Prediction and measurement for grinding force in wafer self-rotational grinding

Abstract

The ultra-precision grinding technology based on a workpiece self-rotational principle is extensively used for silicon wafer thinning in the chip post-processing. Nevertheless, owing to the random nature of diamond grains and the unique machining manner of rotating components, accurate prediction and real-time monitoring of grinding forces remain a significant challenge in wafer self-rotational grinding (WSRG) process. This work establishes an improved theoretical model and builds a novel in-situ measurement system to analyze and detect the grinding force. Firstly, the motion trajectory and contact condition of diamond grains are explored. The grain–workpiece interactions are divided into rubbing, plowing, cutting and brittle stages. Secondly, the grinding force prediction model is proposed, which considers the combined effects of grain wear, grain randomness characteristics, brittle-to-ductile transition, elastic rebound, strain rate and grinding marks. Moreover, piezoelectric force sensors are employed to build the in-situ measurement system. The developed system is proven to simultaneously measure grinding forces online with continuous and stable output in the x, y and z directions. The impacts caused by feed rate, wheel speed and wafer speed are further revealed. Experimental results are consistent well with predicted ones, demonstrating a high prediction accuracy of the proposed grinding force model. Finally, a new index, namely the processing factor, is introduced to calculate the subsurface damage depth and residual stress for various machining parameters. This work not only enhances the understanding of predicting and measuring grinding force, but also provides a new method to evaluate the subsurface defects in brittle material grinding.