Screening of seismic records to perform time-history dynamic analyses of tailings dams: A power-spectral based approach

Time-history deformation analyses of upstream-raised tailings dams use seismic records as input data. Such must be representative the in-situ seismicity in terms a wide range intensity measures (IMs) including peak ground acceleration (PGA), Arias (AI), cumulative absolute velocity (CAV), source-to-site distance and duration, among others. No single IM is sufficient descriptor given demand (e.g. crest settlement) because different records, all them compliant with any IM, can produce very results from insignificant damage to global failure. The brute force, where hundreds set IMs are employed, has proven reasonable workaround this limitation, at least it able probabilistic density function indicators. This procedure, however, requires large number runs, therefore expensive time-consuming. Analyses optimized if an priori simple tool used predict which would yield demand, thus obtaining estimations much fewer runs. In order perform more precise selection, semi-analytical screening procedure presented paper. makes spectral properties record, considering only frequency content not filtered by dam obtain estimate expressed case displacements. validated using analytical numerical models that prove insensitivity constitutive model analysis, applied subjected strong earthquakes. • A new selection based analysis proposed. Constitutive calibration consider cyclic liquefaction com-parisons experimental test presented. calibrate initial relative state pa-rameter suggested. Several dynamic showed analyzed. novel correlation tailing displacement prop-erties discussed.