I motivi per cui è difficile fornire spettri verticali semplificati/standardizzati (legati ovviamente alla variabilità del rapporto H/V che ben conosciamo).

Journal of Earthquake EngineeringVol. 08, No. 02, pp. 175-207 (2004) No Access
THE VERTICAL-TO-HORIZONTAL RESPONSE SPECTRAL RATIO AND TENTATIVE PROCEDURES FOR DEVELOPING SIMPLIFIED V/H AND VERTICAL DESIGN SPECTRA
YOUSEF BOZORGNIA and KENNETH W. CAMPBELL
Cited by:1
Next
View Article
Tools
Share
Abstract
This paper presents the results of a study on the engineering characteristics of vertical ground motion. A ground-motion model for the vertical-to-horizontal ratio (V/H) of peak ground acceleration and 5%-damped acceleration response spectra was derived by using a database of 443 near-source accelerograms recorded from 36 worldwide earthquakes of MW 4.7-7.7. The recordings were all located within 60 km of the seismogenic rupture zone of worldwide shallow crustal earthquakes in active tectonic regions. We found no bias in the V/H estimates from independent analyses of vertical and horizontal response spectra. We also found V/H to be a strong function of natural period, local site conditions, and source-to-site distance; and a relatively weaker function of magnitude, faulting mechanism, and sediment depth. V/H exhibits its greatest differences at long periods on firm rock (NEHRP site category BC), where it has relatively low amplitudes, and at short periods on firm soil (NEHRP site category D), where it has amplitudes that approach 1.8 at large magnitudes and short distances. We propose a tentative simplified model for estimating V/H spectra and two relatively simple procedures for developing a preliminary vertical design spectrum that could be used for practical engineering applications where site-specific spectra (e.g. from a seismic hazard analysis) are not available. Extensive vetting from the seismological and engineering communities will be needed before such simplified spectra are officially accepted and adopted in codified procedures in the United States.

"Data speak for themselves" -Reverend Thomas Bayes 1702-1761
P(Ai|E)=(P(E|Ai)P(Ai))/P(E)