Towards realistic minimum-cost optimization of viscous fluid dampers for seismic retrofitting

• Published in: Bulletin of earthquake engineering Vol. 14; no. 3; pp. 971 - 998
• Main Authors: Pollini, Nicolò, Lavan, Oren, Amir, Oded
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.03.2016; Springer Nature B.V
• Subjects: Civil Engineering; Cost function; Design engineering; Earth and Environmental Science; Earth Sciences; Earthquake dampers; Energy dissipation; Environmental Engineering/Biotechnology; Geophysics/Geodesy; Geotechnical Engineering & Applied Earth Sciences; Hydrogeology; Interpolation; Mathematical models; Nonlinear programming; Optimization; Original Research Paper; Retrofitting; Seismic engineering; Seismology; Structural Geology; Topology; Viscous fluids; Topology optimization; Energy dissipation devices; Viscous dampers; Material interpolation functions; Irregular structures; Seismic retrofitting
• ISSN: 1570-761X; 1573-1456
• DOI: 10.1007/s10518-015-9844-9

This paper presents an effective approach for achieving minimum cost designs for seismic retrofitting using viscous fluid dampers. A new and realistic retrofitting cost function is formulated and minimized subject to constraints on inter-story drifts at the peripheries of frame structures. The components of the new cost function are related to both the topology and to the sizes of the dampers. This constitutes an important step forward towards a realistic definition of the optimal retrofitting problem. The optimization problem is first posed and solved as a mixed-integer problem. To improve the efficiency of the solution scheme, the problem is then re-formulated and solved by nonlinear programming using only continuous variables. Material interpolation techniques, that have been successfully applied in topology optimization and in multi-material optimization, play a key role in achieving practical final design solutions with a reasonable computational effort. Promising results attained for 3-D irregular frames are presented and compared with those achieved using genetic algorithms.