Viscous damping effects on the seismic elastic response of tunnels in three sites

Abstract

Time-domain commercial codes are widely used to evaluate the seismic behavior of tunnels. Those tools offer a good insight into the performance and the failure mechanism of tunnels under earthquake loading. Viscous damping is generally employed in the dynamic analysis to consider damping at very small strains in some cases, and the Rayleigh damping is commonly used one. Many procedures to obtain the damping parameters have been proposed but they are seldom discussed. This paper illustrates the influence of the Rayleigh damping formulation on the tunnel visco-elastic behavior under earthquake. Four Rayleigh damping determination procedures and three soil shear velocity profiles are accounted for. The results show significant differences in the free-field and in the tunnel response caused by different procedures. The difference is somewhat decreased when the soil site fundamental frequency is increased. The conventional method which consists of using solely the first soil natural mode to determine the viscous damping parameters may lead to an unsafe seismic design of the tunnel. In general, using five times site fundamental frequency to obtain the damping formulation can provide relatively conservative results.