SHAKE TABLE TESTING OF STEEL BRACED FRAME CONSIDERING MEMBER FRACTURE

Abstract

 1. Introduction
 The authors have proposed a method to assess the local fracture mechanism of moment frame connections and brace members under cyclic loading directly from global member results, termed the Direct Local Strain Evaluation Method. However, up to now, the accuracy of the proposed method has only been validated under static loading. This research validates the proposed method under dynamic loading with collapse tests.
 2. Test Setup
 A frame composed of a single bay moment frame and a concentric circular hollow section (CHS) brace was assembled for the dynamic collapse test, with LVDT strain gauges installed to monitor fracture of the brace and the beam connections. Initially quasistatic loading tests were carried out to confirm the fundamental mechanical properties of the braced frame specimens. The testing rig included an inertial mass system to simulate PΔ effects large deformation.
 3. Quasistatic and Dyanamic Loading Tests on Braced Frames
 The diameter-thickness ratios of CHS brace of specimens ranged from 42.3 (slender) to 11.8 (compact). In the quasistatic tests, the slender specimens failed by local buckling at the midpoint and ends of the brace member, while local buckling was not observed for the compact specimens. However, when connection fracture governed, the cumulative deformation capacity of the brace was not necessarily determined by just the diameter-to-thickness ratio. In the dynamic loading test, the slender specimens, residual story drift angle was 1/282 rad at initiation of brace fracture, 1/21 rad following fracture of the brace and 1/6 rad once the beam connection fractured.
 4. Numerical Investigation on Simulation of Quasistatic Loading Tests
 A series of FEM analyses were conducted in order to simulate of the quasistatic loading tests, where the member modelled with shell elements. The Direct Local Strain Evaluation Method is largely effective estimating the local strain of the brace member and at the beam connection.
 5. Validity of Direct Local Strain Evaluation Method under Dynamic Loading
 Numerical FEM models were also created to simulate of the dynamic loading test using stick models. The estimated initiation of local buckling was found to be more accurate using truss elements for the brace than by using bending elements. Generally, the post-brace-fracture displacements of a dual brace-moment frame system are much greater than the displacement contribution of local buckling. While local buckling can have a significant effect on the pre-brace fracture displacement, the precise instant that local buckling is initiated has a relatively small effect on the total displacement response of a frame with brace fracture. The proposed method is effectively predicted the braced frame response including the brace fracture under dynamic loading.
 6. Conclusions
 In summary, the following results were obtained:
 1) Local buckling occurred at the midpoint and ends of the brace for slender CHS specimens, while local buckling of compact CHS specimens was not observed in the quasistatic loading tests.
 2) In the dynamic loading test of the slender CHS specimens, the residual story drift angle was 1/282 rad at initiation of brace fracture, 1/21 rad following fracture of the brace and 1/6 rad once the beam connection fractured.
 3) The Direct Local Strain Evaluation Method is effective in estimating the local strain of the brace member and at the beam connection.
 4) The proposed method effectively predicts the braced frame response including the brace fracture under dynamic loading.