1Cr18Ni9 brazed joint

1Cr18Ni9 brazed joint

Journal of Alloys and Compounds 741 (2018) 155e160 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: http:...

2MB Sizes 0 Downloads 21 Views

Journal of Alloys and Compounds 741 (2018) 155e160

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom

Thermal fatigue damage and residual mechanical properties of W-Cu/ Ag-Cu/1Cr18Ni9 brazed joint Chunzhi Xia*, Weiwei Sun, Yi Zhou, Xiangping Xu Provincial Key Lab of Advanced Welding Technology, Jiangsu University of Science and Technology, Zhenjiang, 212003, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 July 2017 Received in revised form 4 January 2018 Accepted 10 January 2018 Available online 12 January 2018

Thermal fatigue properties of W-Cu/1Cr18Ni9 steel brazed joint with Ag-Cu filler metal has been studied in this paper. With the increment of fatigue cycles, fatigue damage of the brazed joint became increasingly serious. There were micro holes and cracks appeared on the interface between brazing seam and base metals, and bending strength decreased from 690 MPa of the original joint to 380 MPa after 200 fatigue cycles. Fracture characteristic of W-Cu/Ag-Cu/1Cr18Ni9 joint changed from ductile fracture of the original joint to mixed ductile-brittle fracture after 200 fatigue cycles under external bending load and internal fatigue damage. The ductile fracture located on the Ag-based solid solution of brazing seam, and the brittle fracture occurred on the W-Cu composite. © 2018 Elsevier B.V. All rights reserved.

Keywords: Ag-Cu brazed joint Thermal fatigue damage Residual mechanical properties Fracture morphology

1. Introduction W-Cu/1Cr18Ni9 steel brazed joint is frequently applied to high temperature and complex environment. Therefore, thermal fatigue property of the joint is an important index to meet the severe requirements. Divertor is one of the most important components in nuclear fusion reactor test device. It is transition region between high temperature plasma and solid structure material [1e3]. At present, W and W alloy are considered as the most suitable plasma facing material, and stainless steel is the most popular structural material in divertor. However, there are great difference in physical properties between plasma materials and structural materials, brazing is considered as one of the most feasible methods to connect them [4e7]. Therefore, the brazed joint between W alloy and stainless steel has gradually become an important component of the divertor. Because the designed joint need to work long time under thermodynamic load, it often require high thermal stress during service and the thermal expansion coefficient between W alloy and stainless steel is quite different [8]. Therefore, thermal fatigue failure has become one of the main failure forms of the brazed joint of W alloy and stainless steel in the divertor. In recent years, considerable interest has been generated in welding and thermal

* Corresponding author. E-mail address: [email protected] (C. Xia). https://doi.org/10.1016/j.jallcom.2018.01.151 0925-8388/© 2018 Elsevier B.V. All rights reserved.

fatigue research of the W-alloy joint [9e12]. In this paper, thermal fatigue property of W-Cu/Ag-Cu/1Cr18Ni9 steel joint were investigated. 2. Materials and methods Our research group have studied the microstructure and properties of vacuum brazed joint of W-Cu composite and 1Cr18Ni9 steel with Ag-Cu filler metal. Process parameters during the vacuum brazing were: brazing temperature T ¼ 855e865  C, holding time t ¼ 30 min, vacuum level superior to 6  103 Pa (optimum technological parameters) [6]. Samples were machined by a linear cutting machine into blocks with sizes of 40 mm  6 mm  5 mm from the brazed joint. A circular hole with a diameter of 5 mm was drilled at the upper end of the 1Cr18Ni9 steel from the brazing seam 15 mm, as shown Fig. 1. The number of thermal fatigue cycles was 0, 100, 150 and 200 times. Thermal fatigue tests were carried out on the thermal fatigue test machine with special clamps. The parameters of thermal fatigue test were as follows: Upper limit temperature qmax ¼ 400  C for 20 min followed by water and cooling for 2min. After thermal fatigue test, the specimens were ground and polished by emery papers, cloths, and diamond grind paste and then etched to reveal microstructure of the brazing seam by mixture solution (HCl: HF: HNO3 ¼ 80 ml: 13 ml: 7 g) for 2e3s. Finally, microstructure feature and fracture morphology of the brazed joint with different cycle times were studied via scanning

156

C. Xia et al. / Journal of Alloys and Compounds 741 (2018) 155e160

Fig. 1. Thermal fatigue specimen.

elector microscope (SEM) and energy dispersive spectrum (EDS), and four-point bending strength of the brazed joint after different cycle times were test using an electronic mechanical testing machine (CMT505). 3. Results and discussion 3.1. Thermal fatigue damage 3.1.1. Microstructure Microstructure of W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint with difference fatigue cycles can be observed in Fig. 2. Previous studied [6] have demonstrated that both brazing seam and substrates of original joint (Fig. 2(a)) formed a good metallurgical bonding with a width about 60 mm. A slight diffusion layer existed in the interface, and no obvious micro cracks or hole defects were observed. A feature area was mainly Cu(Ag, Fe) solid solution and B feature area

was mainly the eutectic Ag(Cu) solid solution. Thermal fatigue defects with difference cycles were listed in Table 1. The length of micro cracks and diameter of holes appeared in the processing of thermal fatigue were used as criterion to show the damage level of different cycles. Subsequently, microstructure of the brazed joints after 80 thermal fatigue cycles was investigated. Without micro cracks or hole defects were observed in the interface of this joint. Compared with the original joint, the microstructure of W-Cu/ Ag-Cu/1Cr18Ni9 steel brazed joint after 100 cycles changed from continuous and dense interface to discontinuous and uneven interface, and even some holes appeared (seen in Fig. 2(b)). Composition of interface between 1Cr18Ni9 steel and brazed seam(C) was analyzed by EDS with a chemical composition of Fe 69.64%, Cu 15.12%, Cr 6.75%, Ag 3.50%, Ni 2.99%. On the basis binary phase diagram of Fe-Cu, main phase in this region was Fe(Cu) solid solution. Therefore, it can be known that no more than 80 cycles of thermal fatigue could be expected in use of this joint. The thermal expansion coefficient of W-Cu, Ag-Cu, and 1Cr18Ni9 steel was list in Table 2. AgeCu and 1Cr18Ni9 steel of the thermal expansion coefficient were taken from literature [13,14], W-Cu was test by thermal mechanical analyzer (TMA 402F3, NETZSCH). Table 2 indicated that the thermal expansion coefficient of all experimental materials increased gradually along with the increment of temperature. The expansion coefficient of 1Cr18Ni9 steel was bigger than that of W-Cu at every temperature value. The expansion coefficient of the Ag-Cu filler metal was between the base materials, which was benefit to relieve the interfacial stress. Distribution of X direction residual stress of the brazing joint was calculated by FEM as shown in Fig. 3, which played an important role in the fatigue defects occurred. It revealed that major tensile and compressive stress was distributed on the 1Cr18Ni9 steel side and W-Cu side, respectively. Fatigue defects was induced by the combined action of residual tensile stress and cyclic thermal fatigue load. During the thermal fatigue test, the joints were subjected to

Fig. 2. Microstructure of the W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint with difference fatigue cycles; (a) Original joint, (b) 100 fatigue cycles, (c) 150 fatigue cycles and (d) 200 fatigue cycles.

C. Xia et al. / Journal of Alloys and Compounds 741 (2018) 155e160

157

Table 1 Thermal fatigue defects with difference cycles. Fatigue cycles

Length of crack or Location Diameter holes/(mm)

0 (original) 80 #1 #2 #3 #4 #5 100 #1 #2 #3 #4 #5 150 #1 #2 #3 #4 #5 200 #1 #2 #3 #4 #5

0 0

No micro cracks or holes found No micro cracks or holes found

5 4 5 2 6 20 15 18 22 25 95 90 101 120 110

Micro holes initialing at interface of 1Cr18Ni9/Ag-Cu

Cracks found at interface of 1Cr18Ni9/ Ag-Cu

Additional micro cracks initialing at interface of W-Cu/Ag-Cu; Lots of cracks paralleling to the interface formed

expand with heat and contract with cold, which resulted in high thermal stress on the joint interface, and the grain boundary weakness occurred at the upper limit temperature of the thermal fatigue test. The dislocation piled up in the weak dislocation slip phase of the junction between brazing seam and base metal, which caused the stress concentration. Meanwhile, there were a small

Table 2 Coefficient of thermal expansion of 1Cr18Ni9 steel, Ag-Cu and W-Cu (106 K1). Materials

20  C

300  C

500  C

800  C

1Cr18Ni9 Ag-Cu W-Cu

18.1 15.1 7.24

18.4 16.2 11.3

19.7 18.8 12.5

26.8 20.5 14.0

Fig. 4. The diagram of microstructure thermal fatigue damage of W-Cu/Ag-Cu/ 1Cr18Ni9 brazed joint.

part of holes for the filler metal spread not enough when the filler metal wetted and filled with the base metal in the brazing process. These holes also caused the stress concentration in the procedure of thermal fatigue test. Moreover, FEM revealed that major tensile and compressive stress were distributed on the 1Cr18Ni9 steel side and W-Cu side, respectively, and the previous study demonstrated the 1Cr18Ni9 steel side interface was weaker than W-Cu side interface [6]. The stress distribution and interface bonding strength played the main roles to fatigue defects initiating along 1Cr18Ni9/Ag-Cu interface firstly. Fe(Cu) solid solution of the junction between brazing seam and 1Cr18Ni9 steel generated a large number of dislocations in continuous action of thermal stress, the grain boundary would separate into micro holes when the stress increased to a certain value, then a large number of dislocations gathered at the grain boundary void, which led to the holes grew up to connect with each other. Microstructure of W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint after 150 cycles can be seen in Fig. 2(c). Compared with the previous joints of 100 cycle, it can be found that obvious micro holes

Fig. 3. Axial stress distribution of steel W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint.

158

C. Xia et al. / Journal of Alloys and Compounds 741 (2018) 155e160

Above all, the variation of microstructural thermal fatigue damage characteristics of W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joints with the increment of fatigue cycles was shown in Fig. 4. The fatigue damage area of the joint was mainly concentrated on the joint interface of the brazing seam and the base metal, in which the fatigue damage on the 1Cr18Ni9 steel side was more serious than that of the W-Cu side.

Fig. 5. Sketch of four-point bending test.

3.1.2. Mechanical properties The residual mechanical properties of the brazed joint with different cycle times was evaluated by using four-point bending strength at room temperature. The sizes of the bending test samples were cut by using a linear cutting machine into blocks with size of 40 mm  6 mm  5 mm and tested on an electronic mechanical testing machine (CMT5205) using a special fixture with a cross-head speed of 0.5 mm/min. Five samples were prepared for the four-point bending test of different cycle times, and a sketch of the four-point bending test was shown in Fig. 5. From a loaddisplacement curve, maximum (s) strength was calculated via the following equation:



3FL bh2

(1)

In this equation, F is the load at the maximum point. L is the span of supporting pins. b (mm) and h (mm) are the specimen width and thickness, respectively. The four-point bending strength of brazing joint was 690 MPa, while the four-point bending strength decreased gradually with the increased of the number of cycles, and after 100, 150 and 200 cycle times the bending strength was 634 MPa, 506 MPa and 380 MPa, respectively (Fig. 6). 3.2. Fracture morphology Fig. 6. Bending strength of brazed joints with different cycles.

existed in the interface of the brazing seam and the 1Cr18Ni9 steel, and some cracks were observed around those holes. With the increment of thermal fatigue cycles, the thermal stress on the interface became greater. The Fe(Cu) solid solution of the junction between 1Cr18Ni9 and brazing seam was formed microscopic cavity at first, and the microscopic cavity grew up under impacting of high thermal stress to form a hole. Meanwhile, the hole further intensified the stress concentration on the joint interface, and induced fatigue cracks appearance at this hole. Even more obvious fatigue cracks with longitudinal distribution were formed on the interface between the brazing seam and the 1Cr18Ni9 steel when W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint underwent after 200 cycles (seen in Fig. 2(d)). The longitudinal distribution of fatigue cracks were not only on the 1Cr18Ni9 steel side but also on the bonding interface of W-Cu composite side, while the thermal fatigue damage in the interface of the 1Cr18Ni9 side was more seriously than that of the W-Cu side. Microscopic cavity formed on the grain boundary of Fe(Cu) solid solution at first in the process of expand with heat and contract with cold, and the microscopic cavity grew up at high thermal stress to form a hole. Subsequently, the hole become crack source to generated fatigue cracks under sustained high stress. Furthermore, the thermal stress in the center of the hole was released for crack initiation and intergranular growth, which led to the hole was no longer grew, and the adjacent holes were connected by cracks propagate along the grain boundaries. The interface of the W-Cu side also produced crack initiation and intergranular growth propagated along the grain boundaries when the thermal stress on the joint surface exceeded the critical value of the crack initiation.

Mechanical properties of the joints were closely related to fracture morphology, so fracture morphology of W-Cu/Ag-Cu/ 1Cr18Ni9 steel brazed joints was observed and analyzed (shown in Table 3). Fracture morphology of the original joint demonstrated that the whole macro section with two different color morphology were dark grey and relatively flat, and a large number of smaller size dimples were densely distributed. The dimple of characteristic area A was Ag(Cu) solid solution and near the dimples was Cu(Ag) solid solution indicated by EDS (Table 4). Under the loading, the Ag(Cu) solid solution with good plasticity formed dimples Ag(Cu). In contrast, the plastic deformation ability of the rich Cu phase was weak, which resulted in the dimples morphology was not obvious. Moreover, the darker surface (characteristic area C) was the fracture surface morphology of 1Cr18Ni9 steel. To sum up, ductile fracture was the main manifestation of rich Ag phase in brazing seam. The joint fracture model was shown in Fig. 7(a). To a certain extent, fracture morphology of the brazed joint has changed after 100 cycles. The section was flat and distinct fracture morphology were not observed. Through further observation of the microscopic section, it can be found that a large number of small dimples were distributed in the micro section, while some of the base metal were torn during the bending test (EDS analysis indicated that the base metal was 1Cr18Ni9 steel). At the same time, several small cracks were observed on the cross section of the base metal, which revealed that the interface between the brazing seam and the 1Cr18Ni9 steel has been obviously fatigue damaged. Therefore, the fracture was formed under the combined action of bending load and fatigue damage after 100 cycles of brazing joint. The fracture mode was mainly ductile fracture of Ag rich phase in the brazing zone, and the joint fracture model was shown in Fig. 7(b).

C. Xia et al. / Journal of Alloys and Compounds 741 (2018) 155e160

159

Table 3 Fracture morphology and fracture diagram with difference fatigue cycles. Cycle times

Macroscopic fracture surface

Microcosmic fracture morphology

0

100

150

200

Table 4 Elements content with different representative areas. Characteristic region

Ag

Cu

Fe

Cr

Ni

W

A B C D E F

80.10 40.05 e 4.55 e 4.62

10.12 52.14 e 3.21 85.10 8.80

9.78 7.81 73.58 69.21 e 75.04

e e 17.78 15.89 e 7.24

e e 8.64 7.14 e 4.30

e e e e 14.90 e

After 150 cycles, fracture morphology of W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joint revealed that the section was flat, and bright compared with former. The microscopic cross section were mainly consisted of dimples and cleavage plane, and the dimples of the cross section morphology was the rich Ag phase in the eutectic structure. A and B areas were W-Cu composite and 1Cr18Ni9 steel revealed by EDS, respectively. Obvious fatigue damaged has taken place on the 1Cr18Ni9 side, and on the W-Cu side fatigue damaged also generated in some degree. The fracture modes was ductile

160

C. Xia et al. / Journal of Alloys and Compounds 741 (2018) 155e160

cycles, the number of holes and cracks increased on the interface between brazing seam and base metal. Further studied indicated that no more than 80 cycles of thermal fatigue could be expected in use of this joint. (2) The mechanical properties of W-Cu/Ag-Cu/1Cr18Ni9 steel brazed joints decreased with the increase of thermal fatigue cycles. The four-point bending strength of brazing joint was 690 MPa, while the four-point bending strength decreases gradually with the increased of the number of cycles, and the bending strength after 100, 150 and 200 cycles was 634 MPa, 506 MPa and 380 MPa, respectively. (3) The W-Cu/Ag-Cu/1Cr18Ni9 steel joints before thermal fatigue test was mainly manifested ductile fracture of rich Ag phase in brazing seam. The fracture was formed under the combined action of bending load and fatigue damage after thermal fatigue test. Moreover, with the number of cycles being increased, the fracture characteristics from ductile fracture of Ag rich phase in brazing seam to mixed mode of ductile fracture of Ag rich phase in brazing seam and cleavage brittle fracture of W-Cu composite.

Acknowledgments This project was supported by the National Natural Science Foundation of China (Grant No.51405205) and the Project Funded by China Postdoctoral Science Foundation (2015M581751) and Postgraduate Practice Innovation Program of Jiangsu Province.

References

Fig. 7. Fracture diagram with difference fatigue cycles.

fracture of rich Ag phase in brazing seam and cleavage brittle fracture of W-Cu composite, and the joint fracture diagram was shown in Fig. 7(c). After 200 cycles, fracture morphology of the W-Cu/Ag-Cu/ 1Cr18Ni9 steel showed the macro section rough. Microcosmic fracture morphology indicated that the section were consist of dimples, cleavage plane and the 1Cr18Ni9 steel section, and the dimples was decreased significantly compared with the former, while the cleavage plane of W-Cu composite and the 1Cr18Ni9 steel section increased, which demonstrated that the fracture position was closer to the base metal zone. Meanwhile, there were more cracks appeared in the cross section of the 1Cr18Ni9 steel. Above all, severe thermal fatigue damage were generated at both W-Cu and 1Cr18Ni9 side. Therefore, the fracture was formed under the combined action of bending load and fatigue damage after 200 cycles of the brazed joint. The fracture mode was characterized by the ductile fracture of a small area of rich Ag phase and the cleavage fracture of a large area of W-Cu composite, as shown in Fig. 7(d). 4. Conclusions (1) The W-Cu/Ag-Cu/1Cr18Ni9 steel joint before thermal fatigue was no obvious micro cracks or hole defects. The brazing zone was mainly composed of eutectic structure rich Ag phase and rich Cu phase. With the increment of fatigue

[1] L. Xu, M. Yan, Y. Xia, J.H. Peng, W. Li, L. Zhang, C.H. Liu, G. Chen, Y. Li, Influence of copper content on the property of CueW alloy prepared by microwave vacuum infiltration sintering, J. Alloys Compd. 592 (2014) 202e206. €schen, G. Pintsuk, Damage and fatigue crack growth of Eurofer [2] J.H. You, T. Ho steel first wall mock-up under cyclic heat flux loads. Part 1: electron beam irradiation tests, Fusion Eng. Des. 89 (2014) 294e301. [3] M.Y. Li, E. Werner, J.H. You, Fracture mechanical analysis of tungsten armor failure of a water-cooled divertor target, Fusion Eng. Des. 89 (2014) 2716e2725. [4] Q.H. Liang, C.Z. Xia, X.P. Xu, J.S. Zou, Research on vacuum brazing of W-Cu composite to stainless steel with Cu-Mn-Co brazing metal, Sci. Eng. Compos. Mater. 22 (2015) 245e250. [5] C.Z. Xia, J. Yang, J.S. Zou, X.P. Xu, Influence of brazing clearance on microstructure and properties of W-Cu/1Cr18Ni9 steel brazed joint with Ni-Cr-Si-B amorphous filler metal, Kov. Mater. 54 (2016) 159e165. [6] C.Z. Xia, Q.H. Liang, J.S. Zou, X.P. Xu, Microstructural investigation of W-Cu composite and 1Cr18Ni9 stainless steel brazed joint with Ag-Cu filler metal, High Temp. Mater. Process. 33 (2013) 293e298. [7] M.F. Wu, P. Zhao, F.J. Wang, Q.X. Hu, Interfacial structure and strength on Ti(C,N) joint using liquid phase diffusion bonding with Ti/Ni interlayer, Mater. Sci. Technol. 30 (2014) 1801e1804. [8] B.A. Kalin, V.T. Fedotov, O.N. Sevrjukov, A.N. Kalashnikov, A.N. Suchkov, A. Moeslang, M. Rohde, Development of brazing foils to join monocrystalline tungsten alloys with ODS-EUROFER steel, J. Nucl. Mater. 367 (2007) 1218e1222. €chl, R. Tivey, [9] V. Barabash, M. Akiba, A. Cardella, I. Mazul, B.C. Odegard Jr., L. Plo G. Vieider, Armor and heat sink materials joining technologies development for ITER plasma facing component, J. Nucl. Mater. 283e287 (2000) 1248e1252. [10] X. Liu, Y. Lian, L. Chen, Tungsten joining with copper alloy and its high heat load performance, J. Nucl. Mater. 455 (2014) 382e386. [11] B.A. Kalin, V.T. Fedotov, O.N. Sevrjukov, A. Moeslang, M. Rohde, Development of rapidly quenched brazing foils to join tungsten alloys with ferritic steel, J. Nucl. Mater. 329e333 (2004) 1544e1548. [12] J.H. Song, K.M. Kim, S.H. Hong, H.T. Kim, S.H. Par, H.K. Park, H.J. Ahn, S.K. Kim, D.W. Lee, High heat flux test of tungsten brazed mock-ups developed for KSTAR divertor, Fusion Eng. Des. 109e111 (2016) 78e81. [13] S.D. Zhu, X.W. Fang, Manual of Mechanical Engineering Material Performance Data, China Machine Press, Beijing, 1995. [14] T. Wang, T. Ivas, W. Lee, C. Leinenbach, J. Zhang, Relief oftheresidualstressesinSi3N4/Invar joints by multi-layered braze structure e experiments and simulation, Ceram. Int. 42 (2016) 7080e7087.