Analysis of the parameters affecting LiDAR intensity and its application in determining rock joint surface alteration

Abstract

Rock mass characterization is required in many rock engineering projects. Among the parameters used to determine the rock mass rating (RMR), the discontinuity of rock mass, which includes the separation and weathering of the discontinuity surfaces, accounts for the largest share. In another method of classifying rock mass, the Geological Strength Index (GSI), the joint alteration factor (𝐽𝐽𝑎𝑎), which also indicates the discontinuity condition of rock mass such as the weathering of the discontinuity and the state of the filling, has a greater influence than any other parameters. There is a tendency lately for the classification of rock mass, which have previously been carried out in a variety of manual ways, to be automated due to the development of photogrammetry and light detection and ranging (LiDAR) technologies. For characterizing rockmass using LiDAR, qualities of rock discontinuities such as joint spacing, waviness, smoothness, and alteration should be determined. Estimating joint spacing, waviness, and smoothness factor by LiDAR have been studied using quantitative point cloud coordinates, but the task of estimating the joint alteration factor is more subjective. As previously mentioned, the effect of the joint alteration factor on the overall GSI is large. This study examines accurate approach for determining joint alteration factor from LiDAR intensity data, which is the return strength of the laser pulse that generated the point. Previous studies have found that the reflective percentages or intensity of LiDAR are high for hard and less weathered rocks and small for more weathered and weak rocks. Therefore, through a number of laboratory experiments, a method of determining the joint alteration factor using LiDAR intensity was formulated by analyzing the factors directly affecting LiDAR. Factors that were not directly related to rock weathering were corrected. Laboratory experiments were performed to assess the impact of the scanning distance, incidence angle, roughness, micro-roughness, RGB color values, water saturation, and the mechanical properties of the rock on the LiDAR intensity and to ascertain how they affect it. As a result, it was concluded that the direct relationship between LiDAR intensity and the joint alteration factor could be obtained by correcting the scan distance, incidence angle, and RGB color values, which are the most influential factors to the LiDAR intensity when determining the joint alteration factor. The separation of a discontinuity and the type of filling material also have a significant influence on the 𝐽𝐽𝑎𝑎 and on LiDAR intensity and this was what the laboratory experiment was also designed to measure. The separation was increased from 1 mm to 6 mm at 1 mm interval to measure the change in intensity. Bentonite and sand were used as a filling material to examine how they affected the intensity. As a result of the experiment, it was concluded that it was possible to estimate which and how much filling material existed through the degree of the change in the intensity in the separation or in the filling position. A comparison of the hand-mapped data of rock alteration and the LiDAR intensity at three sites of rock slope also indicated a good relationship between intensity and joint alteration factor. The LiDAR intensity was high in the case of rock mass that was estimated to have a large GSI joint alteration factor or discontinuity condition within the RMR by means of hand mapping, and the degree to which this was the case was more apparent after correction on distance, incidence angle, and RGB value. Although correcting for each point would be the ideal, it would take significant time and effort. Consequently, for convenient and quick rock mass classification, the average value of the scanning distance, the incidence angle, or the RGB can be used alternatively.

RMR을 산정하기 위해 사용되는 여러 인자 중 불연속면의 틈새, 풍화도 등을 포함하는 불연속면 상태는 가장 큰 비중을 차지한다. 암반 분류를 위한 또 다른 방식인 GSI 산정에 있어서도 joint alteration factor, 즉 불연속면의 풍화도, 충진물 상태 등 불연속면 상태 지수는 joint condition factor를 최대 10배 증가시킬 만큼 다른 어떤 지수보다 그 영향력이 크다. 기존 수기로 많이 진행되던 암반 분류가 photogrammetry, LiDAR 등의 기술 발전으로 자동화 되고 있는 추세이다. 그 중 LiDAR를 이용해 암반 분류를 함에 있어 joint spacing, waviness, smoothness 등은 정량적인 점군의 좌표를 이용해 얻는 연구가 이루어진 바 있으나 joint alteration factor를 산정하는 작업은 보다 주관적으로 산정 시 어려움이 존재했다. 앞서 말했듯이 joint alteration factor가 전체 GSI에 미치는 영향은 크므로, 따라서 LiDAR를 이용해 이를 보다 hand mapping에 가까워지게 구하는 방법에 대해 연구하였다. 기존 연구들을 통해 LiDAR의 반사 강도는 단단하고 풍화가 덜 된 암반에 대해서는 그 값이 크고 많이 풍화되고 약해진 암반에 대해서는 그 값이 작음을 알 수 있었다. 따라서, 여러 실내 실험을 통해 LiDAR에 직접적으로 영향을 미치는 인자를 산정하고 그 중 암반 풍화도와는 직접적으로 관련 없는 인자는 보정해 LiDAR 반사 강도를 이용해 joint alteration factor를 구하는 방법을 연구했다. LiDAR intensity에 직접적인 영향을 주는 인자로 주사 거리, 입사각, 거칠기, 사포를 이용한 미세 거칠기, RGB 색상 값, 암석 기본 물성 (UCS, 탄성파 속도, 공극률), 암석을 이루는 광물 조성, 포화도를 선정하고 얼마나 영향을 미치는지 고찰하기 위해 일련의 실내 실험을 수행했다. LiDAR 반사 강도에 가장 영향을 많이 미치는 요인들인 주사 거리, 입사각, RGB 색상 값을 보정해 반사 강도와 joint alteration factor의 직접적인 관계를 산정했다. 틈새 간격, 충진물의 종류 역시 반사 강도에 많은 영향을 미치는데, 실내에서 틈새를 1 mm부터 6 mm까지 1 mm 간격으로 늘려가며 반사 강도의 변화를 측정했고 충진물로 bentonite, sand를 사용해 각각 intensity에 어떤 영향을 미치는지 살펴보았다. 실험 결과 틈새나 충진물 구간에서 반사 강도가 변하는 정도를 통해 어떤 충진물이 이용되었는지 추측할 수 있다는 결론을 내렸다. 불연속면의 변질도가 다른 세 암반 사면을 LiDAR로 스캔하고 그 반사 강도를 얻어 앞서 언급한 인자들을 보정했다. 그 결과를 hand mapping을 통해 얻은 불연속면 변질도와 비교해본 결과 반사 강도는 불연속면 변질도를 범위로 산정하기에 충분했다.