Exploring the Elastic Properties of Woven Fabric Composites: A Machine Learning Approach for Improved Analysis and Design

Hayat, Khazar; Ullah, Zahur; Siddique, Shafaqat; Ahmad, Zeshan

Abstract

Woven fabric reinforced plastic composites are highly favoured in the aerospace and automotive industries for their exceptional impact resistance and ease of manufacture. To design and analyse these structures, it is crucial to determine their elastic properties of woven fabric composites, which can be estimated through analytical, numerical, or experimental means. In this study, we propose a novel approach that combines machine learning techniques with finite-element methods based multi-scaling analysis methodology to predict the elastic behaviour of woven composites. The method leverages datasets generated from finite element methods based numerical simulations and literature to train and validate models, providing a cost-effective and computationally efficient alternative to conventional homogenization-based finite element method. The approach offers a promising solution to accurately predicting the elastic behaviour of woven fabric composites.

Citation

Hayat, K., Ullah, Z., Siddique, S., & Ahmad, Z. (2023, May). Exploring the Elastic Properties of Woven Fabric Composites: A Machine Learning Approach for Improved Analysis and Design. Presented at 1st International Conference on Modern Technologies in Mechanical & Materials Engineering (MTME-2023), Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Pakistan

• Presentation Conference Type: Conference Paper (published)
• Conference Name: 1st International Conference on Modern Technologies in Mechanical & Materials Engineering (MTME-2023)
• Start Date: May 6, 2023
• Online Publication Date: Jun 13, 2023
• Publication Date: Jun 13, 2023
• Deposit Date: Sep 19, 2024
• Peer Reviewed: Peer Reviewed
• Volume: 381
• Article Number: 01011
• Book Title: MATEC Web of Conferences
• DOI: https://doi.org/10.1051/matecconf/202338101011
• Public URL: https://durham-repository.worktribe.com/output/2859830