An experimental study of nominally symmetric and asymmetric structures isolated with the FPS

This experimental investigation deals with the earthquake behaviour of a nominally symmetric and a mass-asymmetric three-storey structural model isolated with the frictional pendulum system (FPS). Both accidental and natural torsion are evaluated in the structure by using recorded accelerations in all building floors and measured deformations at the isolation level. A 3D-shaking table was used to subject the model to five different ground motions, including impulsive as well as far-field subduction-zone type earthquakes. Results show that the analytical predictions of the earthquake behaviour of the isolated structure, as obtained from a physical model of the FPS, are in close agreement with the true complex inelastic measured behaviour of the FPS. Besides, experimental results also validate previous observations about the importance of accounting for the variability of the normal loads in modelling the earthquake behaviour of FPS isolators. Measured torsional deformation amplifications at the base of the building vary, in the mean, from 2.5% to 6% for the symmetric and asymmetric structural configurations, respectively. In relation to the fixed base structure, the reduction factors for the base shear of the isolated structure are, in the mean, about 3.9 for both configurations. Finally, it is concluded that the FPS is capable of controlling the lateral-torsional motions of mass-asymmetric structures quite effectively by aligning the centre of mass of the superstructure with the centre of pendular and frictional resistance of the isolation system.