Nykirke-Barkåker is one of the railway sections where a double-track high-speed line is being built on the Vestfold line (south west of Oslo). To make room for a new double-track, a rock slope close to the existing track must be adjusted or removed by blasting. Prior to blasting, NGI in collaboration with Geovita was responsible for creating a 3D rock model of the slope, including visualization of the cracks and assessment of risk associated with blasting.

Stability assessment and mapping of cracks

Before blasting the rock slope, a stability assessment is necessary. This involves identifying the fractures and cracks in the rock mass, which in turn helps to identify potential unstable rock blocks, and thus assess the stability of the slope.

NGI and Geovita have used a comprehensive digital workflow to solve this task, with an increasing degree of R&D during the project. The work mainly consisted of the following steps:

  1. Data collection using drones, LiDAR scanners and field measurements
  2. Processing and processing of point cloud (collected data) using software (CloudCompare and Maptek Pointstudio)
  3. Crack detection on 3D model in Maptek
  4. Visualize cracks and potentially unstable blocks in Rhino3D / Grasshopper

Data collection and 3D modelling

To produce a complete point cloud over the entire slope, both photogrammetry (drone) and LiDAR scanning were used. Lots of vegetation and an existing safety net made it challenging to get a complete point cloud only with the help of a drone. Complemented by LiDAR scanning and manual filtration of vegetation, the resulting point cloud became better suited for further treatment. Field measurements of the fracture orientation were also performed in order be able to verify the quality of the fracture classification at a later stage.

Before the point cloud could be transformed into a 3D model, it was processed using the CloudCompare and Maptek Pointstudio software. Here, filtering of point density and noise was performed, and unnecessary parts of the point cloud were removed, for example overlapping parts from the two data collection methods.

The point cloud was then transformed into a mesh, and then assembled into a 3D model. Then, a semi-automatic crack detection was performed using Pointstudio, where similar cracks were grouped into crack sets. A kinematic analysis was performed in Pointstudio, and the risk-exposed cracks were identified.

The cracks were imported into the software Rhino3D / Grasshopper, where one could identify potential unstable blocks using the 3D model and the crack sets, and calculate the volume and weight of the blocks.

3D bergmodel og sprekkegjenkjenning av Bollerudskjaeringen 700

3D rock model and crack recognition of the rock slope

Information flow and improved understanding

From ongoing research at NGI, work is being done to develop a comprehensive automatic workflow, which will facilitate the task of transforming collected field data into a 3D visualization and end up with a recommendation of necessary bolt force and design to secure individual blocks. This will also contribute to a simpler flow of information and an improved understanding for all parties involved in the securing process for construction in rock.