Lapping optimization of additively manufactured ABS-carbon fiber composite parts using the soft computing technique

Abstract

FDM offers cost-effective, rapid component fabrication but often results in poor surface quality, requiring post-processing. Former methods to enhance surface roughness, whether via chemical or mechanical means, had drawbacks including lowered mechanical properties, increased costs, and prolonged processing times. The main contribution of the present study was to explore the effects of different lapping process parameters, such as velocity, pressure, abrasive concentration, and abrasive size, on both the surface roughness of ABS-carbon fiber composite parts manufactured via FDM and the rate of material removal. To accomplish this task, the Box-Behnken design of experiments was executed, and a thorough analysis of variance was conducted to ascertain the level of significance of each parameter with respect to the outputs. The findings indicate that the size of abrasives is the most significant factor that affects the surface roughness and the rate of material elimination. The combination of genetic algorithm and artificial neural network was employed for the purpose of both prediction and optimization. While the total goodness function for MRR was 1.99, it was 1.96 for surface roughness, showing a high precision model. The outcomes indicate that the employment of the optimized parameters recommended by GA-ANN exhibits an outstanding correspondence with the experiment.