Parameters optimization for laser shock peening without coating based on SSA-SVR and GA

Laser shock peening without coating, as a typical surface strengthening technology, can effectively enhance the abrasion resistance of key components of aero-engine, and the reasonable selection of process parameters is the key to improving the abrasion resistance. In this paper, laser energy, spot diameter, overlapping rate, and impact number were used as test factors to construct a prediction model of wear rate using the machine learning method and response surface method (RSM), respectively, and process parameters were optimized with the objective of minimum wear rate. The results reveal that the response surface approach is challenging to depict the complicated, nonlinear connection between process factors and wear rate, and the prediction accuracy and objective optimization results are unsatisfactory. In contrast, the support vector regression (SVR) model optimized based on sparrow search algorithm (SSA) showed better results after being optimized by genetic algorithm (GA). The optimized wear rate is decreased by 20.11 % under the best process parameters and the wear resistance of the material is significantly improved. The proposed optimization method of laser shock peening without coating process parameters can significantly improve the resistance to fretting wear of the material.