An improved approximation technique for the DCOC method of sizing optimization

M. Zhou, G. Rozvany

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

A recently developed optimally criteria method termed DCOC [1, 2] has been shown to improve the efficiency of OC methods significantly. The improvement achieved by this method is due to the fact that the Lagrange multipliers (also termed dual variables) associated with stress constraints are evaluated explicitly and therefore the time consuming computation of generating Lagrange multipliers is greatly reduced. The iterative procedure of DCOC presented in Refs [1, 2] is based on an approximation of the displacement constraints expressed by means of the virtual work principle. The quality of this approximation has been found to be rather poor in certain cases, which causes certain fluctuation during the iterative process. In this paper, a DCOC iterative algorithm using reciprocal linear approximation of the displacement constraints is presented. It is shown by numerical examples that the proposed algorithm considerably improves the iterative performance.

Original languageEnglish
Pages (from-to)763-769
Number of pages7
JournalComputers and Structures
Volume60
Issue number5
DOIs
Publication statusPublished - Jul 10 1996

Fingerprint

Lagrange multipliers
Optimization
Approximation
Stress Constraints
Linear Approximation
Iterative Process
Iterative Procedure
Iterative Algorithm
Fluctuations
Numerical Examples

ASJC Scopus subject areas

  • Computer Science Applications
  • Computational Mechanics

Cite this

An improved approximation technique for the DCOC method of sizing optimization. / Zhou, M.; Rozvany, G.

In: Computers and Structures, Vol. 60, No. 5, 10.07.1996, p. 763-769.

Research output: Contribution to journalArticle

@article{4abdba072ce74235b3eecdcdb0ac1547,
title = "An improved approximation technique for the DCOC method of sizing optimization",
abstract = "A recently developed optimally criteria method termed DCOC [1, 2] has been shown to improve the efficiency of OC methods significantly. The improvement achieved by this method is due to the fact that the Lagrange multipliers (also termed dual variables) associated with stress constraints are evaluated explicitly and therefore the time consuming computation of generating Lagrange multipliers is greatly reduced. The iterative procedure of DCOC presented in Refs [1, 2] is based on an approximation of the displacement constraints expressed by means of the virtual work principle. The quality of this approximation has been found to be rather poor in certain cases, which causes certain fluctuation during the iterative process. In this paper, a DCOC iterative algorithm using reciprocal linear approximation of the displacement constraints is presented. It is shown by numerical examples that the proposed algorithm considerably improves the iterative performance.",
author = "M. Zhou and G. Rozvany",
year = "1996",
month = "7",
day = "10",
doi = "10.1016/0045-7949(95)00436-X",
language = "English",
volume = "60",
pages = "763--769",
journal = "Computers and Structures",
issn = "0045-7949",
publisher = "Elsevier Limited",
number = "5",

}

TY - JOUR

T1 - An improved approximation technique for the DCOC method of sizing optimization

AU - Zhou, M.

AU - Rozvany, G.

PY - 1996/7/10

Y1 - 1996/7/10

N2 - A recently developed optimally criteria method termed DCOC [1, 2] has been shown to improve the efficiency of OC methods significantly. The improvement achieved by this method is due to the fact that the Lagrange multipliers (also termed dual variables) associated with stress constraints are evaluated explicitly and therefore the time consuming computation of generating Lagrange multipliers is greatly reduced. The iterative procedure of DCOC presented in Refs [1, 2] is based on an approximation of the displacement constraints expressed by means of the virtual work principle. The quality of this approximation has been found to be rather poor in certain cases, which causes certain fluctuation during the iterative process. In this paper, a DCOC iterative algorithm using reciprocal linear approximation of the displacement constraints is presented. It is shown by numerical examples that the proposed algorithm considerably improves the iterative performance.

AB - A recently developed optimally criteria method termed DCOC [1, 2] has been shown to improve the efficiency of OC methods significantly. The improvement achieved by this method is due to the fact that the Lagrange multipliers (also termed dual variables) associated with stress constraints are evaluated explicitly and therefore the time consuming computation of generating Lagrange multipliers is greatly reduced. The iterative procedure of DCOC presented in Refs [1, 2] is based on an approximation of the displacement constraints expressed by means of the virtual work principle. The quality of this approximation has been found to be rather poor in certain cases, which causes certain fluctuation during the iterative process. In this paper, a DCOC iterative algorithm using reciprocal linear approximation of the displacement constraints is presented. It is shown by numerical examples that the proposed algorithm considerably improves the iterative performance.

UR - http://www.scopus.com/inward/record.url?scp=0030190803&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030190803&partnerID=8YFLogxK

U2 - 10.1016/0045-7949(95)00436-X

DO - 10.1016/0045-7949(95)00436-X

M3 - Article

VL - 60

SP - 763

EP - 769

JO - Computers and Structures

JF - Computers and Structures

SN - 0045-7949

IS - 5

ER -