Friction Characterization and Mitigation in Large‐Scale 6DOF Multi‐Axial Hybrid Simulation

ABSTRACT In the last few decades, hybrid simulation has become widely used for understanding the response of structural components and systems under extreme loading conditions. Large‐scale, three‐dimensional (3D), multi‐axial testing facilities with six degrees of freedom (6DOF) are versatile testing systems that can be used for testing a variety of structural systems and components. The University of Minnesota Multi‐Axial Subassemblage Testing (MAST) facility, which was originally developed in the 1990s, has had recent upgrades to its 6DOF multi‐axial pseudo‐dynamic hybrid simulation capabilities. One of the well‐recognized challenges in using such 6DOF multi‐axial setups of this size is the friction within the system, especially in the swivels of actuators, which can lead to numerical instabilities in hybrid simulation if not mitigated or compensated for properly. As such, when using such setups, friction effects on the stability of hybrid simulation must be understood. Characterizing the friction within the setup is not only crucial for understanding whether friction effects must be compensated for during the test, but it is also important for adopting the appropriate friction compensation scheme. Given the inherent complexities of such over‐constrained large‐scale multi‐axial setups, due to their size, capacity, and the intricate interaction within the actuators, characterizing the friction within the system is not trivial. This paper provides an overview of the MAST system and its features, a brief review of selected past projects, and the architecture of the newly upgraded hybrid simulation capabilities. The effects of friction on the stability of hybrid simulation are discussed, and commonly used methods for managing friction or compensating for it are presented. The experiments used to characterize the performance of the MAST testing facility, including the internal friction within the system, are presented. These experiments can serve as a framework for internal friction characterization in similar test setups and can be used by laboratories that are new to using 3D, 6DOF multi‐axial test setups. In the end, a suite of validation multi‐axial pseudo‐dynamic hybrid simulation tests was performed, where all 6DOFs of the MAST system were used in the hybrid simulation control loop. The validation hybrid simulations were performed on a three‐story moment resisting frame structure, under the 1994 Northridge earthquake. One of the columns within the building was physically tested at MAST, while the rest of the structure was modeled numerically first in OpenSees, and afterward in Ansys in a repeat test. Results from the pseudo‐dynamic hybrid simulation are presented and compared with purely numerical predictions, which validated the system performance. The hydrostatic friction bearings incorporated within the MAST system ensured negligible friction compared to the capacity of the system, with the friction not exceeding 0.23% for the worst‐case scenario.