Verification of Nonlinear Ground Response Analysis by Vertical Arrays
Date Issued
2012
Date
2012
Author(s)
Chen, Yu-Siang
Abstract
Taiwan is a seismically active region. It is vital for engineers to design for earthquake-resistant buildings and evaluate the performance of existing natural and man-made structures under variety of earthquake scenarios. To perform the aforementioned design or evaluation, engineers must be able to predict the level of ground motion at the site of interest. In the field of geotechnical earthquake engineering, there are two ways to predict ground motions: (1) use of Ground Motion Prediction Equations (GMPEs); (2) theoretical ground response analyses. GMPEs are derived empirically through regression on a database of strong motion. GMPEs can be used to estimate the probabilistic distribution (in terms of median and standard deviation) of an intensity measure, such as peak ground acceleration, based on magnitude, site-source distance, site condition etc. On the other hand, ground response analysis consists of numerical modeling of wave propagation. The analysis would encompass a soil domain of limited dimension and take into account dynamic soil behavior. Soil material models in ground response analysis can be equivalent-linear and nonlinear. In engineering practice, equivalent-linear ground response analyses are more popular due to its ease in input parameter selection. However, nonlinear ground response analyses have the potential to better predict the ground response, especially under large-strain condition, because representation of soil behavior is more accurate.
In this research, performance of nonlinear ground response analyses is evaluated with the use of vertical arrays. It is found that predicted surface response from equivalent-linear and nonlinear analyses is generally the same if weak input motion is used. Moreover, effect of variability of shear-wave velocity on ground motion prediction is studied. It is found that uncertainty of ground motion due to variability in velocity would be underestimated by the First Order Second Moment method, therefore one should use randomized velocity profiles (defined by a layering model and a statistical velocity model that considers standard deviation and correlation between layers) in the uncertainty estimation.
Subjects
Ground response analysis
Nonlinear
Vertical Array
Shear wave velocity
Site Effect
Uncertainty
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-101-R99521129-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):ed1d771d5a52bec4ac6f040ca7a601e3