Estimation of Accuracy of Force-Component-Ratio-based Material Differentiation on a 3-Dimensional Scale on Fluorite-Barite-Concrete-Samples

Research output: ThesisMaster's Thesis

Authors

Abstract

Drilling and Blasting (D&B) is the most common technique used for the excavation of hard rocks. Nevertheless, with the expanding of urban areas, D&B operations get limited due to strong vibrations, noise, blasting fumes and rock disturbance. To avoid those problems, mechanical excavation is an alternative technology. Mechanical excavation provides the opportunity of selective, continuous and autonomous excavation. Optimization on selectivity of mechanical excavation means less dilution of ore, less energy consumption and less need of mineral processing. Material differentiation during the mechanical excavation can provide selective mining. In this research, the force component ratio based material differentiation is investigated. Three fluorite-barite specimens were poured with concrete to create blocks for the cutting tests. The cutting tests were carried out using the full-scale linear cutting test rig HXS 1000-50 in the laboratory of TU Bergakademie Freiberg. With the photos of layers, 3D photogrammetric models are created in MinePlan software. Cutting force data are filtered and processed in Voxler software to create five force component ratios (FCR) and these data 3D modeled in MinePlan software. Optimum thresholds are evaluated. Photogrammetric models and FCR models are statistically compared by using the confusion matrix. For Block 1, Fz/Fx-Interquartile Range indicator shows the most accurate results (accuracy of 0.77, sensitivity of 0.77, specificity of 0.77). For Block 3, Fx/Fres-Aritmatic Mean indicator is better compared to other FCR models (accuracy of 0.81, sensitivity of 0.31, specificity of 0.98).

Details

Translated title of the contributionAbschätzung der Genauigkeit der Kraft-Komponenten-Verhältnis-basierten Materialdifferenzierung auf einer 3-dimensionalen Skala an Fluorit-Barit-Beton-Proben
Original languageEnglish
QualificationMSc
Awarding Institution
Supervisors/Advisors
Award date18 Dec 2020
Publication statusPublished - 2020