Abstract
We propose a novel sensitivity based approach that predicts and explains the accuracy of material parameter identification for a composite plate using the Projected Finite Element Update Method. A typical experiment using the Projected Finite Element Update Method technique involves a plate specimen held at 3 or 4 supports and bent under the application of a point load. Two‐Dimensional Digital Image Correlation is used to measure the pseudo displacements resulting from the projection of out‐of‐plane deflection of the plate onto the image plane. A cost function relating the projected numerical and experimental displacement fields is then minimised to obtain the material parameters. It is shown that the contribution of a specific material parameter in the observed displacement field influences the accuracy of its identification. The contributions from material parameters are first quantified in terms of sensitivity criterion that may be tailored by changing the elements of test configuration such as location of supports, the load application point, and the specimen geometry. Several test configurations are designed by maximising the sensitivities corresponding to individual material parameters. The relevance of proposed sensitivity criterion in these configurations is then validated through material identification in simulated experiments with added Gaussian noise. Finally, a thin CFRP plate is tested under these configurations to demonstrate the practical use of this approach. The proposed approach helps in robust estimation of the in‐plane elastic moduli from a bent composite plate with a simple Two‐Dimensional Digital Image Correlation setup without requiring measurement of the actual plate deflection or curvatures.