Effects of Surface-Hardened Layer Thickness Reduction on the Fatigue Characteristics of Shot-Peened AISI 4340 Steel
Article information
Abstract
Numerous studies have focused on improving the fatigue strength and life of materials by introducing compressive residual stresses into the surface layer through surface treatment processes such as shot peening. This study analyzes the effects that partially removing the shot-peened surface-hardened layer through polishing has on the mechanical properties, surface characteristics, fatigue strength, and fatigue life of AISI 4340 steel. Surface roughness, hardness, and residual stress were measured experimentally. Mechanical properties were evaluated through tensile testing, and ultrasonic fatigue tests were performed using 20 kHz elastic waves. The results show that the fatigue limit decreased by approximately 9.4%, 26.3%, and 25% after a 10%, 15%, and 20% reduction in surface-hardened layer thickness, respectively. Furthermore, the fatigue strength at a life of 1 × 109 cycles (very-high-cycle fatigue) decreased by up to approximately 23.9% after 15% and 20% thickness reductions compared with that before surface layer removal. The findings of this study provide useful data for the optimal design of materials, components, and equipment requiring enhanced fatigue resistance through surface modification treatments.
1. Introduction
Materials used in high-load applications, such as those encountered in marine engineering and equipment, as well as the aerospace, automotive, and military industries, must exhibit excellent toughness, strength, wear resistance, and fatigue resistance. Accordingly, AISI 4340 steel, a low-alloy steel containing nickel, chromium, and molybdenum, is widely employed. It is particularly suitable for power transmission components such as bevel gears used in propulsion and drive systems, including aircraft and wind turbine gear systems that operate under high loads and rotational speeds. From a safety perspective, evaluating the reliability and fatigue life of structural materials commonly used in drive shafts and gears is critically important (Ahn et al., 2020; Bag et al., 2020; Bonora et al., 2010; Jack Champaigne Electronics Inc., 2001; Shin et al., 2000; Sherif et al., 2024; Wagner, 2003). To achieve the required mechanical performance, AISI 4340 steel must undergo precise heat treatment through quenching and tempering. Although this improves strength, wear resistance and fatigue life can be further enhanced through surface modification methods such as shot peening (Jung et al., 2025; Karimbaev et al., 2020; Llaneza & Belzunce, 2015). However, this steel is vulnerable in environments prone to wear or salt exposure because of its relatively low corrosion resistance. Therefore, it is unsuitable for direct exposure to seawater without protection, and surface treatments such as anticorrosion coatings or chrome plating must be applied prior to use (Bonora et al., 2010; Sherif et al., 2024).
Numerous studies have investigated methods for improving the fatigue life of materials, particularly through surface modification treatments (Correia et al., 2023; Smith & Hirt, 1985). Among these methods, shot peening is widely applied. It is typically used to improve resistance to degradation mechanisms such as cracking, fatigue, wear, and corrosion that originate at the surface under tensile stress. Shot peening plastically deforms the surface layer by impacting it with small shots, thereby modifying surface characteristics through work hardening. As a result, extensive research has been conducted on shot peening with respect to fatigue behavior, peening processes, wear characteristics, and resistance to stress corrosion cracking (SCC) (Ahn et al., 2020; Bag et al., 2020; Bagherifard et al., 2012; Gundgire et al., 2022; Jack Champaigne Electronics Inc., 2001; Klotz et al., 2018; Kovaci et al., 2019; Kumar et al., 2019; Lee et al., 1998; Lin et al., 2019; Oguri, 2011; Shin et al., 2000; Trung et al., 2016; Wagner, 2003; Wu et al., 2020; Yan et al., 2022).
The primary purpose of shot peening is to introduce compressive residual stress into the surface layer, thereby delaying the initiation and propagation of fatigue cracks and improving durability. A reduction in the thickness of the surface-hardened layer diminishes this protective effect, leading to shorter fatigue life of both the material and the components. Shot peening also increases hardness through surface work hardening. When the hardened layer becomes thinner, the hardness distribution and maximum hardness value along the depth reduces, weakening resistance to wear and surface damage (Ahn et al., 2020; Bag et al., 2020; Bagherifard et al., 2012; Jack Champaigne Electronics Inc., 2001; Kovaci et al., 2019; Kumar et al., 2019; Lee et al., 1998; Lin et al., 2019; Oguri, 2011; Shin et al., 2000; Trung et al., 2016; Wagner, 2003; Wu et al., 2020). Furthermore, shot peening enhances resistance to SCC (Al-Obaid, 1995; Ralls et al., 2025) and corrosion fatigue (Kim and Cheong, 2012). As the hardened layer becomes thinner, these protective properties weaken, making the material more susceptible to corrosion in aggressive environments such as seawater. Moreover, excessive shot peening (e.g., over-peening) can cause stripping of the surface-modified layer or increase surface roughness, leading to adverse effects (Maleki et al., 2018). As a result, the effective hardened layer thickness may decrease and its quality may deteriorate.
Therefore, maintaining an appropriate surface-hardened layer thickness is essential to ensure the optimal performance of shot-peened materials and components. A reduction in this thickness can significantly compromise the overall reliability and durability of the components.
However, the increased surface roughness caused by dimples after shot peening must be considered when addressing wear, friction, fatigue, impact, and corrosion of treated components (Liu et al., 2017). To reduce this roughness, polishing or stripping processes are commonly applied. Additionally, decreases in hardened layer thickness may also result from defects or external factors during manufacturing. Such reductions can lead to diminished fatigue strength and life. Despite this, relatively few studies have examined how fatigue strength and life change when the thickness of the shot-peened surface-hardened layer is reduced.
Accordingly, this study aims to investigate the reduction in fatigue strength and fatigue life caused by partial removal of the surface-hardened layer in shot-peened AISI 4340 steel, which is widely used in power transmission components such as aircraft and wind turbine gears. After shot peening during manufacturing or maintenance, repair, and overhaul (MRO) processes, a finishing step is applied to remove part of the surface-hardened layer to improve surface roughness. By comparatively analyzing fatigue strength and fatigue life before and after polishing with respect to changes in hardened layer thickness, this study seeks to determine how much fatigue endurance is retained when the thickness of the shot-peened layer is reduced due to degradation.
2. Material and Experimental Methods
2.1 Material and Specimens
The material used in this study was AISI 4340 steel. Its chemical composition is listed in Table 1.
Chemical composition of AISI 4340 steel
Tensile test specimens were fabricated from AISI 4340 steel in accordance with ASTM E8 (ASTM Standards, 2015). The specimen geometry is shown in Fig. 1. The dynamic Young’s modulus and Poisson’s ratio were measured according to ASTM E1876 (ASTM Standards, 2022). A schematic of the measurement system (I-BAT; Mbrosia Co., Korea) is shown in Fig. 2. The measured dynamic Young’s modulus and Poisson’s ratio are presented in Table 2. Based on these values, fatigue test specimens were designed and fabricated; their geometry is shown in Fig. 3. Each specimen underwent heat treatment in accordance with AMS 6414 (SAE International, 1964), which included oil quenching at 816 ± 14 °C for 1 h followed by two-step tempering at 246 ± 8 °C for each 2 h with air cooling. After heat treatment, surface modification was performed by shot peening. The shot peening conditions followed aviation material specifications (SAE International, 2018) and are listed in Table 3. After shot peening, the specimens underwent a baking heat treatment at 246 ± 8 °C for 2 h with air cooling. Following primary precision machining, surface polishing was applied to both tensile and fatigue specimens using sequential micro-precision polishing with #2000, #2500, and #3000 grit papers. The surface-hardened layer produced by shot peening was then partially removed by polishing by 10%, 15%, and 20% to fabricate the final tensile and fatigue specimens for each test condition. Specimens for residual stress measurements were prepared using the same procedure.
Shape and dimensions of the tensile specimen
Measurement system for dynamic Young’s modulus and Poisson’s ratio
Dynamic Young’s modulus and Poisson’s ratio
Shape and dimensions of the fatigue specimen
Shot peening treatment conditions applied in this study
2.2 Experimental Methods
Tensile tests were conducted to evaluate the mechanical properties of shot-peened AISI 4340 steel. The tests were performed using a 1,000 kN universal testing machine (UH-F100A, Shimadzu, Japan) at room temperature with a crosshead displacement rate of 3 mm/min.
Microhardness was measured on the shot-peened surface using a Vickers microhardness tester (Micromet, Buehler, USA). Hardness profiles were obtained along the depth from the surface under a load of 200 g with a dwell time of 15 s.
Residual stress after shot peening was measured using an electrolytic polisher (EP-3, Pulstec, Japan) and X-ray diffraction (XRD) with a residual stress analyzer (Micro-X360S, Pulstec, Japan), based on the JSNDI, EN 15305, and SAE HS-784 standards, using the cosα method.
Fatigue strength and fatigue life were evaluated using an ultrasonic fatigue testing system (Y-UFO, Mbrosia, Korea). Using 20 kHz elastic vibration waves, the maximum displacement was adjusted so that the maximum strain occurred at the center of the gauge section, thereby applying the maximum stress amplitude. A schematic of the ultrasonic fatigue testing system is shown in Fig. 4 (Ahn et al., 2020; Bathias, 2006; Jack Champaigne Electronics Inc., 2001).
Schematic of the ultrasonic fatigue testing system
3. Results and Discussion
3.1 Surface Roughness and Microhardness
Fig. 5 presents the surface roughness measurements of the fatigue specimens prior to ultrasonic fatigue testing. The shot-peened fatigue specimens were divided into Group 1 and Group 2, and the results were averaged. No significant difference was observed between the two groups, and the average surface roughness, Ra, was 1.6698 μm. After polishing 10% of the surface-hardened layer, the average Ra decreased to 0.5848 μm. For specimens polished by 15% and 20%, the average Ra values were 0.5850 and 0.5847 μm, respectively. Compared with the shot-peened condition, surface polishing reduced Ra by approximately 1.0848–1.0851 μm. The surface roughness values for the 10%, 15%, and 20% polished specimens were nearly identical, with only slight differences among them. This reduction is attributed to the removal of dimples created by shot impacts during shot peening. Shot peening forms a hardened surface layer and leaves uneven processing marks in the form of dimples. Although the impacts increase surface hardness and induce compressive residual stress, excessive dimple density can produce a notch effect that reduces fatigue strength and life. Large and deep dimples act as geometric discontinuities that concentrate stress, it becomes the initiation point of new cracks and promoting crack initiation, and decreasing fatigue resistance (Liu et al., 2017; Mattsson, 2025). For this reason, the surface-hardened layer was polished to an acceptable limit, thereby reducing surface roughness. This is particularly important for high-strength materials, which are more sensitive to surface finish.
Average surface roughness of fatigue specimens
Microhardness was measured in the depth direction from the surface for both the shot-peened specimen and the shot-peened specimens with reduced surface-hardened layer thickness. The micro Vickers hardness profiles are shown in Fig. 6. The hardness of the shot-peened specimen ranged from Hv 658–562. After polishing, the hardness ranges were Hv 599–562 for the 10% reduction, Hv 587–560 for the 15% reduction, and Hv 585–560 for the 20% reduction. In the shot-peened specimen, the maximum hardness was observed near the surface, and relatively high hardness was maintained to a depth of approximately 400 μm before converging to a constant value. By contrast, in the specimens with reduced surface-hardened layer thickness, hardness decreased from the surface due to polishing. Although minor deviations were observed, the hardness distribution was generally consistent beyond a depth of 400 μm. The surface hardness increase is caused by shot impacts during peening, whereas polishing partially removes the hardened layer, resulting in a lower near-surface hardness. Because polishing is a micro-scale surface refining process using fine abrasive particles, it does not induce major structural changes. Therefore, the apparent hardness reduction can be attributed to removal of the shot-peened hardened layer and to certain conditions may be the cause such as indentation size effects in shallow microhardness testing (Petrík et al., 2023).
Micro Vickers hardness distributions
3.2 Mechanical Properties
The results of ultimate tensile strength, yield strength, and elongation obtained from tensile tests on shot-peened specimens before and after surface-hardened layer reduction are presented in Fig. 7 and Table 4. The tested specimens included the shot-peened condition and specimens polished by 10%, 15%, and 20%. Three specimens were prepared for each condition, and the average values were reported.
Comparison of ultimate tensile strength and yield strength before and after reducing the surface-hardened layer thickness
Tensile test results
The ultimate tensile strength was highest in the shot-peened condition and decreased slightly as the surface-hardened layer thickness was reduced by polishing; however, the differences were not statistically significant. Because surface cracks can be removed by polishing, tensile strength can often be restored. In this case, the tensile strength remained high due to the surface hardening effect of shot peening. Even with a 20% reduction in hardened layer thickness, the average ultimate tensile strength decreased by only about 1.5%. The yield strength increased primarily due to the work hardening from shot peening and the introduction of compressive residual stress from shot peening. Repeated impacts from spherical shots induce plastic deformation in the surface layer, increasing dislocation density in the crystal structure and enhancing resistance to further plastic deformation. Consequently, the hardness and yield strength of the affected region increase. As the thickness of the hardened surface layer decreased, a slightly greater level of plastic deformation occurred before reaching the maximum tensile strength. As shown in Table 4, elongation increased as the thickness of the surface-hardened layer decreased. Although shot peening significantly improves surface yield strength, this effect is limited to the processed surface layer (Zhan et al., 2012). Accordingly, elongation began to recover as polishing partially removed the hardened layer formed by shot peening.
3.3 Residual Stress Distribution
Fig. 8 shows the residual stress distributions of the shot-peened specimen and the surface-polished specimens after shot peening. The maximum compressive residual stress was −1083 MPa for the shot-peened specimen, −983 MPa for the specimen polished by 10%, −955 MPa for the specimen polished by 15%, and −930 MPa for the specimen polished by 20%. The untreated base material exhibited a maximum compressive residual stress of −456 MPa.
Residual stress distributions of the shot-peened and surface-polished specimens
In the shot-peened specimen, the maximum compressive residual stress occurred at a depth of approximately 100–150 μm below the surface and gradually decreased with increasing depth. Similar trends were observed in the surface-polished specimens. As the degree of polishing increased, the magnitude of compressive residual stress decreased slightly. Although the residual stress was highest in the specimen polished by 10%, no clear or statistically significant differences were observed among the three polished conditions. Nevertheless, even after partial removal of the surface-hardened layer, the compressive residual stress remained significantly higher than that of the untreated base material, demonstrating the sustained effect of shot peening. As shown in Fig. 6, the shot-peening effect in the hardness profile extended to a depth of approximately 400 μm. Therefore, a 15% reduction in the surface-hardened layer corresponds to the removal of roughly 60 μm from the surface. As shown in Fig. 8, the compressive residual stress maintained near-maximum values at depths of 50–100 μm. This indicates that even when the surface-hardened layer is partially reduced by polishing, the beneficial effects of surface modification by shot peening are retained, preventing an abrupt decrease in fatigue strength and fatigue life.
Although surface roughness induced by shot peening dimples was reduced through polishing, the difference in surface roughness among the shot-peened and surface-polished specimens was small. After polishing, fatigue strength and life were expected to improve due to smoother surfaces; however, the shot-peened specimen exhibited the highest fatigue strength and life. This indicates that the high compressive residual stress introduced by shot peening had a stronger influence on fatigue performance than the slight differences in surface roughness. As discussed previously, excessively large or deep dimples can induce a notch effect that promotes stress concentration and crack initiation, thereby reducing fatigue strength and life. Therefore, it will be good that future studies should consider the effects of dimple size.
3.4 S-N Curves
To evaluate the fatigue strength and life of the shot-peened and surface-polished specimens, very-high-cycle fatigue tests up to 1 × 109 cycles were conducted using an ultrasonic fatigue tester. The resulting S–N curves are shown in Fig. 9.
S-N curves
The fatigue limit was 800 MPa for the shot-peened specimen, 725 MPa for the specimen polished by 10%, 590 MPa for the specimen polished by 15%, and 600 MPa for the specimen polished by 20%. The fatigue limit was highest in the shot-peened condition and lowest in the specimen polished by 15%. Compared with the shot-peened condition, the fatigue limit decreased by 9.4% after a 10% reduction in surface-hardened layer thickness, by 26.3% after a 15% reduction, and by 25.0% after a 20% reduction. The slightly higher fatigue limit observed after a 20% reduction compared with a 15% reduction is attributed to greater data dispersion in the 15% condition, likely caused by irregular polishing in the circumferential direction. Therefore, no statistically significant difference was observed between these two conditions. Although surface roughness was reduced by polishing, fatigue strength and life still decreased because the reduction in surface-hardened layer thickness had a more dominant effect.
Sayadi et al. (2024) reported that burnishing as a post-treatment method for additively manufactured metals reduced surface roughness and significantly improved microhardness and fatigue life after near-surface pores were removed. In that study, the fatigue improvement was mainly due to the substantial difference in surface roughness between the shot-peened and burnished specimens. In the present study, however, the surface roughness differences among the polished specimens were minimal. Therefore, whether the slightly higher fatigue limit observed in the 20% reduction condition compared with the 15% reduction condition is partly due to polishing requires further investigation by more precisely controlling surface roughness and introducing larger variations in hardened layer thickness.
The fatigue strength at a very-high-cycle life of 1 × 109 cycles was 775 MPa for the shot-peened specimen, 687 MPa for the specimen polished by 10%, and 590 MPa for both the 15% and 20% polished specimens. Compared with the shot-peened condition, the fatigue strength at 1 × 109 cycles decreased by up to 23.9% after a 15% or 20% reduction in surface-hardened layer thickness. At a high-cycle life of 2 × 106 cycles, the fatigue strength decreased by up to 22.9% after a 20% reduction in surface-hardened layer thickness. The decrease in fatigue strength was not significant when transitioning from the high-cycle to the very-high-cycle regime, indicating that the material can still resist fatigue failure in the very-high-cycle region. The effects of shot peening were maintained because a surface-hardened layer remained even after 20% polishing. Further studies are needed to determine the limiting surface-hardened layer thickness by reducing it further. This study focused only on fatigue characteristics associated with reductions in surface-hardened layer thickness after shot peening; therefore, comparative studies with untreated base material are also required to determine the thickness reduction at which the fatigue limit approaches that of the base material.
3.5 Fatigue Fracture Surface Observation
The fatigue fracture surfaces of specimens tested in the ultrasonic fatigue tests are shown in Figs. 10–12.
Fractography of the fatigue fracture surface of the specimen with a surface-hardened layer reduced by 15% after shot peening (σ: stress amplitude; N: cycles)
Fractography of the fatigue fracture surface of the specimen with a surface-hardened layer reduced by 20% after shot peening (σ: stress amplitude; N: cycles)
Comparison of fractography of the fatigue fracture surface under different stress levels and fatigue lives of the specimens with a surface-hardened layer reduced by 15% after shot peening (σ: stress amplitude; N: cycles)
Fig. 10 presents the fracture surface of a specimen in which the surface-hardened layer was reduced by 15% through polishing after shot peening. In the region corresponding to relatively short fatigue life, surface crack initiation is observed (Fig. 10(a)). By contrast, in the high-cycle fatigue region, characteristic fish-eye fracture features appear (Fig. 10(c)). Figs. 10(b) and 10(d) are scanning electron microscope (SEM) images corresponding to Figs. 10(a) and 10(c), respectively.
Fig. 11 shows the fracture surface of a specimen with a 20% reduction in surface-hardened layer thickness after shot peening. Both Figs. 11(a) and 11(b) exhibit typical fish-eye fracture features in the high-cycle fatigue regime.
Fig. 12 compares fracture surfaces of specimens with a 15% reduction in surface-hardened layer thickness under different stress levels and fatigue lives. At a high stress amplitude of 800 MPa, crack initiation occurs near the surface. At 700 MPa, crack initiation is also observed near the surface at a high-cycle life of 2 × 106 cycles; however, at a longer life of 1.668 × 107 cycles under the same stress, cracks initiate within the interior of the specimen, producing a typical fish-eye pattern. Under lower stress amplitudes in the range of 575–650 MPa, fatigue life extends into the high- and very-high-cycle regimes, and crack initiation occurs predominantly within the specimen with fish-eye morphology. The formation of fish-eye features is attributed to stress concentration around internal defects when the material is subjected to repeated fatigue loading at stress levels much lower than its tensile strength. Under these conditions, these microcracks serve as crack initiation points under fatigue loading, originating at internal defects and gradually propagating, forming a characteristic circular fracture region. This was mainly observed when surface hardening treatment was applied to high-strength steel, as in this study, making internal defects in the material a relatively more vulnerable fracture initiation point. Hence, in shot-peened high-strength steel, such as the material used in this study, surface crack initiation dominates at high stress amplitudes, whereas fish-eye fracture features dominate at lower stress amplitudes (Nishijima & Kanazawa, 1999).
4. Conclusions
(1) The surface roughness was nearly identical for specimens with 10%, 15%, and 20% polishing of the surface-hardened layer produced by shot peening, showing a slight overall decrease compared with the pre-polishing condition.
(2) The shot-peened specimen exhibited an average ultimate tensile strength of 1865 MPa, an average yield strength of 1015 MPa, and an average elongation of 11.0%. After polishing, the specimen with a 10% reduction in surface-hardened layer thickness showed an average ultimate tensile strength of 1860 MPa, an average yield strength of 1070 MPa, and an average elongation of 12.0%. The specimen with a 15% reduction showed 1837.6 MPa, 1496.1 MPa, and 13.9%, respectively, whereas the specimen with a 20% reduction showed 1842.5 MPa, 1530.5 MPa, and 11.9%, respectively.
(3) The micro Vickers hardness of the shot-peened specimen ranged from Hv 658 to 562. After polishing, the hardness ranges were Hv 599–562 for the 10% reduction, Hv 587–560 for the 15% reduction, and Hv 585–560 for the 20% reduction.
(4) The maximum compressive residual stress of the shot-peened specimen was −1083 MPa. After polishing, the corresponding values were −983 MPa for the 10% reduction, −955 MPa for the 15% reduction, and −930 MPa for the 20% reduction. The untreated base material exhibited a maximum compressive residual stress of −456 MPa.
(5) The fatigue limit of the shot-peened specimen was 800 MPa. After polishing, the fatigue limits were 725 MPa for the 10% reduction, 590 MPa for the 15% reduction, and 600 MPa for the 20% reduction. At a very-high-cycle life of 1 × 109 cycles, the fatigue strength was 775 MPa for the shot-peened specimen and 687 MPa, 590 MPa, and 590 MPa for the 10%, 15%, and 20% reductions, respectively. Compared with the shot-peened condition, the fatigue limit decreased by 9.4%, 26.3%, and 25.0% after 10%, 15%, and 20% reductions, respectively. Furthermore, the fatigue strength at 1 × 109 cycles decreased by up to 23.9% after 15% and 20% reductions in surface-hardened layer thickness.
Notes
Seok-Hwan Ahn serves as the Editor-in-Chief of the Journal of Ocean Engineering and Technology but had no role in the decision to publish this article. No potential conflicts of interest relevant to this article were reported.
Acknowledgements
The author acknowledges the support provided by Idewon Company, Republic of Korea.