Premium
A modified operator splitting (OS) method for collaborative structural analysis (CSA)
Author(s) -
Tada Motohide,
Pan Peng
Publication year - 2007
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.2004
Subject(s) - stability (learning theory) , operator (biology) , linearity , stiffness , numerical analysis , scheme (mathematics) , computer science , base (topology) , operator splitting , control theory (sociology) , mathematics , algorithm , structural engineering , engineering , mathematical analysis , chemistry , biochemistry , control (management) , repressor , machine learning , artificial intelligence , transcription factor , electrical engineering , gene
The collaborative structural analysis (CSA) system is capable of performing highly sophisticated structural analyses utilizing the beneficial features of existing individual structural analysis programmes. It requires a time consuming static condensation procedure if adopting an implicit integration scheme. The operator splitting (OS) method, which does not require the tangential stiffness, can be used to improve the system efficiency. Furthermore, the conventional OS method is not able to provide enough numerical stability, particularly for the analyses considering geometrical non‐linearity. Thus, improvement is needed. To this end, a modified OS method is proposed, which treats unbalanced forces in the current step as pseudo‐external forces in the immediate following step. In this paper, first, the conventional OS method is reformulated in an incremental form, and a CSA scheme based on it is proposed. Second, a modified OS method is developed to improve the numerical stability. Third, a fixed‐base steel column with a lumped mass assigned at its top is analysed using the CSA system as a numerical example. It is found that the OS methods are effective for CSA, and the modified OS method exhibits better numerical stability than the conventional one for the analysis considering geometrical non‐linearity. Copyright © 2007 John Wiley & Sons, Ltd.