Physically based damage models for laminated composites

Abstract

Abstract

The computing power that is available for engineering calculation continues to grow at a dramatic pace. Engineers in industry want to have seamless models that can be used to design across the scale range from atoms to structures, including simulation of the manufacturing process. A limited aspect of this wish is the requirement to deal effectively with the progressive growth of microstructural damage in composites and its effect on both property degradation and the catastrophic failure event. This paper reviews progress that is being made at the National Physical Laboratory (NPL) with the development and validation of physically based damage growth models for laminated composites. The review includes:

• (a)

prediction of undamaged ply properties determined from the properties of the fibre and the matrix, with emphasis on comparison of analytical models with each other, and with finite and boundary element solutions;

(b)

discussion of various stress transfer models, and their validation, that have been developed for application to the prediction of the properties of composite laminates having ply crack damage;

(c)

prediction of ply cracking in multiple-ply cross-ply laminates subject to triaxial loading (without shear) and bending;

(d)

prediction of ply cracking in general symmetric laminates subject to combined triaxial loading and in-plane shear loading;

(e)

consideration in a damage mechanics context of progressive ply crack formation in general symmetric laminates subject to thermal residual stresses and general in-plane loading, where an important new methodology is described that results from attempting to develop a continuum damage model from a physically based discrete ply cracking model based on energy concepts;

(f)

discussion of how the models might be integrated into finite element analysis (FEA) systems to enable strain softening in structures to be adequately modelled.

The paper also includes statements concerning the status of the various models in relation to alternative approaches, and to model validation.