The effects of surface pits and intermetallics on the competing failure modes in laser shock peened AA7075-T651: Experiments and modelling

Highlights

• Critical surface imperfections reduce the benefits of LSP by promoting surface crack initiation.

• An FE model was developed to simulate residual stress distribution induced by LSP and its redistribution after the formation of surface pits.

• A lifing framework was proposed for LSP samples, considering the competition between surface defects and compressive residual stresses.

Abstract

The effects of laser shock peening (LSP) on the fatigue life of AA7075-T651 were investigated. The combined influence of surface imperfections (i.e. pits and intermetallics), compressive residual stresses (CRS) and the applied stress on crack initiation sites (surface or subsurface) and the associated fatigue life were investigated. Critical surface imperfections were found to significantly reduce the benefits of LSP in life improvement, by promoting surface crack initiation despite the resisting effects of CRS. To facilitate quantifying the effects of LSP on fatigue life, a finite element (FE) model was developed to simulate residual stress distribution induced by LSP, as well as its redistribution caused by the formation of surface pits. Based on the FE results, a method identifying whether the specified surfaces pits and intermetallics are critical to lead to surface cracking at given stress conditions was proposed, based on the Smith-Watson-Topper method and the Murakami’s model respectively. The interaction between surface imperfections, CRS and the applied loads were taken into account in this method. In addition, a fatigue life assessment framework was proposed based on the prediction of crack initiation sites, which was validated to be reliable in efficiently evaluating the efficacy of the applied LSP in improving fatigue life.

Introduction

Competing failure modes in fatigue refer to distinct failure distributions (i.e. surface, subsurface and internal areas) in the material tested under different conditions. The variety of failure modes have been attributed to the variance in surface conditions, the size and distribution of inclusions, temperature, stress levels, etc [1], [2]. It has been summarised that subsurface and internal failure tends to result in longer fatigue life than the failure originating from the surface. This phenomenon is closely related to the imperfection of material surface which usually contains grooves, dimples, or microcracks introduced by the applied surface treatments [3], [4]. These surface imperfections, as well as inclusions located at the surface, are local stress raisers which favour crack initiation at the early stage of fatigue life [5], [6], [7], [8]. This process is additionally facilitated by the lack of constraints at the free surface of the material [9], [10]. Therefore, surface crack initiation normally leads to low-cycle fatigue (LCF). By contrast, when the stress is reduced to a level being too low to activate surface crack initiation, the failure dominating area is transferred to subsurface or internal sites where cracks initiate due to local strain accumulation. This tends to result in a longer crack initiation process, leading to high-cycle fatigue (HCF) [2], [11].

According to the mechanism of competing failure modes, resisting surface crack initiation by improving the surface quality and reducing the surface stress concentration are effective ways of enhancing the fatigue resistance of material. Laser shock peening (LSP) is a relatively new surface treatment method following this mechanism. Compared to conventional shot peening, LSP uses high power laser pulses to generate a surface compressive residual stress (CRS) distribution in structural components, instead of impacting the surface using particles. Typically, the CRS layer generated by LSP is 1–2 mm, which is about 3 times deeper than that generated by shot peening [12], [13], [14]. In addition, LSP normally leads to a better surface finish with lower surface roughness than shot peening, which is advantageous in retaining the benefits of CRS in improving fatigue life [12], [13].

It has been reported that the CRS generated by LSP is effective in mitigating surface crack initiation and the subsequent crack growth behaviour [15]. At relatively low load levels, surface failure in baseline samples (i.e. without LSP treatment) is usually moved to the subsurface area where the effects of CRS become weakened in LSP samples, achieving 5–10 times life improvement [16], [17]. However, surface crack initiation and growth can be reactivated when the applied load is increased to a level being sufficient to counterbalance the surface CRS. Under this condition, the effects of CRS are limited, degrading the benefit of LSP in life improvement to 2–5 times [16]. It has also been reported that existing surface imperfections or corrosion effects tend to accelerate this degradation by introducing additional stress concentration features at the surface [18], [19], [20].

Although the mechanisms of the competing failure modes occurring in materials treated with LSP have been well understood, investigations on predicting such behaviour and the associated life assessment are still relatively scarce. Some published researches focus on the development of the Kitagawa-Takahashi diagram, which builds up the relationship between defect size and fatigue limit [13], [21], [22], [23]. However, it cannot be used in life prediction at different load levels and the effects of CRS are rarely considered. This impedes the effective evaluation of the efficacy of the applied LSP. In the present study, a modelling-based method is proposed to predict the failure mode (i.e. safety–critical area) in laser shock peened AA7075-T651 at varying load levels, considering the effects of surface pits, intermetallics and CRS. Life assessment is also carried out based on the stress and strain data at the predicted safety–critical area, using the Smith-Watson-Topper method.