6. Conclusions
This work has studied the effectiveness of base-isolation with elastomeric (i.e. HDLRBs) and sliding (i.e. CSSBs) bearings for the seismic retrofitting of r.c. irregular framed buildings, with regard to pulse-type and non-pulse-type ground motions and different site conditions in the near-fault area. To this end, a simulation is conducted in which a fivestorey reinforced concrete (r.c.) benchmark structure of the Re.L.U.I.S. project, characterized by an asymmetric-plan and irregularity due to bays of different length, is retrofitted with regard to four structural solutions for each base-isolation system, diversified assuming two inplan distributions of HDLRBs and CSSBs and moderately-soft and soft subsoil conditions. The records of near-fault earthquakes in recent central Italy and worldwide have been selected and continuous wavelet transforms of the two horizontal components used to identify all the orientations most likely to contain a pulse, in terms of a predictor of the likelihood that a given record is pulse-type.
With the wavelet analyses, false-negative classifications occur when only fault-normal pulse direction is adopted. Specifically, polar plots of a pulse indicator (PI) confirm pulses around this orientation for the L′Aquila and Accumoli EQs, while this orientation does not lie within the range in which pulses are observed for the Ussita EQ. Similarly, in the Kobe EQ pulses are predominant perpendicular to the fault, unlike the Taiwan and Northridge EQs where pulses deviate from the faultnormal orientation. On the other hand, false-positive classifications of pulses are removed due to a change in the PGV threshold. This happened for the L′Aquila (AQV and AQG stations), Accumoli (AMT, NRC and FEMA stations) and Ussita (NOR station) near-fault ground motions, while worldwide the potential pulses of the near-fault EQs exhibit markedly higher values of the pulse indicator than those obtained for the EQs in central Italy.
