Monte Carlo Simulation to Assess Soil Property Uncertainties for Seismic Evaluation of Existing Structures-Paper

Paper

Pdf Summary

The paper presents a probabilistic method for accounting for soil-property uncertainty when evaluating existing structures for seismic performance under ASCE 41. The case study concerns shallow strip footings supporting buildings at a 1969 University of California athletic facility, where subsurface conditions include fill over sandy alluvium and available geotechnical information is limited.<br /><br />Soil friction angle is estimated from California Modified Sampler blow counts. Uncertainty in converting these counts to standard SPT values, correcting for energy, and applying an empirical relationship to estimate friction angle is represented by probability distributions. Monte Carlo simulation samples these uncertainties 4,000 times to produce distributions of soil parameters. For the alluvium, the estimated friction angle has a mean of 43.3° and standard deviation of 4.87°; the fill has a similar mean of 43.4° and standard deviation of 5.53°.<br /><br />The authors also develop an analytical approach using log-normal distributions and numerical integration, finding results nearly identical to the Monte Carlo simulations. The methods can support probabilistic estimates of foundation bearing capacity and can be extended to lateral capacity and seismic earth pressures.<br /><br />The paper emphasizes that reliable estimates of both mean soil properties and their variability are essential for seismic evaluations. Assuming lower soil strength is not necessarily conservative, because changes in soil stiffness and foundation response can also affect structural seismic demands. Probabilistic methods therefore provide a stronger basis for ASCE 41 evaluations than unsupported or improperly quantified variability assumptions.

Keywords

soil-property uncertainty

seismic performance

ASCE 41

shallow strip footings

probabilistic method

Monte Carlo simulation

friction angle

geotechnical uncertainty

foundation bearing capacity

california modified sampler