Scaled Earthquake Resistant Structures

  • Muhammed Atar

Student thesis: Phd

Abstract

Because testing facilities are limited, and small-scale models are more economically viable, small-scale models are commonly employed to evaluate the seismic performance of buildings. Similarity laws are provided that are assumed to characterise the system of interest, and a smaller or bigger model is created from the related circumstances that will behave in a predictable manner if the scaling laws are accurate. Scaled experiments in earthquake-resistant structural testing have a fundamental problem. Unfortunately, dimensional analysis kind of similitude seldom applies to complicated systems, which is especially troublesome when scaling ratios are significant. The present study provides a novel method to reconstruct full-scale behaviour without the use of any additional techniques such as additional mass, makeshift scaling rules, and artificially high accelerations on experimental models, which are common in traditional scaling approaches. This study aims to present two different types of similitude as part of a new scaling theory called finite similitude for earthquake-resistant structures and structural dynamics. This method is not based on dimensionless forms but rather on the assumption that scaling may be seen as an imaginary process in which space is constricted or extended. The projection of the governing physics, defined on a scaled space, onto the original full-scale space lies at the heart of the new method. It is shown here how the notion may be utilised to build experiments, with numerical and analytical studies used to validate the single and two scaled experiments. This research shows how one and two scaled experimentations can be applied to classical linear and non-linear continuous and discrete structural systems, and practical structural dynamics case studies such as high-rise steel buildings equipped with nonlinear-fluid viscous dampers under earthquake loadings. Furthermore, it has been also shown that the presented scaling technique can more easily deal with the scaled experimentations when the geometrical similarity is broken.
Date of Award1 Aug 2022
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorKeith Davey (Supervisor) & Roohoolamin Darvizeh (Supervisor)

Cite this

'