Prediction Of Hole Quality In Drilling Fiber Reinforced Composite Materials, Using Artificial Neural Networks

The main objective of the present work is to develop artificial neural networks (ANNs), with back-propagation training routine, for predicting the hole quality in drilling woven glass fiber reinforced epoxy laminates. Experimental work was performed to produce the data used to develop the required neural networks. Two components dynamometer was manufactured to measure and monitor the thrust force and torque during the drilling processes.The effect of drilling parameters (speed, feed, drill pre-wear, and drill diameter) on hole quality criteria (delamination size, surface roughness, and bearing strength) was studied experimentally. Scanning electron microscope was used to examine the damage of the drilled holes.The results show that, the delaminations associated with push-out (drill exit) are more severe than those associated with peel-up (drill entrance). Thrust force, torque, and delamination size are increased with the increase of the feed and the drill pre-wear value, while the bearing strength of the machined hole was decreased while no clear results about the effect of the cutting speed on the delamination size are observed.