Study of alternatives for obtaining statistical correlation models for modelling the rough honing process

Abstract

In the present paper, in order to model rough honing operations, statistical regression correlation models are presented for surface roughness and material removal rate as a function of process variables. Several different models were studied: linear, quadratic, exponential, etc. In addition, Box-Cox transformations were performed to the models so as to improve their fit. Models were compared taking into account two different criteria: R2(Adjusted) coefficient and R2(Predicted) coefficient. Reduced second order models with Box-Cox transformations were found to be appropriate for modeling average roughness and reduced first order models with Box-Cox transformations for modeling material removal rate.