Accuracy of GPR-based 3D Fracture Surface Geometry Interpretation
Abstract
In the past decade, various nuclear waste management organizations have put significant effort into developing methods that can identify and characterize excavation damaged zone (EDZ) around hard rock excavations. Much of this research has been centered around the use of ground penetrating radar (GPR) to detect fractures parallel to the tunnel profile in the first metres below the excavated surface. Accuracy of the EDZ interpretation is largely affected by the capabilities of GPR to identify said fractures, and to establish their orientation. In order to verify the performance of GPR in this context, a set of tests was carried out on specimens of mostly homogeneous Kuru grey granite from Kuru, Tampere, Finland.
Three slabs of 1 x 1 metre in size and with a thickness of approximately 400 mm were sawcut from intact rock blocks of Kuru granite. These slabs were carefully split approximately at the middle while avoiding excessive damage to the sides of the specimens. This resulted in three specimens with an artificially induced fracture surface. These specimens were then measured using a GPR antenna with a central frequency of 1600 MHz. For each specimen, a total of 18 scanlines (9 in the x-direction and 9 in the y-direction) were measured with a line separation of 10 cm. Along the scanlines a total of 200 traces/m were recorded with 2048 samples per trace. This data was then processed to create 2D profiles of the fracture surface location in respect to the specimen surface. When converting from time to distance, an observed median value of 4.8 was used for the relative dielectric permittivity of Kuru grey granite. After the GPR measurements, the specimens were carefully opened by lifting the top part of the slab. The true geometries of the fracture surfaces were then reconstructed using photogrammetric modelling. A series of photographs were taken using a DSLR and converted to point clouds using VisualSFM software. Further processing and scaling of the point clouds was done using Cloud Compare software. Finally, the results from GPR measurements were compared to the photogrammetrically obtained fracture surface geometry to establish the accuracy of the GPR based fracture surface geometry interpretation.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMNS43D0861K
- Keywords:
-
- 0994 Instruments and techniques;
- EXPLORATION GEOPHYSICS;
- 0999 General or miscellaneous;
- EXPLORATION GEOPHYSICS;
- 1835 Hydrogeophysics;
- HYDROLOGY