Interface debonding characterization by image correlation integrated with Double Cantilever Beam kinematics
KAUST DepartmentPhysical Sciences and Engineering (PSE) Division
Mechanical Engineering Program
Composite and Heterogeneous Material Analysis and Simulation Laboratory (COHMAS)
Permanent link to this recordhttp://hdl.handle.net/10754/564083
MetadataShow full item record
AbstractA procedure is proposed for the identification of spatial interfacial traction profiles of peel loaded Double Cantilever Beam (DCB) samples, from which the corresponding traction-separation relation is extracted. The procedure draws upon recent developments in the area of non-contact optical techniques and makes use of so-called Integrated Digital Image Correlation (I-DIC) concepts. The distinctive feature of the I-DIC approach proposed herein is that the unknown degrees of freedom are not displacements or rotations, but the set of interfacial fracture properties describing the traction profile. A closed-form theoretical model is developed to reconstruct a mechanically admissible displacement field representing the deformation of the adhering layers during debonding in the DCB fracture test. The proposed modeling accounts for the spatial traction profile along the interface between the adherends using few degrees of freedom, i.e. crack tip position, maximum stress and size of the process zone. By minimizing the correlation residual with respect to the degrees of freedom, the full set of interfacial fracture properties is obtained through a one-step algorithm, revealing a substantial gain in terms of computational efficiency and robustness. It is shown that the identified traction profile can be effectively combined with the crack opening displacement to extract the corresponding traction-separation relation, i.e. the key input data for any cohesive zone model (CZM). The proposed procedure is validated by post-processing virtually deformed images generated through the finite element method. The robustness with respect to noisy data, as well as the low sensitivity to the initial guess, are demonstrated.