Prediction and Parametric Effect in Drilling of Natural Fiber Laminate Using Combined Taguchi- Artificial Neural Network Approach

Articles in Press

Document Type : Research Article

Authors

1 Department of Manufacturing Engineering, Annamalai University, Chidambaram-608002, India

2 Department of Manufacturing Engineering, Annamalai University, Chidambaram, 608002, India

3 Department of Mechanical Engineering, Government College of Engineering, Thanjavur, 613402, India

Abstract

The study investigates the prediction and parametric effects in drilling of prosopis juliflora fiber (PJF)-reinforced epoxy resin hybrid composite using combined taguchi-artificial neural network (ANN) approach. The composite was prepared via hand lay-up technique with natural reinforcement including vetiver fiber (VF) and coir pith (CP). The effects of drill bit diameter (DBD), spindle speed (SS), and feed rate (FR) on thrust force (TF) and surface roughness (SR) were evaluated through full factorial design. An ANN model developed using a feedforward backpropagation algorithm successfully predicted the responses. Analysis of variance (ANOVA) results revealed that the regression coefficient (R2) for TF and SR were 96.39% and 95.54%, respectively. The DBD and FR were identified as most significant parameters influencing TF and SR, both significant at the 95% confidence level (p<0.05). The regression plot exhibited a strong correlation (R=0.9883) between the predicted and actual values, while the ANN model achieved a mean squared error (MSE) of 0.089758 within 2 epochs. TF and SR increased with higher DBD and FR but decreased with an increased SS, as indicated by main effect plots. Scanning electron microscope (SEM) revealed drilling induced mechanisms, including fiber pullout, delamination, matrix cracking, matrix debonding, and matrix deformation. The findings demonstrate enhanced machining performance, offering their potential for industrial applications and future research on bio-composite materials.

Keywords

Main Subjects

Mechanics of Advanced Composite Structures

Articles in Press, Accepted Manuscript
Available Online from 22 November 2025

Files

Share

How to cite

Statistics