Variable Damping Profiles for Laser Shock Peening Simulation Using Modal Analysis and the SEATD Method

Abstract

Laser shock peening (LSP) is a surface engineering technique, which aims to increase the fatigue life of various metallic components by inducing compressive residual stress at or near their surface. The finite element method (FEM) is used to identify the most suitable parameters in LSP. Various explicit analyses with artificial material damping are used to attain quasistatic equilibrium between laser shots. Dynamic relaxation (DR) is a well-known conventional technique that uses constant artificial damping to settle an excited model to quasi-static equilibrium. In contrast, the recently developed “Single Explicit Analysis using Time-Dependent Damping” (SEATD) method employs variable damping and performs better in terms of simulation time and accuracy. While recent study has shown that a variable damping profile used in the SEATD technique is beneficial for an LSP set up, identifying the most suitable variable damping profile in general is still ambiguous, given the variety of possible set-ups and boundary conditions. In this paper, a systematic procedure to strive for the best variable damping profile is developed, based on the excited modal parameters of the model. The simulation results are compared with those of an optimum constant damping profile developed using the conventional dynamic relaxation technique, as well as for the best variable damping profile based on exhaustive trial-and-error. The simulation case studies involve circular LSP shot(s) of 5.5 mm diameter spot size applied to Al 2024-T351 aluminum alloy plate under different boundary conditions. Dissipation rates of stain energy, kinetic energy, and total energy and the accuracy of surface residual stresses are investigated to compare the performance of different damping profiles. The results indicate that the proposed method involving modal analysis to systematically identify a variable damping profile, to promote simulation efficiency, appears to work well.