Remarks on the interpretation of current non‐linear finite element analyses as differential–algebraic equations

Abstract

SUMMARYFor the numerical solution of materially non-linear problems like in computational plasticity or viscoplasticity the ÿnite element discretization in space is usually coupled with point-wise deÿned evolution equations characterizing the material behaviour. The interpretation of such systems as di erentialalgebraic equations (DAE) allows modern-day integration algorithms from Numerical Mathematics to be e ciently applied. Especially, the application of diagonally implicit Runge-Kutta methods (DIRK) together with a Multilevel-Newton method preserves the algorithmic structure of current ÿnite element implementations which are based on the principle of virtual displacements and on backward Euler schemes for the local time integration. Moreover, the notion of the consistent tangent operator becomes more obvious in this context. The quadratical order of convergence of the Multilevel-Newton algorithm is usually validated by numerical studies. However, an analytical proof of this second order convergence has already been given by authors in the ÿeld of non-linear electrical networks. We show that this proof can be applied in the current context based on the DAE interpretation mentioned above. We ÿnally compare the proposed procedure to several well-known stress algorithms and show that the inclusion of a step-size control based on local error estimations merely requires a small extra time-investment.

Static condensation within the context of multifield elastoplasticity

“…Compared to conventional methods, like in [1], this approach is characterized by a high numerical effort due to the increased number of unknowns. Hence, it is inevitable to analyze strategies to counter act this phenomenon.…”

“…The time discretization is obtained by applying stiffly accurate RUNGE-KUTTA schemes, cf. [1,4,5]. Thus, in the end a linear system of equations…”

Field Assisted Sintering Technology, Part II: Simulation

“…t n+1 is the time at step n + 1. In this respect, we refer to [8,12,45,46] and the literature cited therein. The problem is solved using a monolithic approach, where the mechanical problem depends only on the heat, the heat equation is controlled by Joule-heating, i.e.…”

Iterative solvers within sequences of large linear systems in non‐linear structural mechanics

“…The number of unknown displacements ranges from n u = 5006 up to n u = 100520 and the corresponding number of internal variables ranges from n Q = 126720 up to n Q = 2965248. For the time integration the method of Ellsiepen is used (see, [20,21]), which is of second order and has two stages. As mentioned in Appendix A the local integration step (8) is very cheap for the POM-model (only function evaluations, see [30]) and expensive in the case of the applied metal plasticity model, where local iterations have to be performed (see [32]).…”

“…The treatment of constitutive models of evolutionary-type within implicit finite elements leads after the spatial discretization to a system of differential-algebraic equations (DAE-system) where the algebraic part results from the discretized weak formulation of the equilibrium conditions and the differential part is the outcome of the assemblage of all constitutive model's evolutionary equations at all spatial integration points -Gauss-points (see, for example, [21,28]). Usually, this is solved by a Backward-Euler method.…”