Numerical-experimental study on the polishing of silicon wafers using coupled finite element-smoothed particle hydrodynamics

The quality and surface roughness of silicon wafers significantly affects the efficiency and quality of follow-up processing. Smoothed particle hydrodynamics is a robust meshless method with good self-adaptability that can be used in the simulation of the polishing process, which has high speed deformation characteristics. In this study, the coupled algorithm of finite element and surface particle hydrodynamic (SPH) has been used to simulate the surface polishing of monocrystalline silicon wafers with a magnetic abrasive finishing process. The effects of machining gap, abrasive particle size, and rotational speed on surface roughness are comprehensively analyzed. In addition, several experiments are carried out on a 3-in.-diameter circular silicon wafer and the results are compared with the simulation results. Our findings show that the decreases in abrasive particle size and also increases in rotational speed significantly deteriorate the surface roughness of the silicon wafer. The obtained results revealed that the machining gap has an optimum condition in which the minimum surface roughness is achieved. According to our results, the best surface roughness value achieved is 63 nm.