Abstract: Since it is difficult for ultrasonic vibration compound machining to get effective cutting mechanism mathematical model through dynamic analysis, and experimental study is shown an effective method to solve this problem, following researches by means of central composite design(CCD) testing are carried out. 4-factor and 3-level SiC wafer ultrasonic vibration compound machining test scheme is designed, and then second-order relational model is established between tangential cutting force, surface roughness, and their main process parameters (wire saw speed, workpiece feed rate, rotational speed, and ultrasonic amplitude) by using response surface methodology. According to multiple quadratic fitting of testing data, quadratic equation of cutting force and surface roughness is obtained. Constrains of actual machining condition upon the parameters are analyzed further. With the goal of improving surface quality (minimized surface roughness) of SiC wafer, the parameters optimization model is established. Particle swarm optimization algorithm and its procedure are designed to solve the model. Test proves that the algorithm could achieve optimized process parameters which satisfy multiple constraints rapidly and effectively.

Key words: Central composite design, Particle swarm optimization, Process parameters, SiC wafer, Ultrasonic vibration