Efficient time stepping for the multiplicative Maxwell fluid including the Mooney-Rivlin hyperelasticity

TL;DR

This paper introduces a simple, efficient, and robust implicit time stepping algorithm for a finite strain Maxwell fluid model combining Newtonian viscosity with Mooney-Rivlin hyperelasticity, suitable for large strains and non-proportional loading.

Contribution

It proposes an iteration-free, unconditionally stable time integration scheme for the multiplicative Maxwell fluid with hyperelasticity, enhancing robustness and computational efficiency.

Findings

The algorithm exactly preserves inelastic incompressibility and other properties.

It achieves accuracy comparable to existing methods like Euler backward and exponential mapping.

Implementation in MSC.MARC demonstrates practical usability in boundary value problems.

Abstract

A popular version of the finite strain Maxwell fluid is considered, which is based on the multiplicative decomposition of the deformation gradient tensor. The model combines Newtonian viscosity with hyperelasticity of Mooney-Rivlin type; it is a special case of the viscoplasticty model proposed by Simo and Miehe (1992). A simple, efficient and robust implicit time stepping procedure is suggested. Lagrangian and Eulerian versions of the algorithm are available, with equivalent properties. The numerical scheme is iteration free, unconditionally stable and first order accurate. It exactly preserves the inelastic incompressibility, symmetry, positive definiteness of the internal variables, and w-invariance. The accuracy of the stress computations is tested using a series of numerical simulations involving a non-proportional loading and large strain increments. In terms of accuracy, the proposed algorithm is equivalent to the modified Euler backward method with exact inelastic incompressibility; the proposed method is also equivalent to the classical integration method based on exponential mapping. Since the new method is iteration free, it is more robust and computationally efficient. The algorithm is implemented into MSC.MARC and a series of initial boundary value problems is solved in order to demonstrate the usability of the numerical procedures.