Article ID: | iaor20111482 |
Volume: | 7 |
Issue: | 1 |
Start Page Number: | 22 |
End Page Number: | 43 |
Publication Date: | Dec 2011 |
Journal: | International Journal of Productivity and Quality Management |
Authors: | Mahapatra S S, Equbal Asif, Sood Anoop Kumar |
Keywords: | neural networks |
In the present work, effect of five factors viz., layer thickness, part build orientation, raster angle, raster to raster gap (air gap) and raster width each at three levels together with their interactions is studied on dimensional accuracy of fused deposition modelling (FDM) build part. Four performance characteristics i.e. percentage change in length, width, thickness and diameter considered in this study are converted into an equivalent response known as grey relational grade. Optimum factor levels are determined for maximisation of grey relational grade using grey‐based Taguchi method. A fuzzy inference system (FIS) is proposed for prediction of overall dimensional accuracy using Taguchi's orthogonal array for developing inference engine. The results of FIS are compared with prediction values obtained through artificial neural network. It has been demonstrated that fuzzy model is able to predict overall dimensional accuracy at all operating condition to a high degree of accuracy.