Experimental and numerical investigation on the seismic effective mass of an industrial steel silo tested on shaking table

Silos are classified as non-building structures whose overall structural response has been extensively studied in the last years. Nevertheless, their dynamic and seismic response still presents open issues, mainly regarding the seismic effective mass of the stored product, the estimation of the fundamental frequency of the filled silo system, and the dependency of the damping ratio on the input nature and magnitude. The main objective of this work is to provide an estimation of the effective mass, taking advantage of a wide experimental shaking table campaign carried out on a full-scale flat-bottom corrugated-wall steel silo filled with soft wheat. The paper initially presents the results of Experimental Modal Analysis for the identification of the vibration frequencies and the calibration of a numerical model considering a no-tension material for the ensiled content. The calibrated model is able to well reproduce the actual response of the filled silo measured during the shaking table tests for both sinusoidal and earthquake inputs and it is thus used to retrieve information about the seismic effective mass beyond the limits of the experimental work. The reconstruction of the global overturning moment at the silo base from the numerical model indicates an average mass participation of 91 % for an industrial silo with a common filling aspect ratio (H/D = 0.9). This highlights that the 80 % value of the effective mass prescribed by Eurocode 8 may reveal to be non-conservative for silos characterised by corrugated wall sections and aspect ratios close to or greater than 1.