The thesis presents a methodology for accurate prediction of the static behaviour of complex elastomeric components. A critical gap in FEA hyperelastic material modelling and industrial manufacturing realities was identified and addressed through a holistic approach to static behaviour prediction. Data obtained from tests performed during this project were combined with data from a previous related project to obtain the multi-deformation experimental data required for defining the parameters of the Arruda-Boyce 8-chain model. A custom tool was developed to facilitate range-specific curve-fitting, capable of improving analysis predictions by up to 300% in comparison with standard ANSYS curve-fitted models. Despite these improvements, the study identified a persistent numerical limitation of statistical hyperelastic models within the low-strain region (0 – 50%). A careful approach to model setup i.e. element choice, mesh size and contact setup was implemented to predict the fatigue-induced tear initiation point in a conical anti-vibration mount with high correlation to physical tests. A statistical approach to validating an analysis predicting the stiffness of the mount was necessitated due to the variability encountered in rubber manufacturing. The investigation of over 250 multi-sourced parts found a reasonably defined mode-range to be the most appropriate measure of central tendency against which FEA predictions will be validated. Using optimised models, a predictive precision of less than 4% was achieved, which is well within the 6% natural variability established during the part’s variability tests. This methodology provides a scalable framework for SMEs in the rubber industry to implement high-accuracy FEA in design improvements, failure investigations and new product development of rubber parts.
| Date of Award | 18 Jun 2026 |
|---|
| Original language | English |
|---|
| Sponsors | Innovate UK & Pendle Polymer Engineering |
|---|
| Supervisor | John Allport (Main Supervisor) & Gina Javanbakht (Co-Supervisor) |
|---|