Loading...
CFD modeling of methane flame, turbulence, and obstacle interaction applied to a longwall coal mine
Nguyen, Thu
Nguyen, Thu
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2020
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The formation of explosive gas zones (EGZs) forming from flammable vapors, gases, or dust pose safety hazards to many industries. An EGZ ignition could occur by faulty electrical equipment, hot streaks from worn bits on a shearer cutting hard rock, rock-on-rock frictional sparks, heat, or fires and result in an explosion, and/or detonation. Agriculture, oil and gas, chemical, and energy industries experience methane-air or organic dust explosion hazards. In many cases, explosions may occur in confined areas with obstacles in the path of flame expansion. By studying the effects of obstacle shape and size on flame propagation and turbulence, a more complete understanding of the interaction of the flame and fluid dynamics has been achieved. Obstacle shape, turbulence model, and spark location were investigated using a single obstacle and flame interaction in the model. Reynolds Averaged Navier Stokes (RANS) models were tested to determine if these simplified turbulence models could capture the flame dynamics and propagation velocities using fewer computational resources compared to the higher fidelity Large Eddy Simulation (LES) turbulence model. Obstacle shapes were varied to examine the impact of shape on methane flame propagation. Results showed that square obstacles caused faster flame propagation around the obstacle compared to hexagons and circles. The square had an average flame speed 26% faster than the circle, and the hexagon was 16% faster than the circle using a k-ω model. Modeling results indicate that variation of the spark location by as a distance as small as 10% of the obstacle diameter can result in a significant difference in flame propagation velocity. Comparing to RANS, an LES turbulence model only increases computational time by 25%. Therefore, the LES turbulence model was used in modeling a simplified 2D longwall mine combustion model. Achieving reasonable computational times while maintaining general flame propagation trends was the main goal of modeling an explosion in a full-scale longwall mine. Investigating different methods of modeling the gob inside the mine showed that a porous media model would not be able to capture effects from turbulence or handle a reacting flow due to a Darcy Flow assumption. As a result, the gob was modeled as a fluid zone with discrete obstacles. Results from the 2D mine model with combustion overpredicted the pressure compared to a 3D model but was still able to track the propagation of the pressure front well when compared to a 3D model. By simplifying the model to 2D, computational time was reduced to three days, compared to three weeks for the 3D model, to simulate a 35ms interval after ignition.
Associated Publications
Rights
Copyright of the original work is retained by the author.