School of Engineering and Materials Science
Research Student Awards
PhD Thesis: Predicting the behaviour of elastomer components using finite element analysis
Author: BUSFIELD, James
Supervisor(s): Craig Davies
In the past, finite element analysis has been used in a variety of ways as a design tool to predict the performance of elastomer components. This thesis aims to provide a fundamental evaluation of the reliability of the technique. The first half of the thesis considers stiffness calculations. The choice of stored energy function, the need to consider finite compressibility effects and the difficulties caused by the Mullins effect are all considered. Four significantly different component geometries are evaluated both experimentally and theoretically using finite element analysis techniques and analytical elasticity based relations where appropriate. The problems tackled are bonded compression mounts, bonded cylindrical bushes, the indentation of flat sheets with a rigid spherical indentor and a practical case study of a non-symmetric bush. A series of guidelines for the use of finite element techniques is derived from this work.
The second part of the thesis concentrates on the prediction of tear rates and fatigue failure times in elastomers. A relationship exists between the magnitude of the stored energy release rate available to drive a crack and the resultant crack growth rate. This basic relationship is said to be a characteristic of the material. The approach has been of limited use in the past as it could only be applied to crack geometries where the stored energy release rate could be deduced from simple analytical relations. It is shown here that the stored energy release rate value can be calculated using a finite element approach and can therefore be derived in principle for a specific small flaw in any complex geometry subjected to an applied load. The stored energy release rate relationships evaluated here include simple geometries such as a crack in a pure shear test piece, the trouser tear test and the more complex geometry found in the inflation of a flat crack under a hydrostatic tension. Having derived stored energy release rate relationships for a variety of test piece geometries and deformation states, an approach is demonstrated that allows fatigue data to be predicted with a reasonable degree of accuracy for filled elastomer components.