Speaker
Description
A high resolution muography survey has been performed in the Királylaki tunnel in Budapest to search for unknown cavities. Preliminary radiographic measurements suggested large density anomalies above the tunnels in a 20--60 m thick cherty dolomite rock. Triangulation helped estimating the location and size of the anomalies, but for excavation, more precise information was needed.
A Bayesian inversion method has been adapted to overcome the underdetermination originating from limited-angle tomographic nature of muography, and it was verified by the muographic measurements and drill samples. A filtering is applied based on back-projected weights to control uncertainty propagation. Relevant Bayesian assumptions made the inverse sufficiently stable, and the stability has been tested by parameter sensitivity analysis and simulated case studies.
Measurements have been performed along a straight line due to the constraints of the tunnel system. Therefore, the underdetermination is reduced by slicing the space to planes intersecting the measurement line, and inversion simplified to 2D reconstructions on multiple tilting planes. The angular resolution of the CCC (Close Cathode Chamber) based gaseous muon detectors enabled a spatial voxel resolution of 1-2 meters, and the 3D distribution of previously unknown karstic crack zones has been obtained for the first time by muography.
The closest points of the density anomalies could be determined which was reachable by a simple core drill technique. Three 5--10 m long core drills validated the existence of the density anomalies. The core samples showed convincing agreement with the inversion, which didn't contain cavities but a low density fractured rock (altered dolomite powder). This result shows the potential of mapping underground crack zones with muography which has possible applications for tunnel construction and maintenance, or hillside-slip risk assessment, among others.
