This work involves simulating the processes of flow and transport in porous media with consideration for geochemical and biochemical reactions. The research is being performed in cooperation with researchers at the Engineering Research and Development Center in Vicksburg, Mississippi.
Abdelkarim Abulaban, Research Associate
Ronnie Daanen, Graduate Student Researcher
Anita Gruber, Graduate Student Researcher
Debasmita Misra, Research Associate
Cam Nguyen, Research Associate
Hung Viet Nguyen, Graduate Student Researcher
Paul Oduro, Graduate Student Researcher
Two specific problems are being addressed. The first deals with the characterization of the dispersion process based on the various levels of spatial heterogeneity that exist in natural porous media. Solute transport is simulated using a particle method in which the solute mass is represented by a very large ensemble of discrete solute parcels (or walkers) that move in the flow field. Both a deterministic step and a random step are imposed to define the parcel pathway. Applying this method, these researchers are characterizing the velocity covariance function (and apparent dispersion coefficient) that results from the transport of a conservative (non-reactive) solute within a steady-state heterogeneous flow field. Relating the velocity covariance function (and apparent dispersion coefficient) to measures of heterogeneity will be of value to practitioners. From this work, a low-resolution model is being developed to simulate solute transport at a much smaller computational effort than required for high-resolution simulations. The low-resolution model will be parameterized using the velocity covariance data derived from the high-resolution simulations.
In both of these problems, it is necessary to perform simulations on highly resolved computational domains, with upwards of 106 grid cells to define the flow field for two-dimensional domains. Three-dimensional flow fields will have much larger requirements. According to the theory for the particle method, the larger the number of parcels, the more accurate the solution. Exact solutions require an infinite number of particles, but in practice the number of particles can be much less, depending on the particular problem being simulated. These researchers have found that a particle number on the order of 106 is generally sufficient. The requirements to perform computations on these highly resolved grids with large numbers of particles naturally leads to the use of the computational resources available at the Supercomputing Institute.
|
|
URL: http://www.msi.umn.edu/about/publications/annualreport/ar2000/depts/AgFoodEnvSci/nieber.html |
|
| This page last modified on Friday, 30-May-2008 16:14:01 CDT | ||
| Please direct questions or problems to help@msi.umn.edu | ||
|
Website related questions or problems should be directed to
webmaster@msi.umn.edu
The Supercomputing Institute does not collect personal information on visitors to our website. For the University of Minnesota policy, see www.privacy.umn.edu. © 2001 by the Regents of the University of Minnesota |
||