Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Aerospace What wind speeds to use in a scaled 1: 64 wind tunnel

  1. Jul 29, 2008 #1
    So I'm using a wind tunnel to show the aerodynamics of a toy truck which scaled 1:64 in the real world. My question is that if i wanted to show that it was going at 64 kilometers per hour in the real world, do i have to set the wind tunnel so that it generates 1 kilometer per hour winds? I'm not sure about what scales to use. Thanks a lot.
  2. jcsd
  3. Jul 29, 2008 #2
    what are you trying to measure, drag force? or just making a visualisation?
  4. Jul 29, 2008 #3
    Uhmm yea im measuring drag force.
  5. Jul 29, 2008 #4
    http://en.wikipedia.org/wiki/Dynamic_similitude" [Broken]

    They have a pretty nice example here which almost applies to your model too.
    Last edited by a moderator: May 3, 2017
  6. Jul 29, 2008 #5
  7. Jul 30, 2008 #6
    Just in case the links left by redargon made you draw a conclusion which seems strange, I'll confirm it: you're out of luck as-is. For your application, you must keep the Reynolds number equal to the real world scenario, which means the wind tunnel speed should be not 64/64 = 1, but 64*64 ~ 4000 km/h. However, to avoid compressibility effects, there is also the Mach number limit of at most 0.2-0.3 ~ 300 km/h. Thus, unworkable situation.

    The way out would be a bigger model, such to be able to keep the wind tunnel speed below Mach limit, but still have it in the order of magnitude of what is needed for equality of Reynolds number. And then apply empirical correction formulas (which I know to exist, but nothing else) to get from experimental to real Reynolds number results.

    Chusslove Illich (Часлав Илић)
  8. Jul 30, 2008 #7
    you could also increase the density and decrease the viscosity of fluid in the experiment to compensate and keep reynolds number constant. By going from air to water, for example at 20°C, you should get a ratio of velocities required that is closer to 4.3, so for a real world velocity of air at 64km/h you could experiment with water at 275km/h. This is also not great and turns your windtunnel into a watertunnel. You could play around with densities and viscosities to find something useful.

    It's no wonder they have such expensive full size wind tunnels to test things like F1 cars. It's one of the "simplest" ways to get reliable results.
  9. Jul 30, 2008 #8
    aww.. man thats exactly wat i didnt want to hear caslav.ilic. so screwed. thanks anyways guys
  10. Jul 31, 2008 #9
    get a bigger model. it should be no probllem to get a 1:24, 1:18 or even a 1:12 scale truck. not to mention the quality of the 1:64 model compared to the 1:1 truck is nonexistant. If you want to make an accurate drag comparison, you need accurate models. This is possibly (of course it is, sarcasm) the reason big aerospace companies spend millions of dollars on windtunnel models.
  11. Jul 31, 2008 #10


    User Avatar
    Gold Member

    So, to measure a 1:64 truck doing a scale 64km/hr, you're saying the wind tunnel would have to run at ~4000km/h?? Something's wrong there... you might want to check your numbers.
  12. Jul 31, 2008 #11
    lets try the Re approach

    since the truck and model will "drive" in air, we'll set them to 1, effectively neglecting them (unless there's reason to believe that the windtunnel speed will be greater than 0.3M)

    lets pick a reference length: say the truck is 8 feet wide
    64 km/h = 58 ft/s
    we'll say Re' is our viscosity and density-less Re

    Re' truck = 8*58 = 464
    Re' truck must equal Re' model
    Re' model = (8* 1/64) *Vtunnel
    Vtunnel = 3712 ft/s = 4073 km/hr, although you could pick a ref. area of 1 foot and get a solution of 64^2

    but, lets look at a bigger model

    Re' truck = 8*58 = 464
    Re' truck must equal Re' model
    use 1/12 model, reference length = .667
    Re' model = (.667) *Vtunnel
    Vtunnel = 696 ft/s = 475, still too fast ~.6ish M

    ok, lets try a different approach, set tunnel at Vmax = 170 mph = 250 ft/s, find ref. length
    ref length ~ 1.9 ft, find scale
    scale approx 1/4

    better (and cheaper) solution
    i've actually done that before, and found out the Cd of my car is calculated on planform, not frontal area

    so basically, follow the link, or get a tunnel with 3 foot wide cross section (considering wall effects) and 190 mph top speed.
  13. Aug 1, 2008 #12
    well, mshinavar checked our numbers and got the same result. I agree, it does seem counter-intuitive, but if we are using the physics correctly, then the values are correct. Is there another approach you were thinking of or something we may have missed somewhere? I think the wiki links that I posted earlier are interesting, but not entirely complete perhaps.

    Thanks for your analysis :approve:

    Don't be discouraged, there are many ways to solve a problem. The howstuffworks link was interesting, and a very simple way to get a rough idea of drag forces. (It obviously neglects friction losses and a whole lot of other stuff, but pretty cool for home experiments) Maybe you could incorporate something similar. Like a scale model free-rolling down a ramp and comparing the measured velocity with a calculated velocity (using physics models of objects with specific weight and friction sliding down slopes of various angles). Just an idea that popped into mind.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook