| Summary: | 碩士 === 國立交通大學 === 土木工程系所 === 105 === In recent years, increase in extreme rainfall is with the global climate changes. The natural disaster happened in mountains increase year by year. Disaster prevention and reduction become very important. One of the base for disaster prevention and reduction is the site investigation. It is a trend to involve non-destructive geophysical method especially surface wave method in large deep-seat landslide case. The shear wave velocity profile obtained from the passive surface wave method would greatly contribute to the investigation planning and the bedrock mapping. It is economic and efficient. However, the application in mountainous area, passive surface wave method violates the horizontal layer model assumption. Thus , the aim of the study is to evaluate the feasibility of passive surface wave method in large deep-seat landslide investigation. The influence of passive source distribution, inclined layer and topography in passive surface wave method are studied by 3D numerical simulation. Three popular passive surface wave methods (Roadside MASW, MAPS and SPAC) are examined. According to the numerical results, source distribution of passive sources would influence the performance of the methods even in horizontal layer earth model. When dominate sources exist, roadside MASW would have higher dispersion curve, MAPS can correct the source out-liner effect and SPAC can perform normally for frequency larger than 5 Hz without any correction. However, the dispersion curve with frequency less than 5 Hz retrieved from SPAC would be usually lower or difficult to identify. For the effect of inclined layer, it would reflect in the distribution of source. Even for the uniform distributed source, it would be forced to reveal as a dominate source case after the waves propagating through the dipping layer. Thus, the examined methods would perform like preview case except for the incident angle correction in MAPS. In fact, the direction for the dominate source is frequency dependent in dipping layer. Currently, limited to the software used, only frequency-free correction is applied. This causes the higher value of dispersion curve in low frequency, which means a shallow bedrock depth would be evaluated. Finally, for the effect of topography, the results show that in simple inclined slope, the effect can be ignored if the geophone spacing set to be the slope distance between geophones instead of horizontal distance and the representing profile should be perpendicular to the slope. Concluding from the numerical studies, MAPS with passive direction correction is the better technique applied in deep-seat landslide inverstigation.
|
|---|
