Abstract
Minalyzer CS is a scanner which in a contactless non-destructive way generates geochemistry, high-resolution images, rock quality designation (RQD), structures, specific gravity and bulk density for drill cores and other drill samples.
The patented scanner is designed for handling large volumes of drill samples and is capable of scanning drill cores directly in core trays. A laser (LiDAR) generates a 3D-model of the topology of the core and trays, enabling the control and precision of the continuous XRF scanning. RQD and structures are also derived based on the 3D-model.
The objective, continuous and consistent nature of the datasets as well as the high data density generated by the scanning technology is paramount in machine learning and deep learning applications and approaches to geology. Machine learning and deep learning have been demonstrated to be effectively used, based on the data from the scanning, for prediction of host rock lithologies.
A cloud-based software www.minalogger.com for visualisation and generation of datasets through digital tools facilitates remote access to a digital twin of the sample. Remote access to data has become critical in order to keep project and operations moving forward when travel has become impossible and/or risky due to the pandemic.
The bulk density can be derived based on measured volume from LiDAR scan of the Minalyzer CS, combined with the weight of the core tray. The method is suitable for friable sediment core where a true representation of the friable or heavily fractured sample through manual measurements and estimates can be error prone. The new method has for a time been tested and applied in live application by Iron ore companies in Western Australia where extensive comparisons between the new method and the traditional have been made. The method has also been tested on known volumes and densities for verification and demonstrate both a high level of repeatability and accuracy.