Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1  Introduction 1.1 Project Background NCHRP Synthesis 428: Practices and Procedures for Site-Specific Evaluations of Earthquake Ground Motions (Matasovic and Hashash, 2011) showed that one-dimensional (1D) equivalent- linear total-stress site response analysis (SRA) is the de facto standard for state department of transportation (DOT) highway facilities at locations where site-specific ground response analyses are conducted in accordance with provisions in the AASHTO LRFD Bridge Design Specifications (2020) (9th Edition) and the AASHTO Design Guidelines for Seismic Bridge Design (2011). How- ever, many users and various DOTs have concerns about the applicability of equivalent-linear analyses for the cases where site-specific SRA is most relevant [i.e., for soft soil sites, liquefiable sites, and sites subjected to strong ground shaking (shaking that induces peak shear strain in excess of 0.5% to 1%; this usually occurs in soft soils when ground surface peak ground accelera- tion exceeds approximately 0.4 g)]. Nonlinear total-stress and nonlinear with excess pore water pressure (PWP) generation and dissipation (i.e., effective-stress) 1D site response analyses are promising alternatives to equivalent-linear analysis. A recent increase in the use of these analyses in engineering practice has been prompted not only by the DOTsâ concerns outlined previously, but also by the opportunity to reduce spectral accelerations by 33%. There is also a recognition that the use of equivalent-linear analysis at liquefiable sites is not always conservative, especially for long-period structures like suspension bridges. Also, many bridge engineers practice build- ing codeâbased structural design and are cognizant of the requirements imposed by building codes [e.g., requirements of the International Building Code (IBC, 2021), with its current reference standard ASCE/SEI 7-22 (ASCE/SEI, 2022)] to evaluate potentially liquefiable sites by means of effective-stress analysis (ESA). Nevertheless, there are concerns about what types of nonlinear models should be used, the lack of clear parameter selection protocols, the lack of consolidation of lessons learned from validation of effective-stress programs with what has been learned from validation of nonlinear total-stress methods (e.g., Kwok et al., 2007), the lack of third-party verification and validation, and the uncertainty with certain interpretations of modeling results. To assist highway facility designers and DOT reviewers, guidelines on the selection of effective-stress numerical models, effective application of 1D ESA, and appropriate interpreta- tion and use of modeling results are needed. In this regard, appropriate use of ESA in engineer- ing practice may lead to a safer, more economical seismic design for various types of highway facilities. C H A P T E R 1
2 Seismic Site Response Analysis with Pore Water Pressure Generation: Guidelines 1.2 Intended Audience and Suggested Background References It is presumed that the reader of this document has a good understanding of geotechnical engineering and a familiarity with the concepts of earthquake engineering, geotechnical earth- quake engineering, engineering seismology, engineering geophysics, and the basic principles of structural dynamics. A reader with this level of understanding should be able to use equivalent- linear and simple (i.e., lumped-mass) nonlinear 1D programs or two-dimensional (2D) pro- grams run in 1D mode using semi-empirical constitutive models (constitutive models that are not strictly based on the theory of plasticity). To use advanced constitutive models (i.e., constitutive models based on the theory of plas- ticity) and to run advanced 1D site response analyses, the reader must have a basic familiarity with the concepts of the theory of plasticity and constitutive modeling. This includes famil- iarity with element testing [i.e., with an exercise performed prior to analysis of the soil profile (system)] to demonstrate that results of advanced laboratory testing can be matched with a selected constitutive model. The latter is an important validation step of both total- and effective- stress SRA. Chapters 2 and 3, including cited references, provide basic and select advanced theoretical information. However, this information may be in abbreviated form in some places. More detailed information about the basic concepts necessary for the understanding and execution of non- linear effective-stress site response analysis is explained in books such as Geotechnical Earth- quake Engineering by Kramer (1996) and Dynamics of Structures by Chopra (2017). Guidance on how to characterize a site, account for uncertainty in site characterization, and perform 1D total- stress SRA is provided in Stewart et al. (2014). The references and list of abbreviations and symbols at the end of this report should further facilitate the use of this document. 1.3 Limitations of This Study In general, this study is limited to vertical propagation of horizontally polarized shear waves (i.e., to 1D analysis, including analysis with 2D software run in 1D mode). It is also limited to nonlinear ESA as applied to the design of highway bridges and other infrastructure founded in or above potentially liquefiable soils. Specific limitations of this study include the following: ⢠The software selection was limited to commercially available nonlinear effective-stress spe- cialty software with widely used constitutive models integrated into the programs. Equivalent- linear analysis is used herein only for calibration and validation purposes. ⢠The analyses were limited to site response in 1D. Therefore, only case histories of predominantly 1D response were selected. ⢠The site exploration program, advanced laboratory testing, and development of constitutive model parameters focus on the undrained response of sands and especially of silty sands. ⢠Limitations of case histories are as follows: moment magnitude (M) range: M 4.9 â M 7.1; approximate site-to-source distance (R) range: 8 km to 70 km; input peak ground acceleration (PGA) range: 0.1 g to 0.6 g; and site-predominant period (Ts) range: 0.2 s to 1.4 s (upper bound, as evaluated herein; several profiles extend beyond the evaluation depth). The limitations outlined here cover many situations commonly found in engineering practice.
Introduction 3  1.4 Study Objectives and Focus The main objective of this study is to provide guidelines for the selection and use of methods for 1D nonlinear seismic site response analysis with excess PWP generation and dissipation (ESA). Based on the information and guidelines presented, a user with adequate background should be able to: ⢠Verify the applicability of the selected software and constitutive models (CMs) (i.e., check whether they meet the requirements set in this study); ⢠If needed (or if required by an agency), perform software/CM validation (i.e., demonstrate that one or more representative case histories can be replicated by the proposed combination of selected software, CMs, and modeling approaches); ⢠Develop a model of the soil profile (system); ⢠Calibrate and validate the model at soil element and system (soil profile) levels; ⢠Perform an analysis; ⢠Extract relevant results from the analysis; ⢠Quantify the effects of site-specific conditions on earthquake ground response; and ⢠If appropriate, recommend a change in code-mandated minimum design ground motions to the extent justifiable by the conducted computations. The secondary objective of this study is to provide a reference that can serve as a basis for a review of reports submitted to DOTs that base design recommendations on the results of 1D non- linear effective-stress SRA. The study focuses on the resolution of the following DOT concerns: (i) What is the applicabil- ity of 1D total-stress equivalent-linear analyses for cases in which SRA is most relevant (i.e., liq- uefiable sites that are subject to strong ground shaking)? (ii) Can the response of representative effective-stress case histories be adequately replicated by the current generation of software and constitutive models? And (iii) how significant is the choice of numerical and constitutive models employed on the calculated response? 1.5 Research Approach The research approach followed in this study was to (i) gather, review, and synthesize avail- able information; (ii) identify data gaps; (iii) fill in data gaps to the extent possible; (iv) perform numerical modeling exercises to support the development of guidelines; and (v) develop and present guidelines. Specific items within this research approach include: ⢠Establish required attributes for software and constitutive models, then perform a literature search to identify analysis platforms that possess these attributes and can be used to meet the project objectives; ⢠Establish required attributes for case histories, then perform a literature search and assemble a suite of case histories with the appropriate attributes that could be used to meet the project objectives; ⢠Develop SRA models for the suite of case histories using the identified numerical modeling software packages and constitutive models, perform SRAs, and compare model outputs to recorded and inferred targets; ⢠Provide general commentary on the capabilities and limitations of the effective-stress SRA approach (e.g., what does a âgood fitâ between measurements and model-predictions look like? How can a state DOT be reasonably assured that a practitioner has performed adequate quality control?); and ⢠Develop guidelines for the implementation of SRA with PWP generation and dissipation in (highway) engineering practice.
4 Seismic Site Response Analysis with Pore Water Pressure Generation: Guidelines 1.6 Data Presentation and Presentation of Results One of goals of this study was to establish a curated set of seven case histories that are repre- sentative of site conditions found in practice (e.g., stratified soil deposits with liquefiable silty sand layers). These case histories had to be relevant for this study (i.e., records of excess PWP were available), easily accessible (i.e., recorded histories of input motion and site response were available in digital format), and sufficiently complete (at a minimum, information required for selection of generic modulus reduction and damping curves was available). Finally, these case histories needed to be representative of 1D site conditions (i.e., suitable for use with effective- stress SRA programs in 1D mode). In other words, data should be presented in a format that allows a practicing engineer using any software that meets the attributes established herein (Tables 4-1 and 4-3) to be able to select a relevant case history, download relevant information, replicate key aspects of that case history, and interpret model outputs, with relatively little effort. In order to establish a curated set of seven case histories, the research team (RT) performed a broad data search (see Section 4.4 and Appendix A-3) and, on consultation with other partici- pants in the project, concluded that the pool of case histories with records of excess PWP had to be augmented. The RT also concluded that one of two case histories with records of excess PWP available lacked relevant site and advanced geotechnical laboratory testing data. Therefore, the research program included the following tasks: ⢠Augment the existing Wildlife Liquefaction Array (WLA) site case histories (single profile, but two recorded time series, including acceleration history from the 1987 M 6.2 Elmore Ranch earth- quake and time series of both acceleration and excess PWP from the M 6.6 Superstition Hills earthquake that occurred 7 hours later) by conducting a supplemental site-specific field investiga- tion, in-situ testing, sampling, routine and advanced laboratory testing, and data interpretation. ⢠Develop a new case history at the broader WLA site. Base the case history on the existing WLA site information [confirm the uniformity of soil profile at two locations by cone penetration test (CPT) sounding], augmented site information (see previous bullet), and select strong-motion and excess PWP information recorded after the 1987 Elmore Ranch and Superstition Hills earthquakes. ⢠Augment the existing WLA site case history of response to the 1987 Elmore Ranch earthquake that lacks recorded PWP response by generating an history of excess PWP in a centrifuge experiment. For that, perform a project-specific centrifuge experiment and validate the results by comparing acceleration response spectra from the Elmore Ranch earthquake and the centrifuge experiment. ⢠Develop a new case history representative of a site outside California shaken by a very high intensity motion. Use an existing ground motion record, perform a project-specific centrifuge experiment using remolded soil, and accompany the experiment with project-specific labora- tory testing of the remolded soil spun in the centrifuge. Additional topics include: ⢠Selection of analytical1 methods and an outline of their limitations; ⢠An overview of input parameters needed to perform 1D nonlinear effective-stress SRA of sandy soils, including the required type of acceleration histories, site characterization infor- mation, material models and parameters, and information required to validate the model; ⢠The process of numerical and constitutive model setup, including constitutive and numerical model validation; ⢠Suggestions on how to interpret the calculations conducted by the numerical modeling of site response; and ⢠Limitations of the approach used herein and how to improve it in future work. Discussions about the uncertainties2 associated with the use of these 1D effective-stress SRAs are also presented in this report. Information used herein in digital format (accelerograms and PWP records) has been posted on DesignSafe3 for public use.