강우재현 모형실험에 의한 불포화 무한사면의 간극수압 특성

Abstract

In infinite slope stability problems, most of soils near surface in an unsaturated state in the beginning of rainfall and they are getting saturated with rainfall from top to depth. Their soil mechanical concepts or behavior characteristics are quite different from those of saturated soils. There are many difficulties in theoretical analysis, interpretation and application to practice, etc. so they were handled as saturated ones in most of designs and engineering decisions, but such an interpretation cannot lead to the rational conclusions in the analysis of the phenomena and behaviors of the considering soils. and consequently it may cause economic inefficiency resulting from excessively safe design. Therefore, for more rational study on the behavior of natural soil, we need to have the understandings of the properties of unsaturated soils and to investigate the soil water characteristics and permeability characteristics considering of negative pore water pressure. In this study, a simulated rainfall system was built, and the unsaturated characteristics were examined by execution of computer simulation and experiment on the soil water characteristic curves under the various rainfall and slope conditions. With the results, the applicability of infiltration behavior under rainfall and soil water characteristic curve models to the unsaturated granitic sedentary deposit slopes was examined. Furthermore, the results of infiltration analysis from numerical model test were compared with those from indoor simulation. Conclusions taken out from this study are as follows. At the results of simulated rainfall system test, the more the intensity of rainfall was high, and the more the slope was steep and the more rainfall lasted long, the more the volumetric water content began to increase earlier and the more the matric suction decreased faster in the wetting process, the more the restoration of matric suction was delayed in the drying process. Consequently, the stability of the unsaturated infinite may be deteriorated by the fast decrease and slow restoration of matric suction. At the results of comparison the volumetric water content and matric suction measured in the wetting process (under rainfall ) with those in the drying process (leaving as it was) of the simulated rainfall system, the volumetric water content showed a difference of 2～5% and matric suction of about 3～10kPa, indicating the occurrence of hysteresis. In addition, the difference was relatively larger in matric suction than in the volumetric water content, and this tells that the hysteresistical behavior is larger in matric suction. The value of unsaturated characteristic of a soil water characteristic curve obtained from SWCC Cell experiment was significantly different from that calculated from a soil water characteristic curve only derived from the physical properties of the soil under consideration. This tells that it is unreasonable to determine the soil water characteristic curve only based on the physical properties of soil and apply it to the analysis of unsaturated soil. When the soil water characteristic curve derived from measurements in the wetting process and the drying process of simulated rainfall system were compared with those from the soil water characteristic curve experiment, both methods produced soil water characteristic curves close each other in the wetting process and the drying process, but in both, there was a difference between results obtained from in the wetting process and those from in the drying process. Thus, when soil water and infiltration characteristic curves are rationally applied to the design and stability analysis considering of the properties of unsaturated soil, it is considered desirable to apply the soil water characteristic curve of the wetting process to the wetting process, and that of the drying process to the drying process. When the infiltration analysis was performed using a verified program under the same state and conditions with the simulated rainfall system, the distributions of hydrostatic pressure and volumetric water content were similar and fell within the range of allowable error. However, there were differences in the values, probably because it was impossible to model perfectly various ground conditions including initial water content, compactness and permeable porous layer under the slope in the program, and because the model slope applied to infiltration analysis was too small in size and therefore any small variation could result in a big difference.