苏中腹地湖相软土土性参数变异性统计描述

Statistical characterization of variability of lacustrine soft soil in central region of Jiangsu Province

基于江苏中部某高速公路工程地质勘察所提供的大量室内试验测试数据,运用随机场理论,将土层剖面模拟为随机场而非传统意义上的随机变量,提出了基于随机场理论、考虑空间趋势分量的土性参数变异性统计方法,系统地统计分析了苏中腹地湖相软土土性参数的变异特性,对比了实验室物理性质指标参数、变形参数和强度参数变异性特征。统计结果表明,该地区软土实验室物理性质指标参数变异性较小,不存在明显的趋势分量,可用传统方法进行统计估计;实验室变形和强度参数存在明显的趋势分量,统计前需对数据进行回归分析和去趋势化处理,从而基于随机场理论进行统计分析。研究不仅提供了土性参数变异性统计的新方法,而且能对苏中地区工程建设提供必要的参考,为区域性软土土性参数随机场模型的建立打下了坚实的基础。

Based on a lot of laboratory testing records from geotechnical investigation of an expressway located in the central region of Jiangsu province and using random field theory to model soil profile as a random field rather than traditional random variable, a statistical method of variability of soil parameters based on random field theory has been recommended, which can take account of the deterministic spatial trend components at the same time. Based on this, the properties of soil variability of lacustrine soft soil in the central region of Jiangsu province were statistically analyzed and the variability characteristics between laboratory physical index, deformation and strength parameters ware compared. It is indicated that the variability of the laboratory physical index parameters in this region is relatively small; no obvious trends or deterministic components can be observed; and traditional statistical method is effective for them too. Correspondingly, the laboratory deformation and strength parameters have obvious deterministic trends. They have to be destrended by regression analysis before statistical estimation, and then estimated by the recommended method due to random field theory. The conclusions of this paper provide not only a new method for statistical estimation of the properties of soil variability, but also prescriptions for engineering or construction in the central region of Jiangsu province. The third but not the least value of this paper is that it can lay a solid foundation for establishing random field model of regional soft soil properties.