Professional Documents
Culture Documents
art ic l e i nf o a b s t r a c t
Article history: We have studied the microstructural evolution in the simulated coarse grain heat affected zone (CGHAZ)
Received 27 April 2015 of novel low carbon microalloyed steel with yield strength of 1000 MPa using electron microscopy, while
Received in revised form the crystallographic characteristics were studied by electron backscatter diffraction (EBSD). The dete-
9 May 2015
rioration in low temperature toughness after the simulated welding cycle was attributed to the formation
Accepted 11 May 2015
of coarse blocky M–A constituent. However, the lower bainite microstructure of the matrix was beneficial
Available online 16 May 2015
for low temperature impact toughness because of similarity in the crystal structure of variants of lath
Keywords: martensite, which were present in high proportion at the high angle grain boundaries. A high fraction of
Coarse grain heat affected zone small M–A constituent also influenced impact toughness.
Low carbon steel
& 2015 Elsevier B.V. All rights reserved.
Ultra-high strength steel
Martensite–austenite constituent
Toughness
http://dx.doi.org/10.1016/j.msea.2015.05.033
0921-5093/& 2015 Elsevier B.V. All rights reserved.
H. Xie et al. / Materials Science & Engineering A 639 (2015) 482–488 483
Table 1
Chemical composition and mechanical properties of base metal.
0.08–0.11 0.4 1.0–1.9 0.12 1.08 0.0013 0.72 0.28 1058 1192 10.1 24
Note: Ceq ¼C þ (Mn þSi)/6 þ (Niþ Cu)/15 þ(Cr þ Moþ V)/5; Pcm ¼C þ Si/30þ (Mn þCu þCr)/20 þ Ni/60 þMo/15 þ5B; Impact toughness of 1/4-Charpy size specimen.
1400 1400
10s 10s
1200 30s 1200 30s
60s 60s
1000 150s 1000 150s
800 800
600 600
400 400
200 200
0 0
0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500
Time(s) Time(s)
Fig. 1. Welding thermal cycle of the simulated CGHAZ specimens. (a) Actual welding thermal cycle of the simulated specimens and (b) the modeled plots of the simulated
specimens.
Block AF
M LB
Packet
LB GB
GB
LB
GB
Fig. 2. Optical micrographs of simulated CGHAZ with different t8/5, 1000 . (a) 10 s, (b) 30 s, (c) 60 s, and (d) 150 s.
484 H. Xie et al. / Materials Science & Engineering A 639 (2015) 482–488
welding thermal simulation was carried out using a thermo-me- 3. Results and discussion
chanical simulator. The specimens were heated to a peak tem-
perature of 1350 °C at a rate of 100 °C/min and then held there for 3.1. The actual temperature of the simulated CGHAZ specimens
1 s, after which they were cooled to 150 °C using different t8/5 time
(time to cool from 800 to 500 °C), and then air cooled to room Plot of temperature vs. time for the CGHAZ simulation is pre-
sented in Fig. 1. The sample temperature was controlled until it
temperature. The selected t8/5 times were 10, 30, 60 and 150 s, to
dropped to 150 °C at which no phase transformation occurs based
simulate different welding inputs, which were 16.2, 28.0, 39.6 and
on the Rykalin mathematical model. As shown in Fig. 1a and b, the
62.7 kJ/cm, respectively, that were calculated using the Rykalin
actual temperature agrees with the modeled plots. Variation in
mathematical model [14].
cooling times is because of different t8/5 time used to simulate
The welding simulations were repeated three times for each
microstructural changes at different cooling rates.
t8/5 time condition, from which the average impact toughness was
obtained. After simulation, the specimens were machined to
3.2. The microstructure of simulated CGHAZ specimens
standard Charpy v-notch specimens and the test was performed at
20 °C using Instron Dynatup 9200 Series drop weight impact
Fig. 2 shows optical micrographs of simulated CGHAZ speci-
tester, consistent with ASTM E23 specification. The fracture sur-
mens at different cooling rates. The figure indicates that micro-
faces were observed using a FEI Quanta 600 scanning electron structural changes dramatically occur with increasing t8/5 cooling
microscope (SEM). time. The microstructure of the specimens subjected to t8/5 of 10 s
The simulated specimens were cut near the monitoring ther- was characterized by refined martensite (Fig. 2a). The prior aus-
mocouple, metallographically polished, and etched with a 4% nital tenite grains comprised of large structural units, i.e. packets and
solution for observing using a Leica DMIRM 2500M optical mi- blocks [15], having a diameter of 25–40 μm. The average packet
croscope and Zeiss Ultra 55 SEM. For EBSD studies, the samples width was 10–15 μm, and the block width was 3–5 μm. This is
were electrolytically polished in a solution of 8% perchloric acid– because of growth of austenitic grains at high peak temperature
alcohol solution. EBSD maps were obtained at a step size of 0.1 μm, and subsequent fast cooling during simulated welding with a low
and analyzed using HKL-Channel 5 software. heat input of 16.2 kJ/cm. The CGHAZ microstructure was observed
LB
M
M
Block
AF
GB
LB
3μm 3μm
Fig. 3. SEM micrographs of the simulated CGHAZ with different t8/5, 2000 . (a) 10 s, (b) 30 s, (c) 60 s, and (d) 150 s.
H. Xie et al. / Materials Science & Engineering A 639 (2015) 482–488 485
to coarsen on increasing cooling time. At t8/5 time of 30 s (Fig. 2b), block consisting of several laths with straight boundaries. At t8/5 of
the microstructure consisted of acicular ferrite, lath bainite, and 30 s, the microstructure comprised of martensite, lath bainite, and
martensite. The martensite fraction decreased with increase in acicular ferrite, as shown in Fig. 3b. Because of low cooling rate, the
heat input, whereas formation of lath bainite and the increase in small phase inside the microstructure exhibited slight coarsening
local C-concentration led to the formation of acicular ferrite within and M–A constituent appeared. At t8/5 of 60 s, (Fig. 3c), the micro-
prior austenite. At t8/5 time of 60 s (Fig. 2c), the microstructure was structure entirely comprised of lath bainite, which was coarser than
predominantly comprised of lath bainite, with width greater than the martensitic structure, although laths had a crystallographic re-
martensite. Hence, grain size was also increased with increase in lationship similar to martensite. At t8/5 of 150 s, the microstructure
heat input [16]. At t8/5 time of 150 s (Fig. 2d), granular bainite and exhibited necklace-like structure of block M–A constituent present
lath bainite microstructure was observed, which was different at the grain boundaries. The M–A constituent and typical granular
from other three steels, with proeutectoid ferrite present at the bainite were present inside the prior austenite grain (Fig. 3d). It was
prior austenite grain boundaries, because of low cooling rate used. also noted that the microstructure was significantly coarse, and the
The M–A constituent is appreciably larger and was 2 μm long spacing between different phases was appreciably large.
and 1 μm wide.
As described above, the microstructure was significantly influ- 3.3. The crystallographic characteristics of the simulated CGHAZ
enced by different heat inputs. Further study of the resulting fine specimens
microstructure was performed using SEM and is presented in Fig. 3.
At t8/5 of 10 s, it was possible to easily identify a packet in the The crystallographic characteristics of the simulated CGHAZ
martensite microstructure containing various blocks (Fig. 3a) and a specimens were analyzed by EBSD and are presented in Fig. 4. The
25μm 25μm
25μm 25 μm
{100} {100}
0 0
Fig. 4. Crystallographic characteristics of the simulated CGHAZ specimens analyzed by EBSD: (a) the orientation image map with t8/5 of 30 s, (b) the orientation image map
with t8/5 of 60 s, (c) grain boundary misorientation distribution with t8/5 of 30 s, (d) grain boundary misorientation distribution with t8/5 of 60 s, (e) experimental 〈100〉 pole
figure containing martensitic variants within the selected austenite grain A, and (f) experimental 〈100〉 pole figure containing bainitic variants within the selected austenite
grain B. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
486 H. Xie et al. / Materials Science & Engineering A 639 (2015) 482–488
5μm 5μm
5μm
Fig. 5. The image quality maps with the austenite distribution of the simulated CGHAZ specimens with different t8/5. (a) 10 s, (b) 60 s, and (c) 150 s.
Small dimples
Fig. 7. SEM micrographs of fracture surfaces of the simulated CGHAZ specimens. (a) 10 s, (b) 30 s, (c) 60 s, and (d) 150 s.
t8/5 times less than and up to 60 s, whereas the reduced toughness 4. Conclusions
observed on exceeding 60 s is due to the formation of blocky M–A
constituents composed of martensitic proportions. It is believed The change in morphology and mechanical properties of the
that the appearance of coarse M–A constituents in specimens with simulated CGHAZ specimens with different t8/5 time were studied.
cooling time of 150 s was the main reason for the dramatic de- The major conclusions are summarized as follows:
creasing in impact toughness of the simulated CGHAZ specimens.
The fracture surface of specimens with different t8/5 time is (1) The microstructure of the simulated CGHAZ specimen changed
presented in Fig. 7, and the fractographs correspond to the radial from lath martensite to lath bainite and then to coarse gran-
regions of specimens. Only highly misoriented boundaries can be ular bainite with increase in t8/5 time. This led to an increase
transformed to cleavage facet boundaries [3], and it is observed and then slight decrease in impact energy.
that the cleavage facet size on the specimen fracture surface was (2) The EBSD analysis indicated that lath bainite retains crystal-
relatively large for samples with a cooling time of 10 s. The facet lographic orientations similar to lath martensite. The orienta-
size was also equal to the packet size, as shown in Fig 7a. On in- tions of variants transformed from prior austenite were close
creasing t8/5 time, the microstructure of the simulated CGHAZ to Nishiyama–Wassermann orientation relationship rather
specimens transformed to lath bainite. Thus, the cleavage facet than the Kurdjumov–Sachs orientation relationship. Orienta-
size was decreased and was separated by boundaries of bainitic tion relationship analysis indicated that there was a relation-
lath, as shown in Fig. 7b. It is well known that the formation of fine ship between the transformed phase and the old phase. The
retained austenitic structure distributed at the grain boundaries of homogeneity of large-angle grain boundaries of lath bainite
prior austenite grain can effectively release stress concentration was better than lath martensite, which benefited the impact
[19]. Examination of fractograph of specimen with t8/5 of 60 s in- toughness of the simulated CGHAZ specimens.
dicated that there was some ductile fracture with large number of (3) The formation of coarse M–A constituent because of low
small dimples (indicated by white arrows in Fig. 7c), which is in cooling rate is the main reason for the decrease in impact
good agreement with the study of Lan [3]. Increasing t8/5 time led toughness of the simulated CGHAZ specimens. The coarse M–A
to the appearance of coarse M–A constituent in the sample, which constituent was nearly fully martensitic, which raises the
was detrimental to toughness, as evinced by cleavage planes on stress concentration at the rim of the prior austenite and
the fractograph of the specimen with t8/5 of 150 s (Fig. 7d). leading to cleavage crack.
488 H. Xie et al. / Materials Science & Engineering A 639 (2015) 482–488
Acknowledgments [7] A. Lambert-Perlade, A.F. Gourgues, J. Besson, T. Sture, A. Pineau, Metall. Mater. Trans.
A 35 (2004) 1039–1053.
[8] J. Moon, S. Kim, J. Lee, B. Hwang, C.G. Lee, C. Lee, Mater. Sci. Eng. A 497 (2008)
The authors appreciate the support from the National High 153–159.
Technology Research and Development Program of China (863 [9] C.L. Davis, J.E. King, Metall. Mater. Trans. A 25 (1994) 563–573.
Program 2015AA03A501), and the National Natural Science [10] A.E. Amer, M.Y. Koo, K.H. Lee, S.H. Kim, S.H. Hong, J. Mater. Sci. 45 (2010) 1248–1254.
[11] P. Mohseni, J.K. Solberg, M. Karlsen, O.M. Akselsen, Mater. Sci. Technol. 28 (2012)
Foundation of China (Grant no. 50527402). RDKM gratefully ac- 1261–1268.
knowledges support from the Department of Metallurgical and [12] S. Moeinifar, A.H. Kokabi, H.R. Madaah Hosseini, Mater. Des. 31 (2010) 2948–2955.
[13] C.W. Li, Y. Wang, Y.H. Chen, J. Mater. Sci. 46 (2011) 6424–6431.
Materials Engineering, University of Texas at El Paso, USA. [14] N.N. Rykalin, Calculation of Heat Processes in Welding, U.S.S.R., Moscow, 1960.
[15] S. Morito, X. Huang, T. Furuhara, T. Maki, N. Hansen, Acta Mater. 54 (2006)
5323–5331.
[16] S. Kim, Y. Kang, C. Lee, Mater. Sci. Eng. A 559 (2013) 178–186.
References [17] P.P. Suikkanen, C. Cayron, A.J. DeArdo, L.P. Karjalainen, J. Mater. Sci. Technol. 27
(2011) 920–930.
[18] P.P. Suikkanen, C. Cayron, A.J. DeArdo, L.P. Karjalainen, J. Mater. Sci. Technol. 29
[1] A. Lambert-Perlade, A.F. Gourgues, A. Pineau, Acta Mater. 52 (2004) 2337–2348.
(2013) 359–366.
[2] S. Moeinifar, A.H. Kokabi, H.R. Madaah Hosseini, Mater. Des. 32 (2011) 869–876.
[3] L.Y. Lan, C.L. Qiu, D.W. Zhao, X.H. Gao, L.X. Du, Mater. Sci. Eng. A 529 (2011) 192–200. [19] C.G. Lee, S.J. Kim, T.H. Lee, S. Lee, Mater. Sci. Eng. A 371 (2004) 16–23.
[20] O. Matsumura, Y. Sakuma, Y. Ishii, J. Zhao, ISIJ Int. 32 (1992) 1110–1116.
[4] H. Xie, L.X. Du, J. Hu, R.D.K. Misra, Mater. Sci. Eng. A 612 (2014) 123–130.
[5] C. Thaulow, A.J. Paauw, K. Guttormsen, Int. Inst. Weld. 59 (1987) 266–279. [21] G. Krauss, Mater. Sci. Eng. A 273–275 (1999) 40–57.
[6] B.C. Kim, S. Lee, N.J. Kim, D.Y. Lee, Metall. Trans. A 22 (1991) 139–149.