Artificial Neural Network (ANN) Approach to Predict Tensile Properties of Longitudinally Placed Fiber Reinforced Polymeric Composites including Interphase

Machine Learning has become prevalent nowadays for predicting data on the mechanical properties of various materials and is widely used in various polymeric applications. In the present study, Artificial Neural Network (ANN), a computational tool is used to predict the elastic modulus of a composite of longitudinally placed fiber-reinforced polymeric composite. The novelty in carried work is that the property prediction is carried out considering interphase and its properties. For this, tensile properties data of Longitudinally Placed Bamboo Fiber Reinforced Polyester Composite (LUDBPC), Longitudinally Placed Flax Fiber Reinforced Polyester Composite (LUDFPC) and Longitudinally Placed Jute Fiber Reinforced Polyester Composite (LUDJPC) has been procured to generate ANN models. The Levenberg-Marquardt training algorithm is used to generate the ANN models as it gives more accurate results compared to other ANN algorithms based on interphase properties data. The validation of ANN models was also carried out based on fresh experimental results of BPC/FPC by doing the fabrication with hand layup technique and testing of composites with a Universal Testing Machine (UTM). The present work signifies that the developed ANN models give accurate results with experimental results for the prediction of elastic modulus of composite (Ecl) and it can be used for the prediction of longitudinally placed fiber-reinforced composite and Ecl of BPC at volume fraction of fiber (vf):22% is 2248.75 MPa and Ecl of FPC at vf:10% is 3210.50 MPa.

Artificial Neural Network (ANN) Approach to Predict Tensile Properties of Longitudinally Placed Fiber Reinforced Polymeric Composites including Interphase Keywords:

Bamboo/polyester Composite , Flax/polyester Composite , Jute/Polyester Composite , Interphase Volume Fraction , Elastic Modulus of Composite