Prediction of surface roughness in abrasive water jet machining of CFRP composites using regression analysis

Abstract

In the present work, two different carbon fiber-reinforced plastics (CFRP), namely, unidirectional (UD) and UD with a woven fabric surface, were selected in order to investigate the surface roughness as a result of machining with an abrasive water jet machine (AWJM). The experiments were conducted based on an L27 orthogonal array by varying the process parameters such as jet pressure (JP), traverse speed (TS), and standoff distance (SOD). Regression models were developed to predict the correlation between the input parameters and the surface roughness for each composite. The experimental results showed that increased pressure with lower TS and SOD produced a good surface finish on the composites. Very low roughness was achieved when machining the UD with woven fabric surface CFRP composite. The developed regression models indicated the significance of the design (at a 95% confidence level) with reduced error. The JP followed by the SOD were found to be the major parameters affecting the surface roughness. Finally, the deformation, delamination, and machined surface morphology were studied using a non-contact three-dimensional surface measurement system and scanning electron microscopy (SEM) images.