In this study, we propose a method, which incorporates faulting into the theory and algorithm based on the computer processing of sedimental and erosional geologic structure. The sedimental and erosional geologic structures can be expressed by the surfaces surround the distributed areas of geologic units, and the logical relation between the distribution of geologic units and surfaces are termed logical model of geologic structure. When we regard the changeable process of geologic structure formed through sedimentation and erosion as the relations between the distribution of geologic units and the surfaces, there is a definite rule of its formative process corresponds with the logical model of geologic structure, and it is expressed as a recursive definition. Concerning the faulting we define the rule which suggests the surface of fault divides a three-dimensional geologic unit and the open space into two areas, and the geologic structures of each area can be preserved. Therefore, faulting can be reasonably included into the recursive definition, which leads logical model of sedimental and erosional geologic structures, and the faulted geologic structure can be expressed as recursive definition. In addition, this recursive definition can lead a logical model of geologic structure cut by plural faults. With introducing a logical model of faulted geologic structure, we propose the faulted geologic map can be generated without any changes of the existent processing system based on a logical model of geologic structure.

View full abstract

We report the study on the evaluation method of ground classification ahead of tunnel face using drilling data such as drilling speed, specific drilling energy and damping pressure and discuss the applicability of the drilling survey system to the tunnel construction field. Moreover, the neural network system on the basis of a learning data has been introduced to analyze the drilling data automatically and objectively. We demonstrate the advantage of the drilling survey system making a precise decision of ground classification without human experiences from the experimental results.

View full abstract

Rock mass failure causes considerable damage and result in serious natural disasters. Hazard management in rock mass failure is clearly very important. Hazard maps using geographical information systems (GIS) which is one of the latest information technology (IT) are likely to be effective in the monitoring and management of potential landslide disasters. To make a hazard map, a precise and effective understanding of the hazard area is of the most importance. Especially, a quantitative estimation of the hazard area and an adequate display on the digital map are indispensable for using GIS. In this study, we describe a Hazard area Estimation System (HES) using the result of multivariate statistical analysis. This system analyzes the hazard area for rock mass failure debris using two factors that are an elevation angle and a spread angle. The process of calculations and analyses in the system is as follows. 1) A digital elevation model (DEM) data is loaded, and the elevation is displayed according to color classification. 2) Failure points are specified by mouse click, and slope conditions such as geology and topography are selectively input. 3) The hazard area is predicted from the input slope conditions. 4) The probability is calculated from the computed hazard area. 5) The probability is classified by color and displayed on the DEM. This system enables the displayed hazard map to be printed for distribution. Moreover, the results can be saved as a csv format file and can be displayed using general GIS software. The main characteristic of this system is simple input process and high-speed estimation, since the system uses previously analyzed results. This system is ideal for use as a first screening process: that is, for prioritizing where countermeasures are most likely to prevent rock mass failure. For slopes where countermeasures are particularly important, actual field investigations and numerical analysis methods should be used, backed by more detailed studies.

View full abstract