A numerical approach to predict the machined surface topography of abrasive belt flexible grinding

Abrasive belt grinding is considered a flexible and precision machining method. The complicated contact status prevents the traditional simulation model from accurately predicting the machined surface topography, so this paper develops a new numerical simulation approach to solve this issue. The abrasive belt surface topography was detected using a 3D appearance optical scanning apparatus and was processed by surface fitting and noise filtering. Based on the Johnson transformation system and filter impulse function, the surface topography generation of non-Gauss abrasive belt was realized. The simulation result showed that it was well consistent with the measured topography. Subsequently, a vibration model was superimposed on the original moving track of abrasive grain in consideration of flexible grinding characteristic to establish the grain kinematic model of abrasive belt. On this basis, the numerical simulation and experimental investigation on machined surface topography and roughness of abrasive belt grinding were conducted. The result revealed that the errors between simulated and measured surface roughness were less than 5%. It is deduced that this numerical approach can be used to accurately predict the machined surface topography of abrasive belt flexible grinding.