An efficient and accurate semi-implicit time integration scheme for dynamics in nearly- and fully-incompressible hyperelastic solids

Edward M. Terrell, Boyce E. Griffith · 2025

Read An efficient and accurate semi-implicit time integration scheme for dynamics in nearly- and fully-incompressible hyperelastic solids by Edward M. Terrell, Boyce E. Griffith (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-14284.txt.

Introduction

Simulating soft solids that stretch a lot, like rubber or tissue, is hard because waves inside them travel at very different speeds. Edward Terrell and Boyce Griffith study time-stepping schemes for such nearly or fully incompressible materials. Simple explicit schemes are cheap but need tiny steps when fast waves are present. Fully implicit schemes allow big steps but cost much more. The pair develops two semi-implicit schemes that sit between those extremes, one based on backward differentiation and one mixing forward Euler steps with that approach.

They prove both schemes are second-order accurate and test them in practice. One scheme, FEBDF2, has a step limit set by the slow shear waves and not by the fast bulk waves, which is the trait users want. The authors also report a weakness: with nonlinear incompressibility rules, the schemes can fail to preserve volume exactly, even under fine refinement.

Catalog record

CREATOREdward M. Terrell, Boyce E. Griffith
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE60 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-14284

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

An efficient and accurate semi-implicit time integration scheme for dynamics in nearly- and fully-incompressible hyperelastic solids Edward M. Terrell, Boyce E. Griffith Abstract The choice of numerical integrator in approximating solutions to dynamic partial differential equations depends on the smallest time-scale of the problem at hand. Large-scale deformations in elastic solids contain both shear waves and bulk waves, the latter of which can travel infinitely fast in incompressible materials. Explicit schemes, which are favored for their efficiency in resolving low-speed dynamics, are bound by time step size restrictions that inversely scale with the fastest wave speed. Implicit schemes can enable larger time step sizes regardless of the wave speeds present, though they are much more computationally expensive. Semi-implicit methods, which are more stable than explicit methods and more efficient than implicit methods, are emerging in the literature, though their applicability to nonlinear elasticity is not extensively studied. In this research, we develop and investigate the functionality of two time integration schemes for the resolution of large-scale dynamics in nearly- and fully-incompressible hyperelastic solids: a Modified Semi-implicit Backward Differentiation Formula integrator (MSBDF2) and a forward Euler / Semi-implicit Backward Differentiation Formula Runge-Kutta integrator (FEBDF2). We prove and empirically verify second order accuracy for both schemes. The stability properties of both methods are derived and numerically verified. We find FEBDF2 has a maximum time step size that inversely scales with the shear wave speed and is unaffected by the bulk wave speed—the desired stability property of a semi-implicit scheme. Finally, we empirically determine that semi-implicit schemes struggle to preserve volume globally when using nonlinear incompressibility conditions, even under temporal and spatial refinement. Keywords— nonlinear elasticity, incompressible materials, finite elements, semi-implicit 1 Introduction 1.1 Background Efficient and accurate numerical methods for simulating dynamic, large-scale deformations in nearly-incompressible or incompressible solids remain scarce, despite the ubiquity of such materials in various scientific fields. Many biomaterials, such as skin and muscle, are mainly comprised of water, and are often approximated as incompressible [ 1 , 2 ] . Various polymers, most notably rubber, also exhibit behavior that can be generalized as incompressible [ 3 ] . Fields that heavily depend on simulations involving such incompressible materials, like physiological modeling or biomedical engineering, would benefit greatly from having access to efficient numerical methods that can resolve incompressible dynamics in solids. Numerous finite element schemes exist for problems in finite elasticity. Because of their utility in structural engineering, particular focus has been applied to developing methods that solve for the equilibrium states of solids under static stresses [ 4 , 5 , 6 , 7 , 8 ] . A broadly-used subclass of static methods, known as quasi-static methods, was developed [ 9 , 10 ] to model elastic deformations under gradual dynamic loads. While extremely effective for analyzing deformations in regimes where inertial effects are negligible, these models cannot resolve finer time-scale phenomena like elastic wave propagation [ 11 , 12 ] . Evaluating static and quasi-static formulations for solid mechanics reduces to solving nonlinear elliptic PDEs. Because of the inherent numerical stiffness associated with resolving nonlinear dynamics in nearly- and fully-incompressible solids, explicit time-stepping methods are often a poor choice due to severe restrictions on time step size [ 13 ] . Implicit methods are capable of capturing the relevant dynamics without crippling time step size restrictions, but they depend on nonlinear solvers [ 14 , 15 ] . This highlights the potential utility of an intermediate form of integrator: a semi-implicit method that is more stable than an explicit method and more efficient than an implicit method. Extant semi-implicit time integration schemes for solving dynamic problems in nonlinear incompressible elasticity come in various forms, though they all involve the introduction of a Lagrange-multiplier pressure field that is used to enforce incompressibility. Lahiri et al. formulated a fractional-step scheme reminiscent of a Størmer-Verlet or leap-frog integrator [ 16 , 17 ] . Gil et al. introduced a three-field Petrov-Galerkin scheme coupled with a total variation diminishing Runge-Kutta method [ 18 , 19 ] . This work also utilizes a variational stabilization technique similar to the dynamic variational multiscale stabilization introduced by Scovazzi et al. [ 20 , 21 ] that permits the usage of linear finite elements for all relevant finite element spaces. Kadapa introduced a semi-implicit scheme based on an explicit form of the Newmark- \beta method [ 15 , 22 ] . These integrators can use time step sizes that scale with the shear wave speed of the modeled material, enabling them to step over dynamics associated with volumetric deformations. However, none of these methods have been demonstrated to achieve the degree of temporal accuracy claimed, nor have their abilities to preserve volume been tested for large deformations. This paper introduces two semi-implicit integration schemes based on the second-order semi-implicit backward differentiation formula (SBDF2), which resembles the standard second-order backward differentiation formula [ 23 ] except that it treats non-stiff force terms explicitly [ 24 ] . Second-order Adams-Bashforth extrapolations are used to push these non-stiff terms into the forward time step [ 25 ] . While SBDF2 is well-studied [ 26 , 27 , 28 ] , its application to hyperbolic problems remains relatively unexplored, and it has not been applied in the context of nonlinear incompressible elastodynamics. The proposed modifications to SBDF2 are unique in that they leverage explicit extrapolations within the volumetric energy to ensure that all relevant forcing terms are calculated at the future time step, maintaining second-order accuracy while minimizing computational cost. Semi-implicit schemes for nonlinear elastodynamics are inspired by methods commonly used in computational fluid dynamics—the most obvious being projection methods of the form popularized by Chorin [ 29 ] . Projection methods require precise enforcement of boundary conditions for the pressure field, which is difficult for general problems in nonlinear elasticity. The schemes proposed in this work distance themselves from the class of projection methods by never explicitly enforcing pressure boundary conditions; they are instead enforced simultaneously with the boundary conditions for the velocity. The key contributions of this paper involve the validation of semi-implicit schemes in their application to problems in nonlinear incompressible elastodynamics. We used both local-truncation error analysis and empirical convergence studies to validate that the proposed schemes are second order accurate for both compressible and incompressible materials. Furthermore, we determined via linear stability analysis that one of the proposed schemes has a time step size that inversely scales with the shear wave speed of the material and is unaffected by the bulk wave speed of the material, which is the standard for semi-implicit schemes. The other proposed scheme is found to be unconditionally unstable despite being derived from a well-established integrator. Lastly, both proposed schemes are found to struggle with preserving volume in incompressible dynamics if a nonlinear formulation for the volumetric energy is used; using a linear incompressiblity condition enables convergence to volume preservation under temporal and spatial refinement. 2 Preliminaries 2.1 Continuum mechanics The aim of the algorithms presented in this manuscript is to efficiently model the dynamic deformation of a domain or body, \Omega_{t}\in\mathbb{R}^{d}\text{ for }d=2 or 3 from time t=0 to T . The state of \Omega_{t} at these time points is denoted by \Omega_{0} and \Omega_{T} respectively.

Read the complete text