Modelling and optimisation of surface roughness from microgrinding of nickel-based single crystal superalloy using the response surface methodology and genetic algorithm

Abstract

Microgrinding is an important method for machining microparts and structures and has received intense focus in the micro-manufacturing field. First, this paper establishes a mathematical model of the maximum undeformed chip thickness and analyses the removal mechanism of nickel-based single crystal superalloy in the microgrinding process. Second, microgrinding experiments perpendicular to the nickel-based single crystal superalloy DD98 (100) crystal plane are designed and carried out using a microgrinding tool; using the response surface methodology, which is based on central composite design, the influence of the spindle speed, feed rate, and grinding depth on the grinding surface roughness is analysed. First-order and second-order prediction models of the surface roughness are established. Then, according to the residual analysis and ANOVA, an accurate model is obtained. The model’s accuracy and scientificity is verified by experiments. Finally, optimum grinding parameters of the second-order model for minimising the surface roughness are acquired using a genetic algorithm. The result shows that the spindle speed has the largest influence on the surface roughness, followed by the feed rate, and the grinding depth has the smallest influence. When the spindle speed is 56.8 kr/min, feed rate is 36 μm/s, and grinding depth is 14 μm; the grinding surface roughness is the minimum (Ra = 563 nm). These provide a significant theoretical and practical reference for the microgrinding process of nickel-based single crystal superalloy.