MTS and the University of Minnesota collaborate to prove the efficacy of Cascade Control, an innovative closed-loop control method designed for conducting high-fidelity hybrid simulation experiments using very stiff substructure test specimens.
Cascade Control combines two layers of closed-loop control to overcome the difficulties encountered when imposing displacement commands from analysis models on highly stiff physical test specimens during hybrid simulation experiments. To evaluate the effectiveness of the new method, MTS engineers (Shawn You, Shawn Gao, and Brad Thoen) collaborated with researchers at the University of Minnesota MAST Lab (Catherine French, Evan Cosgriff, and Paul Bergson) on a series of experiments that employed a range of test specimens, test systems, experimental element types, and test speeds.
Cascade Control features an outer loop that accepts displacement commands from the FEA model, and an inner loop that provides the force control needed to accurately excite test specimens that exhibit little or no displacement. Three test cases, each comparing an FEA simulation with hybrid simulations using cascade and displacement control, were conducted on MTS Landmark load frames and the University of Minnesota's Multi-Axial Subassemblage Test (MAST) system. In all cases, cascade control proved effective for hybrid simulation with stiff specimens.
Cascade Control features an outer loop that accepts displacement commands from the FEA model, and an inner loop that provides the force control needed to accurately excite test specimens that exhibit little or no displacement. Three test cases, each comparing an FEA simulation with hybrid simulations using cascade and displacement control, were conducted on MTS Landmark load frames and the University of Minnesota's Multi-Axial Subassemblage Test (MAST) system. In all cases, cascade control proved effective for hybrid simulation with stiff specimens.
Test Case 1: 1940 El Centro earthquake inputs were applied to an FEA model of a 3D one-story, one-bay frame structure and MTS Landmark load frame testing an aluminum column with axial stiffness of 110 kN/mm. Good correlation was achieved between the FEA and hybrid simulations, however, the vertical force measured using cascade control was smoother than that obtained with displacement control.
Test Case 2: 1940 El Centro earthquake inputs were applied to an FEA model of a 2D one-story, one-bay frame structure and MTS Landmark load frame testing a steel truss with axial stiffness of 120 kN/mm. Good correlation was achieved between the FEA and hybrid simulations, however, at low excitation the axial force measured with displacement control showed higher levels of wave distortion.
Test Case 3: 1994 Northridge earthquake inputs were applied to an FEA model of a 3D three-story, multi-bay frame structure and a 6DOF MAST system testing a steel column with vertical stiffness of 418 kN/mm. Good correlation was achieved between the FEA and hybrid simulations, although it is suspected that cascade control would provide more accurate results on test systems with higher noise than the University of Minnesota's extremely low-friction MAST.
