Multi-response Optimization in Drilling of Carbon Fiber Reinforced Polymer Using Artificial Neural Network Correlated to Meta-heuristics Algorithm. (2016)

• Record Type: Journal Article
• Authors: Shunmugesh, K.; Panneerselvam, K.
• Is Part Of: Procedia technology. Volume 25(2016)
• Page Start: 955
• Page End: 962
• Publication Date: 2016
• Subjects: Drilling -- Carbon Fiber Reinforced Polymer -- Particle Swarm Optimization -- Gravitational Search Algorithm -- Artificial Neural Network; Technology -- Congresses; Technology -- Periodicals; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Technology; Conference proceedings; Periodicals; 605
• DOI: 10.1016/j.protcy.2016.08.187
• Languages: English
• ISSNs: 2212-0173

Abstract

Abstract: This paper aims to optimize the drilling process parameters using artificial neural network (ANN) linked with the most popular meta-heuristics technique such as Hybrid Particle Swarm Optimization Gravitational Search Algorithm (PSOGSA) and Genetic Algorithm (GA). An aerospace grade T300 Carbon Fiber-Epoxy composite laminate of 8 mm thick was made of T300 Polyacrylonitrile (PAN) based Carbon Fiber and two part Epoxy resin was use for this study. The Carbon Fiber used is Bi-directional (BD) with a ply thickness of 0.25 mm and lay-up sequence of [60/90/0/90/90/60/0/60/60/60/60/45/90/90/0/45/60/90/60]. Drilling experiments were conducted on a composite laminate by varying the cutting speed (30, 40 and 50 m/min), feed rate (0.025, 0.05 and 0.1 mm/rev) and drill bit type (HSS, TiAlN and TiN). The experimental results in the form of thrust force, torque and surface roughness obtained are correlated with process parameters through artificial neural network (ANN) and optimized by PSOGSA and GA. The optimization results indicates that the proposed hybrid PSOGSA performances much better than the GA.