03D¥P DV D IXQFWLRQ RI FUDFN-arrest temperature (Ta, o C) in the BM and Weld.
123
Brittle crack-arrest fracture toughness
183
It is reasonable when considering the metals can be more brittle at the lower temperature in general. When comparing the KIa between BM and weld at -Û& (a dotted line in Fig. 2b), the KIa RIDERXW03D¥P 1PP3/2) in the BM significantly decreases to 100 M3D¥P 1PP3/2) in the weld specimen. The results confirmed that the BM has a sufficient crack-arrest capability satisfying the shipbuilding class guideline (IACS, 2013), while it decreases by 58% in the weld. +ORCEV PQVEJ
C
400 face center root
250
200
150
100
0
5
10
15
20
25
30
Distance from weld center (mm)
Charpy energy at -20 oC (J)
Vickers hardness (Hv)
300
+ORCEV PQVEJ
D
face center root
300
200
100
0
Weld
FL
FL + 1 FL + 3 mm
Figure 3. (a) Microhardness (Hv) and (b) Charpy impact energy (J) at -20 oC measured along the face, center, and root lines of the heat-input weld. Charpy energy was measured at weld centerline, fusion line (FL), 1 and 3 mm from the FL, respectively.
Let us discuss about the brittle crack-propagation path and its arrest location. In the BM, Fig. 2(a), the crack propagates a certain distance and deviates from the path, whilst the brittle crack in weld propagates straight along the weld fusion line and arrested far away from the initial notch location. It was similarly observed in the three repeated experiments in welds. It has been known that the crack propagation path can be the result of the interplay between the local material strength and toughness gradients (Bezensek and Hancock 2007). Figure 3(a) shows the local material strength in terms of microhardness in weld. It clearly shows that the lower hardness in the location of the impact notch (~ 13 mm) than those of the weld and BM. In fact, the grain size of the impact region, which includes the weld, FL, and heat-affected zone, was relatively larger than those of the weld and BM (Woo et al. 2013). For the toughness gradient, the Charpy impact energy was experimented in Fig. 3(b). The results show that the FL has the lower toughness than those of near the FL (FL+1mm, FL+3mm), though it is
123
184
higher than weld. Subsequently, the reason of the straight propagation of the crack path along the FL in weld is possibly attributed to the localized degradation of the material strength and toughness in the impact notch region compared to its vicinity. 4. Conclusions. Brittle crack-arrest toughness (KIa) was determined in 50-mm thick steel base metal plates and high heat-input (32 kJ/mm) welds. The largescale brittle fracture experiment provided the KIa as a function of applied load (50 ~ 170 MPa) and crack-arrest temperature (- a Û& 7KH KIa of about 240 03D¥PLQWKHEDVHPHWDODW-Û&VLJQLILFDQWO\GHFUHDVHVE\LQZHOG 03D¥P ,QWHUSOD\ EHWZHHQ WKH ORFDOL]HG GHJUDGDWLRQ LQ PDWHULDO VWUHQJWK DQG toughness, e.g., large grain size, low hardness, and low Charpy impact energy, can cause the straight crack propagation. Acknowledgement. This research activity was supported by the Nuclear Research and Development Program of the Korea Science and Engineering Foundation funded by the Korean government. It is also supported by POSCO project No.20126193. References Masubuchi K. (1980).. Analysis of Welded Structures - Fracture toughness, 1st edn. Pergamon, New York; 336-399. Ripling E.J., Crosley P.B. (1982). Crack arrest fracture toughness of a structural steel (A36). :HOGLQJ Research 65s-74s. Smedley G.P. (1989). Prediction and Specification of Crack Arrest Properties of Steel Plate. International Journal of Pressure VesVHOVDQGPiping 40, 279-302. Priest A.H. (1998). An energy balance in crack propagation an arrest. Engineering Fracture Mechanics 61, 231-251. Roberson T.S. (1951). Brittle fracture of mild steel. Engineering 172, 444. Feely F.J., Hrkto D., Kleppe S.R., Northrup M.S. (1954) Report on brittle fracture studies. :HOGLQJ-RXUQDO 33, 99s-111s. Yoshiki M., Kanazawa T., Itagaki H. (1960). Double tension test with flat temperature gradient. ProcHHGLQJV UG-DSDQ&RQJUHVVRQ7HVWLQJ0DWHULDOV-Metallic Materials 103-106. Pugh C.E., Corwin W.R., Bryan R.H., Bass B.R. (1986). Some advances in fracture studies under the heavysection steel technology program. Nuclear Engineering DQGDesign 96, 279-312. Naus D.J., Keeney-Walker J., Bass B.R. (1989) High-temperature crack-arrest behavior of prototypical and degraded (simulated) reactor pressure vessel steels. International Journal of Pressure VesVHOVDQGPiping 39, 189-208. Bass B.R., Williams P.T., Pugh C.E. (2005). An updated correlation for crack-arrest fracture toughness for nuclear reactor pressure vessel steels. International Journal of Pressure VesVHOVDQG Piping 82, 489-495. Sumi Y. (1990). Computational crack path prediction for brittle fracture in wleing residual stress fields. Interational Journal of Fracture 44, 189-207. Hajjaj M., Berdin C., Bompard P., Bugat S. (2008). Analyses of cleavage crack arrest experiments: influence of specimen vibration. Engineering Fracture Mechanics 75,1156-1170. Inoue T., Ishikawa T., Imai S., Koseki T., Hirota K., Tada M., Kitada H., Yamaguchi Y., Yajima H. (2007). Long crack arrestability of heavy-thick shipbuilding steels. ProcHHGLQJV 17th ,QWHUQDWLRQDO 2IIVKRUH DQG Polar Engineering Conference 3322-3326.
123
G. B. An et al.
Brittle crack-arrest fracture toughness
185
An G.B., Park J.S., Ryu K.M., Yoon Y.C., Kim H.J., Park J.M., Kang K.B. (2009). Brittle crack arrest technique of thick steel plate welds in container ship. ProcHHGLQJV 19th ,QWHUQDWLRQDO 2IIVKRUH DQG 3RODU Engineering Conference 357-361. Priest A.H. (2003). The influence of structural dimensions on crack arrest. Engineering Fracture Mechanics 70, 2421-2437. Bezensek B., Hancock J.W. (2007). The toughness of laser welded joints in the ductile-brittle transition. Engineering Fracture Mechanics 74, 2395-2419. International Association of Classification Societies (IACS). (2013). Rule requirements for use of extremely thick steel plates, s33. Woo W., An G.B., Kingston E.J., DeWald A.T., Smith D.J., Hill M.R. (2013). Through-thickness distributions of residual stresses in two extreme heat-input thick welds: A neutron diffraction, contour method and deep hole drilling study. Acta Materialia 61, 3564-3574.
123
