Undeformed chip width non-uniformity modeling and surface roughness prediction in wafer self-rotational grinding process

Silicon wafers are commonly thinned employing an ultra-precision grinding machine based on a workpiece self-rotational principle. The diamond grains, stochastically located on the grinding wheel surface, remove the workpiece material and generate a difference in the size and shape of undeformed chips. However, an in-depth understanding of the grain-workpiece interaction randomness in wafer thinning process and its link with surface roughness has not been revealed yet. In this paper, a new index, i.e., undeformed chip width, is first proposed to analyze and describe the stochastic characteristics in material removal process as well as predict the surface roughness of ground wafer. The grinding wheel morphology is constructed by the random operation for grain size and position. Then, a 2D topography generation model is established to calculate the undeformed chip width. The non-uniform distribution characteristics of undeformed chip width under different grinding conditions are further analyzed. Finally, the relation between undeformed chip width and surface roughness of ground wafer is presented. The correctness of developed models is verified using a series of grinding experiments.