Statistical distributions of wave velocities and elastic moduli in near-surface unsaturated soils
Statistical distributions of wave velocities and elastic moduli in near-surface unsaturated soils
复制标题
DOI：
10.1016/j.soildyn.2022.107247
复制
发表时间：
2022-03-14
影响因子：
4
通讯作者：
Taylor, Oliver-Denzil S.
中科院分区：
文献类型：
--
作者：
Abdollahi, Masood;Vahedifard, Farshid;Taylor, Oliver-Denzil S.
剪切波速 (v(s)) 和压缩波速 (v(p)) 广泛用于了解近地表地质结构、推导土壤的小应变弹性模量以及进行各种地球物理、岩土工程和地质环境分析。虽然 v(s) 和 v(p) 对含水量或饱和度的依赖性已得到充分认识，但不同饱和度水平内波速测量值和导出的弹性模量的可变性仍有待了解。本研究的主要目的是检查饱和度对非饱和土中测量的波速的统计分布和导出的小应变弹性模量的影响。为此，我们对级配较差的细到中砂进行了 7 个完整的润湿-干燥循环，进行了 360 次超声波实验室测试（广泛的阵列）。使用实验室测量数据以及一系列统计测试来评估 v(p) 和 v(s) 测量值以及导出的弹性变量的统计分布和变异性，包括 v(p)/v(s) 比率、剪切模量 (G)、杨氏模量 (E)、泊松比 (mu) 和体积模量 (K)。结果表明，关于地球物理、岩土工程和地质环境分析中使用的 v(p) 和 v(s) 测量值以及弹性模量的量化的许多假设可能无效。 v(p) 和 v(s) 数据最好分别由对数正态分布和威布尔分布表示，但随后导出的弹性属性可能需要不止一种分布类型才能充分表示不同饱和状态和关系的统计行为。
The shear wave velocity (v(s)) and the compressional wave velocity (v(p)) are extensively used to understand the near-surface geologic structure, derive small-strain elastic moduli of soils, and perform a wide range of geophysical, geotechnical, and geo-environmental analyses. While the dependency of v(s) and v(p) on water content or degree of saturation is well recognized, the variability of wave velocity measurements and derived elastic moduli within different saturation levels remains yet to be understood. The main objective of this study is to examine the effect of degree of saturation on the statistical distribution of measured wave velocities and the derived small-strain elastic moduli in unsaturated soils. For this purpose, 360 ultrasonic laboratory tests, an extensive array, were performed on a poorly graded fine-to-medium sand over seven full wetting-drying cycles. The laboratory-measured data were used, along with a suite of statistical tests, to evaluate the statistical distribution and variability of the v(p) and v(s) measurements and the derived elastic variables-including v(p)/v(s) ratio, shear modulus (G), Young's modulus (E), Poisson's ratio (mu), and bulk modulus (K). The results show that many of the assumptions regarding the quantification of v(p) and v(s) measurements and elastic moduli used in geophysical, geotechnical, and geo-environmental analyses may not be valid. The v(p) and v(s) data are best represented by lognormal and Weibull distributions, respectively, yet the subsequently derived elastic properties may require more than one distribution type to adequately represent the statistical behavior for different saturation regimes and relationships.