Fracture Toughness, Breakthrough Morphology, Microstructural Analysis of the T2 Copper-45 Steel Welded Joints

The performance and flaws of welded joints are important features that characteristics of the welding material influence. There is significant research activity on the performance and characteristics of welding joint materials. However, the properties of dissimilar welding materials and the cracking problem have not been thoroughly investigated. This investigation focuses on the evaluation and analysis of fracture mechanics, including fracture toughness, microstructural analysis, and crack initiation of T2 copper-45 steel dissimilar welding materials. Standard tensile and three-point bending experiments were performed to calculate the ultimate strength, yield strength, and elastic modulus for fracture toughness. The macro/micro-fracture morphology for tensile fracture and three-point bending fracture were analysed. Based on these investigations, it was concluded that the fracture types were quasi-cleavage and an intergranular brittle fracture mixed model. The deflection of the crack path was discussed and it was determined that the crack was extended along the weld area and tilted towards the T2 copper. Finally, the crack propagation and deflecting direction after the three-point bending test could provide the basis for improvement in the performance of welded joints based on experimental testing parameters and ABAQUS finite element analysis.


Introduction
The welding of dissimilar metals has been an area of active investigations for many years. This objective reflects an overall industrial need of increasing importance that is predicated on the technical and economic potential of the process [1]. Dissimilar metals are welded to achieve physical flexibility, but this practice often results in problems that negatively affect the performance of the weld [2]. Many researchers have investigated the effects of the welding method used for different materials that are characterized by different electrochemical [3], thermal [4], optical [5], and mechanical properties [6][7][8][9], especially dissimilar metals [9][10][11][12][13][14][15][16][17][18][19][20][21]. In general, for conventional joints with two dissimilar metals, the primary concern is the potential effect of the unique properties of the materials on the fusing process and further determines the mechanical behavior of the joint [22]. It has been determined that welding defects are highly related to mechanical properties. In the case of keyhole pores, the formation is controlled by the temperature gradient and surface tensions of the liquid/solid interface [23], when the selective laser melting (SLM) defects quantity increase to a certain proportion, the tensile strength, fatigue life, and hardness of the dissimilar joint are dramatically affected [24].

Parameter Test Method for Joint Property
The test materials were prepared while using commercial welding processes and electron beam welding equipment. The model of electron beam welding machine was SEBW and the manufacturer was Guilin Shichuang vacuum CNC Equipment Co., Ltd. (Guilin, China). After investigation, Table 1 shows the parameter cases of some scholars in the electron beam welding of copper-steel. Table 1. Several cases of electron beam welding parameters of copper-steel.

Parameter Test Method for Joint Property
The test materials were prepared while using commercial welding processes and electron beam welding equipment. The model of electron beam welding machine was SEBW and the manufacturer was Guilin Shichuang vacuum CNC Equipment Co., Ltd. (Guilin, China). After investigation, Table  1 shows the parameter cases of some scholars in the electron beam welding of copper-steel. We preferentially adjust the parameters with reference to Table 1. After actual testing, the electron beam welding parameters of T2 Copper/45 steel dissimilar welding materials were as follows: acceleration voltage 80 kV, electron beam 100 mA, vacuum degree 5 × 10 −2 torr, and welding speed 300 mm/min. The surface of the welding sample and the HAZ had a visible dividing line with the weld, which is clearly shown in Figure 1. Table 2 shows the chemical composition of T2 copper and 45 steel.   Figure 2a, and Figure 2b-e show the microscopic topography of welded area via scanning electron microscopy (SEM) after polishing. The magnifications were 50 and 500 times, respectively. Figure 2b displays the iron and copper ends of the weld area. In Figure 2c-e, some pores, microcracks, and the insufficient welding area can be seen. These defects are the important factors that affect the welded bond quality of T2 copper-45 steel.   Figure 2a, and Figure 2b-e show the microscopic topography of welded area via scanning electron microscopy (SEM) after polishing. The magnifications were 50 and 500 times, respectively. Figure 2b displays the iron and copper ends of the weld area. In Figure 2c-e, some pores, microcracks, and the insufficient welding area can be seen. These defects are the important factors that affect the welded bond quality of T2 copper-45 steel. The tensile specimen was obtained by wire cutting according to GB/T228.1-2010 [46] (Metallic Materials-Tensile Testing-Part 1: Method of testing at room temperature). The ultimate strength, yield strength, and elastic modulus were determined while using the INSRON-8801 Servohydraulic Fatigue Testing System (Instron, Darmstadt, Germany) with a loading rate of 1 mm/min. Figure 3 shows the tensile specimen. The processing of the three-point bending specimen was based on GB/T21143-2014 [30] (unified method of test for determination of quasi-static fracture toughness) and GB/T 28896-2012 [47] (metallic materials-method of test for the determination of quasi-static fracture toughness of welds). The sampling orientation of the fracture surface of the fracture toughness specimen in the weld zone was NQ, as shown in Figure 4a, the maximum fatigue preformed twill force was set according to the smaller value of Equations (1) and (2), the maximum fatigue crack stress was calculated as Ff is 1344.7169 N at the last 1.3 mm or 50% fatigue precracking propagation, and the stress ratio r is 0.5. The tensile specimen was obtained by wire cutting according to GB/T228.1-2010 [46] (Metallic Materials-Tensile Testing-Part 1: Method of testing at room temperature). The ultimate strength, yield strength, and elastic modulus were determined while using the INSRON-8801 Servohydraulic Fatigue Testing System (Instron, Darmstadt, Germany) with a loading rate of 1 mm/min. Figure 3 shows the tensile specimen. The tensile specimen was obtained by wire cutting according to GB/T228.1-2010 [46] (Metallic Materials-Tensile Testing-Part 1: Method of testing at room temperature). The ultimate strength, yield strength, and elastic modulus were determined while using the INSRON-8801 Servohydraulic Fatigue Testing System (Instron, Darmstadt, Germany) with a loading rate of 1 mm/min. Figure 3 shows the tensile specimen. The processing of the three-point bending specimen was based on GB/T21143-2014 [30] (unified method of test for determination of quasi-static fracture toughness) and GB/T 28896-2012 [47] (metallic materials-method of test for the determination of quasi-static fracture toughness of welds). The sampling orientation of the fracture surface of the fracture toughness specimen in the weld zone was NQ, as shown in Figure 4a, the maximum fatigue preformed twill force was set according to the smaller value of Equations (1) and (2), the maximum fatigue crack stress was calculated as Ff is 1344.7169 N at the last 1.3 mm or 50% fatigue precracking propagation, and the stress ratio r is 0.5. The processing of the three-point bending specimen was based on GB/T21143-2014 [30] (unified method of test for determination of quasi-static fracture toughness) and GB/T 28896-2012 [47] (metallic materials-method of test for the determination of quasi-static fracture toughness of welds). The sampling orientation of the fracture surface of the fracture toughness specimen in the weld zone was NQ, as shown in Figure 4a, the maximum fatigue preformed twill force was set according to the smaller  (1) and (2), the maximum fatigue crack stress was calculated as F f is 1344.7169 N at the last 1.3 mm or 50% fatigue precracking propagation, and the stress ratio r is 0.5.
Materials 2020, 13, x FOR PEER REVIEW 5 of 15 ( ) In the preceding Equations (1) and (2), the dimensional coefficient ξ is 1.6 × 10 −4 m 1/2 , B is the sample thickness that is shown in Figure 4b, W is the width of the specimen, BN is the net thickness of the specimen and B, BN, W are 13 mm; the span S is 52 mm, the initial crack length a0 is 6 mm, the stress intensity factor coefficient g(a0/W) is 2.29; E is the elastic modulus; and, Rp0.2 is the specified plastic elongation strength of the material in the vertical crack plane 0.2% at the test temperature.
The fatigue crack was prepared while using the constant load method. After this process, the fatigue precracking of the three specimens was: 2.02, 1.96, and 2.04 mm. Figure 5 shows the final specimen of three-point bending.  In the preceding Equations (1) and (2), the dimensional coefficient ξ is 1.6 × 10 −4 m 1/2 , B is the sample thickness that is shown in Figure 4b, W is the width of the specimen, B N is the net thickness of the specimen and B, B N , W are 13 mm; the span S is 52 mm, the initial crack length a 0 is 6 mm, the stress intensity factor coefficient g(a 0 /W) is 2.29; E is the elastic modulus; and, R p0.2 is the specified plastic elongation strength of the material in the vertical crack plane 0.2% at the test temperature.
The fatigue crack was prepared while using the constant load method. After this process, the fatigue precracking of the three specimens was: 2.02, 1.96, and 2.04 mm. Figure 5 shows the final specimen of three-point bending.  ( ) In the preceding Equations (1) and (2), the dimensional coefficient ξ is 1.6 × 10 −4 m 1/2 , B is the sample thickness that is shown in Figure 4b, W is the width of the specimen, BN is the net thickness of the specimen and B, BN, W are 13 mm; the span S is 52 mm, the initial crack length a0 is 6 mm, the stress intensity factor coefficient g(a0/W) is 2.29; E is the elastic modulus; and, Rp0.2 is the specified plastic elongation strength of the material in the vertical crack plane 0.2% at the test temperature.
The fatigue crack was prepared while using the constant load method. After this process, the fatigue precracking of the three specimens was: 2.02, 1.96, and 2.04 mm. Figure 5 shows the final specimen of three-point bending.  Table 3 shows the performance parameters were obtained by standard tensile tests and the results. The displacement-force curve (P-V curve) of the notch opening was obtained based on the three-point bending test that is shown in Figure 6.  Table 3 shows the performance parameters were obtained by standard tensile tests and the results. The displacement-force curve (P-V curve) of the notch opening was obtained based on the three-point bending test that is shown in Figure 6.   After the P-V curve of the three-point bending specimen was shifted, the value FQ of the three specimens was 3286.569, 2727.193, and 3581.864 N.

Characterization Results of Joint Property Parameters
The judgment basis is as follows.
where FQ is the maximum force and Fmax is the maximum force that the specimen can withstand. Given that Fmax/FQ1, Fmax/FQ2, and Fmax/FQ3 are greater than 1.1, Kmax (conditional value of KIC) was calculated while using Equation (4).
The judgment on plane strain fracture toughness KIC is represented, as follows. After the P-V curve of the three-point bending specimen was shifted, the value F Q of the three specimens was 3286.569, 2727.193, and 3581.864 N.
The judgment basis is as follows.
where F Q is the maximum force and F max is the maximum force that the specimen can withstand. Given that F max /F Q1 , F max /F Q2 , and F max /F Q3 are greater than 1.1, K max (conditional value of K IC ) was calculated while using Equation (4).
The judgment on plane strain fracture toughness K IC is represented, as follows. where K Q can be acquired from the three-point bending test, (R p0.2 ) e is the plastic extension strength corresponding to the bias 0.2% at the test temperature, and (R p0.2 ) p is the plastic elongation corresponding to the fatigue precracking offset 0.2%. After the aforementioned judgement, a 0 is the initial crack length, W − a 0 is the difference between the sample width and the initial crack, and K f is the maximum value of the stress intensity factor in the final stage of the prepared fatigue crack. Table 4 presents these parameters. According to results from the data that are contained in Table 4 and Equation (9), the fracture toughness of the welded joint calculated in the three-point bending test is 6.027 MP·m 1/2 .

Fracture Analysis of Tensile Test
After tensile testing, the macroscopic fracture area of the specimen is shown in Figures 7 and 8. The specimen breaks in the weld zone and crack propagation was biased towards the copper interface. The copper can be observed on the fracture surface, which was partially attached to the ends.
where KQ can be acquired from the three-point bending test, (Rp0.2)e is the plastic extension strength corresponding to the bias 0.2% at the test temperature, and (Rp0.2)p is the plastic elongation corresponding to the fatigue precracking offset 0.2%. After the aforementioned judgement, a0 is the initial crack length, W − a0 is the difference between the sample width and the initial crack, and Kf is the maximum value of the stress intensity factor in the final stage of the prepared fatigue crack. Table 4 presents these parameters. According to results from the data that are contained in Table 4 and Equation (9), the fracture toughness of the welded joint calculated in the three-point bending test is 6.027 MP·m 1/2 .

Fracture Analysis of Tensile Test
After tensile testing, the macroscopic fracture area of the specimen is shown in Figures 7 and 8. The specimen breaks in the weld zone and crack propagation was biased towards the copper interface. The copper can be observed on the fracture surface, which was partially attached to the ends.  The microscopic topography of the tensile fracture of the T2 copper/45 steel dissimilar welding materials is shown in Figure 9 via SEM at a magnification of 1000 times for each specimen two-fracture end face, according to the order of the macroscopic fracture morphology in the immediately preceding figures. The microscopic topography of the tensile fracture of the T2 copper/45 steel dissimilar welding materials is shown in Figure 9 via SEM at a magnification of 1000 times for each specimen twofracture end face, according to the order of the macroscopic fracture morphology in the immediately preceding figures. The macroscopic shape of specimen 1 had obvious gloss and irregular geometry, and the shiny surface of the fracture was almost perpendicular to the normal stress, which is associated with the brittle fracture characteristic; and, significant grain-brittle fracture characteristics at the microscopic level. There was a network structure after fracture due to an external force, which was a relatively obvious network brittle phase, as shown in Figure 9b. The reason for this fracture was the brittle precipitation phase on the grain boundary, which results in the formation of a continuous carbide network by allotropes of iron during electron beam welding, which led to a thin layer of brittle fracture splitting.
Brittle fracture also characterized the macroscopic fracture of specimen 2, which appeared as herringbone and radial patterns at the fracture with a shiny surface. There were fluvial, blocky, and spherical structures in the microscopic topography with cleavage steps and tearing ribs, which exhibited the microscopic features of crystal brittleness and cleavage fracture [41,48]. The microscopic topography of the tensile fracture of the T2 copper/45 steel dissimilar welding materials is shown in Figure 9 via SEM at a magnification of 1000 times for each specimen twofracture end face, according to the order of the macroscopic fracture morphology in the immediately preceding figures. The macroscopic shape of specimen 1 had obvious gloss and irregular geometry, and the shiny surface of the fracture was almost perpendicular to the normal stress, which is associated with the brittle fracture characteristic; and, significant grain-brittle fracture characteristics at the microscopic level. There was a network structure after fracture due to an external force, which was a relatively obvious network brittle phase, as shown in Figure 9b. The reason for this fracture was the brittle precipitation phase on the grain boundary, which results in the formation of a continuous carbide network by allotropes of iron during electron beam welding, which led to a thin layer of brittle fracture splitting.
Brittle fracture also characterized the macroscopic fracture of specimen 2, which appeared as herringbone and radial patterns at the fracture with a shiny surface. There were fluvial, blocky, and spherical structures in the microscopic topography with cleavage steps and tearing ribs, which exhibited the microscopic features of crystal brittleness and cleavage fracture [41,48]. The macroscopic shape of specimen 1 had obvious gloss and irregular geometry, and the shiny surface of the fracture was almost perpendicular to the normal stress, which is associated with the brittle fracture characteristic; and, significant grain-brittle fracture characteristics at the microscopic level. There was a network structure after fracture due to an external force, which was a relatively obvious network brittle phase, as shown in Figure 9b. The reason for this fracture was the brittle precipitation phase on the grain boundary, which results in the formation of a continuous carbide network by allotropes of iron during electron beam welding, which led to a thin layer of brittle fracture splitting.
Brittle fracture also characterized the macroscopic fracture of specimen 2, which appeared as herringbone and radial patterns at the fracture with a shiny surface. There were fluvial, blocky, and spherical structures in the microscopic topography with cleavage steps and tearing ribs, which exhibited the microscopic features of crystal brittleness and cleavage fracture [41,48].
A few flaky smooth surfaces existed in the macroscopic fracture of specimen 3 and the entire fracture surface was relatively flat. The cleavage characteristics of trapezoidal and river patterns also appeared in the microscopic morphology, with tiny cleavage steps and tearing ribs that are associated with the cleavage fracture. The defects in the weld area and the impurities of the welding material caused this microscopic appearance.

Fracture Analysis of Three-Point Bending Test
The macroscopic fracture surface is shown in Figure 10 after the three-point bending test and the prepared breach prepared fatigue crack and crack extension zone can be observed. A few flaky smooth surfaces existed in the macroscopic fracture of specimen 3 and the entire fracture surface was relatively flat. The cleavage characteristics of trapezoidal and river patterns also appeared in the microscopic morphology, with tiny cleavage steps and tearing ribs that are associated with the cleavage fracture. The defects in the weld area and the impurities of the welding material caused this microscopic appearance.

Fracture Analysis of Three-Point Bending Test
The macroscopic fracture surface is shown in Figure 10 after the three-point bending test and the prepared breach prepared fatigue crack and crack extension zone can be observed. The purplish-red hue gradually deepens from the top to the bottom in the crack extension zone and the copper attached to the fracture surface gradually increased. It was known that the crack gradually deflected along the copper thereby tearing copper that was attached to the surface, as shown in Figure 11. The purplish-red hue gradually deepens from the top to the bottom in the crack extension zone and the copper attached to the fracture surface gradually increased. It was known that the crack gradually deflected along the copper thereby tearing copper that was attached to the surface, as shown in Figure 11.
The three-point bending fracture was observed at 1000 times magnification while using SEM. As shown in Figure 12, these fractures in the macroscopic image have an obvious shiny surface and irregular geometry. The fluvial, blocky, and spherical structures were readily apparent in the microscopic topography with the cleavage steps and tearing ribs distributed, therefore, it was a typically mixed mode of brittle intergranular and quasi-cleavage fracture. The three-point bending fracture was observed at 1000 times magnification while using SEM. As shown in Figure 12, these fractures in the macroscopic image have an obvious shiny surface and irregular geometry. The fluvial, blocky, and spherical structures were readily apparent in the microscopic topography with the cleavage steps and tearing ribs distributed, therefore, it was a typically mixed mode of brittle intergranular and quasi-cleavage fracture.  The three-point bending fracture was observed at 1000 times magnification while using SEM. As shown in Figure 12, these fractures in the macroscopic image have an obvious shiny surface and irregular geometry. The fluvial, blocky, and spherical structures were readily apparent in the microscopic topography with the cleavage steps and tearing ribs distributed, therefore, it was a typically mixed mode of brittle intergranular and quasi-cleavage fracture.

Analysis of Crack Propagation Direction
The crack propagation path deflection of the dissimilar metal welding materials always deflects to the low strength material region [19]. The attached copper on the fracture area was caused by the deflection tear of the fracture path based on the aforementioned tensile test fracture morphology. Figure 13a shows a schematic diagram of the crack deflection of the standard tensile specimen, which was similar to the three-point bending test of T2 copper/45 steel dissimilar welding materials deflection path. This resulted in the phenomenon of stepwise reduction of the resistance to fracture due to the difference in the toughness between the weld area, HAZ and base metal in the electron beam welding process. Based on the three-point bending test of T2 copper/45 steel dissimilar welding materials cracking failure, the crack path was deflected due to the difference in the toughness, subject to factors, such as pores, micro-cracks in the weld area, crack deflection to T2 copper, as shown in the schematic diagram in Figure 13b. terials 2020, 13, x FOR PEER REVIEW 11 of .

Analysis of Crack Propagation Direction
The crack propagation path deflection of the dissimilar metal welding materials always deflec the low strength material region [19]. The attached copper on the fracture area was caused by th flection tear of the fracture path based on the aforementioned tensile test fracture morpholog gure 13a shows a schematic diagram of the crack deflection of the standard tensile specimen, whic as similar to the three-point bending test of T2 copper/45 steel dissimilar welding materia flection path. This resulted in the phenomenon of stepwise reduction of the resistance to fractu e to the difference in the toughness between the weld area, HAZ and base metal in the electro am welding process. Based on the three-point bending test of T2 copper/45 steel dissimilar weldin aterials cracking failure, the crack path was deflected due to the difference in the toughness, subje factors, such as pores, micro-cracks in the weld area, crack deflection to T2 copper, as shown in th hematic diagram in Figure 13b. Therefore, the crack propagation path of T2 copper/45 steel dissimilar welding materials alway flected to the low strength side of the T2 copper because the strength mismatch between thre gions was comparatively large and the toughness decreases from the weld area to HAZ and the e base metal.
According to the test parameters and conditions, the simulation of crack propagation of th ree-point bending test was performed by ABAQUS, and the results are shown in Figure 14. Wi e increase of the expansion step, the crack expanded along the weld seam position and it w itially biased toward the T2 copper. The crack expanded along the junction until the specimen brok hen the crack extended to the junction of the weld seam area and T2 copper. It is clear that th BAQUS simulation results are consistent with these observations, as represented in Figure 11. Therefore, the crack propagation path of T2 copper/45 steel dissimilar welding materials always deflected to the low strength side of the T2 copper because the strength mismatch between three regions was comparatively large and the toughness decreases from the weld area to HAZ and then the base metal.
According to the test parameters and conditions, the simulation of crack propagation of the three-point bending test was performed by ABAQUS, and the results are shown in Figure 14. With the increase of the expansion step, the crack expanded along the weld seam position and it was initially biased toward the T2 copper. The crack expanded along the junction until the specimen broke when the crack extended to the junction of the weld seam area and T2 copper. It is clear that the ABAQUS simulation results are consistent with these observations, as represented in Figure 11.

Conclusions
For the T2 copper-45 steel dissimilar welding materials that were made by electron beam welding, the joint strength, microstructural analysis, and crack initiation were explored. Based on the standard tensile test, the ultimate strength of T2 copper/45 steel dissimilar welding materials were determined to be 93.73 MPa, the yield strength was 75.37 MPa, and the elastic modulus was 108.86 GPa. It can be seen that the mechanical properties of the weld area are significantly different from those of copper and steel, which causes the strength mismatch between three regions. Through the three-point bending test, the fracture toughness was determined to be 6.027 MPa·m 1/2 , which was lower than that of pure copper (approximately 8 MPa·m 1/2 -10 MPa·m 1/2 ) [49]. This is due to welding defects in the weld area. Some pores and microcracks were found in SEM micro-morphology of the welded area, which directly leads to the reduction of the mechanical properties. Weld defects indicate that, in practical application, the electron beam welding process needs to be optimized, or more suitable welding methods need to be found.
The SEM micro-morphology fracture surface of three-point bending specimen shows that the fracture type was a mixed mode of brittle intergranular and quasi-cleavage fracture. The observation results of macroscopic crack propagation of three-point bending specimen were consistent with the theoretical and ABAQUS analysis, it was concluded that the cracking path was extended along the weld area and biased towards the T2 copper. Moreover, the strength of mismatch and toughness reduction controlled the deflection.

Conclusions
For the T2 copper-45 steel dissimilar welding materials that were made by electron beam welding, the joint strength, microstructural analysis, and crack initiation were explored. Based on the standard tensile test, the ultimate strength of T2 copper/45 steel dissimilar welding materials were determined to be 93.73 MPa, the yield strength was 75.37 MPa, and the elastic modulus was 108.86 GPa. It can be seen that the mechanical properties of the weld area are significantly different from those of copper and steel, which causes the strength mismatch between three regions. Through the three-point bending test, the fracture toughness was determined to be 6.027 MPa·m 1/2 , which was lower than that of pure copper (approximately 8 MPa·m 1/2 -10 MPa·m 1/2 ) [49]. This is due to welding defects in the weld area. Some pores and microcracks were found in SEM micro-morphology of the welded area, which directly leads to the reduction of the mechanical properties. Weld defects indicate that, in practical application, the electron beam welding process needs to be optimized, or more suitable welding methods need to be found.
The SEM micro-morphology fracture surface of three-point bending specimen shows that the fracture type was a mixed mode of brittle intergranular and quasi-cleavage fracture. The observation results of macroscopic crack propagation of three-point bending specimen were consistent with the theoretical and ABAQUS analysis, it was concluded that the cracking path was extended along the weld area and biased towards the T2 copper. Moreover, the strength of mismatch and toughness reduction controlled the deflection.

Conflicts of Interest:
The authors declare no conflict of interest.