Premium
Continuity conditions for finite element analysis of solids
Author(s) -
De Almeida J. P. Moitinho,
De Freitas J. A. Teixeira
Publication year - 1992
Publication title -
international journal for numerical methods in engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.421
H-Index - 168
eISSN - 1097-0207
pISSN - 0029-5981
DOI - 10.1002/nme.1620330411
Subject(s) - finite element method , mathematics , a priori and a posteriori , compatibility (geochemistry) , element (criminal law) , unitary state , constraint (computer aided design) , mathematical optimization , mathematical analysis , geometry , structural engineering , engineering , philosophy , epistemology , chemical engineering , political science , law
In conventional finite element formulations the concept of node—a point where one of the shape functions is unitary and all others are nil—is used to advantage as it simplifies the definition of interelement continuity conditions. This constraint on the definition of the shape functions hinders the formulation of elements with complex shapes and, in particular, of equilibrium elements. In the approach presented herein linearly independent functions are defined within each element irrespectively of the location of the nodes. Interelement continuity conditions are imposed ‘ a posteriori ’, as in hybrid elements. The derivation of the element matrices is based upon the equations expressing equilibrium, compatibility and the constitutive relations without explicitly using variational principles. This results in a wider choice of available funciiuns ami in an easier way to formulate equilibrium elements and/or to use conforming or non‐conforming elements. As the approach used is independent of the choice of basic functions and of the shape of the elements, it is perfectly general. It allows the parallel analysis of kinematically and statically admissible formulations, as proposed by Fraeijs de Veubeke. 1 As the interelement continuity conditions are imposed ‘ a posteriori ’ new variables are used to express this condition.