A comparative study of force models in monitoring the flank wear using the cutting force coefficients

Difficult-to-machine materials, such as nickel-based alloys, are widely utilized in aerospace industry, while their low thermal conductivity and high temperature strength lead to the rapid wear of cutting tool. Tool wear monitoring is regarded as one of the useful methods to guarantee the product quality and maximize the tool utilization. Due to the high nonlinearity and stochastic characteristics of tool wear process, it is difficult to establish a general tool wear monitoring model. This work contributes to find out the most suitable cutting force model by comparing their ability and performance in monitoring the flank wear. The tool wear monitoring is realized through developing the relationship between cutting force coefficients and tool wear, and the coefficients are calculated based on the average cutting forces in different feed rates. Then, correlational analysis is performed to select sensitive coefficients. Finally, the selected coefficients are normalized and then trained by the feed-forward backprop neural network. Experiments are conducted to compare the four different models in cutting force prediction and tool wear monitoring by three criteria. The cutting force model including the shearing forces, the edge forces, and forces due to the tool wear gives the best results. The obtained results also show great suitability for different cutting conditions.