Experimental Study on FGH95 Superalloy Turbine Disk Joint Material by Oblique Laser Shock Processing

: The FGH95 superalloy used for turbine disk manufacturing was strengthened by the oblique laser shock processing (OLSP). The laser energy, beam diameter, and number of impacts were selected as the test factors, and the three-factor three-level oblique laser shock processing orthogonal test was carried out. Based on the analysis of variance and range of the surface residual stress, microhardness, roughness, tensile strength, and yield strength of FGH95 superalloy after LSP, the factor level combination of relative best comprehensive performance was obtained. The results showed that, within the 10% conﬁdence level, the order of inﬂuencing factors with signiﬁcant difference is: laser energy > number of impacts > beam diameter. Compared with the unimpacted sample, the microhardness of the material surface was increased by about 25% after OLSP. With the increase of laser energy, the dimple distribution of the sample was more uniform, smaller in size, and shallower in depth. However, the thickness of the affected layer was very low, and there was no signiﬁcant effect on the tensile properties. In general, and under the test conditions selected in this study, the combination of 8 Joule laser energy, 3 mm beam diameter, and three impacts can be selected to obtain the best comprehensive performance. The results of this paper provide a reference for the OLSP of the FGH95 superalloy turbine disk and other aero engine parts.


Introduction
As the most widely used structural materials in aircraft and aerospace fields, nickelbased superalloys have been widely used due to their excellent tensile and creep properties and high temperature stability after long exposure [1]. The FGH95 alloy is the first generation of precipitation-strengthened high tensile nickel-based powder superalloys, specifically developed for aero engines in China. It is based on the gamma phase (γ-phase), and the content of the gamma prime (γ') enhanced phase accounts for 50-55% of the total volume. It has high thermal stability at 650 • C and can maintain high strength. FGH95 alloy has been widely used in many essential parts of aero engine turbine disks, turboshafts, and turbine disk baffle bearings by alternating cyclic load under a high-temperature working environment [2]. With the required continuous improvements in the working performance of the new generations of aero engines, the various performance indicators of FGH95 superalloys have put forward higher requirements.
In order to improve the performance of the FGH95 superalloy, several scholars have carried out numerous studies using various processes such as powder pretreatment, alloy forming, and heat treatment. The FGH95 superalloy has been studied and analyzed

Objectives and Scope
The main objective was to evaluate the processing of FGH95 superalloy used for turbine disk manufacturing. An orthogonal test was carried out, and variance analysis and range analysis were carried out on the measured index data. Figure 1 shows the complex shape of the mortise area of the turbine disks and turbine blade, in which the structure of the mortise tooth and groove is narrow. Typical problems such as poor beam accessibility and optical path interference occur when the laser is impacted perpendicular to the workpiece surface in the actual machining process, which cannot achieve the strengthening effect. Therefore, the oblique incidence of the laser beam was used to strengthen the impact, to explore the influence of the LSP parameters on the properties of FGH95 superalloy materials, and to optimize the LSP parameters [15]. This provides a reference for the practical processing test of FGH95 superalloy turbine disk by OLSP.
Metals 2021, 11, x FOR PEER REVIEW 3 of 16 structure of the mortise tooth and groove is narrow. Typical problems such as poor beam accessibility and optical path interference occur when the laser is impacted perpendicular to the workpiece surface in the actual machining process, which cannot achieve the strengthening effect. Therefore, the oblique incidence of the laser beam was used to strengthen the impact, to explore the influence of the LSP parameters on the properties of FGH95 superalloy materials, and to optimize the LSP parameters [15]. This provides a reference for the practical processing test of FGH95 superalloy turbine disk by OLSP.

Experimental Materials
The experimental materials used in this study are FGH95 powder metallurgical nickel-based superalloys with the specific chemical compositions shown in Table 1 [16].

Experimental Treatment and Arrangement
First, the FGH95 superalloy material was machined into metal tensile standard samples using the wire cutting process. The dimensions are shown in Figure 2. The surface of the samples to be machined was polished with silicon carbide (SiC) abrasive paper of 240# to 2000# then polished with diamond polish and wool felt polishing cloth to make the surface roughness as uniform as possible. Finally, the sample was inserted into acetone solution for ultrasonic cleaning to remove oil and dirt and then dried with a hairdryer in cold air.

Experimental Materials
The experimental materials used in this study are FGH95 powder metallurgical nickelbased superalloys with the specific chemical compositions shown in Table 1 [16].

Experimental Treatment and Arrangement
First, the FGH95 superalloy material was machined into metal tensile standard samples using the wire cutting process. The dimensions are shown in Figure 2. The surface of the samples to be machined was polished with silicon carbide (SiC) abrasive paper of 240# to 2000# then polished with diamond polish and wool felt polishing cloth to make the surface roughness as uniform as possible. Finally, the sample was inserted into acetone solution for ultrasonic cleaning to remove oil and dirt and then dried with a hairdryer in cold air. Metals 2021, 11, x FOR PEER REVIEW 4 of 16 The laser shock strengthening equipment is the PROCUDO200 laser shock peening system designed and manufactured by LSP Technologies in the United States. As shown in Figure 2, the processing area was 50 mm × 12.5 mm, the ellipse represents the beam area of each impact, and the processing route was S-shaped to ensure the uniform distribution and comprehensive coverage of residual stress in the effective area of the sample. For the sample of multiple impacts, and after each complete route, the sample was re-pasted with black tape to ensure the surface of the sample was not directly ablated by the laser. After the shocking, the samples were ultrasonically cleaned with anhydrous ethanol to remove the residual black tape and any other impurities on the surface.
During the oblique laser shock processing (OLSP), the confining layer was 1 to 2 mm in thickness of flowing deionized water, which extends the expansion time of plasma and increases the amplitude and duration of the impact pressure. The absorption layer was 1 mm thick black tape for laser shock strengthening produced by 3M company in the United States, which can absorb the laser energy and protect the sample surface from ablation simultaneously. The schematic diagram of the OLSP is shown in Figure 3a, and the effect of the sample covered with the black tape after shocking is shown in  The laser shock strengthening equipment is the PROCUDO200 laser shock peening system designed and manufactured by LSP Technologies in the United States. As shown in Figure 2, the processing area was 50 mm × 12.5 mm, the ellipse represents the beam area of each impact, and the processing route was S-shaped to ensure the uniform distribution and comprehensive coverage of residual stress in the effective area of the sample. For the sample of multiple impacts, and after each complete route, the sample was re-pasted with black tape to ensure the surface of the sample was not directly ablated by the laser. After the shocking, the samples were ultrasonically cleaned with anhydrous ethanol to remove the residual black tape and any other impurities on the surface.
During the oblique laser shock processing (OLSP), the confining layer was 1 to 2 mm in thickness of flowing deionized water, which extends the expansion time of plasma and increases the amplitude and duration of the impact pressure. The absorption layer was 1 mm thick black tape for laser shock strengthening produced by 3M company in the United States, which can absorb the laser energy and protect the sample surface from ablation simultaneously. The schematic diagram of the OLSP is shown in Figure 3a, and the effect of the sample covered with the black tape after shocking is shown in Figure 3b. The laser shock strengthening equipment is the PROCUDO200 laser shock peening system designed and manufactured by LSP Technologies in the United States. As shown in Figure 2, the processing area was 50 mm × 12.5 mm, the ellipse represents the beam area of each impact, and the processing route was S-shaped to ensure the uniform distribution and comprehensive coverage of residual stress in the effective area of the sample. For the sample of multiple impacts, and after each complete route, the sample was re-pasted with black tape to ensure the surface of the sample was not directly ablated by the laser. After the shocking, the samples were ultrasonically cleaned with anhydrous ethanol to remove the residual black tape and any other impurities on the surface.
During the oblique laser shock processing (OLSP), the confining layer was 1 to 2 mm in thickness of flowing deionized water, which extends the expansion time of plasma and increases the amplitude and duration of the impact pressure. The absorption layer was 1 mm thick black tape for laser shock strengthening produced by 3M company in the United States, which can absorb the laser energy and protect the sample surface from ablation simultaneously. The schematic diagram of the OLSP is shown in Figure 3a   In this study, and in order to explore the effect of the selected test factors on the index, the OLSP parameters were set as fixed values, with the exception of the test factors. In order to ensure the uniform distribution of residual stress on the sample surface, and also to make sure that the residual stress between the beam areas did not cancel each other, the laser beam lap rate was set to 30%, the laser pulse width was set to 20 ns, the laser wavelength was set to 1064 nm, and the working frequency was set to 20 Hz. The larger the oblique incident angle of the laser beam, the smaller the laser power density received by the material surface and the less noticeable strengthening effect. The oblique incident angle of the laser beam was very small and could not be distinguished from the vertical impact. The included angle between the laser beam and the normal line of the sample surface was set to 30°. The three parameters of laser energy, spot diameter, and number of impacts were selected as test factors through the literature [9][10][11], and three levels were selected for OLSP tests on FGH95 superalloy. The test factors and corresponding level values are shown in Table 2. Due to the high cost and long test period, the orthogonal test was adopted in this study to save manpower and material resources and avoid time-consuming procedures by using the comprehensive test. On the premise of ensuring a small error in the test, the test workload can be greatly reduced and the reasonable and effective arrangement of the test factors can be made. In addition, the variance analysis can be used to process the test data to obtain the influence degree of each factor on each index, so as to obtain the optimal combination to achieve the purpose of the test [17]. Therefore, the three-factor three-level oblique angle laser shock processing orthogonal test was carried out (L 9 (3 3 )), and the orthogonal test schedule is shown in Table 3. The interaction between factors was not considered in this test. In this study, and in order to explore the effect of the selected test factors on the index, the OLSP parameters were set as fixed values, with the exception of the test factors. In order to ensure the uniform distribution of residual stress on the sample surface, and also to make sure that the residual stress between the beam areas did not cancel each other, the laser beam lap rate was set to 30%, the laser pulse width was set to 20 ns, the laser wavelength was set to 1064 nm, and the working frequency was set to 20 Hz. The larger the oblique incident angle of the laser beam, the smaller the laser power density received by the material surface and the less noticeable strengthening effect. The oblique incident angle of the laser beam was very small and could not be distinguished from the vertical impact. The included angle between the laser beam and the normal line of the sample surface was set to 30 • . The three parameters of laser energy, spot diameter, and number of impacts were selected as test factors through the literature [9][10][11], and three levels were selected for OLSP tests on FGH95 superalloy. The test factors and corresponding level values are shown in Table 2. Due to the high cost and long test period, the orthogonal test was adopted in this study to save manpower and material resources and avoid time-consuming procedures by using the comprehensive test. On the premise of ensuring a small error in the test, the test workload can be greatly reduced and the reasonable and effective arrangement of the test factors can be made. In addition, the variance analysis can be used to process the test data to obtain the influence degree of each factor on each index, so as to obtain the optimal combination to achieve the purpose of the test [17]. Therefore, the three-factor three-level oblique angle laser shock processing orthogonal test was carried out (L 9 (3 3 )), and the orthogonal test schedule is shown in Table 3. The interaction between factors was not considered in this test.

Index Measurement and Data Processing
The residual stress on the sample's surface was measured using an AstaTech XL-640 X-ray stress tester. The resolution of the X-ray detector can reach 0.04 • , and the average error of the stress obtained is within ±14 MPa. The Mahr XT20 probe surface profiler was used to measure the surface roughness of the samples. The Mahr XT20 has a relative error of less than 3%, which can obtain abundant surface profile data. The Shanghai temin HVS-1000Z microhardness tester was used to measure the surface microhardness of the samples. The measuring range of the HVS-1000Z is 5 to 3000 HV, with a value error of less than 3% and a minimum measuring unit of 0.05 µm. The Macro GHS50K microcomputercontrolled electronic universal testing machine was used to measure the tensile strength, yield strength, and elongation of the samples. The maximum test force of the Macro GHS50K is 100 KN, the test force error is within 0.5%, and the resolution is ±1/300,000 of the maximum test force. The deformation error is within 0.5%, and the resolution is up to 1/300,000. The Zeiss optical microscope was used to observe the grains and measure the grain size after the corrosion of the samples. The German Zeiss microscope uses a coded 5× objective turntable with magnifications from 2× to 250×. The tensile fracture morphology of the samples was observed using an S-3400N (II) type tungsten filament bench scanning electron microscope produced by Hitachi. The acceleration voltage of the S-3400N (II) is 3 to 30 kV, the resolution of the secondary electron image is 3 to 10.0 nm, and the magnification is 10 to 100,000 times.
The test results were analyzed by variance analysis and range analysis using Data Processing System (DPS v7.03) data analysis software to obtain the influence rule of each factor on each index. The surface roughness of the impact area and the average grain size of the material surface layer should be as small as possible. In addition, the larger the absolute value of the other indicators, the better the performance of the sample, and the better the strengthening effect. The statistics of the test results are shown in Table 4.

Variance Analysis
In order to compare the results of the different indicators under the condition of various factors, the variance of each factor under different performance indicators was first compared, and then the F-test was used to judge the significance. The results are shown in Table 5. In order to visually and clearly represent the significant difference between the indicators under each factor condition, a Venn diagram was adopted to supplement the expression and presentation, as shown in Figure 4.

Range Analysis
The residual stress, surface roughness, microhardness, yield strength, average grain size, and other performance indicators with significance within the 10% confidence level were selected for the range analysis, and the analysis results are shown in Table 6. The results showed that, for the indicators with significant differences, the primary and secondary orders of the factors were as follows: • For residual stress and microhardness, the laser energy is larger than the beam diameter, which is larger than number of impacts. The influence of laser energy was very clear and the range of residual stress and microhardness were 27.69 and 31.09, which was much larger than the other two factors, respectively.

•
For surface roughness and average grain size, the laser energy is larger than the number of impacts, which is larger than the beam diameter. The influence degree of number of impacts and beam diameter was very close. The range of surface roughness was 2.21 and 1.91, and the range of average grain size was 1.33 and 1.23, re- Results of the variance analysis showed that the differences in residual stress, surface roughness, microhardness, and average grain size were significant within the 5% confidence level for different laser energy. For different beam diameters, the difference of residual stress was significant within the 5% confidence level, while the difference of microhardness, surface roughness, and average grain size was significant within the 10% confidence level. For different numbers of impacts, the difference of residual stress was significant within the 5% confidence level, and the difference of microhardness, surface roughness, yield strength, and average grain size was significant within the 10% confidence level, as shown in Table 5 and Figure 4.
Within the 10% confidence level, there was no significant difference in tensile strength, yield strength, and elongation under different levels of other factors, with the exception of the yield strength that was only significantly different under different numbers of impacts. This might be due to the very small value range of laser energy and beam diameter or the because the sample was too thick. In addition, the shallow depth of the plastic deformation caused by the laser-induced shock wave did not adequately change the morphology, structure, and distribution of grain inside the sample, which cannot significantly affect the tensile properties of the material.

Range Analysis
The residual stress, surface roughness, microhardness, yield strength, average grain size, and other performance indicators with significance within the 10% confidence level were selected for the range analysis, and the analysis results are shown in Table 6. The results showed that, for the indicators with significant differences, the primary and secondary orders of the factors were as follows:

•
For residual stress and microhardness, the laser energy is larger than the beam diameter, which is larger than number of impacts. The influence of laser energy was very clear and the range of residual stress and microhardness were 27.69 and 31.09, which was much larger than the other two factors, respectively. • For surface roughness and average grain size, the laser energy is larger than the number of impacts, which is larger than the beam diameter. The influence degree of number of impacts and beam diameter was very close. The range of surface roughness was 2.21 and 1.91, and the range of average grain size was 1.33 and 1.23, respectively. • For the yield strength, the number of impacts are larger than the laser energy, which is larger than the beam diameter. The number of impacts were very obvious, with a range of 137.67, which was much larger than the other two factors. Based on an overall consideration, the primary and secondary orders of the factors in actual machining were as follows: The laser energy is larger than the number of impacts, which is larger than the beam diameter.
The values of K i (i = 1,2,3) of each factor in Table 6, at different performance indicators, were taken as the variation trends of surface residual stress, microhardness, yield strength, and average grain size under different factor levels, as shown in Figure 5a-d. As the variation trends in surface roughness and residual stresses were similar, they are not described in this paper. deepened, and the grain refinement and dislocation that were continuously superimposed [19].
In Figure 5b,d, when the number of impacts increased from 2 to 3, the microhardness and average grain size slightly changed. This may be due to the laser shock, which made the dislocation inside the grain proliferate continuously to form sub-grain boundaries, dislocation walls, dislocation entanglements, and other dislocation structures in the grain, which are annihilated and rearranged, and then the grain is refined [20]. When the refinement reached a certain level, the proliferation and annihilation of the dislocation structure reached a dynamic balance. At this time, whether the number of impacts increased or not, the grain size no longer decreases, and the uniform and stable surface microstructure no longer significantly change the microhardness. In Figure 5, as the laser energy increases, the residual stress and microhardness of samples increase, and the average grain size decreases. The yield strength decreased 59.67 MPa first and then increased 35 MPa. As shown in Figure 5c, the yield strength of the sample only showed some fluctuation in a small range of less than 10% compared with the unimpacted sample (971 MPa), with no significant trend of increase or decrease. It can be concluded that the increase of laser energy had almost no effect on the yield strength, which was also consistent with the results of the statistical significance in the variance analysis. The yield strength represents the limit of the material's resistance to yield. The external forces greater than this limit cause the material to undergo a permanent and irrecoverable plastic deformation. This plastic deformation usually begins with precipitated phases and defects in the middle area of the thickness direction perpendicular to the external force, and the micro-pores in the material continue to grow, extend, and gather, which results in crack formation, which eventually leads to macroscopic yield and permanent deformation of the material [18]. In this study, the shock wave induced by laser energy influence depth was shallow, the laser-induced shock wave completely attenuated before propagating to the middle region of the material thickness direction, and the residual compressive stress layer was very thin and could not inhibit the growth of microvoids in the material or significantly affect the overall plastic deformation and yield capacity of the material. As the beam diameter increased, the residual stress and microhardness of samples increased, the average grain size decreased, and the yield strength decreased first and then increased, which was consistent with the influence of the laser energy. As the number of impacts increased, the residual stress, microhardness, and yield strength of the sample increased, and the average grain size decreased. This was caused by the plastic deformation by the laser-induced shock wave directly accumulated on the material surface, the residual compressive stress layer that deepened, and the grain refinement and dislocation that were continuously superimposed [19].
In Figure 5b,d, when the number of impacts increased from 2 to 3, the microhardness and average grain size slightly changed. This may be due to the laser shock, which made the dislocation inside the grain proliferate continuously to form sub-grain boundaries, dislocation walls, dislocation entanglements, and other dislocation structures in the grain, which are annihilated and rearranged, and then the grain is refined [20]. When the refinement reached a certain level, the proliferation and annihilation of the dislocation structure reached a dynamic balance. At this time, whether the number of impacts increased or not, the grain size no longer decreases, and the uniform and stable surface microstructure no longer significantly change the microhardness.

Fracture Morphology Analysis
The tensile samples of the FGH95 superalloy before and after the laser shock strengthening are shown in Figure 6. Macroscopically, the fracture cross-sections of all the samples had a clear shrinkage phenomenon; thus, all samples showed a ductile fracture.

Fracture Morphology Analysis
The tensile samples of the FGH95 superalloy before and after the laser shock strengthening are shown in Figure 6. Macroscopically, the fracture cross-sections of all the samples had a clear shrinkage phenomenon; thus, all samples showed a ductile fracture. The macroscopic and the microscopic fracture morphology of 10 samples after tensile fracture are shown in Figure 7. When the samples were subjected to unidirectional tensile force, plastic deformation occurred first, causing necking. Then, the cracks expanded in the perpendicular direction to the external force at the necking position. When the cracks reached a certain degree of instability, they rapidly developed to a fracture of about 45° along the direction of the external force, as shown in Figure 7a. The surface of the fracture is shown in Figure 7b, and the fracture was the cross-section of the through-going fault. Figure 7b shows the fracture sample magnified 40 times, indicating that the uneven part in the middle of the sample fracture formed the "bottom of the cup" and produced different degrees of tightening. The section was gray, as shown in the area circled by the dashed lines. The elongation index after fracture was higher, and plasticity was better. Figure 7d was obtained by magnifying the solid circle area of Figure 7c 6000 times. It was found that a large number of torn edges, micropores, and equiaxed dimples appeared on the fracture surface. The white ridges around the dimples were torn edges, and the dimples came in different sizes. The second phase particles can be found at the bottom of some micropores, and their size, shape, and distribution were some of the main factors affecting the size of dimples. When plastic deformation of the specimen occurs, the micropores form in the deformation area. The micropores were small and isolated from each other. Under the action of loading force, the micropores were connected with each other, resulting in sample failure, which was also one of the reasons for crack propagation. The macroscopic and the microscopic fracture morphology of 10 samples after tensile fracture are shown in Figure 7. When the samples were subjected to unidirectional tensile force, plastic deformation occurred first, causing necking. Then, the cracks expanded in the perpendicular direction to the external force at the necking position. When the cracks reached a certain degree of instability, they rapidly developed to a fracture of about 45 • along the direction of the external force, as shown in Figure 7a. The surface of the fracture is shown in Figure 7b, and the fracture was the cross-section of the through-going fault. Figure 7b shows the fracture sample magnified 40 times, indicating that the uneven part in the middle of the sample fracture formed the "bottom of the cup" and produced different degrees of tightening. The section was gray, as shown in the area circled by the dashed lines. The elongation index after fracture was higher, and plasticity was better. Figure 7d was obtained by magnifying the solid circle area of Figure 7c 6000 times. It was found that a large number of torn edges, micropores, and equiaxed dimples appeared on the fracture surface. The white ridges around the dimples were torn edges, and the dimples came in different sizes. The second phase particles can be found at the bottom of some micropores, and their size, shape, and distribution were some of the main factors affecting the size of dimples. When plastic deformation of the specimen occurs, the micropores form in the deformation area. The micropores were small and isolated from each other. Under the action of loading force, the micropores were connected with each other, resulting in sample failure, which was also one of the reasons for crack propagation. The tensile fracture microstructure of 10 samples is shown in Figure 8. Due to the shallow residual stress layer induced by the OLSP (generally within 1 mm), the observation area was selected as close as possible to the edge of the impact surface of the sample to obtain the effect of laser shock strengthening on tensile properties when observing the fracture morphology. Figure 8a-i show the fracture morphology of the samples by different OLSP parameters (corresponding to samples 1 to 9 of the experiment). Figure 8j shows the fracture morphology of the unimpacted sample no. 10. The surface of the unimpacted sample mostly presents equiaxed dimples or slightly elongated dimples, as shown in the dotted circles of Figure 8j. In Figure 8a-d, a gradually transitional elongated dimple was found with a parabolic shape along the internal direction of the material from the impact surface, as shown by the dotted elliptic circle. Compared with the dimple shape of tensile sample fracture morphology strengthened by vertical lasers from Reference [21], it is found that this is due to the surface residual stress caused by the OLSP that was affected by the combined force of unidirectional tension and residual stress during the tensile process, resulting in the elongation of dimple along the stress direction. When the laser energy increased to 6-8 J, the effect of the dimple elongated by force was weakened. The fracture morphology of the sample was mainly manifested in the small size and shallow depth of the dimple, and the dimples were more uniformly distributed, as shown in the solid line circle in Figure 8e-i. This might be due to the grain refinement and deformation hardening caused by the OLSP. The stronger the laser-induced shock wave force, the higher the degree of deformation hardening. Therefore, it was more difficult for necking to occur in the material, and this led to smaller dimple size [22]. The tensile fracture microstructure of 10 samples is shown in Figure 8. Due to the shallow residual stress layer induced by the OLSP (generally within 1 mm), the observation area was selected as close as possible to the edge of the impact surface of the sample to obtain the effect of laser shock strengthening on tensile properties when observing the fracture morphology. Figure 8a-i show the fracture morphology of the samples by different OLSP parameters (corresponding to samples 1 to 9 of the experiment). Figure 8j shows the fracture morphology of the unimpacted sample no. 10. The surface of the unimpacted sample mostly presents equiaxed dimples or slightly elongated dimples, as shown in the dotted circles of Figure 8j. In Figure 8a-d, a gradually transitional elongated dimple was found with a parabolic shape along the internal direction of the material from the impact surface, as shown by the dotted elliptic circle. Compared with the dimple shape of tensile sample fracture morphology strengthened by vertical lasers from Reference [21], it is found that this is due to the surface residual stress caused by the OLSP that was affected by the combined force of unidirectional tension and residual stress during the tensile process, resulting in the elongation of dimple along the stress direction. When the laser energy increased to 6-8 J, the effect of the dimple elongated by force was weakened. The fracture morphology of the sample was mainly manifested in the small size and shallow depth of the dimple, and the dimples were more uniformly distributed, as shown in the solid line circle in Figure 8e-i. This might be due to the grain refinement and deformation hardening caused by the OLSP. The stronger the laser-induced shock wave force, the higher the degree of deformation hardening. Therefore, it was more difficult for necking to occur in the material, and this led to smaller dimple size [22].  Compared with the unimpacted sample, the impacted samples showed a slight increase in tensile and yield strengths, a slight decrease in elongation, and a generally minor change in the size and depth of the dimples. This is due to the peak pressure of the laser-induced shock wave exceeding the dynamic yield strength of the material, causing intense plastic deformation at the impact surface and generating large residual stresses. This caused grain refinement and dislocation entanglement on the material surface, impeding dislocation movement, hardening, reducing plasticity and ductility and inhibiting the occurrence of necking, thus reducing elongation and increasing microhardness. It also inhibited the dimple deepening and elongation near the impact surface to a certain extent, giving the dimple a small size, shallow depth, and evenly distribution, which prolonged the elastic deformation phase and thus increased the tensile and yield strengths to a certain extent [23]. However, due to the shallow strengthening layer, there was a little strengthening effect on the middle area in the thickness direction of the sample, with no significant effect on the overall tensile strength and elongation of the material. This is also consistent with the conclusion of the variance analysis that the tensile strength and elongation were not significant under all factors' conditions.

Conclusions
Under different level factors, the FGH95 superalloy tensile sample was strengthened by OLSP through the orthogonal test method. After shocking, all performance indicators were improved compared with those without processing. The specific conclusions were as follows: Compared with the unimpacted sample, the impacted samples showed a slight increase in tensile and yield strengths, a slight decrease in elongation, and a generally minor change in the size and depth of the dimples. This is due to the peak pressure of the laser-induced shock wave exceeding the dynamic yield strength of the material, causing intense plastic deformation at the impact surface and generating large residual stresses. This caused grain refinement and dislocation entanglement on the material surface, impeding dislocation movement, hardening, reducing plasticity and ductility and inhibiting the occurrence of necking, thus reducing elongation and increasing microhardness. It also inhibited the dimple deepening and elongation near the impact surface to a certain extent, giving the dimple a small size, shallow depth, and evenly distribution, which prolonged the elastic deformation phase and thus increased the tensile and yield strengths to a certain extent [23]. However, due to the shallow strengthening layer, there was a little strengthening effect on the middle area in the thickness direction of the sample, with no significant effect on the overall tensile strength and elongation of the material. This is also consistent with the conclusion of the variance analysis that the tensile strength and elongation were not significant under all factors' conditions.

Conclusions
Under different level factors, the FGH95 superalloy tensile sample was strengthened by OLSP through the orthogonal test method. After shocking, all performance indicators were improved compared with those without processing. The specific conclusions were as follows: The results of variance analysis showed that only the yield strength was significantly different under different numbers of impacts within a 10% confidence level, and there was no significant difference in tensile strength, yield strength, or elongation under different levels of other factors. This is because the laser energy of the shock wave induced by residual stress produced by hardening layer is very shallow, about 1 mm, and the growth of the ductile fracture of microporous materials can significantly affect the material's ability to resist plastic deformation and overall yield, so the tensile strength, yield strength, and elongation of tensile performance do not show an obvious difference between the various levels. The significant difference of residual stress, surface roughness, microhardness, and average grain size is very obvious, and these factors should be considered as the main performance indexes.
The results of range analysis showed that the primary and secondary orders of the factors affecting each index were as follows: laser energy > number of impacts > beam diameter. The analysis results of the fracture morphology showed that laser shock strengthening could make the dimples more uniformly distributed, smaller in size, and shallower in depth. However, compared with the unimpacted sample, there was little difference in the entire performance, and the thickness of the affected layer was extremely limited and could not affect the overall tensile properties of the material. This is consistent with the conclusion that there is almost no significant difference in tensile properties under various factors.
Combined with the above analysis and results, with the factor level combination of 8 J laser energy, 3 mm spot diameter, and three impacts, the FGH95 superalloy sample can obtain the best comprehensive performance, which provides a reference for the OLSP of FGH95 superalloy turbine disks and other aeronautical parts. The results are only applicable to the test conditions and parameter value range in this paper, and different processing conditions of FGH95 turbine disk joint materials need further study and optimization.
Finally, this experiment only focused on the influence of laser energy, spot diameter, and number of impacts on the performance of the FGH95 superalloy sample, while other factors such as beam lap rate and laser pulse width need to be further analyzed and explored.