

Delamination Growth in Damage Tolerance Composite Structures
Abstract
The susceptibility of composites to damage has proved to be a critical factor in structural design. Despite their apparent superiority to metals they develop inherent problems; the difficulty to detect internal damages that require expensive NDI (non destructive inspections), such as ultrasonic scans, to detect for instance interlaminar delaminations. This is one of the most common failure modes for composite structures, especially for compression loads reducing drastically the structural capability. Thus, many investigations are analysing composite delaminations onset and growth, and there is a growing interest in developing FEM tools able to perform accurate simulations. In general it has been found that the simulation tools are complicated and computationally expensive, and there is in many cases a considerable difference in the structure behaviour between the simulation and the tests. Many tools are based on the Virtual Crack Closure Technique (VCCT), which is widely applied for computing the energy release rates based in FEM analyses results. This technique allows both prediction of delamination onset and growth under static loads, however the technique is dependent of the FEM methodology used (mesh size for instance) and the parameters of the non – linear analysis. When referring to commercial FEM codes, there are many limitations on the control of the delamination growth algorithm, on the robustness and convergence and on the processing time. This paper explains in detail an “in-house†tool developed at INTA, that based on the Modified VCCT theory and using non-linear static solution (MSC.Nastran SOL 106) as base code, simulates the delamination growth of a composite structure with initial delaminations. The FEM methodology required for this purpose is also explained in detail, and a comparison between simulation and test results is given, showing an accurate correlation. The investigation also covers different mixed module failure criteria (linear, power law, K