Prediction of grinding parameters based on specific surface integrity of hard-brittle materials

The selection of appropriate processing parameters is essential for achieving high-quality grinding of hard-brittle materials. This paper presents a prediction method of grinding parameters (grinding depth, feed rate, and wheel speed) to achieve specific surface integrity parameters (surface roughness, chipping layer depth, and subsurface damage depth). The method integrates the Optimum Latin hypercube design, a surface integrity model, and a genetic algorithm back propagation neural network (GA-BPNN) model. The surface integrity model considers the inclination effect of subsurface cracks and the elastic recovery of workpiece material. The GA-BPNN model is developed for constructing the relationship between grinding parameters and surface integrity parameters. To validate the prediction method, grinding experiments are performed on fused silica samples. The elastic recovery of fused silica is determined by indentation experiments. The surface roughness, chipping layer depth, and subsurface damage depth of the ground samples are measured. The results show that the surface integrity model has an average relative error below 7.5%, and the regression correlation coefficient for the GA-BPNN model exceeds 0.8. The prediction method achieves a maximum relative error of less than 6.5%. The influences of grinding parameters and material elastic recovery on the subsurface crack inclination and surface integrity parameters are investigated theoretically. The results show that increasing the grinding depth or feed rate, or decreasing the wheel speed or elastic recovery coefficient, leads to an increase in the maximum instantaneous inclination angle. The instantaneous inclination angle decreases while the instantaneous peak-valley value, chipping layer depth, or subsurface damage depth remains nearly unchanged with an increase in the elastic recovery coefficient. The research is conducive to evaluating the ground surface integrity and optimizing the grinding process to achieve efficient and low-damage machining of brittle materials.