Click on the image to download a larger version.
This simulation (the last one of its kind ran on the CM-5) was performed to demonstrate the ultimate capabilites of our numerical software and tools in performing these kinds of simulations. It is of extremely large scale and took quite a long time to perform. The simulation is of 1000 spheres sedimenting in a liquid-filled tube. More information about this simulation will be include here when the simulation is actually completed.
My entirely new (MPI-based) fluid-object interaction flow solver is expected to be up-and-running on the new T3E-1200 (the replacement to the CM-5) sometime in February. This new solver will be much faster, handle larger-scale problems, more accurate (reduced projection errors), easier-to-use, more efficient, and more portable than the CM-5 solver. I hope that with this new solver, the "demonstration" phase of this work can be completed, and actual important large-scale applications will be solved.
Last updated 12-30-97