Project Overview
This PhD project focuses on the development of advanced Machine Learning (ML) techniques to generate accurate and efficient constitutive models for complex smart materials – particularly Electroactive Polymers (EAPs). Traditional modelling approaches for such materials often rely on computationally intensive homogenisation schemes within Finite Element Methods (FEM), which can limit their practicality for design and optimisation tasks.
To address this, the project employs Gaussian Process Regression (GPR) to learn constitutive behaviour directly from data, eliminating the need for costly multi-scale modelling procedures. These data-driven models have been successfully validated through a series of academic studies, demonstrating their capability to predict the response of materials under complex deformations. The models offer significant advantages in terms of generality, flexibility, and integration into simulation frameworks.
In parallel, the research will explore the application of ML techniques for the optimisation and control of EAP-based devices. As EAPs are increasingly used in soft actuators and sensors, there is a growing demand for model-informed control strategies that can handle their nonlinear, coupled behaviour. The goal is to develop intelligent design and control pipelines that accelerate innovation in soft robotics and smart material systems.
This work forms part of a wider research effort to combine computational mechanics, materials science, and Machine Learning to enable the next generation of functional, flexible, and adaptive engineering systems.
Project Aims
– Model large deformations for a multi-physics coupled problem
– Develop accurate complex constitutive models via accurate Machine Learning
– Integrate constitutive metamodels within FEM to accelerate homogenised multi-scale modelling
– Investigate optimisation and control algorithms for EAP modelling
Project Team
- Professor Antonio J Gil
- Dr Mokarram Hossain
- Dr Rogelio Ortigosa
- Dr Jesús Martínez-Frutos
- Dr Anil Bastola
- Rollo Pattinson
- Elena Kingston
Collaborators
Swansea University; Universidad Politécnica de Cartagena (UPCT)
