Statistical analysis of magnetorheological ball end finishing parameters using response surface methodology

Abstract

Bidisperse magnetorheological polishing fluid (BMRPF) was used to gently finish work-piece surfaces on ball end magnetorheological finishing (BEMRF) tool. BMRPF was synthesised with optimum composition obtained by vibration sample magnetometer (VSM) test and magnetorheological characterisation which gives maximum magnetisation and yield behaviour. A five-levels three-factors central composite design of experiments (DOE) was used to develop an experimental plan for investigating process parameters' effect on percentage reduction in surface roughness (%ΔRs). Response surface methodology (RSM) was used to predict profile and surface roughness. Experiments were conducted on mild steel and surface roughness before and after experiments were measured. An effective second order response surface model for %ΔRs was obtained with significant factors and their interactions. Physical interpretation of process parameters' effect on %ΔRs was analysed and optimum values were obtained. Experiment was further conducted with optimum process parameters, and %ΔRs was calculated then compared with maximum %ΔRs obtained by regression model.