Interface friction effects on scaling a vertical spring-viscous damper isolation system in a shaking table test

A vertical spring had a changeable horizontal stiffness to avoid resonance under a seismic displacement, and a viscous damper had a perfect energy dissipation function to reduce the seismic displacement. They were joined together to achieve a seismic isolation system considerably reducing the seismic acceleration response, however, their displacement should be paid more attentions under an earthquake or in a shaking table test. Because the interface friction in the system affected the scaling process of the system in the shaking table test and the corresponding displacement responses, the system was numerically scaled before a real shaking table test. The comparison between the displacement responses of scaled models and prototype models shows that the spatial friction variation and the gravity distortion are the error sources of the relative and residual displacements of scaled models. The error of scaled relative displacement is reduced by the means of increasing PGA, damping constant, spring constant or reducing vertical spring length with zero stress, however, the error of scaled residual displacement is still chaotic under these means. Only if the error sources above are eliminated fundamentally will the error of scaled models be avoided. The conclusions are validated correct by a real shaking table test.