Research on roughness prediction model of robotic sanding carbon fiber reinforced plastics

Abstract By taking a pneumatic eccentric grinder and round sandpapers as research objects, the modeling method is proposed to accurately predict surface roughness of robotic sanding carbon fiber reinforced plastic parts. A four-factor, four-level orthogonal experiment was conducted which took four process parameters into account: pressing force, feeding speed, grit size, supply air pressure. The multiple linear regression analysis then applied create prediction model. Furthermore, accuracy validation results show that model’s error within ±9%.