Derivation of floor acceleration spectra for an industrial liquid tank supporting structure with braced frame systems

Past seismic events have shown that industrial facilities may suffer greatly from earthquake-induced actions, which can cause simultaneous damages to different key apparatus, initiating major/multiple accidental chains. Therefore, the evaluation of seismic demand acting upon structural systems and equipment is of utmost importance when assessing or designing a complex industrial plant, which will be inevitably exposed to seismic hazards during its lifetime. Furthermore, it is noteworthy that, if modern seismic design procedures are able to successfully limit damage to the main structural elements, the evaluation of secondary structural and non-structural components may imply less trivial considerations because of the important role played by them (e.g. storage of chemical/hazardous substances) and the lack of accurate code-compliant approaches applicable to each specific case. As such, this paper is chiefly concerned with the derivation of floor acceleration spectra for a special concentrically braced frame supporting a cylindrical storage tank, the latter being a basic strategic component widely used in several industrial applications and the former one being one of the most common forms of lateral-force resisting systems for plant structures. Firstly, a fibre-based finite element model of the supporting frame itself was developed within an open source platform and then nonlinear time history analyses were performed assuming a set of 47 natural ground motions scaled to 8 seismic intensities. These results, used to assess the tank as an uncoupled system, were additionally compared with those obtained by a second analysis run, in which the interaction between the supporting structure and the tank was explicitly accounted for by modelling them together. A well-known analytical model, consisting of two uncoupled single degree of freedom systems for the impulsive and convective components of motion, was considered in this case to reproduce the response of the tank. The floor spectra resulting from these two approaches were compared together so as to quantify trends and differences in the observed estimates. Sensitivity to the viscous damping was examined as well. A comparison was finally derived between the nonlinear dynamic analyses of both coupled and uncoupled systems and the analytical methods of current Codes (ASCE 7-10, EC8) and recent research proposals.