Most gas turbine equipment uses nickel-based superalloy, Inconel 718. The alloy's response to fatigue influences its service life, wear, oxidation and corrosion when exposed to thermomechanical loads. Most of these mechanisms originate from the surface and are very sensitive to the microstructure. Shot peening is one of the widely used surface treatments to enhance the microstructural properties of the surface. One of the key outcomes of shot peening is Compressive Residual Stress (CRS), which helps in improving fatigue life. But CRS is observed to relax and re-distribute under various thermomechanical loads. The amount of relaxation depends on the strain hardening during the shot peening process, which is difficult to measure due to its dependency on strain rate gradients. Finite Element Modeling (FEM) combined with the Discrete Element Method (DEM) has been extensively used for decades to simulate the shot peening process. Still, minimal studies focused on microstructural changes during shot peening that potentially influence strain hardening. This warrants a need for modeling the shot peening process at a microstructural level, capturing the effects of grain size and dislocation density, for accurate estimation of relaxation of CRS.
This work aims to study and model the micromechanics during the shot peening process and subsequent cyclic loading. First, a dislocation density-based crystal plasticity model is developed to model the target surface. All the complex cyclic deformation characteristics of IN718, (1) the Bauschinger effect, (2) Mean stress relaxation, and (3) Cyclic softening, are included in the rate-dependent material model. The model can also capture the grain size effects and reduction in precipitate size due to dislocations shearing through them. Next, the constitutive model is used to simulate the shot peening process. A new approach is proposed, which uses the results from a single shot and feeds all the state variables as initial conditions for the next single-shot simulation. This approach reduces the model size as well as the simulation time. A comprehensive analysis is conducted to study the influence of (1) single load-unload case and (2) cyclic load on the relaxation of CRS.