Delamination research: progress in the last two decades and the challenges ahead

Paul Robinson
Department of Aeronautics, Imperial College London, United Kingdom

Delamination has long been recognised as a significant weakness of fibre reinforced, polymer matrix laminated composites – indeed it is often referred to as the Achilles Heel of what, in many other respects, can be justifiably termed high performance structural materials. This weakness initiated a considerable research activity which has addressed a range of aspects associated with this problem. This paper will summarise the progress of this research over the past two decades and identify some of the remaining challenges.

The development of test methods to characterise the resistance to delamination has seen considerable progress. Research has focussed primarily on test methods for determining the critical energy release rates in Modes I, II and mixed mode I/II, with a smaller effort directed to Mode III. The double cantilever beam test devised for Mode I has achieved national and international standard status but, perhaps surprisingly, no single accepted test has been developed for Mode II.

The ability to reliably measure the interlaminar toughness, at least in Mode I, and therefore rank competing materials has lead to the development of composites with improved delamination resistance. These materials developments have not simply concerned improvements of the matrix; the addition of reinforcement through the thickness of a laminate has also been investigated in a variety of forms including ‘2.5D’ fabrics, stitching, z-fibres, and tufting. These additions can considerably improve the resistance to delamination growth but this improvement is often at the expense of other mechanical properties – particularly the in-plane compression strength. (The development of such materials has also had a consequence for research into interlaminar test methods. These through-thickness reinforced materials can exhibit large fibre-bridged zones in the wake of the delamination front and this can invalidate existing test procedures).

Research has also been conducted to develop analysis methods to model the growth of delamination so that engineers can assess the susceptibility to, and consequences, of delamination during the design of laminated composite structures. Some simple formulae exist for particular situations but to properly represent the complexity of practical structures it is necessary to use finite element (FE) analysis. Early FE-based approaches focussed on the evaluation of energy release rate (most commonly by the virtual crack closure method but others have also been proposed). This energy release rate could then be compared to the critical values measured in tests to decide whether the delamination under investigation would grow. Strategies for automatically advancing the crack front within the FE model were also proposed but for practical structures, in which, for example, delaminations might grow under stiffeners, very complex re-meshing strategies are required. An alternative which can avoid these remeshing difficulties is the use of ‘cohesive zone’ or ‘interface’ elements which have been developed more recently. These elements are incorporated in the FE model at interfaces between plies and have a constitutive law which ensures the correct amount of energy is absorbed as the adjacent plies separate. This approach has become widely adopted and is available in a number of commercial FE packages.

Despite the considerable research that has been performed it is clear that there is still some way to go before all the tools to address the problem of delamination can be considered routine. The methods to fully characterise the interlaminar toughness are not yet available. As noted above an international standard for Mode II testing is yet to be established and to date these interlaminar test methods have nearly exclusively examined unidirectional materials with the fibre reinforcement lying in the crack growth direction, although there have been a number of research papers attempting to drive this issue. The modelling tools and choice of interface element laws still need further development – it is clear that for some situations there is a wide choice of form of the interface constitutive law and this needs to be fully resolved. Real delamination growth often progresses with migration. 
 

<< back