Development, modeling, identification and simulation of a small shaking table system

This comprehensive study presents the results of a comparison between the analytically predicted and the experimentally identified dynamics of a small shaking table system used in the undergraduate structural dynamics laboratory at UCLA. An uni-axial, servo-hydraulic, displacement-controlled shaking table system was completely developed, including the design and implementation of a digital controller using the reviewed control theory. The system consists of a 2 x 4 ft slip table, controlled by a small servo-hydraulic actuator of 3 in. stroke and 1.5 in. bore diameter. The dynamic performance of the system was studied and a series of linearized analytical dynamic models (in the form of the total shaking table transfer function) was developed from basic principles. The analytical models incorporate the inherent dynamic characteristics of the various components of the shaking table system and their dynamic interaction.; By adding one component at a time, a series of analytical models of escalating sophistication resulted, fully explicating the effect of each added component. The most important modification was the addition of the reaction frame flexibility. The reaction frame has the lowest natural frequency of all of the system components which affects the dynamic properties of the shaking table system. A major feature controlling the dynamics of medium to large size shaking tables is the oil column frequency. This research found that for this small size actuator, its high oil column frequency does not influence the dynamic behavior of the system in its operating frequency range, and that the slip table mass does not determine the oil column frequency, as it usually does, due to the inherent compliance of the linkage between the slip table and the actuator arm.; Adding the flexibility of the reaction frame to the analytical model of the shaking table system yielded more realistic values of the system parameters obtained through least squares fitting of the experimentally identified shaking table transfer function with the analytical model. The numerical simulations performed with the analytical models showed the effectiveness of simple models in capturing the total shaking table transfer function and how more complex models resulted in only minor improvements.