Most geotechnical earthquake engineering problems can be broadly divided into two categories: response problems and failure problems. Response problems involve the effects of geomaterials on the amplitude, frequency content, and duration of earthquake ground motions; little or no permanent soil deformation is involved (fig. 1).
Fig. 1. Structural Damage in Mexico City, 1985. Amplification by underlying soft clays led to a five-fold increase in ground motions in some parts of the city.
The tendency of soil to amplify ground motions has long been known, and nonlinear amplification effects are now well recognized by seismologists and geotechnical engineers. Failure problems involve permanent deformations of soil masses, generally through transient or long-term exceedance of the shearing resistance of the soil. Both types of problems must be addressed to advance the development of performance-based geotechnical earthquake engineering.
Both response and failure problems frequently involve the interaction of geomaterials with structures such as buildings, bridges, and piers. Because the interaction generally takes place through the foundation of the structure, these problems are frequently referred to as "soil-foundation-structure interaction" (SFSI) problems. SFSI problems were noted as among the most pressing contemporary problems by practitioners at a recent PEER Consultants Forum. Other important problems, however, involve the response and/or failure of soils in the absence of nearby structures (e.g., stability of slopes adjacent to highways) or in situations in which the structure has little effect on the seismic performance of the soil (e.g., stability of soft soil beneath a bridge approach embankment). Knowledge of this type of "free-field" behavior is also required for solution of SFSI problems. The further development of performance-based geotechnical earthquake engineering will require improved tools for prediction of free-field and SFSI behavior.
Geotechnical engineers, despite their lack of formal discourse on performance-based engineering, have dealt with issues of serviceability for many years. For geotechnical problems, serviceability is strongly related to permanent deformations. The development of performance-based geotechnical earthquake engineering will require improved procedures for estimation of permanent deformations. Such procedures will necessarily involve nonlinear analyses of soils and soil-structure systems that go well beyond the simple sliding block models that have been used for more than thirty years.
As described by Helmut Krawinkler, PBEE will be developed in a reliability-based framework to allow uncertainties to be treated in a rational and objective manner (PEER Center News, January 1999). Because geotechnical problems usually involve natural geomaterials, the contributions of the geotechnical components to overall system uncertainty are likely to be significant. Natural soils, depending on their depositional environment, can exhibit strong spatial variability even over relatively short distances. In characterizing soil, the geotechnical engineer is usually required to estimate the soil properties based on measurements at a relatively small number of discrete sampling locations. This spatial variability can significantly increase the uncertainty of the characterization process.
Fig. 2. A shear-flexible soil bin being prepared for soil-pile interaction testing on the Richmond Field Station shaking table.
Fig. 3. The geotechnical centrifuge at the Center for Geotechnical Modeling at UC Davis. With a 9 m radius and the ability to accelerate a 4,500 kg payload to 50 g, the Davis centrifuge is one of the largest in the world.
Studies of ground motions, site response, and ground failure undertaken by others have not been oriented toward multilevel performance prediction, and have generally not included uncertainties and spatial/temporal variabilities. The explicit consideration of these factors is a distinguishing feature of PEER's Hazard Assessment Thrust Area.
The strategic plan for the Hazard Assessment Thrust Area calls for a coordinated sequence of research efforts that will allow geotechnical and ground motion hazards to be realistically accounted for in improved tools and procedures for performance evaluation and performance-based design. These research efforts will consist of one or more individual projects directed toward (1) ground motion characterization including development of ground motion catalogs, (2) improved tools and procedures for free-field performance prediction, (3) improved tools and procedures for soil-foundation-structure interaction analysis, and (4) development of performance-based design procedures for soil improvement and foundation remediation.
Research in the area of ground motions began in Year 1 and is continuing with interaction with earth scientists (through the Southern California Earthquake Center) and PEER Business and Industry Partnership researchers. A series of related projects that include testing and modeling of soil stress-strain behavior and prediction of free-field performance began in Year 1; these projects are continuing in Year 2, along with several new projects dealing with site characterization and SFSI. Over a ten-year period, the Hazard Assessment research is expected to lead to significant advances in the prediction of ground response, ground failure, and SFSI effects, and to ultimately provide a sound technical basis for the consideration of ground shaking hazards in PBEE.
Steven L. Kramer, Professor
Department of Civil and Environmental Engineering
University of Washington