多元统计
航程(航空)
联合概率分布
统计
阿太堡极限
数学
岩土工程
领域(数学)
地质学
工程类
含水量
航空航天工程
纯数学
作者
Jianye Ching,Kok‐Kwang Phoon
标识
DOI:10.1139/cgj-2013-0353
摘要
This paper constructs a 10-dimensional multivariate probability distribution covering 10 clay parameters. The parameters are the liquid limit, plasticity index (PI), liquidity index, effective vertical stress, undrained shear strength, sensitivity, and three piezocone test parameters. A CLAY/10/7490 database is compiled in a companion paper for this purpose. The database consists of 7490 data points from 251 studies. The number of data points associated with each study varies from 1 to 419 with an average 30 data points per study. The clay properties cover a wide range of overconsolidation ratios (but mostly 1∼10), a wide range of sensitivity (S t ) (sites with S t = 1∼tens or hundreds are fairly typical), and a wide range of PI (but mostly 8∼100). The constructed multivariate probability distribution can be used as a prior distribution to derive the joint distribution of design parameters based on limited but site-specific field data. Note that the entire joint distribution of the 10 clay parameters is derived, not marginal distributions or simply means and coefficients of variation. These multiple design parameters can be updated from multiple field measurements, which is more useful than updating one design parameter using one field measurement that is typical in current practice. This paper also demonstrates that it is practical to build multivariate probability models by combining available bivariate models, which are prevalent in the geotechnical engineering correlation literature. The proposed approach circumvents the need to collect multivariate data, which are rarely found in typical site investigation programs.
科研通智能强力驱动
Strongly Powered by AbleSci AI