Lagrangian vortex modeling is essentially a method where the model follows the fluid. I've been playing with vortex models attempting to arrive at a 'stable' configuration with a multivortex tornado. This is like in the 2d applet on the right, except this is in 3d. A large issue I'm running into is that when the vortex lines become concentrated, that is when the vortex strengthens, this causes the lines of vorticity to tilt above. The tilt induces a down draft which eventually dissipates the strengthening at the surface. The tilt I believe is similar to the complex affect of vortex break down. For example, in dust devils it is sometimes observed that there is a down draft at the top of the dust devil where the concentrated rotation dissipates.
Below are some examples of simulations I have produced. The lines represent locations where vorticity is concentrated.



