Improving seismic site response analysis for non-standard sites

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2019
Thesis identifier
  • T15439
Person Identifier (Local)
  • 201679615
Qualification Level
Qualification Name
Department, School or Faculty
  • Earthquake engineering aims to design and analyse structures to avoid damage in minor earthquakes and to prevent collapse in major earthquakes. An important aspect of assessing the ground motions that could occur at the basis of a structure is how the seismic waves will be affected by travelling through soil deposits near the surface. In fact, these near-surface layers act as a filter modifying (amplify/de-amplify) the input motion and leading to modified ground motions at the ground surface. Hence, the study of the effect of these near-surface layers on seismic waves is important for earthquake engineers designing and analysing structures in earthquake-prone regions.The main issues connected with assessing earthquake ground motions are the uncertainties connected with many aspects of this phenomenon: from the knowledge of the fault rupture generating the seismic waves, through the overall travel path of these waves, to the modification of the waves near the surface. In this work, I concentrate my attention on the uncertainties connected with the near-surface soil deposits. In particular, I focus on aspects of site response analysis, i.e. the modelling of the modulation of seismic waves by soil deposits. The main goal of the thesis is to improve site response analysis for "non-standard" sites, which are those for which the general assumptions of shallow (<50m deep) perfectly horizontal layers breakdown.;Firstly, I focus on the importance of the stratigraphy; in particular, I undertake an analysis of the possible two-dimensional effects of quasi-horizontal layers. In general, there are two macro-worlds of interests: a) sites with marked one-dimensional characteristics (flat and parallel layers) and b) sites with pronounced two-dimensional characteristics (valleys and canyons). In this thesis, I conduct parametric site response analyses for sites at the boundaries between these two worlds after carefully checking the validity of my calculations by comparing the results of various software packages. Based on these analyses I propose guidelines on the best way to analyse such sites.Secondly, I carry out a detailed study on the uncertainties related to poor knowledge of the geotechnical characteristics of the soil, in particular its stiffness (shear-wave velocity). The main question I try to answer is when the lack of information on these characteristics has a significant impact on the computed site amplification. In particular, I focus on uncertainties connected with the deep portion (>50m) of soil profiles. Very often, it is difficult to have sufficient and/or reliable in-situ measurements of the characteristics of this part of soil deposits. Therefore, it is common to take advantage of the quarter-wavelength method to assess site amplifications for these profiles because it does not need a detailed shear-wave velocity profile. In this thesis, I develop a procedure to generate randomized profiles for use within the standard site response analysis technique used for shallow deposits. This procedure uses basic information obtained from a large database of real shear-wave velocity profiles, such as length of the deep portion of the profile, the thickness of the layers and the presence of softer layers below the seismic bedrock. Finally, I compare the site amplifications using this procedure and those computed using the quarter-wavelength method and draw conclusions about when our lack of knowledge is important.
Advisor / supervisor
  • Douglas, John
  • Tarantino, Alessandro
Resource Type
Date Created
  • 2019
Former identifier
  • 9912780291202996