OPTIMIZATION OF MACHINING PARAMETERS IN MILLING HYBRID SISAL COTTON FIBER REINFORCED POLYMER COMPOSITE USING ARTIFICIAL NEURAL NETWORK MODEL

Abstract

This thesis explores the use of artificial neural networks (ANNs) to optimize surface roughness (Ra) and material removal rate (MRR) in CNC milling of hybrid sisal-cotton fiber composites, aiming to fulfill the demand for automated and efficient machining solutions. Surface roughness is a key indicator of product quality, while MRR is crucial for production efficiency. The study proposes an integrated approach using ANN and Genetic Algorithm (GA) to determine optimal machining parameters for minimizing Ra and maximizing MRR. Implementing a standardized hand lay-up method, the research establishes a predictive model for Ra and MRR through mechanical testing and ANN modeling. The model, validated with high correlation coefficients (0.98 for Ra and 0.99 for MRR), undergoes further refinement using genetic algorithms to enhance accuracy. Key factors influencing surface roughness and MRR are identified, with depth of cut being the primary factor for Ra and feed rate for MRR. The developed ANN model proves valuable in simulating and predicting machining outcomes, enabling process optimization before actual production. The analysis confirms that the ANN-GA approach effectively predicts optimum machining parameters.