Investigation of the effect of sanding parameters on surface roughness of S235 steel and PP plastic by response surface method and artificial neural networks

In this study, the surface roughness values obtained as a result of sanding the surfaces of S235 and polypropylene (PP) materials with different sanding parameters were investigated. Sanding size (P60, P100, P180, and P320) and sanding time (15, 30, 45, and 60 seconds) were used as sanding parameters. The effect ratios on material roughness were determined by analysis of variance. It was concluded that sanding size was the most effective parameter on surface roughness. Mathematical models and artificial neural networks were created for prediction. Surface roughness decreased with increasing sanding size number and sanding time. In the applied parameters, it was measured that there was a decrease in surface roughness of over 80% for both materials. Optimum parameters were determined based on the criterion of minimizing the surface roughness (Ra) value. As a result of the optimisation with Response Surface Method (RSM), a 60-second sanding process with P320 sandpaper was determined as the optimum condition for achieving the minimum Ra value.