Do locking-free finite element schemes lock for holey Reissner-Mindlin plates with mixed boundary conditions

Mark Ainsworth, Charles Parker · 2025

Read Do locking-free finite element schemes lock for holey Reissner-Mindlin plates with mixed boundary conditions by Mark Ainsworth, Charles Parker (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-2506-21999.txt.

Introduction

Mark Ainsworth and Charles Parker ask whether locking-free finite element schemes still lock for Reissner-Mindlin plates with holes and mixed boundary conditions. Their 2025 paper revisits plate discretizations in this harder setting.

Guided by the de Rham complex, they derive conditions under which a scheme gives locking-free, optimal convergence rates. The usual assumptions for clamped or simply supported plates appear naturally. New conditions also arise from holes in the domain or from mixed boundary conditions. The good news, they show, is that many popular schemes in current use do satisfy all of the conditions. So those schemes remain locking-free here.

Catalog record

CREATORMark Ainsworth, Charles Parker
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE133 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-21999

Browse the full library · About the Tower of Babel

Full text

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

Do locking-free finite element schemes lock for holey Reissner-Mindlin plates with mixed boundary conditions? Note: The second author acknowledge that this material is based upon work supported by the National Science Foundation under Award No. DMS-2201487. Journal: Computer Methods in Applied Mechanics and Engineering Mark Ainsworth, Charles Parker Abstract We revisit finite element discretizations of the Reissner-Mindlin plate in the case of non-simply connected (holey) domains with mixed boundary conditions. Guided by the de Rham complex, we develop conditions under which schemes deliver locking-free, optimal rates of convergence. We naturally recover the typical assumptions arising for clamped, simply supported plates. More importantly, we also see new conditions arise naturally from the presence of holes in the domain or in the case of mixed boundary conditions. We show that, fortunately, many of the existing popularly used schemes do , in fact, satisfy all of the conditions, and thus are locking-free. 2020 MSC 65N30, 65N12, 74K20, 74S10 1 Introduction Although the connection between differential forms, exterior calculus and de Rham exact sequences to finite element analysis goes back to Bossavit Bossavit98 in the seting of approximation of Maxwell’s equations, the recognition that finite element exterior calculus provides powerful tool for understanding discretisation schemes is, thanks to the efforts of Arnold2010 , Arn18 , ArnFalkWin06 , Hiptmair02 now widely understood. This connection seems particularly powerful in the case where high order finite elements are considered and has led to a number of results. As far as our own work is concerned, these ideas have been exploited to establish that various families of mixed finite elements are uniformly inf-sup stable AinCP19StokesI , AinCP21LE which, in turn, paves the way for high order, uniformly stable schemes for incompressible flow that are pointwise mass conserving, and to explain locking of finite element in linear elasticity AinCP21LE . On a different tack, ideas from exact sequences were used AinCP23KirchI to construct C^{1} (or H^{2} -conforming) finite element approximations without having to implement families of elements with built-in C^{1} continuity. The same ideas were shown in AinCP24KirchII to lead to computational procedures for constructing C^{1} finite element approximations using elements for which it is known that there is no local basis (meaning that they cannot be implemented using classical finite element methodology). In the current work, we revisit the classical Reissner-Mindlin (RM) plate model. The literature on finite element approximation for RM is enormous and one may reasonably question what, if anything, new can be said about finite element schemes for RM? While RM is of interest in its own right as a computational model for thin elastic plates, part of the interest in the problem arises from the fact that it constitutes a coupled system of elliptic partial differential equations in which there are two distinct scales corresponding to the O(1) diameter of the plate and its thickness t\ll 1 . As such, the model may be viewed as a prototype for more complex multiscale problems. The presence of two distinct scales leads to difficulties in the numerical discretization in the form of locking whereby a finite element scheme converges, if it converges at all, at a sub-optimal rate until the mesh size h is reduced to O(t) . Nowadays, various schemes are available that have been proved to not lock in this way BoffiBrezziFortin13 , Braess07 , Brezzi1991MITC . The analysis and development of many of these schemes makes use of ideas from discrete de Rham sequences but often predates the widespread adoption of ideas from finite element exterior calculus. Virtually all of the existing literature on finite element methods (conforming or otherwise) for RM considers, for simplicity, the case of a clamped, simply connected plate Ain02 , Arn93 , Arn07 , Arn05 , Arnold89 , Arn97RM , BatheBrezzi85 , BatheBrezzi87 , Behrens11 , Bosing10 , Bramble98 , Brezzi89 , Brezzi86 , Brezzi1991MITC , Calo14 , Carstensen11 , Chen25 , Chinosi06 , Veiga13 , DiPietro22 , Duran92 , Falk2000 , Falk2008 , Hansbo11 , Huang24 , Huoyuan18 , Lovadina05 , Pechstein17 , Peisker92 , Sky23 , StenbergSuri97 , Ye20 , Zhang23 . While this certainly simplifies the analysis, practical applications where one might wish to use the RM model are more likely to involve all types of boundary condition including stress-free, simply supported, and clamped on complicated domains with holes. The mention of domains with holes raises warning flags for any analysis that is based on exact sequences, discrete or otherwise. In particular, the structure of the continuous exact sequence on domains with holes is complicated by the presence of mathematical structures known as harmonic forms which have an entirely different structure from those that appear in the simply connected case. These are discussed in detail later in the manuscript but, for present purposes, their presence means that arguments valid in the simply connected case are no longer necessarily valid in the more general case of domains with holes. Equally well, when mixed boundary conditions are prescribed, similar considerations apply and again, arguments that are used in the case of pure clamping and no longer necessarily valid in the case of mixed boundary conditions. This raises the danger whereby schemes developed and analyzed in the case of simply connected domains with pure clamping may not perform as expected on practical problems. In the current work, we revisit finite element discretizations of the RM plate in the case of non-simply connected domains with mixed boundary conditions. Our approach starts from exact sequences and the de Rham diagram. Firstly, we discuss the problem of developing schemes for RM at a rather intuitive level guided by the de Rham sequence. One outcome of this is that one sees the typical assumptions, which form the starting point for many existing works, materialize naturally from intuitive arguments based on exact sequences. More importantly, one also sees new conditions arise naturally from the presence of discrete harmonic forms in the case of domains with holes or in the case of mixed boundary conditions. Respecting these new conditions is necessary if one wants to have locking-free schemes or if one wants to be able to use the existing schemes with confidence for such problems. These conditions are absent from existing analyses with the result that, as mentioned above, one cannot be sure that the schemes will be valid for practical problems. Having derived the new conditions following heuristic arguments guided by exact sequences, we then provide rigorous mathematical analyses showing that the conditions are sufficient to deliver locking-free, optimal rates of convergence provides that the conditions are satisfied. This naturally leads to the question of whether existing schemes satisfy the full set of conditions. The good news, though hardly obvious at a glance, is that many of the existing popularly used schemes do , in fact, satisfy the full set of conditions. While the analysis needed to arrive at this conclusion is rather involved, the outcome should be what one expects given that the schemes have certainly been used by practitioners to approximate problems on domains with holes and mixed boundary conditions without reports of failure. In a sense, the fact that many existing schemes do in fact satisfy the full set of conditions may be attributed to good fortune. However, it is

Read the complete text