This article explores the effects of parameters such as cutting speed, force, polymer wheel hardness, feed, and grit size in the abrasive belt grinding process to model material removal. The process has high uncertainty during the interaction between the abrasives and the underneath surface, therefore the theoretical material removal models developed in belt grinding involve assumptions. A conclusive material removal model can be developed in such a dynamic process involving multiple parameters using statistical regression techniques. Six different regression modelling methodologies, namely multiple linear regression, stepwise regression, artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), support vector regression (SVR) and random forests (RF) have been applied to the experimental data determined using the Taguchi design of experiments (DoE). The results obtained by the six models have been assessed and compared. All five models, except multiple linear regression, demonstrated a relatively low prediction error. Regarding the influence of the examined belt grinding parameters on the material removal, inference from some statistical models shows that the grit size has the most substantial effect. The proposed regression models can likely be applied for achieving desired material removal by defining process parameter levels without the need to conduct physical belt grinding experiments.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited