Failure prediction of composite structural components via laser-measured displacement based material recognition using Machine Learning algorithms

Project: Research GrantsResearch

Project Details


"The aim of this research proposal is the generation of novel machine learning (ML) algorithms to predict the failure of advanced composite materials used in aerospace industry. Composite materials, such as carbon-fiber composites, are widely used in the aerospace industry however, due to their composite nature, these materials exhibit anisotropic properties. Due to their anisotropic properties and the constantly evolving mechanical loads in these materials, the classic mechanical models used to predict failure are not accurate. To develop more accurate prediction models, we will test several carbon-fiber composites to determine their failure stress and strain. By using contact-less laser displacement sensors during the test we will determine the displacement and deformation of several points in the samples throughout the test. Then we will train Neural Networks so that given different displacement and deformation profiles at different stress or strain conditions our model will be capable of predicting when the sample will fail and the location of the failure. Additionally, correlation between the material properties of each composite and failure stress, strain will be identified to provide the aerospace industry better information of potential failure locations in their vehicles. Completion of this project will provide the aerospace industry a novel tool to better diagnose potential problems to prevent failure in-flight and to better understand the failure mechanics of these advanced composite materials. The results obtained from this project will be used to create a novel research proposal that will be used to apply for funding in competitive programs both in Spain as well as in Europe.

The second member of the research team is a candidate for the PhD programme who is awaiting a PhD scholarship. The reserved budget item corresponds to recruitment during the months prior to the award of the scholarship."
Effective start/end date1/01/2031/12/20