3 resultados para Size-disparity correlation
em University of Washington
Resumo:
Thesis (Ph.D.)--University of Washington, 2016-06
Resumo:
The relationship between saturated hydraulic conductivity (Ks) and grain-size distribution was evaluated for 49 sites underlain by either glacially over consolidated or normally consolidated fluvio-glacial deposits in the Puget Lowland. A linear regression comprising pairs of grain-size analyses and pilot infiltration tests predicts Ks with a 1 sigma uncertainty of a factor of about 3.5 with 70% of the population variance accounted for. The correlation coefficient R^2 of about 0.90 shows that there is a strong correlation between the grain-size distribution and Ks. In contrast, a widely applied analysis proposed by Massmann (2003) explains only 20% of the population variance for normally consolidated materials with an R^2 of only 0.15. That analysis entirely fails to explain the population variance for over consolidated materials. The method developed in this study is recommended for determination of Ks for fluvio-glacial deposits of the Puget Lowland.
Resumo:
This study evaluates two methods for estimating a soilís hydraulic conductivity: in-situ infiltration tests and grain-size analyses. There are numerous formulas in the literature that relate hydraulic conductivity to various parameters of the infiltrating medium, but studies have shown that these formulas do not perform well when applied to depositional environments that differ from those used to derive the formulas. Thus, there exists a need to specialize infiltration tests and related grain-size analyses for the Vashon advance outwash in the Puget Lowland. I evaluated 134 infiltration tests and 119 soil samples to find a correlation between grain-size parameters and hydraulic conductivity. This work shows that a constant-head borehole infiltration test that accounts for capillarity with alpha approximately 5m^-1 is an effective method for calculating hydraulic conductivity from our flow tests. Then, by conducting grain-size analysis and applying a multiple linear regression, I show that the hydraulic conductivity can also be estimated by log(K) = 1.906 + 0.102D_10 + 0.039D_60 - 0.034D_90 - 7.952F_fines. This result predicts the infiltration rate with a 95% confidence interval of 20 ft/day. The results of study are for application in the Puget Lowland.