Shin et al. 2004

From Composite Systems
Jump to: navigation, search

This paper focuses on the experimental and analytical behavior of concrete filled tubing column to H-beam welded moment connections with external T-stiffeners. Parameters tested were strength ratio of the horizontal stiffener to beam flange and strength ratio of the vertical stiffener to the beam flange. The ABAQUS finite element package was used to simulate the experimental behavior.

Experimental Study, Results, and Discussions

Six full-scale specimens were tested. Ratios of the horizontal stiffener strength to beam flange strength were 70% and 130% and ratios of the vertical stiffener to beam flange strength were 70%, 100%, and 130%. Length and height of stiffeners were also varied. Horizontal stiffeners were welded to the beam flange using double-V-groove welds and to the column flange using complete joint penetration single bevel groove welds. The vertical stiffeners were welded to the horizontal stiffeners using double bevel groove welds and to the column corner using single bevel groove welds.

Tests were conducted on specimens fabricated as cantilevers attached to CFT column stubs. There was no floor slab on the beam. The test setup was designed to simulate the boundary conditions of a beam to column connection subassembly in a moment resisting frame under typical lateral loading. The column was assumed to be pin supported at mid-story point and the beam was loaded at an assumed pin at its midspan.

The failure modes of the connections included shear failure in the horizontal stiffener, tensile failure in the vertical stiffener, and beam flange failure and local buckling in beam. Tests found that vertical stiffeners only effectively transfer loads introduced from beam to column while the horizontal stiffeners plays a more critical role in the ultimate moment and deformation capacity.

The finite element model used in this study was and eight node brick reduced integration element. Contact surfaces were used in consideration of the filling effect of the concrete. Material non-linearity was accounted for in the monotonic analyses through an isotropic classical plasticity model based on the Von Mises yield criterion and associated plastic flow. Isotropic hardening was assumed for monotonic analyses and kinematic hardening was assumed for cyclic analyses. The kinematic hardening rule considered non-linearity and so the Bauschinger effect was adopted in this study. Comparison with experimental results found that the ultimate load and initial stiffness were well represented in the analysis.


  • Shin, K. J., Kim, Y. J., Oh, Y. S., and Moon, T. S. (2004). “Behavior of Welded CFT Column to H-beam Connections with External Stiffeners.” Engineering Structures, 26(13), 1877–1887. doi:10.1016/j.engstruct.2004.06.016