Click for next page ( 32


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 31
21 and Rix 2005). Cost was often cited as a reason for not this is not the case among their consultants and research- conducting parametric studies, but a number of the survey ers. The most commonly used program for equivalent-linear respondents indicated that they subcontract this work to analysis is SHAKE2000, followed by DEEPSOIL. specialists/experts. Nonlinear Total Stress Analysis SITE RESPONSE ANALYSIS The survey responses reflect a broad range of opinions as to when a nonlinear (NL) site response analysis is needed. Equivalent-Linear Models Most respondents concur that a site response analysis is needed when dealing with softer soil sites (i.e., with classes The survey responses reflect a broad range of opinions as E and F), though, as with equivalent-linear analyses, others to when an equivalent-linear (EL) site response analysis is reported that they use them for site classes C and D classes as needed. Most respondents concurred that a site response well. There were no clear criteria regarding what PGA input analysis is needed when dealing with softer soil sites (classes would trigger a site response analysis. Some respondents E and F); some indicated that they use them for C and D indicated that they use it only for high PGA levels (PGA > classes as well. There were no clear criteria on the input PGA 1.0 g) whereas others indicated a lower PGA threshold (0.4 that would trigger a site response analysis. Some respon- 0.5g). Most respondents recognized that for more important dents said that they used it for low levels of seismic hazard, structures (e.g., base isolated and long structures), an NL while others reported that they used it when the hazard is site response analysis is warranted. There was a consensus high. Most recognized that a nonlinear (NL) site response amongst respondents that an NL site response analysis could analysis is warranted for more important structures (e.g., be used when computed shear strains exceed 1%. A num- base isolated or long structures). ber of respondents said that NL analyses are used to reduce demand on a structure. Equivalent-linear analysis is a robust approach that has been used in the profession for the past 40 years. The lit- The responses reflect the occasional confusion about the erature (e.g., Hashash et al. 2010) indicates that 1-D EL use and role of site response analysis and the need for greater site response analysis can always be performed regardless understanding and guidelines. The use of strain-based crite- of whether the higher dimension or nonlinear site response ria for switching to nonlinear analysis is reasonable as this analyses are included. One-dimensional EL site response reflects when nonlinear soil behavior becomes important. analyses provide the key characteristics of a site and the However, a 1% shear strain threshold is likely too high, as propagated ground motion. They can be used to flush out many soils would be at or near failure at this level. Studies errors that might inadvertently be introduced in more by the authors have shown that nonlinearity in soil behavior advanced site response analyses. can affect site response at strains as small as 0.1% or lower. A PGA or ground intensity measure on its own probably will Equivalent-linear analyses require characterization of not be sufficient because strain levels in soft soils can be dynamic soil properties by modulus reduction and damping quite high even for what appears to be a low level of shaking. curves. A number of relationships for such curves are avail- Better guidance is clearly needed on when a nonlinear site able in the literature. In the 1970s, separate curves were pre- response analysis is necessary. sented for sands and clays before the effects of soil plasticity and effective confining pressure on soil behavior were well Nonlinear site response analyses use constitutive models understood. The responses showed, however, that the origi- that represent nonlinear soil behavior. Models are param- nal sand and clay curves are still widely used in contempo- eterized in different ways, so comparison is much more rary practice. Other widely used curves include Vucetic and difficult than for equivalent-linear site response analyses. Dobry (1991), EPRI (1993), and Darendeli (2001). The survey responses indicate that the choice of nonlinear soil model was closely tied to the choice of nonlinear site The available literature does not provide a systematic evalu- response software, because some nonlinear software does ation of the various modulus reduction and damping curves and not offer multiple soil models. Modified hyperbolic models their applicability. Indeed, the rigorously documented curves of with Masing rules were the most widely reported nonlin- Darendeli (2001) and Menq (2003) have not been yet published ear soil models. The respondents indicated that they tried to in the refereed literature. All these curves are generic in nature calibrate their models to match both modulus reduction and and can be selected based on soil index properties and stress damping curves. history. Only one respondent indicated that they directly mea- sure soil dynamic properties in the laboratory. Nonlinear Effective-Stress Analysis The survey indicates that 1-D, equivalent-linear analyses The survey responses reflect a broad range of opinions as to are by far the most commonly used by DOTs surveyed, but when an NL effective-stress site response analysis is needed.