Computational Technology Resources - CCP

Keywords: assemblies, error estimation, quantity of interest.

Summary

A common practice in the framework of industrial design is to use finite element codes to model assemblies. In this paper a study is presented examines the quality of such simulations. Some work has been developed in the past for controlling finite element simulations with contact but the measure introduced was global. Mechanical criteria are often local, hence a global measure is not sufficient.

The path followed is to obtain an error in a quantity of interest and to extend previous work developed for linear analysis. The final objective is to control meshing in the framework of assemblies, starting from the local error measure defined in this work.

For more than thirty years the global discretization error has been extensively developed. More recently, research has focused on goal-oriented error estimation, i.e. the estimation of the error on specific outputs of interest, which may be relevant for design purposes. Several techniques have been proposed for goal-oriented error estimation, and particularly for linear problems based on the resolution of a dual problem. Extensions of this method have also been proposed for different non-linear problems. Here a different approach is proposed that results in a lower cost. The results are not proved mathematically but remains very sharp, which is interesting from a practical point of view. Here the constitutive error concept is used to estimate the finite element error. This method relies on the construction of the so-called admissible fields.

The first point of the work, described in this paper, is to revisit the construction of the stresses that check equilibrium. Particular attention is paid to the construction at the boundary of the studied domain; especially for boundaries that are part of the contact zone. From a practical point of view, some information can be missing as a result of file transfer between the codes and are not available for a posteriori error estimation.

The second point is to develope a practical tool for estimating error on the bais of stress. Different methods are proposed and numerical tests are carried out in order to illustrate the proposed methodology.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £65 +P&P)

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Proceedings ISSN 1759-3433 CCP: 99 PROCEEDINGS OF THE ELEVENTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: B.H.V. Topping Paper 297 A Practical Tool for Estimating Errors of Stresses in Assemblies S. Pavot¹, E. Florentin¹, P. Pasquet² and S. Guinard³ ¹LMT-Cachan (ENS Cachan / CNRS / UPMC / PRES UniverSud Paris), France ²SAMTECH, Massy, France ³EADS-IW, Toulouse, France doi:10.4203/ccp.99.297 purchase the full-text of this paper Full Bibliographic Reference for this paper S. Pavot, E. Florentin, P. Pasquet, S. Guinard, "A Practical Tool for Estimating Errors of Stresses in Assemblies", in B.H.V. Topping, (Editor), "Proceedings of the Eleventh International Conference on Computational Structures Technology", Civil-Comp Press, Stirlingshire, UK, Paper 297, 2012. doi:10.4203/ccp.99.297 Keywords: assemblies, error estimation, quantity of interest. Summary A common practice in the framework of industrial design is to use finite element codes to model assemblies. In this paper a study is presented examines the quality of such simulations. Some work has been developed in the past for controlling finite element simulations with contact but the measure introduced was global. Mechanical criteria are often local, hence a global measure is not sufficient. The path followed is to obtain an error in a quantity of interest and to extend previous work developed for linear analysis. The final objective is to control meshing in the framework of assemblies, starting from the local error measure defined in this work. For more than thirty years the global discretization error has been extensively developed. More recently, research has focused on goal-oriented error estimation, i.e. the estimation of the error on specific outputs of interest, which may be relevant for design purposes. Several techniques have been proposed for goal-oriented error estimation, and particularly for linear problems based on the resolution of a dual problem. Extensions of this method have also been proposed for different non-linear problems. Here a different approach is proposed that results in a lower cost. The results are not proved mathematically but remains very sharp, which is interesting from a practical point of view. Here the constitutive error concept is used to estimate the finite element error. This method relies on the construction of the so-called admissible fields. The first point of the work, described in this paper, is to revisit the construction of the stresses that check equilibrium. Particular attention is paid to the construction at the boundary of the studied domain; especially for boundaries that are part of the contact zone. From a practical point of view, some information can be missing as a result of file transfer between the codes and are not available for a posteriori error estimation. The second point is to develope a practical tool for estimating error on the bais of stress. Different methods are proposed and numerical tests are carried out in order to illustrate the proposed methodology. purchase the full-text of this paper (price £20) go to the previous paper go to the next paper return to the table of contents return to the book description purchase this book (price £65 +P&P)
Back to top	©Civil-Comp Limited 2023 - terms & conditions