The Earthquake Engineering Online Archive

Estimation of uncertainty in geotechnical properties for performance-based earthquake engineering

Jones, Allen L.; Kramer, Steven L.; Arduino, Pedro

PEER-2002/16, Pacific Earthquake Engineering Research Center, University of California, Berkeley, 2002-12, 100 pages (400/P33/2002-16)

Prediction of the performance of structures during earthquakes requires accurate modeling of the geotechnical components of soil-structure systems. Geotechnical performance is strongly dependent on the properties of the soil beneath and adjacent to the structure of interest. These soil properties can be described by using deterministic and probabilistic models. Deterministic models typically use a simple discrete descriptor for the parameter of interest. Probabilistic models account for uncertainty in soil properties by describing those properties by using discrete statistical descriptors or probability distribution functions. This report describes sources and types of uncertainty in geotechnical engineering practice and introduces the basic concepts and terminology of the theory of probability. Statistical parameters and the probability distributions most commonly used to describe geotechnical parameters are reviewed. The report then presents tabulated statistical parameters for soil properties that have been reported in the literature; both laboratory- and field-measured parameters describing moisture-density, plasticity, strength, and compressibility characteristics are presented. The theory of regionalized variables, including concepts of autocorrelation, variograms, and stationarity are presented, along with tabulated values of parameters describing spatial variability that have been reported in the literature. Finally, procedures and tools for estimation and simulation of spatially variable soil properties are presented.

Available online: http://peer.berkeley.edu/publications/peer_reports/reports_2002/0216.pdf