Complicated air flow in front of a fan

I started this video to show why a cowling around the fan wouldn't help, but as I experimented, things got more interesting.

[ryk]

Seems like adding cowling could increase efficiency for an exhaust fan where there is concern for what air is being drawn in. There may be applications where you don't want to draw air from the sides, only from the back; maybe a concern of system efficiency versus local efficiency.

[fzigunov]

Aerodynamicist here. Just wanted to help with some points to perhaps improve your understanding of your fan!
1. Yes, Bernoulli principle applies to turbulent flow; otherwise it would be quite useless. It doesn't apply to *compressible* flow, but the air speeds of you fan are way below sound speed (M<0.2), so it does indeed apply to your flow.
2. The reason you get flow inwards as you observed in 1:37 is because of "jet entrainment" (you can google it). Turbulent mixing as well as viscous diffusion draws air from the sides of the fan towards the jet core. This is how "bladeless fans" work.
3. The technique you used in 5:40, seeding the air flow with sawdust, is called "smoke flow visualization" (and usually it's done with smoke from a fog machine). In labs we use cameras and lasers to quantitatively get the motion of the particles, what we call Particle Image Velocimetry (PIV). PIV gives us the entire velocity field from the experiment, like the ones you see in simulations, and is the state of the art. Spoiler alert, when you compare experimental flows with simulations, simulated flows "look right" but the details (that matter) are usually wrong. This is why if you talk to any experimental aerodynamicist, you'll find they're quite incredulous of CFD results.
4. The probe you "mocked up" at 6:10 is what's called a "vorticity probe" and yes, it captures the "spinning" motion of the fluid. A fan produces a swirling jet downstream of it, as you observed. That's quite expected, though. We don't use vorticity probes anymore as Particle Image Velocimetry gives so much more data.
5. The air "sticks" to the floor at 6:40 because of Coanda effect. The jet of air from the fan sucks air from the sides (entrainment effect), but the presence of the floor acts as a wall and creates a lower pressure zone, which "sucks the jet" towards the floor.

All in all, it's really cool to see you're interested in this topic. Doing experiments like that is a lot of fun!!

[JKDangerfield]

Hey Matthias, I believe you are seeing a venturi effect in front of that tube where as the flow across the tube opening causes a lower pressure within the tube. The faster the fan blows across the face of the tube, the lower the pressure inside... Drawing in the ribbon from the other end.

[vinny]

0:33 "and even at the sides it seems to suck in"

That's the problem; the air that it sucks in there comes from the nearest area of high pressure: the front of the fan.


Cowling prevents the air that is pushed out at the front from just flowing around the blades to the back and get sucked in again.
That is why prestty much every kind of efficient design for a fan includes cowling.

[npatel]

Imagine all the things you’d witness if Matthias was your neighbor.

[reilly]

These ‘nerd out’ videos are by far my favorite videos threads of yours. Your genuine pleasure and interest (and knowledge) is just awesome. Keep up the happiness!!

[jamiekawabata]

"Now, people are going to argue Beronoulli principle and all that..."

Proceeds to demonstrate Bernoulli principle.

[tomallen]

I got my engineering degree 45 years ago with my masters in hydrodynamics, I find these video experiments fascinating. In school we had a large lab and used various show & tell models to demonstrate the physics principles you are showing here. Turbulent air flow is difficult to model but you did a great job of demonstrating the effects.

[jakewitte]

I can only imagine your wife and kids listening to you down in the basement screaming and air movement over an incredibly loud fan 😂

[QuadDoc]

I teach all these concepts in my aviation courses at a technical education school. It’s cool to see you excited about it enough to learn about it and teach others. Not too many people are educated about these things. Keep up the great videos!

[EeekiE]

Testing claims—even sensible and reasonable sounding ones—often leads to extremely interesting results. Love these kinds of experiments.

[iamgerg]

The efficiency a cowling imparts is specifically when you want the net flow of air to move in a single direction. By putting a cowling you won't restrict the airflow significantly you will simply restrict lateral air flow that is the pressure will be lower behind the fan for the same pressure in front of the fan. If you put a cowling I doubt you would notice more airflow, but simply more "thrust" for the same amount of airflow.

[olivercouch]

The cowling reduces tip vortices, allowing the ends of the fan to contribute to air flow - this widens the stream and improves the efficiency of the fan.

Trust me I'm an engineer ;) (mech eng)

[feudiable]

Now we really need so see experiments with a cowling - ideally fabricated on the big new bandsaw:)

[BaconSniffer]

Thanks for all your recent uploads! You’ve been missed. Wish you and your family a Happy Christmas 🎄

[shift]

4:10 what you call vacuum is just the static pressure (air is moving, so static pressure is lower because of bernouilli : height + speed + pressure = cste)

[wendywillms]

The fan plays a B Flat from the musical scales, in case you wanted to know.😂

Dear Matthias,
Love your channel. I think you should count on the Venturi effect on your experiment: the more you increase flow speed, the more you decrease pressure. I think what's happening here is that the high-speed flow in front of the blade is a low-pressure spot and the ambient pressure of static air push and narrow it as you going farther from the fan. That's also why your pipe is always sucking air wherever you place it in the flow. As long as one end is in the flow, the pressure drops.
About the fact that the flow seems to go closer to the ground, it could be ground effect: it's less turbulent for the flow to run closer to a surface because the vortices are smaller. So it chooses the easiest way.

[erictheviking]

This bloke is fantastic. I love how enthusiastic he is about finding the stuff out. ANd, of course, he is so good at making the apparatus. I wish I had his skill and know how, but then he's spent all those years earning it. Great job, fella.

[teedawg]

I would spent way less times reading comics in science class if I was taught like this!