LIFETIME ANALYSIS OF A COMPOSITE FLYWHEEL ENERGY STORAGE SYSTEM.

Abstract

This thesis is concentrated on the long-term fracture of thick unidirectional glass and carbon fibre composites subjected to transverse stress. The objective was to develop a methodology for predicting the long term lifetime of a composite rotor used as part of a flywheel based energy storage system. The flywheel design is based on accommodating high hoop stresses induced during the high speed rotation. However, the different Poisson's ratios of the constituent materials in the rotor result in a complex stress distribution with significant stresses introduced in a direction transverse to the fibres. The possibility has been raised that the lifetime of the rotor will be limited by crack growth in this transverse direction, originating from defects (pores, cracks etc) that can be introduced into the rotor during its manufacture. The approach explored in this work has been to adopt a fracture mechanics based methodology whereby the rate of crack growth in a thick composite is measured as a function of an applied stress intensity. The basic fracture parameters for the material were measured such that the time taken for a crack to grow to a size sufficient to cause failure under an operating stress could be calculated. The materials were also examined to characterise the nature, size and extent of inherent defects. The stress distribution in the rotor under operating conditions was modelled using finite element analysis. The combination of information on inherent defects, stress directions and crack growth rates enable predictions to be made concerning the likely lifetime of the composites. Proof stress diagrams were also constructed in order to demonstrate an approach to product quality assurance testing. The end point of the work was to identify critical manufacturing defect sizes that could be tolerated under the specified operating conditions. The methodology developed for lifetime predictions was critically assessed and considered to be generally acceptable. The work did however raise some concerns regarding the applicability of a conventional fracture mechanics approach applied to heterogeneous composite systems where the size of the cracks are very small. It is recommended that future work should concentrate on studying this area with an emphasis on crack nucleation studies rather than on further crack propagation work.