Related Experiment Video
Updated: Nov 29, 2025
10:24
Neutron Radiography and Computed Tomography of Biological Systems at the Oak Ridge National Laboratory's High Flux Isotope Reactor
Published on: May 7, 2021
2.6K
HRSBallast: A high-resolution dataset featuring scanned angular, semi-angular and rounded railway ballast
André Broekman1, Jacobus Oostewald Van Niekerk1, Petrus Johannes Gräbe1
1Department of Civil Engineering, University of Pretoria, South Africa - University of Pretoria, Lynnwood Road, Hatfield, Pretoria 0002, South Africa.
Data in Brief
|November 18, 2020
Summary
A high-resolution dataset of digitally scanned railway ballast (HRSBallast) was created from field and lab samples. This dataset aids granular media simulations, wear modeling, and deep learning applications.
Area of Science:
- Geotechnical Engineering
- Materials Science
- Digital Imaging
Background:
- Railway ballast degradation impacts track performance and safety.
- Accurate geometric data is crucial for simulating ballast behavior.
- Existing datasets lack the resolution and scope for advanced modeling.
Purpose of the Study:
- To present a high-resolution dataset of digitized railway ballast samples.
- To provide a reference for granular media simulations and digital asset creation.
- To capture geometric changes and material loss during simulated wear.
Main Methods:
- Digitally scanned 108 ballast samples using a VSLAM-based scanner (40-micrometre accuracy).
- Collected samples from a South African heavy haul coal line and a local quarry.
- Performed iterative hydraulic actuator tests on quarry samples, scanning before and after each test.
Main Results:
- Created the HRSBallast dataset with diverse geometric features (angular, semi-angular, rounded).
- Captured pre- and post-test geometries, including fractured samples.
- Quantified geometric changes and material attrition from wear tests.
Conclusions:
- HRSBallast provides a valuable, high-fidelity resource for research.
- Enables advancements in discrete element method (DEM) simulations and wear modeling.
- Facilitates the development of synthetic datasets for deep learning in railway engineering.
