Homemade wind tunnel?
 

Why couldn't you use a peanut oil fog machine and a big fan for home wind tunnel testing?

Note: Wind tunnel assistant not included.

justpassntime -

Would you have to "scale" the wind speed to the scale of the model or ?!?!?!?

CarloSW2

I wish I had the space to build a full size one. Maybe build like a small steel building in the backyard of some future house and put a fan that could simulate 60mph winds on my actual car or something...

My dad actually has a 5 foot (blade) diameter fan that can sustain 35 mph wind. Runs on a 2hp electric motor; barley starts on a 4kw generator but man is it nice when working outside. Hmmm

I've thought about building a scale model wind tunnel. Just can't seem to find a model of the Jeep Cherokee.

A good example of a scale model wind tunnel setup can be found here.

Try the toy section...
 

Try the toy section at your local department store. The model wouldn't have to be a "model".

Check this out VINTAGE METAL TONKA JEEP CHEROKEE WAGONEER TOY TRUCK - eBay (item 140245515768 end time Jul-06-08 18:21:57 PDT)

Each of those would put out turbulent flow, and together it'd be even worse. You can't put that fog through a small wand, so how are you going to target the area -- have it on a stand?

Yes, if one wanted to get real drag numbers. Look up "Reynolds number."

Yes, the model would have to be a model. Toys are extremely inaccurate, as proven by the link! ;)

Do want wind tunnel or data?
 

justpassntime,is there some specific info your seeking? I know you're aware of Reynold's Number effects, and how it effects model scaling in wind tunnels.A 1:24 Scale model would require a minimum of 480-mph wind in the test section to just simulate a turbulent boundary layer on your model .Anything short of that would yield meaningless data for you.If there's something specific you'd like to know about your car or potential projects,we may be able to help you out.I have a small tunnel I built for public demos but it's not a serious research tool.And for me,I can intuit and tuft-test in full-scale,faster and cheaper than I ever could with models.And the results are always valid 'cause its full-scale.Let us know.
.

I know alot of you are DIY'ers, has anyone thought of retro fitting a spray booth as a wind tunnel? Or is it to small?

That would seem logical, plus you get the added benefit of being able to paint in it, if you maintain it properly.

It probably wouldn't be too small in cross-section, but you wouldn't have sufficient empty space in front to ensure that the oncoming air was flowing properly.

spray booth
 

A spray booth would make a great test section.They are smooth-walled and have recessed lighting.Frontal area of model would be limited to 5% of test-section area,so size of booth would determine maximum size of model.Remember,without a turbulent boundary layer your results would be garbage.In full-scale you need at least a 20-mph air velocity for proper Reynold's Number.At 50% you'd need 40 mph.25% -80-mph,and so on.Pope has a great book on" low-speed wind tunnel design."It's worth checking out.The re-circulating design is cheapest to operate,although if using smoke you've got to design for that.

According to my calculations, based on stuff I just looked up all over the internet, if 20mph air at 1:1 scale is enough to observe the relevant turbulence, 30mph water will work at 1:18 scale; a scale at which models of most of our cars are widely and cheaply available. That takes like 40,000gph at a 6 inch by 6 inch cross section, so a pump would be pretty expensive. A powerboat would probably be cheaper, although whatever cord you dragged your model by and the boat's wake would keep the water pretty turbulent.

But if you could hoist a lot of water 30 feet in the air, and get a few seconds of flow at the correct speed at ground level--assuming you could keep the flow laminar and drag-free--you'd have the correct ratio. The challenge, then, would be keeping the model from being immediately pulverized.

We don't have any waterfalls here in Florida, because we don't have any elevation changes. Does anybody who lives near one wanna give this a try?

It's a Cherokee... just use a few cardboard boxes :D

water
 

water has 833X the density of water so small models are just fine under water.The trick is how you 'work' inside the test section.
A model can be drilled and tapped with dye outlets for food coloring joined to a common reservoir outside the tunnel.The colorant ejection velocity must be matched to the flow velocity to prevent a jet.Once calibrated,selective ports can be selected for release and observing/photography.When the water becomes saturated with colorant,chlorine bleach is injected to clear up the water.
------------------------------------------------------------------------- An underwater tow tank can also be constructed in which models are towed underwater from a traveling overhead crane ( as at the US NAVY David Taylor Tow Basin in Maryland) .Dye injection and tufts can be used for flow imaging.

Thanks, aerohead! I thought the relevant figure for the Reynolds Number was the viscosity, 13x that of air. If we can really move the water 833x slower than air for a model of the same size, that's great.

My thought was to ignore the flow imaging and just do a/b testing: Put a scale on the tow rope and see how much drag I get on my car model (or any similar bluff body). Then put on miniature turbolator tape in various locations and see if the scale registers a lighter reading for the same water speed.

Then, rig up a waterproof stepper motor to a spoiler and maybe a pump to test the devices in this crazy-looking Chrysler patent: Motor vehicle with flow-influencing devices to reduce air resistance - US Patent 6378932 Description

Rn
 

khafra,you're correct about the Reynolds number.Kinematic viscosity is the operative parameter on that one.I do have an SAE paper on underwater tow-testing of 1:24-scale Formula-1 models and they're size was not an issue.Which is great as far as model making goes.

I think I'm missing something--if you want to see how a F1 car behaves at 130mph in air, you could put a 1:1 model in 10mph water; but a 1:24 would need 220-something mph water, right? I must be scaling one of these dimensions incorrectly.

Which is easier way to get nice visual results towing the car vs flowing water? I would guess towing but how long do you have to make the tank to get nice flow and are the rotating wheels valid due to reynolds number?

If I want to get reasults for 1:24 truck for speed 50 mph do I tow the truck
50mph/833=0.06 mph
(50mph*24)/833=1.44 mph
or something else?.

If I build flowing water tank is this good design like here on the this video or is it better to have only one flowing channel vs two returns?
http://www.youtube.com/watch?v=L0ki-FtPYrA