Design of grinding factors based on response surface methodology

The presented paper describes a systematic methodology for empirical modelling and optimization of the plunge centreless grinding process. The assessment of microgeometric quality defining quantity is supported by post-process surface roughness measurements. The design of grinding factors is based on response surface methodology, which integrates a design of experiment, regression modelling technique for fitting a model to experimental data and basic optimization. Central composite response surface design has been employed to develop a second-order surface roughness model. The model has been fully constructed by determination of its structure and regression coefficients. The final goal of experimental study focuses on determination of optimum centreless grinding system set-up and operating conditions for minimization of surface roughness. The computer-aided single-objective optimization, solved by non-linear programming and genetic algorithm, is applied. The results of two different optimization approaches for determination of optimal operating conditions are compared. Finally, further research directions are presented.