Dynamic Experimentation And Theoretic Analysis Of Soil-Box Foundation-Frame Structure Interaction System

For current structure seismic design, the first-phase work is still elastic analysis. Calculation of structural natural-frequency is the important content of elastic analysis, and its value directly influences the seismic action in seismic design.The natural frequency of soil-box foundation-frame structure is obtained by puisateon test and traction excitation test. The results of two test methods are coincident with each other, which afford accurate test results for theoretical calculation next step. Velocity sensors fixed on soil in two directions, the effects of upper-structure on soil are studied through traction test. In soil-structure dynamic interaction system when ambient soil bonds to foundation concrete well, the velocity component of ground movements will give priority to the direction of upper-structure vibration. The frequency composition of velocity for longitudinal test points of structure is mainly low frequency and for transverse test points of structure is mainly high frequency.FEM programs Patran and Nastran are adopted to perform natural frequency analysis and traction excitation analysis for the large-scale frame model. In natural frequency analysis the effect of some factors are taken into consideration, such as boundary conditions, element type, beam-plate offset, ratio of reinforcement, stiffness of upper-lower structure, meshing density and distribution of soil stiffness. Theoretical results shows that without consideration of beam-plate offset and structural reinforcement, the natural frequency calculated by traditional design method for beam-plate model based on rigid-soil assumption is always lower than test results. In case of small seisms, the seismic design carried out according to current code is not safe. Under this condition the interaction of upper-lower structure can not be considered in the calculation of natural frequency. The adjusting factors of the first-order frequency for different kinds of grounds are given. The conic expression between reinforcement ratio of beam and column and the first-order frequency of structure is also advanced. It is pointed out that when soil-structure interaction is considered, the range of soil size in FEM modeling is not specified but related to the distribution of soil stiffness. When the stiffness of soil increases, the range selected can be decreased. The stiffness of soil around foundation affects low-order frequency of structure more and the stiffness of deep-layer soil affects high-order frequency of structure more.FEM computational results of traction excitation show that when the vibration amplitude of upper structure is not large and soil bonds to foundation concrete well, theupper-lower structure can transfer vibration to soil smoothly. The acceleration amplitude of model without foundation is larger than model with foundation. The acceleration amplitude of upper structure is larger than lower. The acceleration amplitude of points on soil near foundation is close to that of points on foundation. The amplitude of acceleration parallel to tractional direction is about ten times as that of acceleration in vertical direction for test points on soil.