An alternative method for pillar stability analysis based on machine learning assisted geomechanical simulations

Date

2023-8

Author

Duman, Dilan

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

129
views

54
downloads

Cite This

Coal reserves remain to be one of the major sources to provide the global energy demand. Weak rock characteristics make instability a vital concern in terms of occupational health and safety in coal operations. As a support element, pillar is a critical structure in an underground mine plan. Although various methodologies have been developed for designing a geotechnically stable pillar, structural and geological variations may be misleading. This study makes use of a 2D plane-strain based numerical modeling scheme to investigate the effects of variations in geomechanical properties, field stress ratio, coal seam angle, and mine depth. Numerical outputs like horizontal and vertical displacements, yielded element thicknesses, percentage of yielding in pillars, and maximum principal stress in abutments will be used to establish correlations between the control variables. The aforementioned variables are acquired through finite element analysis and documented. Furthermore, the yielded span percentages are computed and recorded for each individual pillar encompassed within the analysis. Through this approach, a comprehensive database is established. Besides the conventional statistical methods, a supervised machine learning model was trained using the database and an alternative method for analysis of stability in underground mine pillars is proposed. The study outputs have potential to provide a safe and user-friendly scheme for inexperienced technical staff assisting the conventional design methods.

Subject Keywords

Coal Mining, Pillar Stability, Abutment Pressure, Numerical Analysis

URI

https://hdl.handle.net/11511/105354

Collections

Graduate School of Natural and Applied Sciences, Thesis