Full scale dynamic tests on dissipating antiseismic devices

Experimental results of full scale tests on an elasto-plastic dissipating device are reported. Tests have been carried out on the SRMD test facility of the Powell Structural Research Laboratories, located at the University of California San Diego. This test facility allows real time 6 degree of freedom dynamic characterization of antiseismic devices to be installed on long span bridges. Characterization of these devices has already been carried out through quasi static tests; the SRMD test facility allowed to verify the dynamic properties of the bearings under real time dynamic conditions. The device being tested is composed of two main components: a central shock transmission unit and 4 layers of C shaped elasto-plastic dissipating components. During the seismic event the shock transmission unit engages the C-shaped elements which undergo plastic deformations. Displacement controlled tests have been carried out by imposing a sequence of sinusoidal excitations with increasing level of the maximum displacement. In order to verify the influence of velocity and frequency of the excitation on the device, a first group of tests has been performed for different values of these parameters, keeping the maximum displacement below the yielding value. The behavior of the bearing at the increase of the maximum displacement has been verified through the following sequence of tests carried out with cycles of loading at 25%, 50%, 100% and 125% of the design displacement. A seismic test has also been conducted imposing a displacement time history characterized by a peak displacement of the order of the design value. The residual capacity of the device after a seismic event has been verified performing the last test with a maximum displacement equal to 140% of the design value. The behavior of the device during the entire sequence of tests has been analyzed through the evolution of its characteristics parameters as defined by AASHTO: effective stiffness, equivalent viscous damping and energy dissipated per cycle. This evolution has shown that frequency and velocity of the excitation have slight influence on the system characteristics, hence in service conditions the stiffness and damping of the bearing can be assumed constant. At the increase of displacement the effective stiffness decreases while the equivalent damping ratio increases. During the tests carried out with maximum displacement equal respectively to the design value and to 125% of the design value two C-shaped elements broke. The effect of this failure on the behavior of the device has been analyzed as well. The force-displacement relation exhibited by the device during the sequence of test have been used to define a very simple elasto-plastic model of the device. This kind of model may be implemented in finite element analysis of structures isolated with devices of this nature. The response of the model to the displacements time history used for the experimental tests has been compared with the response of the real device. The comparison indicates that this simplified model is able to reproduce very well the evolution of the effective stiffness but strongly overestimates the dissipative capacity of the real device. A more refined model is hence needed in order to take into account this aspect of the device response.