2A: Experiments and analysis of delamination of carbon epoxy composites


It is well known that large scale bridging in fracture of layered composites is one of the most important toughening mechanisms. The resulting resistance to fracture, however, is dependent upon loading, specimen geometry, layout and microstructure rendering its modeling very challenging. In this presentation, experimental results and modeling of fracture in CFRP are presented. The experimental part consists of monotonic tests of inter-, intralaminar fracture of unidirectional specimens as well as load-controlled fatigue of interlaminar specimens. The modeling part involves an iterative scheme to calculate traction-separation relations, due to large scale bridging (LSB), using strains from embedded sensors, parametric finite element simulations and optimization.

The results demonstrate an important effect of specimen thickness in the fracture response under monotonic and fatigue loads and allow to deduce scaling relationships due to LSB. The obtained traction-separation relations are employed in cohesive zone simulations to predict very well the corresponding load-displacement and fracture resistance curves (R-curves) for various thicknesses. Micromechanical observations and analysis show that when separation-dictated cohesive response is present (i.e., adhesive joints), no effects of specimen's stiffness is expected. When fiber bundles, are involved in the formation of closing tractions, i.e. LSB in CFR, an important variation should be expected, both in R-curve behavior and traction-separation relations. The later effect is attributed to the fibers bundles, in the bridging zone, loaded in traction and bending which varies with loading type and specimen's stiffness.

To elucidate further these observations, computational micromechanics models are developed to predict the specimen thickness effects on bridging. Data reduction and analysis shows that if the traction-separation relation is enriched with the local crack opening angle, the observed experimental response can be easily reproduced thus, suggesting cohesive relations with two-kinematic parameters as a physically sound model. The results of these works point to traction-separation-angle relation as a structural property.



[1] L. P. Canal, M. Alfano & J. Botsis, 'A multi-scale based cohesive zone model for the analysis of thickness scaling effects in fiber bridging', Composites Science and Technology, 139, 2017, pp. 90-98.

[2] Pappas & J Botsis, 'Variations on R-curves and traction-separation relations in DCB specimens loaded under end opening forces or pure moments', International Journal of Solids and Structures, https://doi.org/10.1016/j.ijsolstr.2019.11.022.

[3] G. Pappas, J. Botsis, 'Towards a geometry independent traction-separation and angle relation due to large scale bridging in DCB configuration', submitted.


Prof. John Botsis

EPFL, Switzerland