Empirical modelling of grinding force based on multivariate analysis

The force exerted during the grinding process is a key influential factor for grinding quality and the design of grinding fixtures. In this research, a model of the grinding force for aerospace alloys was built using an empirical approach. A robust design of experiments that included orthogonal arrays and the signal-to-noise ratio were tightly integrated to acquire reliable force data. In order to obtain clear and correct force signals, an optimized matching method to select parameters and filters was put forward based on a correlation function. The results verified that a wavelet filter gives much better accuracy. A set of empirical models of the grinding force for superalloy CMSX4 were build up using multivariate analysis and these models are characterized by the process parameters such as depth of cut, feed rate, wheel speed as well as wheel diameter. They have been verified through further experiments and the correlation coefficients between these models and experimental data are around 0.91.