Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal

Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal

Accepted Manuscript Title: Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal Author: Wenl...

798KB Sizes 1 Downloads 21 Views

Accepted Manuscript Title: Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal Author: Wenlin Mo, Xiaobing Hu, Shanping Lu, Dianzhong Li, Yiyi Li PII: DOI: Reference:

S1005-0302(15)00126-7 http://dx.doi.org/doi: 10.1016/j.jmst.2015.08.001 JMST 540

To appear in:

Journal of Materials Science & Technology

Received date: Revised date: Accepted date:

26-11-2014 20-1-2015 28-1-2015

Please cite this article as: Wenlin Mo, Xiaobing Hu, Shanping Lu, Dianzhong Li, Yiyi Li, Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal, Journal of Materials Science & Technology (2015), http://dx.doi.org/doi: 10.1016/j.jmst.2015.08.001. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effects of Boron on the Microstructure, Ductility-dip-cracking, and Tensile Properties for NiCrFe-7 Weld Metal Wenlin Mo1, 2, Xiaobing Hu1, Shanping Lu1, 2, , Dianzhong Li1, Yiyi Li1 1

Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy

of Sciences, Shenyang 110016, China 2

Key Laboratory of Nuclear Materials and Safety Assessment, Institute of Metal Research, Chinese

Academy of Sciences, Shenyang 110016, China [Received 26 November 2014; Received in revised form 20 January 2015, Accepted 28 January 2015] *Corresponding author. Prof., Ph. D.; Tel.: +86 24 23971429; Fax: +86 24 83970095. E-mail address: [email protected] (S.P. Lu).

The distribution of boron and the microstructure of grain boundary (GB) precipitates (M23(C, B)6 and M2B) have been analyzed with their effects on the susceptibility of ductility-dip-cracking (DDC) and tensile properties for NiCrFe-7 weld metal, using optical microscopy (OM), secondary ion mass spectroscopy (SIMS), scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The results show that boron segregates at GBs in NiCrFe-7 weld metal during the welding process. The segregation of boron at GBs promotes the formation of continuous M23(C, B)6 carbide chains and M2B borides along GBs. The addition of boron aggravates GB embrittlement and causes more DDC in the weld metal, by its segregation at GBs presenting as an impurity, and promoting the formation of larger and continuous M23(C, B)6 carbides, and M2B borides along GBs. DDC in the weld metal deteriorates the ductility and tensile strength of the weld metal simultaneously. Key words: Boron; M23C6; M2B; ductility-dip-cracking; tensile properties 1. Introduction Nickel based Alloy 690 (Ni‒30Cr‒10Fe, mass fraction, wt%), is replacing Alloy 600 (Ni‒16Cr‒9Fe) as steam generator (SG) tubes in pressurized water nuclear reactors (PWR). Alloy 690 has superior resistance to intergranular stress corrosion cracking (IGSCC) and intergranular 1 / 16

Page 1 of 16

corrosion (IGC) in various environment[1‒3]. Welding material matching for Alloy 690 is AWS A5.14 ER NiCrFe-7 (also, FM-52), which has been widely used for joining Ni‒Cr base alloys. Compared with FM 62, FM 82, and that of type 304 stainless steel, NiCrFe-7 weldment shows good resistance against stress corrosion cracking (SCC) in high temperature water, and oxygenated and deoxygenated environment[4‒6]. NiCrFe-7 is applied to manufacture the key components in nuclear plants, such as nuclear PWRs and SGs[7,8]. It is also used in dissimilar welding between several Ni-base alloy, stainless steel and carbon steel in building nuclear power plants[9]. However, it has been pointed out that NiCrFe-7 weld metal is susceptible to ductility-dip-cracking (DDC) under heavy restraint conditions such as welding thick components[10‒13]. DDC is a solid-state hot cracking, caused by GB embrittlement at homologous temperatures that ranges from about 0.5 to 0.8 of the alloy melting points, which is observed at grain boundaries (GBs)[14,15]. DDC has been investigated extensively from various perspectives, such as chemical composition[15, 16], elements segregating at GBs[17‒20], precipitation behavior[12], GB migration[14, 21‒25], peak temperature and cooling rate [26] and so on. Filler metal composition, such as S, P, H, C, Ti, Ta, Mo, La, Ce, and Nb, has a strong effect on DDC formation in austenitic weldments. Segregation of S, P, and H to GBs is a dominant factor for DDC aggravation in Ni-base alloys, whereas Ti, Ta, Mo, La, Ce, and Nb addition to a filler metal reduces the sensitivity of DDC[13,17,20,27‒31]. As an essential micro-alloying element, boron is often added to nickel base superalloys to reduce the GB surface energy to improve interfacial cohesion at GBs, which would increase the fracture resistance of these GBs[32‒34]. Using secondary ion mass spectroscopy (SIMS), Thuvander and Stiller studied the effect of 55 ppm boron addition to Alloy 690, and found boron segregates to GBs which limits grain growth[33]. Boron segregates at GBs easily by two mechanisms namely equilibrium segregation[35,36] and nonequilibrium segregation[37,38]. In equilibrium segregation, solute atoms that diffuse to GBs are actually bound to GB sites[35,36,39]. Nonequilibrium segregation occurs during cooling of the material from the heat treatment temperature. It requires the formation of solute-vacancy complexes and a concentration gradient of these complexes between the grain interiors and GBs. Under the concentration gradients the complexes diffuse to GBs. This diffusion causes excessive solute atoms to concentrate in the vicinity of GBs, after the annihilation of vacancies at GBs. The degree of nonequilibrium segregation has been found to depend on the starting temperature, cooling rate, and bulk concentration of solutes[36,39,40]. Vishwakarma and 2 / 16

Page 2 of 16

Chaturvedi[41] found nonequilibrium segregation of boron could not be suppressed by water quenching. Karlsson and Norden[42] have measured GB boron levels of up to 18 at.% in stainless steel using an atom probe. During the segregation, B might further react to nucleate borides, such as M3B2, M5B3 and M2B phase[43]. Unlike M3B2 and M5B3 phases, M2B phase is always heavily faulted[43,44]. M2B phase has dual structural character, namely the C16 structure (Strukturbericht notation C16, space group I4/mcm) with the lattice parameter:

a=0.52 nm, c=0.43 nm and Cb

structure (Strukturbericht notation Cb, space group Fddd) with the lattice parameter:

a=1.47 nm,

b=0.74 nm, c=0.43 nm[44,45]. The role of boron in superalloys has been generally believed to be beneficial especially to mechanical properties such as creep parameters[32]. However, there is still no common agreement on the role of boron on weldability of superalloys. Special metal corporation (SMC) added boron to FM-52M to reduce the tendency for DDC[46]. Several works in superalloys found that the addition of boron would cause heat-affected zone (HAZ) microfissuring[36,41,47]. The boron addition affects the wetting characteristics along the GBs and lowers the solidification temperature, increasing HAZ liquation cracking susceptibility. Still, very little research has been undertaken to study the influence of boron on the microstructure and DDC for NiCrFe-7 weld metal. This study is aimed to investigate the effects of boron on the microstructure, DDC, and tensile properties for the ER NiCrFe-7 weld metal. 2. Experiments Two types of filler metals (B0 without boron addition, B1 with 46 ppm boron) were fabricated based on an AWS ER NiCrFe-7 filler metal (AWS A-5.14). The diameter of the filler metals is 0.9 mm. The chemical compositions are listed in Error! Reference source not found.. The welding process was performed in a flat position. The base metal is carbon steel, 500 mm × 125 mm × 25 mm in size. A single V groove at 20° on both sides was generated in a butt joint. Surfacing was applied to the three carbon steel sides to prevent element migration, especially C, as shown in Error! Reference source not found.. After surfacing, butt welding was performed using cold-wire feed (CWF) multiple semi-automatic gas tungsten arc welding (GTAW) and the inert gas was argon. About 55 weld passes in sixteen layers were filled for each weld. The weld metal made 3 / 16

Page 3 of 16

by B0 filler metal was named as WM-B0 (without boron addition), whereas the weld metal made by B1 filler metal was named as WM-B1 (with boron addition). The crack length and width were measured by optical microscopy (OM) on the cross sections of each of the welds. Ten cross-sections were cut randomly along the welds. The differential thermal analysis (DTA) tests were carried out on an SETSYS Evolution18 integrated thermal analyzer developed by SETARAM Corporation. During the test the specimens were heated to 1100 °C at a rate of 80 °C/min, and heated to 1450 °C at a rate of 10 °C/min. The initial melting point was measured through the first peak on the melting curve. Mechanical grinding and polishing were applied to prepare the samples for microstructure observation, followed by electrolytic etching using a 10 g H2C2O4 + 100 ml H2O reagent under the applied potential 10 V DC for 20‒30 s. The as-welded microstructures were observed using OM, scanning electron microscopy (SEM), and transmission electron microscopy (TEM, FEI® F20) operating at 200 kV. Chemical analysis was performed using an ICP spectrometer and SIMS. Tensile test was performed on AG-X 250 KN tensile testing equipment at room temperature. The geometries and sizes of the tension specimen are shown in Error! Reference source not found.. Equilibrium calculations for the ER NiCrFe-7 alloy were performed using the software Thermo-Calc (Version S), and the TTNi8 database developed by ThermoTech was used for the calculations. The elements considered for the multi-component calculations were Ni, Cr, Fe, C, Si, Al, Mn, Ti, and B. 3. Results 3.1. Ductility-dip-cracking Error! Reference source not found. shows the DDC in the WM-B1 weld metal. DDC is a solid-state

hot cracking, which is observed along GBs because of GB embrittlement[14‒16]. Table 2 Statistics of the length and number of DDC in the weld metals shows the statistics of the length and number of DDC in the weld metals, and WM-B1 has the significant longer crack length and more crack number than that of WM-B0. The longest crack approaches a limit of about 3 mm. This limit might be a combined effect of the size of the weld metal and the size of the grains in the weld metal. Table 3 Statistics of the DDC width in the weld metals 4 / 16

Page 4 of 16

shows the statistics of the DDC width in the weld metals, and WM-B1 has wider cracks than those in WM-B0. Therefore, weld metal with boron is more sensitivity to DDC. 3.2. Segregation of boron Error! Reference source not found. shows the distribution of boron in WM-B1 measured by SIMS. It can be seen that boron segregates at GBs and DDC. During welding, the segregation of boron at GBs is mainly nonequilibrium in nickel based alloys[36]. The segregation of boron at GBs would influence the microstructure and the properties of the GBs. The addition of boron would decrease the initial melting point and increase the solidification range of the weld metal, as shown in the vertical cross-section of the phase diagram (Error! Reference source not found.). The DTA test for the WMs shows the initial melting point is 1361 °C for WM-B0

and 1355 °C for WM-B1, as shown in Error! Reference source not found.. The segregation of boron at GBs decreases the melting point of GB material and increases the wettability of the GB surface[36]. Even in low boron (11 ppm) content nickel based alloys[34], the segregation effects at GBs were observed. The segregation of boron at GBs potentially increases the solidification range of the material in the GB region, which adversely influences the weldability of the nickel based alloys. Therefore, the addition of boron should be carefully controlled for better weldability. During the nonequilibrium segregation, once the boron-vacancy complexes have formed and accumulated at GBs during cooling, they might further react to nucleate borides in order to lower the GB energy[39]. The formation of stable borides may cause a depletion of these complexes in the vicinity of GBs, which will serve as a driving force for further accumulation. The segregation of boron at GBs promotes the precipitation of M23C6 carbides and M2B borides at GB in WM-B1 weld metal. 3.3. M23C6 Error! Reference source not found. shows the morphology of M23C6 carbides. The presence of M23C6 carbide was confirmed by selected area electron diffraction (SAED) pattern at the axis zone [110], as shown in Error! Reference source not found.(c, d). M23C6 carbide has a cube-on-cube orientation relationship with oneγmatrix grain, as shown in Error! Reference source not found.(d). The 5 / 16

Page 5 of 16

M23C6 carbide contains more Cr and Fe and less Ni than that in the matrix, and the grain with a cube-on-cube M23C6 has a Cr-depletion region, as shown in Error! Reference source not found.(e). WM-B1 has more and larger M23C6 carbides than those in WM-B0 along GBs, and small discrete GB M23C6 particles in WM-B0 turn into a wide continuous carbide chains in WM-B1, as shown in Error! Reference source not found.(a, b). Boron is known to increase the thermodynamic phase stability and

precipitation kinetics of M23C6, and boron segregation at GBs likely contributed to the improving stability of M23C6 over MC[48]. The equilibrium phase calculation shows that boron will participate in the formation of M23C6 and form M23(C, B)6 (Error! Reference source not found.(b)), which increases the initial precipitation temperature and the equilibrium quantity of M23C6 (Error! Reference source not found.(a)). The high boron content of M23(C, B)6 is most likely the result of nonequilibrium

segregation during cooling[36]. The mobility of boron in γ matrix is much higher than that of carbon[32]. Consequently, at the early stages of M23(C, B)6 formation the flux of boron from γ towards the growing M23(C, B)6 particle is higher than that of carbon, leading to a high boron concentration in M23(C, B)6[32]. In the welding process, there is no sufficient time for M23C6 precipitates fully, but M23(C, B)6 can form even at high cooling rates during welding. The higher the initial precipitation temperature of M23C6, the faster the precipitation speed. Therefore, the addition of boron increases the initial precipitation temperature and the equilibrium quantity of M23C6, which promotes the formation of M23C6 in the weld. 3.4. M2B As shown in Error! Reference source not found., it is known that boron segregates at the GBs. Considering the low spatial resolution for SIMS, it is impossible for us to determine the existing forms of the precipitated phase that contains boron. In order to clarify the exact existing form of boride, TEM analyses were performed with high spatial resolution. Error! Reference source not found.(a) displays one blocky precipitate along the GB with heavily faulted feature. This feature is clearly seen in Error! Reference source not found.(b) with a higher magnification. Based on this striped contrast inside the precipitated grain, we can infer that this precipitate is the M2B phase. Error! Reference source not found.(c) displays the chemical information for the M2B phase. It is found that the M2B boride is

composed mainly of Cr, and a small amount of Fe, Mn, Ti. 6 / 16

Page 6 of 16

M2B phase has dual structural character, namely the C16 and the Cb structure, and the lattice parameters for the two structures are closely related. As for the exact structural characteristic of M2B phase precipitated in NiCrFe-7 alloy, it is a little complicated. Error! Reference source not found.(a‒d) schematically displays a quarter of the simulated electron diffraction pattern (EDP) along [001] Cb, [010]Cb, [1 1]C16, and [1

]C16, respectively. Considering the secondary diffraction, which results

from the dynamical effect, it is seen that the EDP of [001]Cb and [010]Cb are the same as that of [1 1]C16, and [1

]C16, correspondingly. For identifying the exact structure, the only choice is tilting

the grain to another zone-axis. Due to the heavily faulted feature of M2B grain, the real-space high resolution TEM imaging technique is more useful than the reciprocal space electron diffraction analysis. In Error! Reference source not found.(a‒d), HRTEM images are acquired from single one M2B grain located at the edge of

electron transparent area. Error! Reference source not found.(a) and (c) correspond to one border area of the grain but with different tilting angles. Error! Reference source not found.(b) and (d) correspond to interior of the grain with different tilting angles. The tilting angle for Error! Reference source not found.(a) and (c) are the same as that of Error! Reference source not found.(b) and (d), respectively. And Error! Reference source not found.(a) can be labeled as [013]Cb direction. While for Error! Reference source not found.(b), it is complicated because we can take it as [010]Cb or [1

]C16 directions because

of the indistinguishability as discussed for Error! Reference source not found.. However, after tilting the grain along [100]Cb about 30°, Error! Reference source not found.(d) is acquired from the proximate area as Error! Reference source not found.(b). Error! Reference source not found.(d) can be labeled as [011]Cb unambiguously. So Error! Reference source not found.(b) should be labeled as [010]Cb. Similarly, Error! Reference source not found.(c) can be labeled as [001]Cb or [1 1]C16. However, considering the

correspondence between Error! Reference source not found.(a) and (c), Error! Reference source not found.(c) should be labeled as [001]Cb. In other words, when one border area is tilted from [013]Cb

(Error! Reference source not found. (a)) to [001]Cb (Error! Reference source not found. (c)), the interior is tilted from [010]Cb (Error! Reference source not found. (b)) to [011]Cb (Error! Reference source not found.(d)). Actually, the intersection angle between [001]Cb and [011]Cb, [013]Cb and [010]Cb is about

60°. Thus it can be determined that the M2B grain displayed in Error! Reference source not found.(a) contains a 60° rotation twin for Cb structure. Besides the rotation twin in Cb structure, nano-scaled intergrowth between C16 and Cb is also determined. On the basis of atomic image in Error! Reference 7 / 16

Page 7 of 16

source not found., which is obtained from the same grain as Error! Reference source not found., the

orientation relationship for the intergrowth can be indexed as [001]C16//[011]Cb and (110)C16//(100)Cb. 3.5. Tensile properties Error! Reference source not found. shows the results from a tensile test for the two WMs. And WM-B0 has better tensile properties than WM-B1. Error! Reference source not found. shows the morphology of the ruptured sections using SEM. There are two types of morphology in the weld metal ruptures, as shown in Error! Reference source not found.(a) and (b). One has equiaxed dimples (Error! Reference source not found.(e) and (f)), another type does not have dimples, but a rough face with sidestep shape (Error! Reference source not found.(c) and (d)). The rough face is along GBs, and the tensile specimen rupture is initiated at GBs (Error! Reference source not found.), which is caused by DDC in the WMs[15]. WM-B1 has more DDC in the weld metal, which has more rough face area. Therefore, the addition of boron deteriorates the ductility and tensile strength of the WM-B1 simultaneously because more DDC forms in WM-B1. 4. Discussion Researchers have proposed a variety of mechanisms of DDC, including impurity elements embrittlement, such as the segregation of S, P and other impurity elements at GBs, causing the loss of GBs ductility, thereby leading to DDC[17‒20], GB sliding

[14,21‒25]

, and precipitation-induced

cracking (PIC), such as M23C6 PIC[49]. In essence, DDC is caused by GB embrittlement[49], and elements segregation at GBs and precipitates along GBs would affect GB embrittlement and DDC. The segregation of boron at GBs would cause GB embrittlement and deteriorate hot ductility of a nickel based alloy[41,48], making them more susceptible to DDC. The general effect of increasing boron concentration at GB is to increase DDC tendency, with boron having a highly significant effect, in contrast to the minor influence of C[34]. Other trace elements such as S[50], P[51], Bi[52] would also segregate at GBs by nonequilibrium segregation mechanism and cause GB embrittlement, making the alloys susceptible to DDC. Boron plays a key role in the kinetics of dissolution and formation of the M23C6, and was found to have an effect on the DDC susceptibility[48]. Researchers[12,53,54] inferred that DDC is caused by 8 / 16

Page 8 of 16

the macroscopic thermal and solidification stresses induced during welding and the local stresses generated during GB precipitation of partially coherent M23C6 carbides. Previous research[49] reported that DDC is initiating at the M23C6/γ interface at GBs and then propagates along GBs. M23C6 has a cube-on-cube orientation relationship with one grain (Error! Reference source not found.(d)). DDC initiates at the incoherent M23C6/γ interface, because the incoherent M23C6/γ

interface has weak interface strength and higher stress concentration[49,55‒57]. Together with the thermal and solidification stresses induced during welding, the incoherent M23C6/γ interface is prone to crack initially, which exacerbates DDC in NiCrFe-7 alloy at elevated temperatures. WM-B1 has more and larger M23C6 than WM-B0 at GB (Error! Reference source not found.(a) and (b)), which corresponds to longer and wider DDC in WM-B1 (Table 2 Statistics of the length and number of DDC in the weld metals and Table 3 Statistics of the DDC width in the weld metals ). Furthermore, small discrete GB M23C6 particles in WM-B0 turn into wide continuous carbide chains in WM-B1. Previous researches have shown that continuous intergranular carbide chains along GBs promote the nucleation and growth of intergranular crack[14,58]. Therefore, the addition of boron will promote the formation of larger and continuous M23C6 along GBs, which aggravates DDC in the WM-B1. Besides M23C6, M2B also precipitated along the GBs, and the M2B/γ interfaces are incoherent without a fixed orientation. As stated above, the incoherent interface of the intergranular precipitate/γ is prone to crack initially, which aggravates DDC in the weld metal. Compared with the M23C6/γ interface, the M2B/γ interfaces are all incoherent without a fixed orientation. Furthermore, the M2B grain is heavily faulted. Therefore, it could be inferred that the M2B/γ interfaces are more prone to crack than M23C6/γ interfaces. The precipitation of M2B should be carefully controlled for minimizing DDC in the weld metal. Therefore, the addition of boron in NiCrFe-7 weld metal would aggravate GB embrittlement and cause more DDC in the weld metal, by its segregation at GBs presenting as an impurity, and promoting the formation of larger and continuous M23C6 carbides, and M2B borides along GBs. As a result, more DDC in WM-B1 deteriorates the ductility and tensile strength of the WM-B1 simultaneously. 9 / 16

Page 9 of 16

5. Conclusions During welding, boron segregates at GBs. The segregation of boron at GBs promotes the formation of M23(C, B)6 by increasing the initial precipitation temperature and the equilibrium quantity of M23C6. Compared with the weld metal without boron addition, the weld metal with boron addition has more and larger M23C6 carbides along GBs. The small discrete GB M23C6 particles in the weld metal without boron addition turn into a wide continuous M23(C, B)6 carbide chains in the weld metal with boron addition. Besides M23C6 carbide, M2B borides also precipitated along the GBs in the NiCrFe-7 weld metal with boron addition. And the grains of this phase possess the heavily faulted characteristics. Based on the atomic TEM imaging technique, it is found that the M2B boride has dual structural character, namely the C16 and Cb structure. Moreover, the rotation twin for Cb structure and nano-scale intergrowth of C16 and Cb structure are also identified in the heavily faulted areas. The addition of boron in NiCrFe-7 weld metal aggravates GB embrittlement and causes more DDC in the weld metal, by its segregation at GBs presenting as an impurity, and promoting the formation of larger and continuous M23C6 carbides, and M2B borides along GBs. DDC in the weld metal deteriorates the ductility and tensile strength of the weld metal simultaneously. Acknowledgments The authors are grateful for financial support by the National Natural Science Foundation of China (No.51474203) and Key Research Program of the Chinese Academy of Sciences (No. KGZD-EW-XXX-2). The authors also wish to recognize the assistance provided by China First Heavy Machinery Co. Ltd. in the welding process. References [1] F. Huang, J. Q. Wang, E.H. Han, W. Ke, J. Mater. Sci. Technol. 28 (2012) 562‒568. [2] Z. Zhang, J. Wang, E.H. Han, W. Ke, J. Mater. Sci. Technol. 28 (2012) 785‒792. [3] T.Y. Kuo, H.T. Lee, Mater. Sci. Eng. A 338 (2002) 202‒212. [4] M. Casales, V.M. Salinas-Bravo, A. Martinez-Villafane, J.G. Gonzalez-Rodriguez, Mater. Sci. Eng. A 332 10 / 16

Page 10 of 16

(2002) 223‒230. [5] R.S. Dutta, R. Tewari, P.K. De, Corros. Sci. 49 (2007) 303‒318. [6] J. Crum, R. Scarberry, J. Mater. Energy Syst. 4 (1982) 125‒130. [7] S.J. Zinkle, G.S. Was, Acta Mater. 61 (2013) 735‒758. [8] T. Allen, J. Busby, M. Meyer, D. Petti, Mater. Today 13 (2010) 14‒23. [9] H.T. Lee, S.L. Jeng, C.H. Yen, T.Y. Kuo, J. Nucl. Mater. 335 (2004) 59‒69. [10] R. Qin, Z. Duan, G. He, Metall. Mater. Trans. A 44 (2013) 4661‒4670. [11] A.J. Ramirez, J.C. Lippold, Mater. Sci. Eng. A 380 (2004) 259‒271. [12] G.A. Young, T.E. Capobianco, M.A. Penik, B.W. Morris, J.J. McGee, Weld. J. 87 (2008) 31s‒43s. [13] K. Nishimoto, K. Saida, H. Okauchi, Sci. Technol. Weld. Joining 11 (2006) 455‒461. [14] A.J. Ramirez, J.C. Lippold, Mater. Sci. Eng. A 380 (2004) 245‒258. [15] W.L. Mo, S.P. Lu, D.Z. Li, Y.Y. Li, Mater. Sci. Eng. A 582 (2013) 326‒337. [16] W.L. Mo, S.P. Lu, D.Z. Li, Y.Y. Li, J. Mater. Sci. Technol. 29 (2013) 458‒466. [17] K. Nishimoto, K. Saida, H. Okauchi, K. Ohta, Sci. Technol. Weld. Joining 11 (2006) 471‒479. [18] K. Saida, Y. Nomoto, H. Okauchi, H. Ogiwara, K. Nishimoto, Sci. Technol. Weld. Joining 17 (2012) 1‒8. [19] K. Saida, A. Taniguchi, H. Okauchi, H. Ogiwara, K. Nishimoto, Sci. Technol. Weld. Joining 16 (2011) 553‒560. [20] T.E. Capobianco, M.E. Hanson, in KAPL (Knolls Atomic Power Laboratory (KAPL), Niskayuna, NY, 2005. [21] D.J. Lee, Y.S. Kim, Y.T. Shin, E.C. Jeon, S.H. Lee, H.J. Lee, S.K. Lee, J.H. Lee, H.W. Lee, Met. Mater. Int. 16 (2010) 813‒817. [22] A.J. Ramirez, J.C. Lippold, in T. Böllinghaus, H. Herold (Eds.), Hot Cracking Phenomena in Welds, Springer, Berlin Heidelberg, 2005, pp. 19‒41. [23] A.J. Ramirez, C.M. Garzon, in: T. Böllinghaus, H. Herold, C. Cross, J. Lippold (Eds.), Hot Cracking Phenomena in Welds II, Springer, Berlin Heidelberg, 2008, pp. 427‒453. [24] N.E. Nissley, J.C. Lippold, in: Proceedings of the 7Th International Conference on Trends In Welding Research, Callaway Gardens Resort, Pine Mountain, Georgia, USA, 2006, pp. 327‒332. [25] M.G. Collins, A.J. Ramirez, J.C. Lippold, Weld. J. 83 (2004) 39S‒49S. [26] N.E. Nissley, Intermediate Temperature Grain Boundary Embrittlement in Nickel-base Weld Metals, Ph.D. Thesis, The Ohio State University, 2006. 11 / 16

Page 11 of 16

[27] K. Nishimoto, K. Saida, H. Okauchi, K. Ohta, Sci. Technol. Weld. Joining 11 (2006) 462‒470. [28] A.J. Ramirez, C.M. Garzon, in: Hot Cracking Phenomena in Welds, Berlin, Germany, March 5‒6, 2007, pp. 427‒453. [29] S.L. Jeng, H. Lee, W.P. Rehbach, T.Y. Kuo, T.E. Weirich, J.P. Mayer, Mater. Sci. Eng. A 397 (2005) 229‒238. [30] V. Shankar, T.P.S. Gill, A.L.E. Terrance, S.L. Mannan, S. Sundaresan, Metall. Mater. Trans. A 31 (2000) 3109‒3122. [31] R. Vallant, in: T. Böllinghaus, H. Herold (Eds.), Hot Cracking Phenomena in Welds, Springer Berlin Heidelberg, 2005, pp. 141‒164. [32] D. Tytko, P.P. Choi, J. Klower, A. Kostka, G. Inden, D. Raabe, Acta Mater. 60 (2012) 1731‒1740. [33] M. Thuvander, K. Stiller, Mater. Sci. Eng. A 281 (2000) 96‒103. [34] N.L. Richards, M.C. Chaturvedi, Int. Mater. Rev. 45 (2000) 109‒129. [35] D. McLean, Grain Boundaries in Metals, Clarendon Press Oxford, 1957, pp. 340. [36] W. Chen, M.C. Chaturvedi, N.L. Richards, Metall. Mater. Trans. A 32 (2001) 931‒939. [37] X.L. He, M. Djahazi, J.J. Jonas, J. Jackman, Acta Metall. Mater. 39 (1991) 2295‒2308. [38] X.L. He, Y.Y. Chu, J.J. Jonas, Acta Metall. 37 (1989) 2905‒2916. [39] W. Chen, M.C. Chaturvedi, N.L. Richards, G. McMahon, Metall. Mater. Trans. A 29 (1998) 1947‒1954. [40] L. Karlsson, H. Norden, H. Odelius, Acta Metall. 36 (1988) 1‒12. [41] K.R. Vishwakarma, M.C. Chaturvedi, Mater. Sci. Technol. 25 (2009) 351‒360. [42] L. Karlsson, H. Norden, Acta Metall. 36 (1988) 35‒48. [43] X. Hu, Y. Zhu, N. Sheng, X. Ma, Sci. Rep. 4 (2014) 1‒9. [44] X.B. Hu, Y.L. Zhu, X.L. Ma, Acta Mater. 68 (2014) 70‒81. [45] I. Goldfarb, W.D. Kaplan, S. Ariely, M. Bamberger, Philos. Mag. A 72 (1995) 963‒979. [46] S.M. Corporation, www.specialmetalswelding.com. [47] M. Chaturvedi, W. Chen, A. Saranchuk, N. Richards, Superalloys 718, 625, 706 and Various Derivatives (1997) 743‒751. [48] N.E. Nissley, J.C. Lippold, Weld. J. 88 (2009) 131s‒140s. [49] W.L. Mo, S.P. Lu, D.Z. Li, Y.Y. Li, Metall. Mater. Trans. A 45 (2014) 5114‒5126. [50] C.L. White, J.H. Schneibel, R.A. Padgett, Metall. Trans. A 14 (1983) 595‒610. [51] M. Miller, J. Horton, W. Cao, R. Kennedy, J. Phys. IV 6 (1996) 241‒246. 12 / 16

Page 12 of 16

[52] Z. Lei, R. Schlesiger, R. Chellali, D. Baither, G. Schmitz, Mater. Des. 34 (2012) 155‒158. [53] F.F. Noecker, J.N. DuPont, Weld. J. 88 (2009) 62s‒77s. [54] W.L. Mo, X. Zhang, S.P. Lu, D.Z. Li, Y.Y. Li, Acta Metall. Sin. 51 (2014) 230‒238 (in Chinese). [55] B.C. Peng, H.X. Zhang, J. Hong, J.Q. Gao, H.Q. Zhang, Q.J. Wang, J.F. Li, Mater. Sci. Eng. A 528 (2011) 3625‒3629. [56] J.Q. Chen, H. Lu, C. Yu, J.M. Chen, M.L. Zhang, Sci. Technol. Weld. Joining 18 (2013) 346‒353. [57] X. Li, Effects of Misorientation on the Tensile and Fatigue Properties of SRR99 Superalloy Bicrystals at Room Temperature, Master Thesis, Graduate University of Chinese Academy of Sciences, 2012. [58] P. Venkiteswaran, M. Bright, D. Taplin, Mater. Sci. Eng. 11 (1973) 255‒268.

Figure captions: Fig. 1. Schematic diagram of weld joint and tensile specimen. Fig. 2. DDC in the WM-B1. Fig. 3. SIMS images showing the distribution of boron at GB (a) and DDC (b) in WM-B1. Fig. 4. Vertical cross-section of the phase diagram for NiCrFe-7 alloy with B addition calculated by Thermo-Calc software. Fig. 5. DTA test results of B0 (a) and B1 (b) weld metal. Fig. 6. M23C6 morphology: (a) SEM image of WM-B0, (b) SEM image of WM-B1, (c) TEM image of WM-B0, (d) SAED of (c) , and (e) line scan of Ni, Cr, and Fe contents across the M23C6 at the GB. Fig. 7. Effect of boron on the mass fraction of M23C6 (a) and elements in M23(C, B)6 (b) as a function of temperature for NiCrFe-7 alloy calculated by Thermo-Calc software (B0 without B addition, B1 with 46 ppm B addition). Fig. 8. M2B morphology: (a) M2B, (b) amplification of (a), and (c) EDS of (b). Fig. 9. Schematic displaying a quarter of the of simulated electron diffraction pattern (EDP) along the [001]Cb (a), [010]Cb (b), [1 1]C16 (c) and [1

]C16 (d), respectively. The solid square in (c)

represents the pattern resulted from dynamics diffraction. Fig. 10. HRTEM image acquired from single one blocky M2B grain located at the edge of the electron transparent area. (a) and (c) correspond one boarder area of the grain but with different 13 / 16

Page 13 of 16

tilting angles. (b) and (d) correspond to the interior of the grain with different tilting angles. (a) and (b), (c) and (d) correspond to the same tilting angle. Fig. 11. HRTEM displaying the nano-scale intergrowth between C16 and Cb structure with the orientation relationship of [001]C16//[011]Cb and (110)C16//(100)Cb. Fig. 12. Tensile test for the weld metals. Fig. 13. Fractographs: (a) WM-B0, (b) WM-B1, (c) amplification of area c in (a), (d) amplification of area d in (b), (e) amplification of area e in (a), (f) amplification of area f in (b), (g) amplification of area g in (a), and (h) amplification of area h in (b). Fig. 14. Longitudinal section of the rupture for WM-B1.

14 / 16

Page 14 of 16

Tables: Table captions: Table 1 Chemical composition of the filler metals (wt%)

Filler metals

Ti

Al

Mn

C

Si

Fe

Cr

Ni

B ‒

B0

0.61 0.29 0.62 0.023 0.11 9.13 29.80 Bal.

B1

0.60 0.35 0.67 0.026 0.11 8.90 29.69 Bal. 0.0046

Table 2 Statistics of the length and number of DDC in the weld metals

Weld metals Longest crack length (μm) Crack length per section (μm) Crack number per section WM-B0

2432

15363

21.6

WM-B1

3178

57693

96

Table 3 Statistics of the DDC width in the weld metals Crack width Weld metals

100‒150 μm 150‒200 μm >200 μm

WM-B0

0

0

0

WM-B1

12

6

3

15 / 16

Page 15 of 16

16 / 16

Page 16 of 16