RSR
Platinum Member
Pics or it didn稚 happen.
I post these here with the following requested stipulations. Please do not repost/share/etc. These are ongoing investigations.
Shown are snapshots (sequentially labeled A - D), extracted from a video of 2-D CFD simulations of the unsteady (varying in time) particle transport during production of an aspirated f sound. (Think of how you make an "f" sound when saying the word "huff". That is breathy, as opposed to a stop consonant, as when you make an "f" sound when saying a word like "far"). Unfortunately, the whole video (much more interesting) is waaaay too large to upload here. The simulation progresses over a time of 0.4 seconds. The simulation was run with an inlet pressure boundary condition of 600 Pa. This is an estimate of our lung pressure (what forces air out) during comfortable speech. The mouth geometry (small stuff on the left) is an idealized geometry of the vocal tract cross-sectional area when saying "f". The dots are the particles in the flow, which for this early work, was just a uniform distribution between 1 micron (10^-6 m) and 50 micron (5 x 10^-5 m), colored blue to red, respectively. The particles were inserted at the back of the mouth with no initial velocity to replicate them being shed (i.e., produced) at the vocal folds (cords). Velocity fields are not shown, but would be nice to superimpose so you can see both the velocity magnitudes and particle motions at the same time. (Note to self: Tell post-doc to do that).
What does it show?
1) Larger particles lag the smaller particles as the flow accelerates. (Not surprising)
2) As the flow progresses, a vortex ring is produced at the exit of the mouth, which propagates at the leading edge of the puff of air. This is important, and arises for many speech utterances. Vortex rings (think like a smoke ring you can blow) are an annulus of rotating fluid that can self-advect as a coherent structure for very long distances. If you go to youtube and search for videos (e.g., vaping smoke ring) all of the cool kids that vape practice blowing smoke rings. These videos are actually great for teaching fluid dynamics, as they clearly show how once formed, a vortex ring will stay intact, and propagate very long distances. In the video, you can see the rotation of the vortex rings, but these are still snapshots, so ... I've denoted where they are with the overlaid red vectors.
3) Notice, the smaller droplets (that are more likely to penetrate deep into your lungs and infect you) become entrained in the rotational part of the flow. This is kind of like looking at a river and noticing a small, light leaf on the surface of the water will follow all of the little eddies and swirls. A large log, in turn, has enough momentum, it just follows the path of the main river. The droplets do the same thing. The bigger ones (red) follow a largely-ballistic trajectory (straight out, slowly falling to ground). The small ones get entrained in, and follow the smaller eddies/currents.
4) The domain size (from mouth exit to right edge) is 0.75 m (~2.3 feet) It's a pressure-release boundary condition on the right hand side, meaning, it's as if the room keeps going (Particles can pass through the surface) but the simulation stops there. Even 2 feet after leaving the mouth, the particles are nowhere close to slowing down, and the larger ones have maybe just started to settle. It is expected that the larger droplets will eventually fall to the ground, but the smaller ones will stay entrained in the vortical structures of the flow. (Again, watch youtube. Vortex rings can propagate the length of a room). But, those results need to be confirmed with simulations over a larger domain, and with realistic background flow conditions/turbulence intensities. It takes MUCH longer to run larger simulations, so initially we are making sure the particle dynamics look right, and benchmarking the velocity fields with the experimental particle image velocimetry (PIV) measurements in steady flow. I'd show the PIV measurements, but they're pretty boring. In steady flow, it's basically just a slot jet, so nothing too interesting. But, it's important to be able to compare the velocity predicted by the CFD measurements with the experiments to make sure they match, before proceeding.
View attachment ParticlesTrajectories.jpg
