A Distributed-Plasticity Analytical Framework for Cyclic Modeling of SMA Angle Connections
Please login to view abstract download link
Accurate prediction of the cyclic response of angle connections requires capturing the spatial spread of inelasticity and the evolving internal force redistribution along the angle legs, which are not represented adequately by conventional hinge-type idealizations. This work presents a distributed-plasticity modeling framework for shape memory ally (SMA) angle connections, formulated to describe inelastic deformation as a continuous field rather than being concentrated at a predefined rotational spring. Closed-form relationships are derived to link curvature, flexural rigidity, and the moment distribution along the connection, while enforcing deformation compatibility. The formulation is coupled with an SMA superelastic constitutive description that accounts for asymmetric tension-compression transformation behavior, enabling the model to track the evolution of phase transformation under cyclic loading and its impact on stiffness and hysteresis. The proposed approach yields direct predictions of hysteretic moment-rotation responses, including stiffness transitions and the associated recentering tendencies. Finite element simulations and cyclic experimental results are used to assess the framework, demonstrating that it reproduces the measured response trends with good agreement. Parametric analyses over diverse angle geometries further identify a geometry-governed rigid-body translation trend in the normalized equivalent-length-displacement relationship, providing practical guidance for geometric calibration and phase-dependent boundary conditions. Transferability is examined by benchmarking against experimental datasets of steel angle connections, showing that the same kinematic structure remains applicable when the SMA constitutive description is replaced by a conventional elastoplastic representation.
