Title: Neural surface roughness models of CNC machined Glass Fibre Reinforced Composites

Authors: S. Alexandrakis, P. Benardos, G-C. Vosniakos, N. Tsouvalis

Addresses: National Technical University of Athens, School of Mechanical Engineering, Iroon Polytexneiou 9, 15780 Zorgafou, Athens, Greece. ' National Technical University of Athens, School of Mechanical Engineering, Iroon Polytexneiou 9, 15780 Zorgafou, Athens, Greece. ' National Technical University of Athens, School of Mechanical Engineering, Iroon Polytexneiou 9, 15780 Zorgafou, Athens, Greece. ' National Technical University of Athens, School of Naval Architecture & Marine Engineering, Iroon Polytexneiou 9, 15780 Zorgafou, Athens, Greece

Abstract: CNC machining of parts from pre-made Glass Fibre Reinforced Composites (GFRCs) blocks started gaining ground. However, wrong cutting conditions result in poor surface quality, delaminations or other damaging effects. In this work, a computational tool is developed to help improve machinability of these parts by accounting for surface quality. Artificial Neural Network models trained with data obtained through Taguchi-style designed experiments predict surface roughness obtained. GFRC blocks made from D.E.R.321 epoxy resin, CHEM.93-1-74, PC12 stabiliser and Woven Roving (500 gr/m² and 800 gr/m²) were CNC machined. Microscopy and image analysis studies enrich the ANN models with machined material macro-structural characteristics.

Keywords: GRFC; glass fibre reinforced composites; CNC machining; surface roughness; ANN; neural networks; image analysis; Taguchi mathods; machining composites; machinability; surface quality; microstructure.

DOI: 10.1504/IJMPT.2008.018986

International Journal of Materials and Product Technology, 2008 Vol.32 No.2/3, pp.276 - 294

Published online: 27 Jun 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article