심부 터널 주변 과응력 암반 거동의 평가와 예측에 관한 수치해석적 연구

Abstract

터널 시공기술의 발전과 교통, 전력, 용수 등에 대한 수요에 힘입어 1000 m 이상 지하를 굴착하는 터널 건설 사업들이 증가하고 있다. 심부 터널의 굴착은 주변 암반에 발생하는 과응력으로 인한 파석, 암석 파열, 압출과 같은 현상들이 발생할 가능성이 크다. 이들은 터널 안정성을 크게 저하시키는 주요한 위험 요인이 된다. 따라서 심부 터널의 성공적인 건설을 위해서는 계획 단계에서 이러한 현상들의 가능성과 규모를 합리적으로 예측하고 적절한 대책을 수립하여야 한다. 과응력 암반에서 발생하는 특별한 현상들은 파석, 암석 파열과 같이 급작스럽게 발생하는 취성 파괴와 큰 변형이 서서히 발생하는 압출 거동의 두 가지 범주로 크게 나뉘며 이들은 서로 다른 모델을 통하여 연구되어 왔다. 따라서 심부 터널 주변 암반의 거동을 해석하고 평가하기 위해서는 두 가지 범주의 현상 중 한가지를 전제한 검토를 별도로 수행하여야 한다. 본 연구에서는 Hoek-Brown 파괴기준을 기반으로하여 손상 임계, 취성 한계를 합성한 복합 거동 수치모델을 통하여 두 가지 범주의 현상을 한번에 해석하고 검토할 수 있도록 하였다. 심부 암반의 거동을 결정하는 여러 인자 중 자료 제한적인 조건에서 쉽게 추정하고 다른 변수들을 추출할 수 있는 확장성을 지닌 다섯 가지 기본 변수로 심도, 최대주응력비, 암석의 일축압축강도, 재료 상수 𝑚i, GSI를 선정하고 500∼3000m의 심부암반 조건의 변수를 몬테카를로 방법을 이용하여 조합한 1000개의 해석 사례에 대한 수치해석 매개변수 연구를 실시하였다. 결과 분석을 통해 취성 파괴와 압출 거동이 발생할 수 있는 임계 조건을 기본 변수로 표현하여 수식화하였으며 기본 변수에서 추출된 최대접선응력-암반강도비와 터널축소량, 소성반경 그리고 취성 파괴 깊이의 상관식을 유도하였다. 또한 심부 현장 암반의 조건을 편리하게 기술하기 위한 다섯 자리의 숫자로 구성된 현장암반조건코드를 고안하고 이를 활용한 현장암반조건 지수를 개발하여 매개변수 해석 결과의 상관도 분석에 활용하여 그 가능성을 확인하였다.|Thanks to development of tunnel construction technology and demands for transportation, power generation and water for irrigation, various projects of over 1000 m deep tunnels are on increasing. In deep tunneling, it is highly probable that special phenomena such as spalling, rock burst and squeezing are developed due to overstress in the ajacent rock mass. These are the major risk factors reducing tunnel stability significantly. For the successful completion of deep tunnel projects, therefore the possibility and magnitude of these phenomena should be reasonably predicted, and appropriate measures should be established at the planning stage. Special phenomena occurring in overstressed rock masses are divided into two categories: sudden brittle failure such as spalling and rock burst, and squeezing in which large deformation occurs slowly. These have been studied through peculiar models respectively. Therefore, there was no choice but to perform a numerical analysis with one model of the two categories in order to analyze and evaluate the behaviour of rock mass around a deep tunnel. In this thesis, by adopting the complex behaviour numerical model combining the damage threshold and the brittle limit based on Hoek-Brown criterion, two categories of phenomena was allowed to be analyzed and reviewed at the same time. Among factors governing behaviour of rock mass in a great depth, the depth of a tunnel, the maximum principal stress ratio, uniaxial compressive strength of intact rock, a material constant 𝑚i and GSI were selected as basic parameters which can be easily estimated in a data-limited condition and have scalability to extract other required parameters for the study. A numerical parametric studies was performed for 1000 cases which consist of parameters generated through the Monte Carlo method with the condition of depth from 500 m to 3000 m. By analysing the results, thresholds triggering brittle failure and squeezing were formulated as equations with the basic parameters. And correlations between the maximum tangential stress ratio to the strength of rock mass extracted from the basic parameters and the tunnel closure, the plastic radius and the depth of brittle failure were induced. In addition, IRCode, a five digit code system, named the in-situ rock mass condition code was contrived for conveniently describing the condition of in-situ rock masses. And using them the in-situ rock mass condition index, IRC was developed and the practical use possibility of it was indentified throungh application in the correlation analysis of the resuls of the numerical parametric study.; Thanks to development of tunnel construction technology and demands for transportation, power generation and water for irrigation, various projects of over 1000 m deep tunnels are on increasing. In deep tunneling, it is highly probable that special phenomena such as spalling, rock burst and squeezing are developed due to overstress in the ajacent rock mass. These are the major risk factors reducing tunnel stability significantly. For the successful completion of deep tunnel projects, therefore the possibility and magnitude of these phenomena should be reasonably predicted, and appropriate measures should be established at the planning stage. Special phenomena occurring in overstressed rock masses are divided into two categories: sudden brittle failure such as spalling and rock burst, and squeezing in which large deformation occurs slowly. These have been studied through peculiar models respectively. Therefore, there was no choice but to perform a numerical analysis with one model of the two categories in order to analyze and evaluate the behaviour of rock mass around a deep tunnel. In this thesis, by adopting the complex behaviour numerical model combining the damage threshold and the brittle limit based on Hoek-Brown criterion, two categories of phenomena was allowed to be analyzed and reviewed at the same time. Among factors governing behaviour of rock mass in a great depth, the depth of a tunnel, the maximum principal stress ratio, uniaxial compressive strength of intact rock, a material constant 𝑚i and GSI were selected as basic parameters which can be easily estimated in a data-limited condition and have scalability to extract other required parameters for the study. A numerical parametric studies was performed for 1000 cases which consist of parameters generated through the Monte Carlo method with the condition of depth from 500 m to 3000 m. By analysing the results, thresholds triggering brittle failure and squeezing were formulated as equations with the basic parameters. And correlations between the maximum tangential stress ratio to the strength of rock mass extracted from the basic parameters and the tunnel closure, the plastic radius and the depth of brittle failure were induced. In addition, IRCode, a five digit code system, named the in-situ rock mass condition code was contrived for conveniently describing the condition of in-situ rock masses. And using them the in-situ rock mass condition index, IRC was developed and the practical use possibility of it was indentified throungh application in the correlation analysis of the resuls of the numerical parametric study.