Animation of a 2-D tornado simulation

This page contains a new rendering of a simulation from my article from 1994:

These simulations are conducted in what could be called the "Fiedler Can":

The black contours show the fixed buoyancy field confined to the vicinity of the central axis of the can. This image also shows the vertical velocity early in the integration, at t=5. A vortex ring is migrating toward the outer wall. A central vortex has not yet "spun up", and the central vertical velocity has stabilized with maximum value of 0.78. The formal "thermodynamic speed limit" is 1.0. The figures below will show the domain only with 0<r<0.5.

Here is a noteworthy suction vortex simulation from the 1994 article.


u and p are the minima in the radial velocity and the pressure. v and w are the maxima in the azimuthal and vertical velocity. The Fortan simulation is with the code used for the 1994 article. The Fluent simulation is with the commerical software Fluent.

The simulation with Fluent 6.3

Note Fluent has a bug: u and v do not go to 0 on the rotation axis. This may be inconsequential. u and v seem to be asymptoting towards 0. Perhaps the nonzero values on the axis do not feedback into the solution.

t=25.2, the time of maximum intensity

Below are some shots from the Fluent simulation at t=25.2, the time of the maximum intensity in the suction vortex. These features compare well with the t=24.9 in the above Fortran simulation.