SensIT – Verifiering och prognostisering av tekniska funktionskrav på tunnelinfattning av betong – sensorbaseras prognosmetod med artificiell intelligens

Research Project, 2021 – 2024

The project analyzes how sensor-based monitoring can be used to identify and track loads and damage states in shotcrete linings in rock tunnels. The study adopts a structural health monitoring perspective in which Distributed Optical Fibre Sensing (DOFS) is combined with experimental studies, numerical modeling, and data analysis to enable continuous verification of the structural condition.

The project includes an experimental series in which typical load cases in rock tunnels were reproduced at laboratory scale. The test setup consists of two concrete layers, where a fibre-reinforced concrete layer simulates the shotcrete lining and an underlying concrete layer represents the rock mass. Loads corresponding both to local block loads and to distributed loads caused by loss of adhesion between the rock and the shotcrete have been applied. All specimens were instrumented with distributed fibre optic sensors to record strain fields in the structure.

The results show that distributed fibre optic sensing can capture both global deformation and local damage processes in shotcrete linings already at relatively low load levels. Strain patterns and crack development can be identified and linked to different load types and failure mechanisms. The experiments also show that the adhesion at the interface between the shotcrete and the underlying material has a major influence on the failure process and residual load-bearing capacity. Differences in the surface preparation of the substrate can lead to clearly different failure modes and load-transfer mechanisms.

The project results provide an important foundation for the development of advanced tunnel monitoring systems, where experimental data can be used to calibrate numerical models and create synthetic databases for machine learning-based analysis. In the longer term, this enables the development of digital tools for load identification, early damage detection, and prediction of the structural condition.

The scientific results of the project are presented in several publications as well as in a licentiate thesis, in which experimental studies of loads and failure mechanisms in shotcrete linings are reported.