Abstract
Biaxial mechanical testing combined with the virtual fields method (VFM) has been increasingly employed to characterize hyperelastic materials. In our preliminary tests, we observed that when uniform-type virtual fields (VFs), originally developed for uniaxial tensile configurations, were applied to biaxial loading conditions, the nodes along the edges of the selected region of interest (ROI) exhibited lateral virtual displacements. These displacements generated additional external virtual work (EVW) owing to the inherent shear stresses along the ROI edges; however, since these stresses were experimentally inaccessible, this EVW contribution could not be reliably quantified and thus biased the identified parameters. Furthermore, the magnitude of this parasitic contribution was found to depend on the ROI size, as the shear stresses along the ROI edges changed correspondingly. These observations underscore the need for a systematic investigation into the effects of ROI selection and the underlying shear stresses on parameter identification. In parallel, although a trigonometric-type VF has been proposed for parameter identification, its underlying working mechanism remains insufficiently understood. To address these gaps, this study evaluated five cruciform ROIs of different sizes using the VFM in conjunction with full-field deformation data obtained from biaxial tests. Material parameters were extracted from each ROI under both the Mooney--Rivlin and Ogden hyperelastic formulations. The influence of ROI selection was assessed based on the capability of the identified parameters to reproduce the experimental load–displacement responses and the strain fields. The identified parameters obtained from uniform-type VFs alone were compared with those derived from a combination of uniform- and trigonometric-type VFs. The results demonstrate that 1) ROI3, with a span ratio of 0.49, effectively mitigates the adverse effects of additional EVW induced by shear stresses and represents the optimal compromise for parameter identification; 2) the introduction of the trigonometric-type VF reduces the condition number of solving a linear system of equations, thereby enhancing the numerical stability of parameter identification. Overall, this study provides a comprehensive methodological framework for investigating the role of ROI selection and the underlying shear stresses on the parameter identification of hyperelastic materials under biaxial loading.