The problem under consideration is the earthquake impact on structures. The subject of the performed research is the efficiency of seismic base isolation using layers of predominantly natural materials below the foundation, as well as the development of a numerical model for seismic analysis of structures with such isolation. The aseismic layers below foundation are made of limestone sand - ASL-1, stone pebbles - ASL-2, and stone pebbles combined with layers of geogrid and geomembrane - ASL-3. The experimental research methodology is based on the use of shake-table and other modern equipment for dynamic and static testing of structures. Experiments were conducted on the basis of detailed research plan and program. Efficiency of the limestone sand layer - ASL-1 was tested on cantilever concrete columns, under seismic excitations up to failure, varying the sand thickness and intensity of seismic excitation. Influence of several layer parameters on the efficiency of stone pebble layer - ASL-2 was investigated. For each considered layer parameter, a rigid model M0 was exposed to four different accelerograms, with three levels of peak ground acceleration (0.2 g, 0.4 g and 0.6 g), while all other layer parameters were kept constant. On the basis of test results, the optimal pebble layer was adopted. Afterwards, the optimal ASL-2 efficiency was tested on various model parameters: stiffness (deformable models M1-M4), foundation size (small and large), excitation type (four earthquake accelerograms), and stress level in the model (elastic and up to failure). In the ASL-3 composite aseismic layer, the optimal ASL-2 is combined with a thin additional layer of sliding material (geogrid, geomembrane above limestone sand layer), in order to achieve greater efficiency of this layer than that of the ASL-2. A total of eleven different aseismic layers were considered. To determine the optimal ASL-3, the M0 model was used, like for the ASL-2. On the basis of test results, the optimal ASL-3 layer was adopted (one higher strength geogrid at the pebble layer top). The optimal ASL-3 is tested on various model parameters, analogous to the optimal ASL-2. A numerical model for reliable seismic analysis of concrete, steel, and masonry structures with seismic base isolation using ASL-2 was developed, with innovative constitutive model for seismic isolation. The model can simulate the main nonlinear effects of mentioned materials, and was verified on performed experimental tests. In relation to the rigid base - RB without seismic isolation, model based on the ASL-1 had an average reduction in seismic force and strain/stress by approximately 10% at lower PGA levels and approximately 14% at model failure. Due to the effect of sand calcification over time, the long-term seismic efficiency of such a layer is questionable. It was concluded that the aseismic layers ASL-2 and ASL-3 are not suitable for models of medium-stiff structure M3 and soft structure M4. In relation to the RB without seismic isolation, the M1 (very stiff structure) and M2 (stiff structure) based on the ASL-2 had an average reduction in seismic force and strain/stress by approximately 13% at lower PGA levels and approximately 25% at model failure. In relation to the RB without seismic isolation, the M1 and M2 based on the ASL-3 had an average reduction in seismic force and strain/stress by approximately 25% at lower PGA levels and approximately 34% at model failure. In relation to the RB without seismic isolation, the ASL-2 and ASL-3 did not result in major M1 and M2 model displacements, which was also favourable. It is concluded that the ASL-2 and especially ASL-3 have great potential for seismic base isolation of very stiff and stiff structures, as well as small bridges based on solid ground, but further research is needed. In addition, it was concluded that the developed numerical model has great potential for practical application. Finally, further verification of the created numerical model on the results of other experimental tests is needed, but also improvement of the developed constitutive models.
Identification of Track Stability Model Parameters Based on Numerical Experiments
