Mechanical and informational modeling of steel beam-to-column connections

Jun Hee Kim, Jamshid Ghaboussi, Amr S. Elnashai

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

The behavior of beam-to-column connections in steel and composite frames strongly influences their stability and strength. This is particularly true in response to sever dynamic actions, where the changes in stiffness and damping influence both supply and demand. This means that it is necessary to accurately model the stiffness, strength and ductility of connections in seismic assessment and analysis for design. Starting from the current state-of-the-art, two different approaches, mechanical and informational, are presented to model the complex hysteretic response of bolted beam-to-column connections. The basic premise of the article is that not all features of response are amenable to mechanical modeling; hence, consideration of information-based alternatives is warranted. First, a component-based mechanical model is proposed where each deformation source is represented with only material and geometric properties. Second, a neural network approach is examined to extract an informational model directly from the experimental test data. Finally, the merits and drawbacks of the two approaches are discussed. The results presented in this article indicate that the two models demonstrate good capabilities of predicting complex hysteretic responses in certain cases. There is still, however, room for improvement. Such improvement may be achieved through combining the best features of each approach in a hybrid mechanical-informational modeling environment.

Original languageEnglish
Pages (from-to)449-458
Number of pages10
JournalEngineering Structures
Volume32
Issue number2
DOIs
Publication statusPublished - 2010 Feb 1

All Science Journal Classification (ASJC) codes

  • Civil and Structural Engineering

Fingerprint Dive into the research topics of 'Mechanical and informational modeling of steel beam-to-column connections'. Together they form a unique fingerprint.

  • Cite this