2. Introduction to Residual Stress Impingement Methods
In general, RSI methods aim to impart compressive residual stresses into a metal surface that will offset external applied, or existing residual, tensile stresses. By reducing these stresses, the crack initiation and/or propagation rate may be decreased due to reduced stress intensity being present at the site of a small defect, or at a crack tip. These compressive residual stresses have a maximum effect while the defect remains within the compressively stressed material layer, and so the effective penetration depth achieved by a given RSI method is key. The most commonly utilized RSI treatments all have advantages as well as drawbacks that affect the final state of the alloy surface microstructure, the magnitude and depth of the imparted compressive stress, the resultant surface finish, and the efficiency of its application. Selected treatment methods will be briefly summarized.
Shot peening is one of the oldest RSI methods that are still commonly utilized today and consists of firing high-hardness shot (often glass, metal, or ceramic, depending on the target and goal) at a metal surface to impart compressive residual stress (CRS). This process is demonstrated schematically in
Figure 1A. The SP process is controlled by the Almen intensity and coverage, where the Almen intensity reflects the effect of spot size, shot hardness, speed, flow rate, and impact angle
[10]. An advantage of SP is its relative ease of application compared to some other methods, and the deep knowledge base associated with the history of its use
[11]. One downside of the method is that, due to the aggressive bombardment of a surface with shot, the process parameters must be carefully controlled to achieve the desired surface roughness and consistent surface coverage, which is not as uniform as with other methods
[12]. The overtreatment of a surface during SP may result in brittle cracking in the deformed surface layer, folds that may conceal defects, and even embedded shot, all of which could aggravate corrosion or crack formation
[13]. The impact of shot on the alloy surface leaves dimples that form a gradient of compressive residual stress that commonly reaches 0.25–0.50 mm in depth
[14][15]. The ultrasonic nanocrystal surface modification (UNSM) process uses a tungsten carbide-tipped applicator that is pressed down onto the alloy surface with a specific load and vertically vibrates at an ultrasonic frequency while moving systematically around the alloy surface
[16]. The UNSM process is schematically illustrated in
Figure 1B. This treatment is computer-directed; therefore, a uniform surface coverage is achieved that results in a low hardness increase and a low surface roughness increase in a variety of alloys
[16]. UNSM commonly forms a nanocrystalline grain structure below the alloy surface, and the CRS depth tends to be similar to that seen with SP
[16].
Figure 1. Schematic representations of (A) the shot peening process, and (B) the ultrasonic nanocrystal surface modification process.
Laser shock peening utilizes laser energy bombardment to impact a surface water layer and create plasma pulses through an underlying material, which drives in the CRS
[6][15]. The LSP process is schematically illustrated in
Figure 2A. The advantages of LSP include the fact that this process is computer-controlled, and each laser shot is measured and the output energy is recorded, making the process highly traceable and repeatable. The surface coverage is uniform and the CRS depths are consistent. CRS depths typically range from 0.75–1.25 mm when no ablative layer is used (
Figure 3)
[8]. When an ablative layer is used, the maximum residual stress is similar to that which is achieved without an ablative layer but the maximum penetration depth is reduced; thus, this treatment method is less common
[17]. The LSP process is tuned through the laser spot size and power density applied to the alloy surface, as well as the beam overlap, all of which impact the CRS depth and final surface roughness
[6][7]. One disadvantage of LSP is that this method must be utilized in controlled settings, and strict control of the surrounding area is required due to the hazard of the laser, which can make LSP one of the more expensive RSI methods to deploy. Overtreatment through LSP occurs through the application of excessive power density, which can actually form tensile stress in the surface layer and can cause melting
[18][19]. However, this is easily avoidable through preliminary research on best practices for treating a given alloy type. Lastly, LPB utilizes a hydraulically pressed bearing to apply force to an alloy surface without applying heat or causing significant microstructural deformation (
Figure 2B)
[20].
Figure 2. Schematic representations of (A) the laser shock peening treatment process (shown without the damping layer), and (B) the low plasticity burnishing process.
Figure 3. Impact of LSP applied 1–3 times on 316L steel weldments, compared to the residual stress profile without treatment. LPwC—laser peening without ablative coating
[8]. Reprinted with permission from
[8]. Copyright 2020, Metals Journal in MDPI.
This process induces minimal changes to the alloy surface profile and is proven to impart similarly high magnitude compressive residual stresses as are achieved with LSP, to a depth of roughly 1 mm or greater. An example dataset from Inconel 718 is illustrated in
Figure 4, which compares the CRS depth and magnitude achieved by SP, LSP, and LPB
[4][21].
Figure 4. A comparison between the residual stress profiles achieved by SP, LSP, and LPB in Inconel 718
[21]. Reproduced with permission from
[21]. Copyright 2003, Journal of Engineering for Gas Turbines and Power on behalf of ASME.
Due to the use of a hydraulically loaded ball bearing, this process is best used on open surfaces, where it can easily be utilized in a CNC machine to complement a typical machining process before the part is completed. The benefits and risks of the SP, UNSM, LSP, and LPB processes have varying impacts on the ability to improve corrosion performance in alloys of varying hardness, which will be discussed in the following section.
3. Corrosion Minimization
Corrosion is a highly surface-sensitive phenomenon; the effect of surface roughness on corrosion susceptibility is well documented in the literature
[22][23]. Residual stress impingement imparts compressive stress into the alloy surface, which can improve corrosion resistance through a combination of work-hardening and grain refinement. However, RSI treatments can also increase the surface topography, which may reduce the benefit of the compressive stress and refined microstructure. Previous works have demonstrated that there is a relationship between the target alloy’s passivation mechanism, the surface topography caused by the RSI, and the alloy corrosion resistance
[4][24][25].
This relationship is evident when viewed across the hardness spectrum. Considering the effects of SP, firstly on low-hardness alloys, Curtis et al. evaluated shot-peened 2024-T351 (137 HV) via potentiodynamic analysis in 3.5 wt % NaCl and measured a 5-fold increase in the corrosion current density (i
corr) compared to the as-polished sample
[26]. After 24 h in this solution at an open circuit, the 2024–T351 experienced an increased pitting rate after SP
[26]. Similarly, the SP treatment of 7075-T651 (175 HV) increased the alloy’s surface roughness from 0.32 μm to 5.81 μm, after which Zupanc and Grum measured a 2.5-factor increase in i
corr in 0.1 M NaCl
[27]. When evaluating AISI 430 steel (162 HV) after SP, Peltz et al. observed a 10-fold increase in i
corr in 0.05 M NaCl, due mainly to the increase in surface roughness from 0.02 μm to 3.18 μm
[28]. These examples demonstrate that the RSI-induced roughness in low hardness alloys detrimentally affects the passivity and corrosion resistance.
In the medium-hardness AISI 304 stainless steel (240 HV), Iswanto et al. demonstrated, in intravenous Otsu-Ringer lactate solution, that the pitting rate initially increased by as much as 20 times when SP was conducted for 5 min but decreased as SP was conducted for longer periods of time to achieve better coverage and more plastic deformation across the alloy surface
[29]. Treating the 304 SS for 40 min more than doubled the surface hardness to reach 496 HV, and the pitting rate decreased from the as-polished rate of 0.042 mpy to 0.036 mpy
[29]. In 316L stainless steel (220 HV), Peyre demonstrated via potentiodynamic analysis that both SP and LSP (8 GW/cm
2) achieved similar improvements in the alloy’s pitting resistance, and both improved the i
corr in 0.5 M NaCl, despite slightly rougher surface finish as well as martensite formation following SP
[30]. Interestingly, the open-circuit potential of the SP-treated 316L was also roughly 100 mV greater than the LSP-treated 316L
[30]. Various authors have attributed the LSP-induced improvement in corrosion resistance to slight melting and resegregation at the alloy surface, such that the high-energy, work-hardened material is less exposed to aggravate corrosion reactions
[18][31]. Considering high hardness alloys, Cuifini et al. demonstrated in 300–325 HV super duplex stainless steel that SP treatment resulted in a 3- to 4-fold increase in mass loss through salt fog cabinet exposure
[32]. Overall, these findings support the need to understand the alloy’s passivity and the dependence on surface morphology before applying RSI, especially when considering techniques such as SP that can cause significant roughness and plastic deformation.
UNSM, despite achieving a more uniform surface finish than SP, can also increase corrosion susceptibility due to the added dislocation density, except in specific circumstances. In the low-hardness AZ31B (60 HV), Hou et al. demonstrated in simulated body fluid (SBF) and 0.1 M NaCl that a 2-fold increase in i
corr occurs following UNSM
[33]. With Alloy 600 (180 HV), however, UNSM treatment below the critical amplitude was demonstrated by Kim and Kim in 1 wt % NaCl to improve the alloy passivation by creating a reactive nanocrystalline surface with low roughness, which also reduced the pitting susceptibility
[34]. However, higher-amplitude treatments increased the surface roughness and created crevice-forming features that promoted aggressive chemistry formation, oxide rupture, and pitting
[34]. On 4140 steel, in the annealed (183 HV) and nitrided conditions (450 HV), UNSM treatment decreased the corrosion resistance of the steel in alkaline, neutral, and acidic 3.5 wt % NaCl solutions
[35]. In contrast, when Li et al. evaluated UNSM on 304SS (240 HV) in 3.5 wt % NaCl, the data revealed increased nobility and passivity, as well as improved pitting resistance
[36]. Closer inspection via transmission electron microscopy (TEM) and surface analysis showed that the UNSM treatment created a cleaner surface with fewer MnS inclusions, and the nanocrystalline surface layer showed better Cr distribution, such that the passive film achieved greater Cr enrichment and improved stability
[36]. It is worth noting that Kim evaluated UNSM on 316L (220 HV) in 3.5 wt % NaCl, however, and demonstrated that the improvement in pitting following UNSM depends on the level of sensitization present in the alloy; when sufficiently sensitized, UNSM can actually accelerate the pitting attack
[37]. This comparison between 4140, 304SS, and 316SS demonstrates that a strong passivation mechanism in the underlying alloy may assist the RSI to improve corrosion resistance; however, the underlying alloy metallurgy, such as a highly sensitized state, can reverse this trend. In relatively high-hardness Ti-6Al-4V (380 HV), Cao et al. observed increased pitting susceptibility and a 2-fold increase in i
corr in SBF after applying UNSM treatment
[38]. The disparities between alloys and variable RSI-related corrosion improvement demonstrate the need for a more microstructural-based understanding of why improvements are achieved in some alloys, but not in all.
LSP has a more consistent track record of improving corrosion resistance when not overly applied. In the low hardness 5083–H112 (72 HV), Yang demonstrated that LSP achieved the greatest improvement in the surface corrosion resistance when lower power density was applied for a smoother surface finish, estimating that repassivation in 3.5 wt % NaCl was more stable with less topography
[39]. In 6082–T651 (85 HV), LSP evaluations across the power density range of 5.7–15.8 GW/cm
2 demonstrated that this alloy is less sensitive to power density than 5083 in dilute NaCl, since nearly all LSP treatments within the study achieved a similar reduction in pitting susceptibility, despite a 5-fold increase in surface roughness (0.72 μm to 3.74 μm in the L-direction)
[19]. Trdan and Grum later demonstrated that LSP improves the polarization resistance of 6082–T651 by 25-fold, expands the passive electrode voltage region on the potentiodynamic curve, and decreases i
corr as much as 10-fold compared to untreated 6082–T651 in 0.6 M NaCl
[18]. In 7075–T6 (175 HV), Aravamudhan demonstrated, through potentiodynamic polarization, that LSP reduced i
corr by 2–3 times in 3.5 wt % NaCl
[40]. Pitting was observed to occur preferentially near the valleys formed during LSP treatment on the 7075–T6, where chemistry could more easily acidify, and the magnitude of valley formation depended on the power density selection
[40]. Considering LPB, Cao demonstrated that LPB reduced mass loss in AZ31B (60 HV) in 5 wt % NaCl over 7 days’ immersion, and the corrosion rate was more consistent than that measured on the non-treated samples
[41]. The main cause for this improvement was hypothesized to be the smaller grain size and reduction of intermetallic phases near the alloy surface, as well as the smooth surface finish and aligned crystalline orientation generally caused by the LPB, all of which promote corrosion resistance in Mg (but this is likely different in other alloys, especially regarding the effect of grain size)
[41]. These collective findings demonstrate that the impact of RSI methods on corrosion susceptibility is largely dependent on alloy, the intensity of surface treatment, and the resulting microstructure. The impact of microstructural changes and surface deformation on the intermetallic presence and oxide stability will also play a significant role in the final corrosion susceptibility. If properly applied, these results demonstrate that specific RSI/alloy combinations and processing could reduce a component’s tendency to corrode over a service life, which is an added benefit to well-known CRS-induced fatigue life improvements.