Stochastic cohesive interface analysis of layer debonding

Interfacial debonding in layered structures is recognized as a critical failure mode of such structural forms. The way uncertainty of input parameters affects the debonding mechanism is still an open question. In this paper, a methodology for the stochastic analysis of debonding in layered structures is developed. The layered structure is modeled using the cohesive interface approach while parameter uncertainty is represented by random fields of material and interfacial properties. To analyze the influence of uncertainty, the methodology integrates the cohesive interface approach into the perturbation based stochastic finite elements method. A representative model problem of a bar bonded through a cohesive interface to a rigid substrate and subjected to debonding is looked at. The proposed methodology is developed through the stochastic analysis of displacement profile, debonding length, and reaction force at the displaced end. The formulation is followed by numerical results demonstrating the capabilities of the method in describing the influence of input uncertainty on the structural response. The results are validated by comparison with a reference MCS analysis conducted for this purpose. Opposed to the MCS, which heavily draws upon computational power, the proposed method is expandable and can be applied to more general cases with a tolerable increase in complexity. The current investigation thus establishes a stochastic analysis tool for assessing uncertainty in more advanced models for debonding of layered structures.