Research On Estimation Of Soil Liquefaction Under Fuzzy Mathematics

Earthquake-induced soil liquefaction can destroy the construction. In order to know the suspectibility of soil liquefaction, the proper method should be employed to evaluate the soil liquefaction. The traditional evaluation methods of soil liquefaction, which belong to a kind of deterministic method, are mostly obtained from the macroscopic seismic hazard and laboratory experimentts. In fact, the occurrence of sandliquefaction is a complicated process, and has many influence factors, in which there are great uncertainties. Besides random uncertainties, in the evaluatition of soil liquefaction exist the other sort of uncertainties, i.e fuzzy uncertainties, because that many aspect of soil liquefaction have not been clearly revealed such as evaluation criterion, the influence of earthquake, the mechanism of soil liquefaction and so on. In the paper, the fuzzy uncertainties, which exist in assessing of the discriminant equation and the result of the evaluation of soil liquefaction, are reasearched and the fuzzy set is employed to assessing the reliability of soil liquefaction.Besides conveniency of operation, the cone penetration test can obtain continuous profile of soil and their parameters, which having high reliability compared with standard penetration test and shear wave velocity test. According to the field testing liquefaction data of cone penetration test from 19 strong earthquakes around the world, the fuzzy uncertainties in construction of the soil liquefaction discrimiant equation are analysized. The uncertainties in field testing data cause that they paly different the role in the construction of the discriminate equation. So in order to obtain more accurate discriminant equation, the fuzzy uncertainties must be included, that is to say, the field testing data should be considered as a kind of fuzzy point data. The fuzzy pattern reconginition set is employed to fit the uncertainties in the field testing data, and then the uncertainties are reflected in the liquefaction discriminant equation by logistic regression. The discrimnant equation based on the fuzzy point data has more stable property, and the result of evaluation more reasonable.Evaluation result of soil liquefaction with traditional methods is deterministic. Virtually, due to lack of the comprehensive research on the soil liquefaction mechanism, the fuzzy uncertainties also exist in the result of the evaluation of soil liquefaction, such as the geology environment, soil parameter, and stress condition around field soil and so on. On considering some drawbacks of the safety factor method, by taking into account the randomness of the distribution of the characteristics of soil in space and the effect of diversified loads and the illegibility of safety rules, the fuzzy random reliability method is applied to analyze the reliability of soil liquefaction. Based on a great deal of fied testing data, the fuzzy membership function is established, which has distinct engineering significance.Then the author deduced its fuzzy random limit state equation and calculated the fuzzy liquefaction probability according to the MCS method. The research reveals that the fuzzy liquefaction probability can systematically, quantificationlly consider the fuzziness and randomness of soil liquefaction compared with the traditional methods, and the decision-making based on the analysis of the fuzzy reliability will more agree to the property of the engineering.