This study evaluates the feasibility of a new class of partially restrained connections using shape memory alloys. In the martensitic form, shape memory alloys (SMAs) have the ability to recover large residual deformations by heating the alloy above its transformation temperature. The proposed connection consists of four large diameter NiTi SMA bars connecting the beam flange to the column flange and serve as the primary moment transfer mechanism. Two full-scale connections were tested using the SAC loading protocol. The connections exhibited a high level of energy dissipation, large ductility capacity, and no strength degradation after being subjected to cycles up to 4% drift. Following the initial testing series, the tendons were heated to recover the residual beam tip displacement. After initiating the shape memory effect within the tendons, the connections were retested, displaying repeatable and stable hysteretic behavior. An additional test was performed under dynamic loading to examine the strain r...

The results of tests on 48 T-stub specimens are presented and discussed. These tests were carried out as part of a SAC Phase II project in order to provide insight into the behavior, failure modes, and ductility of these components. The main variables tested include the size of the T-stub, the gauges of the bolts, and the type and diameter of the bolts. The primary intent was to develop a large database to calibrate simplified models suitable for design. In order to develop these models, a comprehensive instrumentation system that could identify the different components of deformation in the individual T-stubs was utilized. Most of the T-stubs failed by net section fracture through the stem and by tension fracture of the bolts, but generally the failure was after significant plastification had occurred. Other failure modes observed included bolt shear and block shear. The results indicate that current design equations provide conservative estimates of the ultimate strength of the T-stubs but that they are...

The results of 48 T-stub component tests and eight full-scale T-stub connection tests are used to develop and calibrate a comprehensive T-stub connection stiffness model. The model, based on basic component spring theory, incorporates deformations from tension bolt elongation, bending of the T-stub flange, elongation of the T-stem, slip of the T-stub relative to the beam flange, bearing deformation of the T-stem, and bearing deformation of the beam flange. Nonlinear material models, for both the tension bolts and the T-stub base material, are included. A sophisticated slip/bearing routine accounts for the misalignment of shear bolt holes and variable slip conditions. Comparisons of multilinear model predictions with experimental data show that the model accurately predicts connection stiffness response and deformation at failure.