Evaluation of Sensitivity-Based Virtual Fields for Non-Linear Parameter Identification Including DIC Filtering Effects

Abstract

Recently, the issue of automatically defining virtual fields for non-linear constitutive models has been resolved (Marek et al., Comput Mech 60:409–431, 2017), relying on a new sensitivity-based approach hereby reducing the influence of noise on the parameter identification. These new set of fields act as weighting factors in the identification process emphasizing the impact of measurement regions with high signal to noise ratios. Conclusions in (Marek et al., Comput Mech 60:409–431, 2017) were drawn based on a numerical example involving small strain plasticity. In this presentation, the performance of these newly defined fields is studied when applied to digital image correlation (DIC) measurement data, hereby including the DIC filtering effects through synthetic image deformation (Rossi et al., Strain 51:206–222, 2017). Results are presented for both large strain isotropic plasticity and hyperelastic material models.