Effect of surface strengthening by ultrasonic shot peening on TC17 alloy

This paper investigates the effects of ultrasonic shot peening (USP) on surface morphology, microstructure, and residual stress field distribution of TC17 alloy under different process parameters. The aim is to reveal the surface strengthening mechanism of TC17 alloy caused by USP. The results suggest that the use of the 2.5 mm projectile diameter leads to an increase in surface roughness, plastic deformation, and a deeper grain refinement layer compared to the 1.5 mm projectile diameter. Additionally, it results in a greater depth of the compressive residual stress layer and maximum compressive residual stress. The crack initiation sites under two projectile diameters are located below the compressive residual stress layer. The USP treatment introduces compressive residual stress on the surface, inhibiting the initiation of surface cracks, and the deeper compressive residual stress layer offsets the early fatigue failure caused by higher roughness.


Introduction
Due to its exceptional specific strength and fracture toughness, TC17 alloy is extensively utilized in crucial components of aero-engines, including compressor disks and centrifugal impellers [1][2][3].However, these components endure severe working conditions, such as thermal stress and cyclic vibration, leading to potential deformation and fracture, thereby diminishing the service life and reliability of aero engines.Consequently, various processes can be employed to enhance the surface-related properties and enhance the fatigue performance of components [4].Among these processes, ultrasonic shot peening (USP) stands out as a sophisticated surface treatment technique that has garnered increasing attention due to its ability to induce large-depth compressive residual stress onto the surface of metallic components.The compressive residual stress produced by USP has proven effective in enhancing the fatigue properties of titanium alloys by refining surface grains, delaying crack initiation, and reducing crack propagation rates [5].
Nevertheless, the improper selection of shot peening parameters, such as employing large shot or excessive peening, may lead to the generation of surface defects, including overlaps, scales, microcracks, and imperfections in surface roughness.These defects can significantly diminish the high cycle fatigue strength of the treated component [6,7].Therefore, it is essential to understand the influence of the shot peening process on the surface state to improve the high cycle fatigue life of shot peened parts.Presently, research on USP predominantly focuses on surface morphology, microstructure, residual stress, and fatigue performance [8,9].Studies have demonstrated that USP induces highly random and repetitive impacts within a specific tooling chamber, resulting in improved surface quality and a deeper compressive residual stress layer [10][11][12].Furthermore, research has suggested that increasing the intensity of USP elevates the average roughness but decreases the dispersion of roughness distribution [13,14].
Numerous studies have substantiated that the enhanced fatigue properties of titanium alloys following shot peening arise from the refinement of the subsurface microstructure and the introduction of localized compressive residual stresses [15,16].In light of the intricate interplay between fatigue performance and microstructure, several researchers have employed conventional shot peening to delve into the connections among microstructure, grain orientation, grain boundary changes in the alloy, and fatigue properties.Their findings underscore the significant role of microstructure characterization in the exploration of surface strengthening [17][18][19][20][21][22][23].However, research concerning the effects of USP on TC17 alloy remains relatively limited.
This study utilizes a combination of surface profilometry, optical microscopy (OM), scanning electron microscopy (SEM), electron backscatter diffraction (EBSD), microhardness testing, and x-ray diffraction (XRD) techniques for a thorough characterization of the USP-affected layer.The investigation explores the impact of various USP process parameters on the morphology, microstructure, microhardness, residual stress field, and fatigue properties of TC17 alloy.Additionally, it elucidates the fatigue failure mechanism of TC17 alloy induced by USP.

Materials
The test material utilized in this study is β-forged TC17 alloy, with its chemical composition outlined in table 1.The heat treatment process is solution treatment (at 800 °C for 4 h, followed by air cooling) and aging treatment (at 650 °C for 8 h, followed by air cooling), resulting in a uniform α + β two-phase microstructure, as illustrated in figure 1.The mechanical properties of TC17 alloy post heat treatment are provided in table 2. To ensure a consistent initial surface condition, the specimens' surfaces underwent meticulous polishing using waterproof abrasive paper.Subsequently, the specimens underwent ultrasonic washing with anhydrous ethanol at ambient temperature for a duration of 5 min.

Ultrasonic shot peening treatment
Figure 2(a) illustrates the equipment utilized for USP strengthening.The ultrasonic shot peening chamber adopts a polymerized chamber, and the strengthening effect of this polymerized chamber on the target area of the specimen is more pronounced compared to that of a conventional cylindrical chamber.In figure 2(b), the dynamic process of projectile impact is depicted.For the USP treatment in this study, zirconia ceramic projectiles with a density of 6.00 g cm −13 and a microhardness of 1250 HV were utilized.The specimens measured 50 mm × 60 mm × 5 mm.The shot peening Almen intensity was set to 0.2 mmA, employing 1000 projectiles, and the duration of shot peening was 1800 s.The Almen intensity refers to the arc height when the Almen test piece is used for calibration, following the SAE AMS 2580A standard document.The vibration system operated at a frequency of 20 kHz, utilizing projectile diameters of 1.5 mm and 2.5 mm, with amplitudes set at 60 μm and 80 μm.Detailed process parameters for the USP treatment are provided in table 3.

Characterization
The surface roughness (Ra) was evaluated employing a super-field depth three-dimensional visualization profilometer (VHX-900).Measurements were performed at three distinct locations on each specimen, and the average values were calculated.For crosssectional observation, specimens underwent corrosion treatment with Kroll reagent and were examined using an Olympus GX71 metallographic microscope.EBSD data were acquired using a JEOL JSM-7800F field emission scanning electron microscope.In-depth microhardness distribution measurements at various locations in the subsurface section were carried out using a microhardness tester (HXS-1000A) with a 25 gf load applied for a holding time of 10 s.The residual stress field was assessed using a French MRX x-ray stress meter, and the residual stress distribution along the depth was obtained by corroding and delaminating the samples through the electrochemical delamination method.In addition, a rotational bending fatigue test  was conducted on a QBWP-10000 rotary bending fatigue testing machine, following the GB/T 4337-2015 standard for fatigue testing at room temperature on a cylindrical test bar with a working section diameter of 6 mm.The test operated at a frequency of 80 Hz and a stress ratio of R = −1.Fatigue fracture analysis was performed using a Hitachi Su3500 scanning electron microscope.

Theoretical and experimental analysis of surface pits in USP
The sinusoidal harmonic signal emitted by the ultrasonic generator is: where A denotes the amplitude and ω denotes the angular frequency.It is deduced that: According to the formula (2), the maximum initial velocity is: where f represents the frequency of ultrasonic vibration.
The initial theoretical shot peening velocity for the 60 μm amplitude process parameter is computed as 7.5 m s −1 , whereas for the 80 μm amplitude, it is 10.1 m s −1 .These values represent the average velocity during the collision process.To compare the theoretical results with the experimental distribution of the crater diameter, establishing the relationship between the impact velocity and the crater diameter is crucial.For this purpose, the model proposed by Johnson (formula (4)) is employed in this research [24].The model is suitable for low-velocity elastoplastic indentation between the projectile and the surface.Both the projectile and the surface are assumed to be made of isotropic materials without any initial residual stress.The vertical impact area above the vibrating head can be considered the maximum velocity due to the low chamber height and more vertical impacts of the projectile in the normal direction, which can be replaced by V max instead of V.The process of the projectile impacting the specimen is illustrated in figure 3, where impact areas 1 and 2 represent high-pressure stress areas (blue), while the overlapping areas represent low-pressure stress areas (yellow).
where f num is the theoretical pit diameter, D is the projectile diameter, ρ is the projectile density, V is the  normal impact velocity of the projectile, and σ y is the yield stress of the impacted material.The roughness of the shot peening surface is measured by the magnitude of the Ra value (contour average arithmetic deviation), as shown in the formula (5) [25]: where l is the sampling length, and Y = f(x) is the peak height within the sampling length.Figure 4 illustrates the mean values of theoretical and actual pit diameters under various process parameters of USP.The average pit diameters for groups A to D were 0.18 mm, 0.31 mm, 0.24 mm, and 0.37 mm, respectively, accompanied by corresponding standard deviations of pit sizes of 0.00803, 0.01082, 0.00877, and 0.00693.The standard deviation of pit sizes was below 0.011, indicating a high level of consistency between theoretical and actual values of the average pit diameters.Furthermore, an increase in the diameter of the 1.5 mm projectile to 2.5 mm resulted in a 61.9% increase in the average pit diameter, while an increase in amplitude from 60 μm to 80 μm led to a 24.5% rise in the average pit diameter.
Figure 5 illustrates the macrostructure of surface pits on TC17 alloy following the application of various process parameters of USP, along with the distribution of surface morphology based on three-dimensional image analysis using ImageJ.The sample surface exhibits numerous pits after undergoing USP, attributed to the high-energy impact of the projectiles, causing significant strain and plastic deformation during the shot peening process [26,27].In the three-dimensional distribution map, the horizontal axis is denoted by x, the vertical axis by y, and the depth by z.The size of surface pits is more prominent when using a larger diameter projectile for shot peening, while the depth of surface pits is deeper with a higher amplitude for shot peening.The three-dimensional image results reveal that the size of pits in group A is the smallest, while group D exhibits the largest and most uniform pit size distribution, with group C displaying better uniformity than  group B. The pit density for groups A to D as follows: 28.9%, 33.2%, 39.5%, and 45.6%, respectively.This increase correlates with the rise in projectile diameter and amplitude intensity, aligning seamlessly with the macroscopic morphology of the surface.Under the process conditions of groups C and D, the pit size exhibits a low standard deviation, the distribution of pits is uniform, and the pit density is significant.In contrast to groups A and B, there is a notable 37.1% increase in density, highlighting the high consistency of shot peening surface morphology.Consequently, the process parameters of group C (1.5 mm projectile diameter and 80 μm amplitude) and group D (2.5 mm projectile diameter and 80 μm amplitude) are selected for further analysis in surface strengthening characterization.The results of the roughness tests indicate that the surface roughness Ra of specimens subjected to USP with 1.5 mm and 2.5 mm projectile diameters are 0.352 μm and 0.451 μm, respectively.Notably, the roughness of the 2.5 mm specimen is 28.1% higher than that of the 1.5 mm specimen.

Microstructure
Figure 6 illustrates the microstructures of crosssectional views of specimens exposed to various treatments, exhibiting a characteristic basketweave microstructure.While the cross-section of the specimen with no shot peening shows no signs of plastic deformation, USP yields a gradient distribution of grain size in the surface plastic deformation area, indicating a more distinct plastic layer for surface strengthening.Notably, the presence of a distinct boundary between the deformed layer and the inner material matrix was not observed [28].Moreover, an increase in projectile diameter significantly strengthens the distribution of the plastic layer and enhances the gradient distribution of plastic deformation.The gradient distribution of the microstructure in the plastic layer indicates that the grain refinement layer depths for projectile diameters of 1.5 mm and 2.5 mm are 65 μm and 90 μm, respectively.The grain refinement layer depth of 2.5 mm exceeds that of 1.5 mm by 38.5%, widening the distribution of the grain deformation layer as the projectile diameter increases.Previous studies have also reported grain refinement resulting from USP in other materials [29,30].

Strain gradient Figures 7(a) and (d) present Inverse Pole figure (IPF)
diagrams of the TC17 alloy cross-section treated with projectile diameters of 1.5 mm and 2.5 mm, respectively.Distinct colors in the diagrams indicate varying grain orientations within the specimen cross-section.The USP treatment leads to a significant increase in the degree of grain refinement in the surface area, accompanied by observable grain rotation and intense plastic deformation.With increasing depth, the grain changes induced by the USP treatment gradually diminish, revealing a gradient distribution of grain size from the surface layer to the subsurface layer.
Figures 7(b) and (e) display the Kernel Average Misorientation (KAM) distribution diagrams of the TC17 alloy cross-section deformation layer treated with projectile diameters of 1.5 mm and 2.5 mm, respectively.These diagrams provide a qualitative assessment of the homogenization level of plastic deformation induced by USP treatment.Regions with higher values in the KAM diagrams indicate a more pronounced degree of plastic deformation or a greater density of defects [21].The plastic strain induced by shot peening is observed at depths of 60 μm and 90 μm below the surface for the 1.5 mm and 2.5 mm projectile diameters, respectively.The distribution range of the plastic strain layer aligns consistently with the depth of the grain refinement layer, demonstrating uniform transverse distribution of plastic deformation.
Figures 7(c) and (f) illustrate grain boundary distribution diagrams of the TC17 alloy cross-section treated with projectile diameters of 1.5 mm and 2.5 mm, respectively.The diagrams depict yellow, red, and blue grain boundaries, representing low-angle grain boundaries (LAGB) with a grain boundary angle of 2°− 15°, high-angle grain boundaries (HAGB) with a grain boundary angle > 15°, and twin grain boundaries (TGB), respectively.The results indicate that, under both projectile diameters, twin crystals are predominantly distributed in the surface and subsurface areas, which correspond to the grain refinement areas.The LAGB is densely distributed, and the α phase is widely distributed in this region.The twin density distribution results reveal that the number of twins increases by approximately 25.6% when the projectile diameter is increased from 1.5 mm to 2.5 mm.This finding suggests a higher degree of plastic deformation in the surface area of the specimen treated with a projectile diameter of 2.5 mm.

Microhardness
Figure 8 illustrates the in-depth microhardness distributions for the TC17 alloy subjected to various surface treatments.The in-depth microhardness distributions follow a similar pattern after different processes, gradually decreasing with increasing depth and eventually reaching and remaining at the level of the matrix microhardness (400-410 HV 0.2 ).A rapid decline in microhardness is observed within the depth range of 10-70 μm.All surface microhardness values This observation indicates that increasing the projectile diameter effectively enhances microhardness.The observed increase in microhardness is primarily attributed to grain refinement and the heightened dislocation density resulting from the high-speed impact of the projectiles [31,32].depths, is observed in the specimens after shot peening.The maximum compressive residual stress is situated at a certain distance from the surface, gradually decreasing until reaching the residual stress level of the bulk material.Similar characteristic curves have been documented in prior studies on diverse alloys subjected to shot peening treatment [28,33].For the 1.5 mm projectile diameter, the surface compressive residual stress measures −542 MPa, with the maximum compressive residual stress approximately −604 MPa located at a depth of 25 μm from the surface layer.The compressive residual stress layer extends to a depth of 150 μm after shot peening.In contrast, for the 2.5 mm projectile diameter, the surface compressive residual stress is −608 MPa, and the compressive residual stress layer extends to a depth of 200 μm.The maximum compressive residual stress is −656 MPa, located at a depth of 40 μm.The surface residual stress value for the 2.5 mm projectile diameter exceeds that of the 1.5 mm diameter by 12.2%.Furthermore, both the distance between the maximum residual compressive stress and the surface, and the depth of the compressive residual stress layer, exhibit an increase in the case of the 2.5 mm projectile diameter when compared to the 1.5 mm projectile diameter.The depth of the compressive residual stress layer experiences a notable 33.3% increase, while the maximum compressive residual stress sees an 8.6% rise, with the maximum depth extending to 15 μm into the subsurface layer.The compressive residual stress layer extends to the subsurface by approximately 100 μm, based on the grain refinement layer.The augmentation in projectile diameter results in a more profound surface deformation layer due to heightened kinetic energy, contributing to an escalation in both the magnitude of the compressive residual stress and the depth of the affected layer.This observation aligns with the findings reported by Robertson [34].

Fatigue behavior
Figure 10 illustrates the high cycle fatigue S-N curves of TC17 alloy with various surface treatments.The fatigue limit of the specimen reaches 702 MPa when the projectile diameter is 2.5 mm, marking a noteworthy 3.2% increase compared to the 1.5 mm diameter limit of 680 MPa.The ultrasonic shot peening samples, with projectile diameters of 2.5 mm and 1.5 mm, exhibit fatigue limit enhancements of 12.7% and 9.1%, respectively, in comparison to the unpeened sample with a limit of 623 MPa.The introduction of a compressive residual stress field emerges as a primary factor in enhancing fatigue performance, effectively inhibiting the initiation and propagation of fatigue cracks.Furthermore, ultrasonic shot peening induces the refinement of surface grains and an increase in distortion.This phenomenon obstructs the movement of dislocations within the deformation layer, causing their accumulation between the deformation layer and the matrix interface.Consequently, this accumulation serves as a barrier to the initiation of fatigue cracks on both the surface and subsurface, ultimately leading to an extension of the fatigue life [35][36][37].For every stress amplitude, the fatigue life of the 2.5 mm projectile diameter surpasses that of the 1.5 mm projectile diameter, with the difference in fatigue life becoming negligible when the number of cycles is less than 10 4 .These results suggest that the compressive residual stress field induced by USP exhibits a more pronounced inhibitory effect on crack initiation under low-load and high-cycle conditions compared to high-load and low-cycle conditions.This observation finds support in the research conclusion of Wu et al. [38].
Figure 11 illustrates the high-cycle fatigue fracture morphology of TC17 alloy featuring projectile diameters of 1.5 mm and 2.5 mm.Fractures under both   direction, smoothly reaching the side of the fracture.Under the influence of high cyclic loading, the area of crack initiation appears smooth and flat.Notably, both crack nucleation sites are situated below the compressive residual stress layer.In the case of the 1.5 mm projectile diameter specimen, the crack nucleation depth is 200 μm below the surface, while the 2.5 mm projectile diameter specimen's crack initiation site is approximately 260 μm from the surface.Comparatively, the fracture surface of the 1.5 mm projectile diameter appears relatively rough, and the crack initiation site is closer to the surface than that of the 2.5 mm projectile diameter.

Conclusions
In this paper, the effects of surface strengthening by different USP process parameters on TC17 alloy were investigated and compared.The main conclusions are summarized as follows: (1) The prediction model for crater morphology is established, achieving accurate crater prediction.
No defects, such as material stacking, local deformation, or cracks, are introduced by USP on the surface of TC17 alloy.When the projectile diameter is constant, increasing the amplitude from 60 μm to 80 μm enhances the consistency of USP surface morphology.
(2) As the projectile diameter increases from 1.5 mm to 2.5 mm, the surface roughness experiences a 28.1% increase, and the depth of the grain refinement layer grows by 38.5%.The impact of USP on grain refinement gradually diminishes with increasing depth, resulting in a gradient distribution of grain size from the surface to the subsurface of the sample.
(3) After ultrasonic shot peening, the microhardness increased significantly, and the microhardness was distributed in a gradient.From the surface to the internal matrix, the microhardness gradually decreased.When the diameter of the projectile increases from 1.5mm to 2.5mm, the work hardening of the surface layer of the material is more obvious, the surface microhardness value increases by 5.6%, and the depth of the hardened layer increases by 16.7%.
(4) With the projectile diameter increasing from 1.5 mm to 2.5 mm, the surface residual compressive stress value rises by 12.2%, the depth of the residual compressive stress layer expands by 33.3%, and the maximum residual compressive stress increases by 8.6%.In comparison with the surface residual compressive stress and the depth of the residual compressive stress layer, the maximum residual compressive stress is less influenced by the projectile diameter.
(5) After ultrasonic shot peening, the crack initiation position is located below the residual compressive stress layer.The fatigue limit increases by 3.2% when the projectile diameter increases from 1.5 mm to 2.5 mm.The positive effect of increasing the depth of the compressive residual stress layer on fatigue performance is more significant compared to the negative effect of higher surface roughness caused by larger projectile diameters.

Figure 4 .
Figure 4. Mean values of theoretical and actual pit diameters under various process parameters of USP.

Figure 5 .
Figure 5. Macroscopic morphology and three-dimensional distribution map of surface pits on TC17 alloy after different process parameters of USP.

Figure 9
illustrates the residual stress distribution curves along the layer depth for various treatment processes.A 'hook' type distribution of compressive residual stress field, characterized by high values and

Figure 8 .
Figure 8. Microhardness distribution along the layer depth of TC17 alloy.

Figure 9 .
Figure 9. Residual stress distribution along the layer depth of TC17 alloy.

Figure 10 .
Figure 10.High cycle fatigue S-N curve of TC17 alloy.