Continuous fiber composites are much more complex than metal, with respect to failure in particular. If they are so-called unidirectional (UD), they involve stacks of several plies, each ply characterized by a single fiber orientation. Hence they fail because of various mechanisms taking place at the ply level (matrix cracking, fiber breakage, fiber-matrix debonding) or between the plies (delamination). These mechanisms remain not fully understood and are investigated through experimental and virtual testing.
This complexity is usually not captured by simulation so that UD composite material properties are currently obtained only from physical testing, requiring high investments in time and money. The replacement of a fair amount of real tests by simulation requires the development of an accurate model for progressive failure in order to obtain predictive simulations of plain, open-hole or filled-hole coupon scenarios among others.
To accurately predict the properties of UD composites, Digimat advantageously combines micromechanics, deriving composite properties from constituent properties through mean-field homogenization, and progressive failure. At the phase level, Digimat is employed to define and reverse-engineer the matrix – e.g. epoxy – and fiber – e.g. carbon – stiffnesses. At the ply level, it exploits a Hashin failure criterion to apply a mechanism sensitive to matrix and fiber failure. In addition, it enables a stiffness reduction according to the Matzenmiller-Lubliner-Taylor model.
Digimat is then coupled to a finite element solver to provide the solver with the material properties. In the example of the simulation of a quasi-isotropic open hole tensile test, Digimat’s progressive failure enables to account for the failure sequence involving damage initiation in 90° plies and ultimate failure after failure of 0° plies. Such coupled analyses can be run for implicit or explicit solvers, both available within MSC Nastran for instance. Taking into account the specific requirements of test standards and the systematic collection of experimental data, Digimat enables a high level of automation for the purpose of screening material properties.
Continuous fiber composites have rapidly spread across aerospace components for their lightweighting capabilities but pose design challenges because of their complex properties. In particular, their failure behavior is not easily characterized and requires new tools to be realistically simulated. In that respect, micromechanical material models allied to progressive failure provide an in-depth understanding of the composite behavior at the constituent – matrix or fiber – level and a directionally selective stiffness degradation. Hence they pave the way for a reduction in experimental testing in favor of virtual testing.