FWIW, a vertical wind tunnel says $2-5mil for a stationary one.
Source
Any thoughts?
It depends on what you’re using it for. A vertical “wind tunnel” used for indoor skydiving is different from a scientific wind tunnel that will produce repeatable data. If you were building a wind tunnel for engineering you would need to minimize variance as well as purchasing high-quality force balances and other testing equipment. Of course if you can do the engineering in house (a la Specialized or universities), the cost comes down significantly. An indoor skydive tunnel wouldn’t have any of those requirements and should be much cheaper.
Say this wind tunnel is for testing bicycle rider and corresponding equipment aerodynamics. What do we think would be the cost for a reputable, legitimate wind tunnel? In other words, what did it cost specialized to build their tunnel? Forget the engineering phD overhead…just the pure capital cost of the equipment itself. Presumably operating expenses (electricity?) are very low…
Why not a portable one ?
We have about 5 LSWTs in town and none of them look like the materials or construction were particularly expensive ( lots of plywood and a fan). They aren’t small, so the building would be the big expense.
You should check Dr. Coggan’s (might be mis-spelled) small scale tunnel. I think there is a write-up on TriRig.
I was wondering the same thing but with a slightly different but I think a significant difference.
If you did not care about absolute results but only relative results how much would it cost?
Meaning you would not be able to measure the " real "drag with any degree of accuracy but could you tell if A was better or worse than B? You wouldn’t be able to tell how much A was better or worse than B but just that it was. No one would be able to look at your numbers and repeat them.
Also you wouldn’t have to worry to much about calibration since you would be testing A and B in the same session. You couldn’t with any degree of accuracy take data for A on one day and compare it to B data from a month later. To many variables to control for that.
If you take away those 2 things. Absolute data and data repeatability over time, how cheap could you go ?
A box fan, some sheet rock, a platform and a scale ? I’m sure that is way to simple and some one will tell me why in about 30 seconds that I am a dumb ass. I would think the real trick is to get the air flow to be laminar and not turbulent coming out of the fan.
So, like Virtual Power. I think that’s right. But then why aren’t more shops doing it…
Just like with power meters, if you can produce reliable relative results you can also produce absolute results.
Calibration is easy, you just make a table converting X drag to Y drag. You could do with with excel and some spheres of known size in an afternoon.
The hard part is the consistency. Air speed needs to be the same and smooth everywhere in the tunnel and from day to day. The tunnel needs to be large enough to not have air bouncing off the walls and interacting with the bike and rider, you need to be able to accurately turn the bike to given yaw angles, and so on and so on.
I’ve often thought about making one.
First thing you need and probably the cheapest one is a room to do it in. So a room wide enough so that it has at least 3 feet above you and to the sides. It has to be long enough to get laminar airflow from it’s source in front of you so probably at least 15 feet with air straighteners. Then long enough behind you so that the air can come off of you cleanly and not have any back drafts, that would probably be another 15 feet.
You need an adjustable air movement source capable of ~15 m/s
You need to be able to measure air density so a: barometer, temp meter, humidity.
A way to measure air velocity so a: hot wire anemometer, kiel probe, Brandt nozzle, or other LFE. Depending on the type of air flow meter you may need a nozzle inlet pressure sensor too.
A platform to mount the bike on and measure drag, I don’t know how to do this with a bike, I’ve only done it with cars, I have some ideas but have never seen a diagram on how others do it. I can imagine the platform on an arm pivoting on an vertical axle with the opposite arm pressing against a load cell. I have some other ideas with linear bearings too.
Calibrating the load cell is fairly straight forward
Calibrating for air density and changing from ACFM to SCFM is a little more difficult but needed to compare results from different days or locations.
Setting up runs would require a cal run then a test run.
So probably with a few thousand dollars you could have a half arsed wind tunnel.
jaretj
I assume it is like most things, the cost for a Y% accurate result is X. From there a 1% improvement in accuracy is associated with a substantially larger than 1% cost impact. The closer you want to try and get to 100% accurate, the steeper the cost curve slope.
I thought about making one too. Since I have absolutely no knowledge of such things, my initial diagrams (now lost) basically consisted of a box fan, a fog machine, and a video camera. Part of me still thinks that set-up could work.
that’s a fairly safe assumption. law of diminishing returns, essentially.
if it’s as cheap as people are suggesting, why aren’t there more of them? does 1 wind tunnel in CA and 1 in NC really meet the needs of the entire triathlon population, of both vendors & consumers? plus whatever industries also use wind tunnels? (eg recent post in here about Under Armour going into the wind tunnel to finalize fabric selections for speed skaters in Sochi).
I visited Steven Vogel’s lab at Duke NC years ago, whilst setting up my own flow lab, and I got to see a Harris Hawk hovering in a wind tunnel two feet in front of my nose. Remarkable. His books on fluid flow are well worth reading.
It’s probably just a high-risk thing. The percentage of cyclists that are interested in doing a wind tunnel test is probably small enough (and existing tunnel availability cheap enough) that the market demand is satisfied (or so hard to measure that they can’t stomach the investment).
A bike shop would have to dedicate a fairly large amount of floor space to something that’d only be generating revenue a small percentage of time. They’d have to pay for the capital investment, train a staff member to run it, etc. And if you do it on the cheap, then people might not trust the data they get out of it.
And from a bike-shop perspective: what if it ends up giving potential customers objective evidence that paying for the $8k bike over the $1k bike gives almost no benefit? That’d be a very bad investment 
Put mini wind turbines INSIDE said wind tunnel capturing all the power within said wind tunnel produced wind. Sell idea to Obama energy dept as clean energy source reducing our independence on foreign oil BOOM! Millions of $ in loans that you can fritter away and not repay. Get rich AND have a sweet wind tunnel to play with all your tri buddies. You can thank me later…
I don’t know if Mr. Chu is gonna buy that one =)
Put mini wind turbines INSIDE said wind tunnel capturing all the power within said wind tunnel produced wind. Sell idea to Obama energy dept as clean energy source reducing our independence on foreign oil BOOM! Millions of $ in loans that you can fritter away and not repay. Get rich AND have a sweet wind tunnel to play with all your tri buddies. You can thank me later…
If you could just put air evenly through a tube big enough to put you and your bike in and then find a way to measure the force on your body/bike that would be close enough to measure differences between setups.
Getting air pressure from a local weather station would probably be good enough for air density.
I just don’t know the setup for measuring the drag (and not the queen version Does anyone have any sort of diagrams?
jaretj
I just don’t know the setup for measuring the drag (and not the queen version Does anyone have any sort of diagrams?
Get a big slab of granite, mount the bike to the slab, then float the whole thing on a big pool of mercury. Tie a long string between the granite slab and the loose front tooth of a six-year-old kid and turn on the fan. Use the whine of the kid as an indicator of the amount of drag force on the bike. You may need a few runs to calibrate correctly.
Well, what your really just trying to do is have a really large section of ductwork. So I think you could probably use a single blower assembly like the one you’d find in a large HVAC Plenum. Bakcward incline ans have good effciencies and fairly wide fan curves.
So lets see, if we have a “duct” that’s lets say 8’ tall and 6’ wide. That’s 48 sqft. 30mph is 2640 ft/minute. So you need 126,720CFM. 4 plenum fans stacked together would work. I looked up some specs. They would have 30HP motors each. you could use vairable speed dirves, but it would be a lot cheaper to use inlet dampers with a servo or stepper motor. You could use motor current for approximate air velocity calculations, then use a airflwo sensor in the chamber for more precise calculations. Then you just need a table with good strain gauges to meaure the change in load.
But I think the example I showed gives you an idea that even a small wind tunnel, needs really big fans for even 30mph. The fans shown above would need almost 60kw when running. That’s around $6-8/hr just in electricity at most utility rates. Then you probably don’t want to roast the poor bastard in the wind tunnel, so now you want more expensive belt driven fans that aren’t in the airstream generating heat.
Honestly, thsoe fans plus, electric switchgear, controls, instrumentation, I’ll guess you’re looking at maybe $200k minimum in equipment. That’s before programming, a building, land. Add another $200k for that. Bare minimum of $700k I’d say inclulding purchasing a commercial lot. That’s just for a tiny 6’x8’ wind tunnel.
I should have said resolution and not accuracy.
To measure large differences between A and B will cost much less than trying to measure small differences between A and B. That has more to do with resolution than accuracy.
Measuring differences in body position between A and B is way different than measuring the differences between wheel A and wheel B. Manufactures are of course more interested in the later so they have to have access to much better resolution in the tunnel.