Optimizing SLM process parameters using FE analysis for distortion mitigation

Abstract

The investigation aimed to determine the optimal value of process parameters to minimize distortion while reducing the overall printing time. To achieve this, different combinations of process parameters are selected, and 25 sets of FE experiments are prepared using an orthogonal matrix. The coupled thermomechanical analysis with the element birth technique is used to mimic the behavior of Selective Laser Melting (SLM). Initially, a mesh convergence study is performed to reduce computation time with the least sacrifice in results. To minimize the temperature gradient, simulations are performed at different levels of bed temperature ranging from 303 K to 843 K. The minimum distortion was observed for a higher bed temperature of 843 K, as at this temperature, the gradient in temperature is minimum, resulting in the least development of residual stresses. The interactive behavior between laser power and scanning speed is studied, revealing an approximately linear trend with respect to distortion while keeping all other parameters constant. Additionally, it was observed that laser power and layer thickness also exhibited a similar linear trend with respect to distortion. To identify the optimal process parameter value within the design space, the Multi-Island Genetic Algorithms (MIGA) technique is employed, leading to a remarkable minimum distortion of 0.45 mm for a specific set of input process parameters.