Experimental and numerical investigation of a vertical vibration isolator for seismic applications

In near-fault seismic zones, the vertical acceleration experienced during a strong event can be greater than horizontal acceleration. Methods to reduce horizontal acceleration are applied in various forms and are in common use. However, methods to reduce vertical acceleration, and practical protection systems for these applications, remain elusive. One strategy to protect structures, which has been demonstrated to be effective in situations where the excitation is horizontal, is to isolate the structure. For vertical excitations, this is difficult due to the need to maintain sufficient stiffness and strength in the direction of gravitational loads. The need to maintain high stiffness for gravity loading while allowing flexibility for isolation during earthquakes has led to research on the use of High-Static-Low-Dynamic Stiffness Systems (HSLDSS) and in particular Quasi-Zero Stiffness Systems (QZSS), which have zero equivalent stiffness in the equilibrium position. Although effective, the QZSS is sensitive to mistuning and prone to large deformations for relatively small increments in static load for building applications. This paper presents the results of an analytical and experimental study in which a HSLDSS isolation system carrying a payload is subject to vertical base excitation using sinusoidal as well as actual, scaled earthquake signals. Static loading tests are also presented. This isolation system consists of rigid rotating arms, horizontal and vertical springs and a vertical damper. By a suitable selection of parameters this could also serve as a QZSS. Results show that both the QZSS and HSLDSS can significantly reduce the magnification of the force as well as the transmission of the acceleration and that the HSLDSS retains stiffness at the equilibrium position. The numerical model includes friction and is solved using direct integration of the equation of motion. Experimental results from a scale model agree well theoretical predictions