Abstract

The work presented is part of a research program conducted to study the response of built-up T-stubs under cyclic loading. The goal of this research is to demonstrate that the failure modes and limit states that control the behavior of built-up T-stubs are the same that govern the hot rolled T-stub behavior, therefore current design recommendations for double split tee moment connections are appropriate when built-up T-stubs are used. The behavior of built-up T-stubs subjected to cyclic loading through the stem is studied through an experimental program and numerical analyses: twenty one-half scale T-stub specimens were tested under cyclic loading and a 3D finite model previously developed was adapted to simulate the cyclic response of the specimens under loading conditions similar to the experimental conditions. The experimental results showed overstrength with respect to that predicted by current design recommendations and differences between the expected and actual failure modes observed. Only one specimen presented weld damage triggering a brittle failure, caused by a defective weld. The model reproduced adequately the load-deformation response and failure modes of the specimens and can be used to study other configurations. It was concluded that the weld has no effect on the response of the T-stub and, consequently design recommendations for Double Split T connections can be applied when using built-up T-stubs. (C) 2014 Elsevier Ltd. All rights reserved.