EACE Journal — Vol. 13, No. 1
Comparative Study of Different Probabilistic Methods in the Analysis of Shallow Foundation Safety
Tewodros Gemechu, Eskndir Getachew, Henok Fikre
Abstract
With the aim of proposing better analysis methods for designing shallow foundations, this research compares different types of probabilistic analysis methods used to analyze foundation safety. Both deterministic and random variable input data have been used for this analysis which include the N values obtained from the SPT result (number of blows), the width of foundation, depth of foundation, and load of the building. The spatial interpolation (kriging) method estimates the soil data at unsampled locations to increase the quality of the data set for the probabilistic analysis. The variable N is treated as a random variable among the input variables, while the rest are treated as deterministic values. Foundation safety was evaluated for different foundation widths, and then compared with the target reliability index or probability of failure, and then, for each method, the proper width was chosen. The foundation width obtained from the probabilistic methods, the first-order reliability method, first-order second moment, and Monte Carlo simulation gives a similar result. In addition, the coefficient of variation has a significant effect on the value of the probability of failure. In general, the probabilistic method gives an excellent framework to incorporate the uncertainties of input variables.
Read Full Article (PDF)