Analysis and optimization of a hybrid testing methodology using a small-scale set-up

Abstract

Hybrid testing provides an efficient and less costly way to explore the response of structural systems to realistic dynamic or seismic loading. However, the required equipment to execute hybrid tests are high-cost tools. To get insight in the hybrid testing methodology, a small-scale setup has been developed in this project. An Arduino UNO controls the system that imposes the displacement to a linear actuator. Connecting the small-scale setup , i.e., the Arduino UNO, to MATLAB allows imposing a time history to the physical substructure. A load cell measures the restoring force which will be communicated to MATLAB by the Arduino UNO. Numerical integration based on the Gravouil-Combescure scheme with Classic Lagrange Multipliers (CLM) determines the displacement for the next time step. This paper describes hot spots of the methodology and the results of a demonstrative experimental test. The experiment consists of a 4 degree of freedom (DOF) numerical model combined with a 1 DOF physical specimen. The installed linear actuator only has one gearing option, which leads to a possible overshooting loop. Interesting conclusions can be drawn from the analysis of the small-scale setup in view of its future upscaling and implementation of the hybrid test method at laboratory scale. Firstly, the linear actuator requires a non-negligible amount of time to reach the imposed displacement which imposes boundary conditions in the MATLAB directives. Secondly, a velocity-controlled actuator is essential in the exploitation of hybrid testing. Thirdly, the displacement tolerance influences the stability of the system. If one increases the displacement tolerance, the risk of an overshooting loop decreases. However, the accuracy might be influenced. Good balance must therefore be found between stability and accuracy.