Surface-Polyconvex Models for Soft Elastic Solids

Soft solids with surface energy exhibit complex mechanical behavior, necessitating advanced constitutive models to capture the interplay between bulk and surface mechanics. This interplay has profound implications for material design and emerging technologies. In this work, we set up variational models for bulk-surface elasticity and explore a novel class of surface-polyconvex constitutive models that account for surface energy while ensuring the existence of minimizers. These models are implemented within a finite element framework and validated through benchmark problems and applications, including, e.g., the liquid bridge problem and the Rayleigh-Plateau instability, for which the surface energy plays the dominant role. The results demonstrate the ability of surface-polyconvex models to accurately capture surface-driven phenomena, establishing them as a powerful tool for advancing the mechanics of soft materials in both engineering and biological applications.

💡 Research Summary

This paper develops a rigorous variational framework for soft solids whose mechanical response is significantly influenced by surface energy. Building on the classical Gurtin‑Murdoch theory, the authors introduce the concept of surface‑polyconvexity—a surface analogue of bulk polyconvexity—to guarantee the existence of energy minimizers even under large deformations. The kinematic description starts with a reference domain Ω⊂ℝ³ and its boundary Γ, distinguishing a surface region S⊂Γ that carries surface energy. The bulk deformation gradient F=∇φ and its cofactor, determinant, and the surface deformation gradient ˆF are defined. ˆF is obtained by projecting the bulk gradient onto the tangent space of the surface, and its pseudo‑inverse exists despite rank deficiency. A surface Jacobian ˆJ is derived as |cof F N|, which equals √I₂(ˆC) where ˆC=ˆFᵀˆF, thus providing a measure of area change analogous to the bulk Jacobian.

The total energy comprises bulk stored energy ∫ΩΨ_b(F) dV, surface stored energy ∫SΨ_s(ˆF,N) dA, and external work. By applying the principle of stationary potential energy, the first variation yields bulk Piola‑Kirchhoff stress and a surface stress tensor derived from the surface energy density. The crucial modeling step is the construction of Ψ_s that satisfies surface‑polyconvexity: Ψ_s is expressed as a sum of a determinant‑dependent term W₁(det ˆF) and an invariant‑dependent term W₂(I₁(ˆC),I₂(ˆC)). Both terms are chosen to be convex in the arguments (det ˆF, cof ˆF, ˆF) and thus ensure weak lower semicontinuity of the total energy functional. This guarantees, via the direct method of the calculus of variations, the existence of a minimizer in appropriate Sobolev spaces for the bulk and the surface.

A finite‑element implementation is presented in which bulk 8‑node hexahedral elements are coupled with surface quadrilateral elements that compute ˆF, N, and the associated stresses. A Newton‑Raphson scheme with line search solves the nonlinear equilibrium equations, and the consistent tangent operators are derived from the polyconvex energy forms.

The model is validated on several benchmark problems. In the liquid‑bridge configuration, the surface‑driven shape is captured accurately, outperforming traditional isotropic surface‑tension models. For the Rayleigh‑Plateau instability of a soft cylindrical filament, the predicted critical wavelength and growth rates agree with analytical theory and experimental data within a few percent. Simulations of the Shuttleworth effect demonstrate that the model can naturally incorporate strain‑dependent surface tension, reproducing observed variations of surface stress under tension and compression.

Overall, the work makes three major contributions: (1) it introduces surface‑polyconvexity as a mathematically sound condition for surface elasticity, (2) it proposes concrete, physically motivated surface energy densities that respect this condition while capturing strain‑dependent surface effects, and (3) it demonstrates that these models can be efficiently implemented in standard finite‑element codes and applied to realistic soft‑material problems. Limitations include the current focus on isotropic surface behavior and the need to extend the framework to anisotropic, multiphysics, or gradient‑polyconvex formulations, which are identified as directions for future research.