| Summary: | 博士 === 國立臺灣大學 === 土木工程學研究所 === 92 === This study establishes a Unit-Pore-Throat Ensemble Model (UPTEM). It mimics the soil pore structure by many pore-throat units of different sizes which are not connecting to one another. Methodologies are developed to estimate the parameters of size distributions of pores and throats. They utilize the water-air soil experiment data of capillary pressure (Pc) and saturation (s) during primary drainage, primary imbibition and secondary drainage processes. The objective is to make the UPTEM hydraulically equivalent to the actual water-air-soil system. Provided with the UPTEM parameters, the Pc-s scanning loops of water-NAPL two-phase fluids within the same soil can be inferred without doing the characteristic-curve experiments.
The equilateral triangle throats and cubic pore morphology of pore-throat units enable the UPTEM to simulate the ink-bottle effect. The simple geometry and interconnected pore-throat edges make easy the positioning of fluid-interface balanced with the capillary pressure using Laplace Equation, as well as calculating the critical position and capillary pressure for interface displacement. Knowing the interface position, conductivity in addition to wetting- and non-wetting-phase saturations can be computed. In direct problems, UPTEM can reproduce the hysteresis effect and quantify the trapping ratio on macro-scale. For the displacement criteria are derived from fluid mechanic principles and morphology of pore-throat units, UPTEM can demonstrate the effect of contact angle on characteristic curves explicitly. The ensemble design of pore-throat units avoids the difficulties caused by size effect and the need to estimate joint distribution of network pore-throats in inverse problems. For the same reason, the size distribution parameters are more directly linked to the Pc-s curves.
Fitting the drainage and imbibition Pc-s curves with UPTEM was tried with application of different ‘large-throat’ and ‘aspect ratio’ distribution types. It is shown that using a simple one-to-one relationship between the two variables, the model can match the data reasonably well. The drawbacks of the UPTEM model are tedious calculation and over-simplified parameter estimation methodologies. Future research is needed for further improvement.
|
|---|
