Designing for Wind Tunnel Test Section Velocity

[jnelson33]

Please copy and paste my post if that assists responding.

I am trying to work backwards by selecting a target test chamber velocity to design a wind tunnel around. I want to revamp our current system as minimally as possible which means using the same fan and the current settling chamber and laminizer. The entrance and some of the test chamber section can be seen in this thread.

I understand all the area and pressure relationship stuff just fine. To me, deciding on the test section velocity by using the settling chamber and Bernoulli Principle equations seems partially erroneous since the laminizer is in there in the real world and affects free flowing air movement. The current setup has no entrance diffuser/section currently....this isn’t my design and I know it’s coefficients are wrong and its aerodynamics are questionable…

Regardless, this wind tunnel is used for heat exchanger studies, so that's why I'm looking to achieve a target velocity. 100 percent of the air is to be moved through the core for heat exchanger studies, and for that reason our test chamber now has the opportunity to be resized since it was designed incorrectly for the intended purpose.

Since this no doubt will require some fabrication work, I want to make sure the speed increase is viable (+10 MPH, currently 35 MPH). I'm probably just going to flow sim this in Solidworks, but I was also thinking I could just use the known test chamber speeds from our first tests to Bernoulli it out using the dimensions of the new, smaller cross section, higher speed test chamber. This method should be good for an estimate because it does not require my transition angles, correct? Do y’all know any other hacks like that?

Since I know Q (volumetric flow) of the fan and current dimensions of the test section and settling chambers transition, I’ve also got a sure fire way to roughly figure it out using advertised specs. But why on Earth is nobody teaching stuff this way? Why have the useless conceptual equations never been evolved and presented better for designing a wind tunnel??

 

[rb1957]

actually I don't think you're "working backwards". To achieve your increased air flow in the test section, using the same fan, you'll need to reduce the size of the test section. (in my mind working backwards would mean I've set the test section parameters (size, airflow) and will select a fan to achieve these).

This will change almost the entire tunnel.

There are texts and such "out there" but wind tunnels is such a small field that they may be hard to find. I'd seek out a nearby wind tunnel (like at a university or college) and talk to someone there. A key (the key ?) to tunnel design is the uniformity of the airflow across the test section.

another day in paradise, or is paradise one day closer ?

 

[3DDave]

Most wind tunnels aren't operated with the amount of obstruction you are looking to test with, limiting design info further.

For certain, flow out of a centrifugal blower isn't uniform, which is why I would have expected the blower to be on the down-wind side as a suction system so the inlet can have a simple grid straightener. It also avoids pre-heating the air from the work done by the blower which means the inlet temp will also be uniform, making temperature measurement easier.

 

[jnelson33]

Rb1957 I suppose I meant backwards from the methods I was loosely referencing from my undergrad lab material as they wanted us to make a wind tunnel by sizing the test section vs settling chamber first. Figures the school way is the sillier way since these loons never got into the real world…plus it would have merits for a different intent.

I know UCF has at least two wind tunnels as I saw them a few times while in undergrad and got to use them too. I’d love to make some senior students do this study for me haha. At the same time, I’ve been pretty successful with convincing my boss to revamp the erroneous, so now we have some cash sunk into this thing and it’s like a weird elephant in the room. Most people just really don’t understand how !@#$ing crazy and expensive testing is: This system still needs another $1K-$2K to become close enough to the Kays & London setup, and we’re not a heat exchanger company lol.

I’m having such an ethics struggle with trying move forward knowing that the measurements won’t be valid, nor the method correct for understanding heat exchanger performance. I’ve actually found and contacted a company that received a $10K grant for a wind tunnel to rate their heat exchangers, but it was gonna be hard to get them to test ours it seemed.

 

[jnelson33]

3DDave I have this as a constant thought in my brain as well, perhaps Kays and London were just using it as the most analogous term for their wind box setup for their heat exchanger trials haha? Regardless, the one company I mentioned above with the $10K wind tunnel is following the Kays and London method that 100 percent of air must move through the core.

My boss just bought the fan reactively to replace another home fan for this other hysterical method that was devised before I got to the company; it’s hard out here for an engineer sometimes. I would’ve set up the tunnel as a down draft as well, and I would’ve optioned a bladed fan driven by pulleys so the speeds can be changed.

 

[rb1957]

your mgmt are bitching over another $2k ?? good grief, they'll spend more discussing what flavour of coffee to have in the break room (the exec break room, that is) !

I don't think you can work with a bad wind tunnel (or any other apparatus) without at least saying here's the results, and here's the error bands ! I think you need to work up some tests to show how poorly it works.

another day in paradise, or is paradise one day closer ?

 

[jnelson33]

rb1957 most people just really aren't ready for a scientific measuring trial as they're both very accomplishing and very annoying to carry out. Right now the system was set up with a YouTube science video level of planning, aka just for fun lol. You'll see on some of my other posts that I'm encountering heavy ideologue resistance in general here, despite me being the only one who has the education to understand it at face and deeper levels. The fact that I have to even get on here to spend time away from just fixing it is the funny part, but it is the company money at the same time...

Once I get a few flow sim images (pending my 10 year old personal laptop can handle it on a scale model) I think it should be easier to see how poorly air flows. We don't have flow sim in Inventor, and there's also no way that would be paid for either. In the end, even with convincing results, I still won't be able to do anything about it most likely

 

[IceMan30]

Apologies up front if I missed something in the other replies...

You can certainly use Bernoulli and other similar simplifications(VA_1 = VA_2) for a low speed setup like that. If you're going that route, you could also calculate for a top end around 60mph, and your losses will put you in the ballpark or just over. Just keep in mind that before your rework, with a known Q your speed is what it is in low speed tunnels.

One of the best references for low speed tunnel design and methodologies is "Low Speed Wind Tunnel Testing" by Jewel Barlow. Basically the industry standard.

 

[btrueblood]

Pushing air at higher velocities means you need to get to higher differential pressures, which means more fan horsepower and more CFM simultaneously. You are probably fan-limited not section limited right now. The idea of a pulley-driven fan is not bad, as you could always find a bigger motor. But, worst case, you might get there by adding more cheap leafblowers too...