On modelling of damage evolution in textile composites on meso-level via property degradation approach

The continuous damage mechanics (CDM) approach is a popular tool for modelling of damage evolution in textile composites on the meso-level. It is based on the assumption that a material with defects can be replaced by a fictitious material with no defects but with degraded elastic constants. In such way the presence of defects is only reflected in the material elastic properties and damage evolution is recorded through the loss of these properties. The CDM approach incorporated into finite element analysis often predicts unphysically wide damage zones and in some cases failure across yarns – findings that are not supported by experimental data. The current work is geared toward identifying the source of inconsistencies between experiment and modelling by revisiting basic assumptions of CDM. A test problem is proposed to illustrate a break down of the CDM approach where a single crack-like defect in a yarn is modelled as an inhomogeneity with elastic constants reduced according to Murakami–Ohno model. It is shown that CDM in combination with local stress analysis of failure may predict a different direction of damage evolution as well as an incorrect failure mode in comparison with the crack problem. We also investigate whether the Murakami–Ohno model adopted for calculation of properties of a fictitious inhomogeneity contributes to the unphysical results. For this we compare contributions of a crack and an inhomogeneity into material elastic response. A new property degradation procedure is suggested (referred here as an effective elastic response model) where the size of an inhomogeneity and properties of the surrounding material are taken into account.