PSI - Issue 44

3

Angela Chiecchio et al. / Procedia Structural Integrity 44 (2023) 11–18 Angela Chiecchio/ Structural Integrity Procedia 00 (2022) 000–000

13

More specifically, the BMSE is obtained by the following steps. (1) Within the adopted seismogenic model, a selection of the sources lying within a representative large distance (say 100km) from the site is performed. (2) For each of these sources, a proper magnitude, distance and ε range are defined. The magnitude range is set such that the lower ( M min ) and the upper limits ( M max ) are the same as adopted in the seismogenic source model. The distance range is fixed between a minimum distance R min, and a maximum distance R min +100 km, where R min is the minimum distance from the site to the selected source. The epsilon range is defined within a standard range between -2 and 2. (3) For each source, and for each ( M w - R - ε ) triplet, the squared normalized difference between the GMM prediction (Sa GMM ) and the UHS target (Sa Target ) is computed for each structural period (Eq. 1). An error index for each ( M w - R - ε ) triplet is therefore calculated as in Eq. (2), by minimizing the error indexes over all periods. (4) Based on the previous steps, a BMSE is assigned to each seismogenic source contributing to the seismic hazard at the site, provided that the error (2) is lower than a value of 0.3, that was preliminary identified as a representative threshold, based on the sensitivity analyses performed. At this preliminary stage or research, only median values were considered, i.e., the search was limited to  =0, to support the immediate use of BMSE for practical applications.    , R, ε,    =      ,, ,               (1)  , R, ε = min ,, { mean  diff , R, ε,    } (2)