Scholarlink

<p>Building 3D geological models from field and subsurface data is a typical task in engineering geology studies. Owing to limited modeling data and subjective interpretation errors, these models may contain substantial structural uncertainties related to regional tectonic interfaces and joint fissure surfaces. However, the uncertainties are commonly disregarded, and systematic uncertainty analysis is rarely performed. This paper presents an uncertainty analysis framework to quantify and locate spatial structural uncertainty for rock tunnelling. Geological uncertainties are classified into three different types: inherent natural variability, data and measurement errors, and imprecise knowledge. The uncertainties are modelled using a combination of stochastic simulation and the spatial interpolation technique. For inherent natural variability, a disturbance based on Monte Carlo (MC) random sampling is carried out in the Kriging procedure, and the disturbance is limited to interpolation. For the two other types of uncertainties, the raw data disturbance based on MC sampling is addressed before the Kriging process. The differences between the realizations enable us to evaluate the range and degree of geological uncertainties. The probability distribution function at each location can be obtained by assuming that all the realizations have equal probabilities. Moreover, the results of the uncertainty simulation can be visualized in several ways, ranging from borehole histograms to probability maps. The proposed method is illustrated through stochastic modeling and geological uncertainty analysis of the Longtoushan rock tunnel in China. Numerical results indicate type2 uncertainty curves exhibit greater fluctuations in the middle and bottom section of the tunnel rock mass, type3 uncertainty curves show larger fluctuations at unsampled locations, whereas significant fluctuations are virtually absent at sampled points. Integrating analyses of both uncertainties can eliminate the shortcomings inherent in considering only one type of uncertainty. The case study demonstrated that this approach provides insight into model quality and offers an effective way to calculate the probability field for different rock types.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Geological uncertainty assessment for rock tunnelling using spatial stochastic modeling

  • 期刊论文

  • 发表日期：

    2026年6月18日

查看全文
  • Yandong Li,
  • PeiNan Li,
  • Jiaxiao Wang,
  • Xiaojun Li

摘要

Building 3D geological models from field and subsurface data is a typical task in engineering geology studies. Owing to limited modeling data and subjective interpretation errors, these models may contain substantial structural uncertainties related to regional tectonic interfaces and joint fissure surfaces. However, the uncertainties are commonly disregarded, and systematic uncertainty analysis is rarely performed. This paper presents an uncertainty analysis framework to quantify and locate spatial structural uncertainty for rock tunnelling. Geological uncertainties are classified into three different types: inherent natural variability, data and measurement errors, and imprecise knowledge. The uncertainties are modelled using a combination of stochastic simulation and the spatial interpolation technique. For inherent natural variability, a disturbance based on Monte Carlo (MC) random sampling is carried out in the Kriging procedure, and the disturbance is limited to interpolation. For the two other types of uncertainties, the raw data disturbance based on MC sampling is addressed before the Kriging process. The differences between the realizations enable us to evaluate the range and degree of geological uncertainties. The probability distribution function at each location can be obtained by assuming that all the realizations have equal probabilities. Moreover, the results of the uncertainty simulation can be visualized in several ways, ranging from borehole histograms to probability maps. The proposed method is illustrated through stochastic modeling and geological uncertainty analysis of the Longtoushan rock tunnel in China. Numerical results indicate type2 uncertainty curves exhibit greater fluctuations in the middle and bottom section of the tunnel rock mass, type3 uncertainty curves show larger fluctuations at unsampled locations, whereas significant fluctuations are virtually absent at sampled points. Integrating analyses of both uncertainties can eliminate the shortcomings inherent in considering only one type of uncertainty. The case study demonstrated that this approach provides insight into model quality and offers an effective way to calculate the probability field for different rock types.