Behavior and modeling of nonprismatic members having T-sections

Balkaya C.

JOURNAL OF STRUCTURAL ENGINEERING-ASCE, vol.127, no.8, pp.940-946, 2001 (Peer-Reviewed Journal) identifier identifier

  • Publication Type: Article / Article
  • Volume: 127 Issue: 8
  • Publication Date: 2001
  • Doi Number: 10.1061/(asce)0733-9445(2001)127:8(940)
  • Journal Indexes: Science Citation Index Expanded, Scopus
  • Page Numbers: pp.940-946


This paper emphasizes the discrepancies in conventional methods of analyzing nonprismatic members having T-sections. In practice, the change in the location of the centroidal axis over the length of nonprismatic members is ignored in structural analysis. Commonly used portland Cement Association tables for fixed-end moment and stiffness factors are based on the assumptions of a straight-line centroidal axis and variable rectangular cross section: However, discontinuities in the centroidal. axis produce strong coupling between the end moments and horizontal thrust. Nonprismatic members behave similar to an arch, and this creates an axial force that is neglected in conventional methods. Moreover, beams are considered as T-beams in the,structural models because of the effect of RC slabs. Consequently, nonprismatic T-sections require special consideration in structural analysis. The purpose of this paper is to present the behavior of nonprismatic T-section beams based on a 3D finite-element investigation. Based on the results of parametric studies, a simple, practical 2D beam model with an effective length concept is proposed for the analysis of nonprismatic members having T-sections. In the proposed model, the effective length of the 2D beam element representing the haunch evolves as 75% of its actual length. It-is concluded that the deviation in results from the results of the finite-element analysis is within 15%. Additionally, different modeling schemes in the analysis of frames with nonprismatic members using the proposed 2D beam representation are investigated.