Potentials of FEA-simulation for elastomer stress softening in engineering practice

Since L. Mullins in 1969 first reviewed the stress softening effect in rubber materials, it has been subject of continuous research within the field of elastomer materials science. Depending on the material properties, this stress softening, commonly referred to as the Mullins effect, can drastically affect the mechanical behaviour. Due to its complex nature, the Mullins effect is still considered as a challenge for modelling and simulation. Several models have been introduced for its prediction, most of which are based on a phenomenological approach and particular assumptions and simplifications. During this project the potentials of the commercial finite elements suite Abaqus for the simulation of the Mullins effect were investigated. This was not only undertaken with contemplation of the implemented material model itself, but instead with particular regard to its practical applicability. Practice relevant techniques for the modelling of rubber components with stress softening behaviour were introduced. Several examples of relevant use cases were indicated, corresponding simulations were carried out, and the results were evaluated. It was found that the stress softening effect can be predicted with a sufficient accuracy for a broad range of engineering applications. It was further shown that, at least for simple applications, this is even possible with a limited amount of calibration test data. However, the investigations also showed that in some cases extensive test data preparation and modelling effort might be necessary. Furthermore, the simulation results were found to follow several simplifications that contrast with the actual material behaviour.