Predictive modelling of ice-shelf calving – the dominant mechanism of mass loss from polar ice sheets – remains limited by phenomenological criteria in current ice-sheet models. A physically robust description requires a framework capable of capturing both short-term elastic response (weeks to months) and long-term nonlinear viscous flow (years to centuries). We propose a finite-deformation viscoelastic framework based on logarithmic (Hencky) strain measures, with the logarithmic strain rate as the primary kinematic variable. Unlike conventional multiplicative decomposition approaches, the formulation preserves an additive structure at the strain-rate level, providing a key conceptual and numerical advantage. This choice enables a clear separation between elastic, plastic and viscous contributions, while avoiding intermediate configurations and Lie/Oldroyd-type objective rates, which are known to induce numerical stiffness and algorithmic complexity. The use of the logarithmic strain rate provides a natural and physically meaningful measure of deformation rate under large strains, ensuring objectivity, frame indifference, and a direct link to stress power. From a computational standpoint, the additive structure significantly simplifies time integration, improves numerical robustness at large time steps, and facilitates the consistent incorporation of nonlinear rheologies. Classical Glen-type viscous flow laws, traditionally expressed in terms of deviatoric stress and strain-rate tensors, can be consistently reformulated within the logarithmic strain-rate framework. This provides a thermodynamically sound pathway for incorporating nonlinear viscous behavior while preserving objectivity under finite deformations. Moreover, this formulation enables a straightforward extension toward anisotropic viscosity, which is essential for capturing the strongly directional deformation patterns observed in ice shelves, especially within shear margins where strain localization and fabric-induced anisotropy are prominent. The framework is thermodynamically consistent, suitable for finite-element implementation, and provides a flexible, efficient basis for resolving stress and strain-rate fields near calving fronts. This work represents a promising step toward physically grounded, numerically robust ice-shelf models supporting improved calving criteria.