Density-based topology optimization of multi-condition peening pattern for laser peen forming

Abstract

Laser peen forming (LPF) is a novel forming process with high flexibility but faces challenges in determining process conditions to produce complex geometries. This study proposes a process planning method to determine peening patterns with multiple process conditions based on topology optimization, providing an effective approach for complex shaping of LPF and liberating the high flexibility of the process. The process planning is formulated as a partial differential equation (PDE)-constrained optimization problem, employing the eigen-moment as the design variable. The formulation is discretized using numerical methods, and a continuous optimization model is derived via density-based topology optimization for efficient resolution. The ordered Heaviside interpolation is proposed specifically for LPF process planning to promote intermediate densities to the desired values. The globally convergent method of moving asymptotes (GCMMA), a well-established solver for topology optimization, is used to resolve the process planning of a flexible forming process. The process planning model designs the forming strategy of the dome and cylinder geometries using different numbers of process conditions. The results demonstrated that process planning with multiple process conditions is superior to using a consistent process condition. Finally, experiments are conducted using the optimized peening patterns to form the two geometries. The deformed results perform a high level of conformity to the desired objectives with a maximum error of 0.473 mm and 0.833 mm, respectively, validating the effectiveness of the proposed process planning method.