Intelligent Grinding Process via Artificial Neural Networks

Abstract

Surface roughness of the ground parts and the grinding forces are two important factors for the assessment of the grinding process. The surface roughness directly influences the functional requirements of the workpieces and the grinding forces are an important criterion for the achievable material removal rate. The establishment of a model for the reliable prediction of surface roughness and grinding forces is a key issue. This work deals with design of appropriate control strategy for prediction of grinding forces and surface roughness as one of the important indicators of the machined surface quality via applying Artificial Neural Networks (ANNs) through special sensors integrated into the machine tool. A micro-grinding process of Ti6Al4V was chosen. The model was verified by various experimental tests with different grinding and dressing parameters. It was found that the predictions made by the ANN model matched well with the experimental results.