With metallic structures, there are well known methods to assess the effect that defects will have on structural performance. However, for composite materials there are limited unified and standardised methods to understand the effect that a defect may have. The challenge is therefore to develop a theoretical and numerical approach to assessing complex and interacting damage phenomena in composite materials.
Fibre reinforced composites have been adopted widely by the automotive and aerospace industry sectors, where design requires both material strength and lightweight. Due to the combination of plies of different orientations and the complex geometry of components that can be obtained from these materials, the design analysis is most often performed by finite element modelling.
Specific continuum damage models (CDM) were used by TWI for the prediction of the strength of composite joints, where parameters quantifying damage mechanisms such as fibre breaking or buckling, resin cracking in tension and compression, and delamination, where calculated and material properties were degraded in order to model progressive damage . In this model, the critical parameters and degradation of properties were identified from experimental testing, and were defined as state variables of the volume elements.
In the proposed project, CDM models will be refined to be applied to a phasefield model, where for a given strain field applied to a volume element, separate stress fields for the fibres and resin would be calculated, depending on the behaviour of each constituent and the local damage condition. The local field calculation will be done in an enriched element, either using a super-element method or a user-developed behaviour subroutine. The model may also be combined with X-FEM when cracking is predicted.
The behaviour and properties for damage prediction will be determined from a set of mechanical tests, and modelling and experimental work on a joint geometry will be performed for assessing the capability of the model to predict fracture for different loading conditions.
The interest of designing an element able to partition the stress between the two main components of the composite material is to be able to quantify damage in each constituent separately. This can be used when analysing the part under a sequence of different loading cases, where potential damage caused by one loading case can be cumulated and taken into account for estimation of the damage by the subsequent loading cases. This capability was not available in the damage calculation tool based on Hashin's model .
 TWI CRP Report 13182.01/2004/1197.2