wy92

Spring 2009 Mid-Semester Contributions

During the first half of the Spring 2009 semester, I completed/started the following simulation experiments:

• Resolved the discrepancy between 2D model and 3D model with periodic boundary conditions
• Completed preliminary simulations using Reynolds stress model in 2D settings
• Created a finer 3D mesh and started simulation with k-epsilon realizable turbulence model
• Set up an account on Swanson SGI server for parallel computing and figured out the LINUX configuration(still need to figure out the appropriate command for parallel computing)

Fall 2008 Contributions

During the second half of Fall 2008 semester I changed the project from modeling sedimentation tank to 3D simulation of the flcculation tank. A series of preliminary simulation were carried out in order to have an idea of the scope of the problem and how to approach it. The preliminary simulations could be catergorized into two topics: investigation of convergence in 3D systems, and comparing 3D results with 2D results to understand the disctinction of the key underlying assumptions behind these models and the validity in terms of closenes to the real flow.

Based on the results and reflections from preliminary simulations, I came up with a clear road map about how to do this project to best contribut to AguaClara.

   During the first half of Fall 2008 semester I have been working on the CFD team. I'm specifically working on simulating the flow field in the sedimentation tank to optimize the design. I spent the first week understanding the problem I'm going to solve (Problem Specification) and doing trade-off thinking on whether to  adopt computational or experimental approach. I also learned to document my work with wiki. During the next 5 weeks, I completed a "quick-and-dirty" tutorial of CFD and FLUENT. I learned about the underlying fluid mechanics and numerical methods of CFD techniques, and practiced using Gambit and Fluent with basic simulation problems. And now I'm ready to jump in and create a first mesh. I will start from simulating of  2D cross-section, and will work towards better approximation by iterations of trials.

So far, I can come up with more detailed semester goals which are both challenging and feasible, and make more sense. 

By reasonably reducing the 3D flow field to 2D and further decomposing the 2D cross-section, I now have a starting point to create the first geometry and the first mesh.