Paper No. 5
Presentation Time: 2:40 PM
LATTICE-BOLTZMANN MODELS OF SPELEOGENIC PROCESSES
WALSH, Stuart, Department of Geology and Geophysics, University of Minnesota, 310 Pillsbury Drive SE, Minneapolis, MN 55455-0219, ALEXANDER, Scott, Geology and Geophysics, Univ of Minnesota, Pillsbury Hall, Minneapolis, MN 55455 and SAAR, Martin, Department of Geology and Geophysics, University of Minnesota, 310 Pillsbury Dr. SE, Minneapolis, MN 55455, sdcwalsh@umn.edu
Lattice-Boltzmann simulations are capable of modeling a wide variety of complex fluid mechanical problems that may be difficult or impossible to simulate with other modeling methods. For example, lattice-Boltzmann simulations can reproduce complex boundary geometries, miscible and imiscible fluid flow, and buoyancy-induced convection due to solute and thermal gradients. This makes these methods particularly attractive for a wide range of geofluid-mechanical applications, including simulation of speleogenesis. This presentation will discuss two separate lattice-Boltzmann models for reproducing speleogenic processes: the first is an explicit method that simulates the formation of features on the cave scale; while the second is an implicit method capable of simulating development of larger-scale systems.
The explicit method employs a solute-induced buoyancy model that captures host rock dissolution and subsequent solute transport. While useful for simulating the development of small-scale features, the computational demands of this explicit method restrict the large-scale application of this approach. Instead, we simulate the development of larger-scale features using an implicit method based on a partial-bounceback model. This model employs a probabilistic method to simulate host rocks of varying permeability and is particularly suited for systems with rapidly varying heterogeneities and anisotropies.
Applications of these two models will be discussed in the context of reproducing small-scale features and large-scale karst permeabilites. Better models of the formation of these features will result in more accurate predictions of contaminant transport through these systems.