Optimization and performance of metafoundations for seismic isolation of small modular reactors

• Published in: Computer-aided civil and infrastructure engineering Vol. 38; no. 12; pp. 1558 - 1582
• Main Authors: Guner, Tugberk, Bursi, Oreste S., Erlicher, Silvano
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc 01.08.2023
• Subjects: Earthquake dampers; Earthquakes; Multiple objective analysis; Nuclear reactors; Optimization; Parameter sensitivity; Reinforced concrete; Search algorithms; Seismic isolation; Seismic response; Springs (elastic); Stiffness; Structural models; Vertical loads; Viscous damping
• ISSN: 1093-9687; 1467-8667
• DOI: 10.1111/mice.12902

This paper aims to study the seismic mitigation of a typical nuclear small modular reactor (SMR) where extreme loading conditions are considered by the safe shutdown earthquake. For this purpose, to reproduce the main dynamic properties of the reactor's reinforced concrete system, a detailed structural model was synthetized, also taking into account the presence of the reactor pools. Thus, to protect the reactor from strong earthquakes, finite locally resonant multiple degrees of freedom metafoundations were developed; and resonator parameters were optimized by means of an improved frequency domain multivariate and multiobjective optimization procedure. Also, the stochastic nature of the seismic input was taken into account. It is proposed: (i) a linear metafoundation endowed with multiple cells, linear springs, and linear viscous dampers; and (ii) a foundation equipped with additional nonlinear vertical quasi‐zero stiffness (QZS) cells. QZS cells were obtained by horizontally precompressed springs in an unstable state with vertical springs in parallel. With this arrangement, additional flexibility and dissipation against nonsymmetrical modes of the SMR and vertical seismic loadings are proposed. It was shown in both cases, how each metafoundation was successfully optimized via a sensitivity‐based parameter grouping strategy and a hybrid grid searching algorithm. Thus, the performance of the optimized metafoundations was assessed by means of frequency and time history analyses; and finally, results were compared with an SMR endowed with both rigid foundation and conventional base‐isolation solutions.