Head-on vortex ring collision 

It appears the shadow itself is a projected Oct tree But it's incorrect You'd have to have a very specific circular light to even get it close. It hasn't diffused enough per octal , and it's as though the dye has no more viscosity I think because the dye is a collection of particles, without any mesh and without that it cent cast shadows, And so, thusly the separate projected octtree. That is the shadow itself An approximation. Cuda cores. I'm sure that's where the magic sits

balls

Debussy got me acting unwise

Yes!

Successfully simulating this seems like a pretty major achievement!

@SmarterEveryDay

in his video i think he said this cannot be done in 2D simulation because there's vertical interactions or something so this isn't trivially done stuff on computers at all

yes

Such hydrodynamic instabilities are characterized by a so-called `fasted growing mode', which determines this magic number (e.g. here its 16). For this setup, this mode in governed by the Reynoldsnumber (Re, see top right), and higher Reynolds numbers (i.e. relatively larger primary vortices) give more secondary vortices, upto a point where these subvortices interact which each other to generate a less organized turbulent behaviour. If you are still interested: Apart from the most unstable mode, there also needs to be a disturbance that triggers the instability. In this simulation, this is likely to stem from the crude implementation of the injection phase.

The shadow☠️

You may a look in my PhD thesis: repository.tudelft.nl/islandora/object/uuid:e45cea45-8915-4a11-b8fd-389cb3e19d22 But I doubt that it provides more info than (the links in) the description.

@@zemmelzoltan thanks a lot

16 is indeed a suspicious number for a mesh with cubic cells. However, the instability is bound to create an integer number of secondary rings and this number depends on the relative size of the vortex rings (quantified via the Reynolds numbers), but also some initial perturbation that may be present in the primary rings. So the answer to your question: Maybe, but 16 also happens to be a reasonable number for this setup.

The vortex rings have a special property that they entrain a volume of fluid with them, this volume is said to be "inside the separartix", where the separatrix is an imaginary surface thatencloses the entrained volume. At the start locations you see a collection of tracers that were not entrained by the initial vortex rings, as these were placed outside the separatrix.

Your question is indicative of a misconception regarding the role of the particles in the simulation: The quadtree (i.e. octree in 3D) grid structure is used to solve for the evolution of the the flow field. This approximates the continuous vectorfield of fluid velocities. Indeed, this simulation ignores the exact effects of individual fluid element (molecules) collisions. The particles are there to mimic dye, by passively advecting with the flow field. There is no interaction between these particles. The apparent interaction is a result of the resolved vortex dynamics. I hope it helps...

​@@zemmelzoltan I don't understand everything in your answer but I think I get the idea. I thought it's a silmulation of particles and their interactions but it's a numerical approximation of a vectorfield (flow field?). Thanks for the explanation :)

Bruh quantisation? You could just ask what defines the number of the secondary rings😂. Sorry, not to sound agressive, but after reading science books for my studies such overcomplicated descriptions just trigger me

💀

Hallo David, Thank you It is a visualization of a computer simulation of fluid flow. The flow configuration mimics the famous vortex ring lab setup of Lim & Nickels. The movie shows 400000 flow tracer particles and a slice of the adaptive grid structure that was used for the computation of the flow evolution. The computations can take a while (about 100 core hours). But you can attempt reproduce/improve it as all the finer details are in the link in the description.

@@zemmelzoltan Thanks, keep up the great work!

@@zemmelzoltan What did you use to make it?