The main findings of the joint research initiative "Numerical Simulation in Tunnelling" and a follow-on project on efficient numerical simulation models for tunnelling are presented.

Die wichtigsten Resultate des Forschungsschwerpunkts „Numerische Simulation im Tunnelbau" und einem Folgeprojekt ueber effiziente numerische Simulationsmethoden werden prasentiert

Les resultats les plus importants du point principal de la recherche "simulation numerique dans la construction de tunnels" et un projet consecutif sur des methodes de simulation numeriques efficaces sont presents

Introduction

Numerical simulation methods can be applied to determine the tunnel construction method/sequence and ground support in order to ensure stability or reduce settlements. They can also be used for assisting the decision making process during the construction, for example in the case of a tunnel being driven with the New Austrian Tunnelling Method (NATM), where support requirements may be adjusted according to prevailing geological conditions. A survey of the state of the art, carried out about 5 years ago, identified a number of shortcomings of numerical simulation methods used for tunnel construction. These shortcomings related to the collection of input data (geological survey), the efficiency / user friendliness of the numerical methods and their limitations. As a consequence an Austrian joint research initiative (JRI) was established, where all major research groups working at three Universities in Austria were networked. The JRI, which has been supported by the Austrian national group of the ISRM, finished at the end of 2002 with an international review. The research findings have been published in book form by Springer (Beer, 2003). In a follow-on project to the JRI we are investigating the feasibility of using Boundary Element methods for modelling the sequential excavation of a tunnel in an visco-plastic rock mass. This method has the advantage over the FEM that the mesh is reduced by one dimension (surface instead of volume elements) and that no truncation error is introduced. First results, which will be shown in the paper, are encouraging and we are confident that in the future three-dimensional non-linear analysis of tunnel advance will be performed in a very short time on a laptop PC.

Input data

The accuracy of numerical simulation predictions relies heavily on the quality of the input data. In tunnelling the input consists of the description of geological features, of the material behaviour, of the tunnel shape, the pre-excavation stress field, the excavation/construction sequence and the characteristics of the ground support.

Geological features

The description of geological features is important since these may influence ground behaviour significantly. Very often we have the situation that the geological information is patchy because only a limited exploration was carried out due to cost reasons. Geological exploration methods range from core drilling to seismic methods. Sometimes exploratory tunnels are constructed. A way of getting geological information is the observation of the tunnel face after each excavation step. Usually this is done by an experienced geologist making sketches after each blast. This method is not only dangerous because the face is still unsupported but also fairly subjective.

This content is only available via PDF.
You can access this article if you purchase or spend a download.