On the identification of rigid body boundary conditions from full-field measurements

This study presents a method to identify Dirichlet boundary conditions (BCs) and constitutive parameters from heterogeneous experimental data, solving a multi-objective optimization problem. The BCs are parameterized as remote rigid body motions leading to a very low number of unknowns and allowing for faster convergence. This methods enables for accurate models of material and structural tests to be obtained. A comparison with classical methods of BC determination is introduced, demonstrating the advantages of the proposed approach. The identified models are evaluated by introducing a validation metric based on experimental data and their uncertainties. The approach is illustrated by using a multi-instrumented tensile test on a well-characterized material. Beyond this validation example, the introduced method paves the way for improved model updating of large-scale tests. • A reduced model for boundary conditions identification method is proposed. • Boundary conditions are accurately determined outside the measurement area. • A quality metric is introduced to better evaluate and compare models. • The study accounts for measurement uncertainties.