Calibration of a large shaking table equipment for testing geotechnical physical models under plane strain conditions

Experimental tests carried out on physical models under laboratory-controlled conditions can provide a valuable insight into the seismic performance of geotechnical systems and are extremely valuable to check the predictions of numerical analyses. In this vein a new shaking table apparatus with a large flexible soil container has been set up in the laboratory of Geotechnical Engineering of the University of Messina. The apparatus consists of a shaking table connected to a servo-hydraulic actuator, a large shear stack container for the soil and an automated system for soil deposition. The actuator can excite the low friction shaking table by applying horizontal displacements reproducing real and artificial seismic motions characterized by amplitude, frequency and energy contents varying in wide intervals. The large soil container consists of a laminar shear box which has been assembled to accurately model a soil layer of infinite lateral extent; under horizontal shaking the container deforms in plane strain conditions according to a shear beam mode, thus reproducing a free-field condition under vertically propagating shear waves. The soil deposition system consists of a hopper that can be moved back and forth above the soil container allowing sand pluvial deposition; the velocity of the hopper, the width of its lower opening and the soil falling height can be set to attain a desired relative density in the soil model. The paper describes this new experimental facility pointing out its peculiarities and provides the results of some static and dynamic calibration tests demonstrating that: (i) the plane strain conditions can be satisfactorily assumed for the soil specimen in the laminar box; (ii) the servo-hydraulic control system is capable to accurately reproduce a prescribed acceleration time-history at the shaking table platform, regardless the amplitude, frequency and energy content of the desired motion; (iii) no significant boundary effects arise in the portions of the soil specimen placed near the end-walls of the laminar box; (iv) the shaking table tests carried out with the new apparatus are repeatable.