Tunnel Static Behavior Assessed by a Probabilistic Approach to the Back-Analysis

Abstract

Problem statement: The steps of the validation procedure of the project of a tunnel are briefly illustrated in this study, starting from the geological and structural surveys on the excavation walls and the measurement of the physical and mechanical parameters during the excavation works. Unfortunately, however, the knowledge of the rock mass, which is fundamental to the project, is usually approximate before the study is started. This knowledge improves considerably once construction of the tunnel is started, when it is possible to have direct access to the rock and analyze its behavior in relation to the excavation and support works. Approach: The measurement of displacements and stresses in the rock mass and in the support structures represents a different methodology for the evaluation of the Geotechnical characteristics of a rock mass and therefore also of the support work conditions. To correctly interpret the measurements it is necessary to make use of a more complex procedure, called back-analysis, that, starting from an estimation of the unknown parameters of the rock mass obtained through a preliminary characterization, integrated and modified by sampling of the rock mass during the construction stage and by the performed stress and displacement measurements, is able to define the unknown parameters of the rock mass. Results and Conclusion: Back-analysis in engineering in the rock field occurs, however, in an uncertainty context, which complicates the problem. The preliminary estimation of the Geotechnical characteristics of the rock mass has in fact a degree of reliability that is a function of the intensity of the preliminary investigations. The performed measurements present a certain precision in relation to the various typologies of error that can occur. The final result of the back-analysis therefore also consists in the definition of the Geotechnical parameters of the rock mass that are considered to be of influence in the problem under examination, with a certain reliability and precision that is obviously greater than that relative to the initial estimation of the same parameters. The purpose of this study is to present a global approach to back-analysis in a probabilistic context that is aimed to obtain a reliable calibration of the parameters of the rock mass that are necessary to study the behavior of the support structures.