Poor Man's Aero: If I Only Had a Brain

[Jack Olsen]

If I hadn't gotten myself kicked out of college, I'd be able to take advantage of web pages like Flow Illustrator. As it stands, it's like I'm a monkey scratching himself with a slide rule. But still.

Click on the image to see the video:



If you're an aerodynamicist who can take a look at this site and tell me what my Reynolds Number and Velocity ought to be, I could start playing around with all sorts of crazy ideas. Until then, I suspect it's just some cool pictures.

[Flieger]

Is that a water tunnel?

You can't scale everything at the same time with fluids, which is why F1 guys want to test full scale like the Laurel Hill Tunnel if it was not banned by the regulations. Generally the most important non-dimensional number to match is Reynolds number, which is the unitless ratio between the inertial and viscous forces. If you have compressibility effects from high speed > Ma .3 then you'd look at Mach number too.

Reynolds Number is VD/nu. V is velocity, nu is the kinematic viscosity of the fluid- look it up for your particular fluid and temperature. D is the "characteristic length" and in pipe flow you would use the inside diameter of the pipe. For wings you would use the chord.

Re(chord) for air at 90F (since you run at WSIR), 8" chord (no idea if that is representative of your wing), 120 mph (Turn 8?) = 6.74x10^5

For reference, the kinematic viscosity I used was 1.74x10^-4 ft^2/s
funny thing is that air is actually more viscous than water on a pound for pound basis.

Freshwater, 70F = 1.052x10^-5

If you have a 1/8 scale chord then you need to run a water tunnel at 85 ft/s or 58mph! The velocity required is linear with scale, so full scale would be 7 mph.

[Dave at Pelican Parts]

That is not a tunnel at all. It is Flow Illustrator, a very lightweight CFD type of program. The consensus among people who actually know aero well on another BBS I frequent was that this was good mostly for entertainment purposes. That is, it could give you some idea of what goes on with the air when it goes over a section of your car, but that the simplifications required to make it workable were big enough that using it for making real predictions about actual air flow wasn't really practical.

A lengthy discussion about it: Flow Illustrator - a simple windtunnel simulator - Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com

The conclusion that was eventually reached about the Re number for the program was that one should play around with it and see what produced interesting results. Try setting it in the 1-million range, for instance. Go up and down an order of magnitude and see how that affects the results.

--DD

[Jack Olsen]

I'm sure you're right, Dave. I tried a range up to 8 million for the Reynolds numbers, and got what seemed the most 'real' at their default of 5000 (which seems crazy) and the velocity adjusted to .06.

Here's an example of the chassis with just an S bumper, then with a ducktail, then with my splitter and wing, and finally with the wing adjusted with 4° more angle. Grabbing one frame (even if it is at the same point in each video) is probably misleading. But this does show some reasonable-seeming changes in airflow.



The downside? Probably no meaningful or useful data.

The upside? It costs nothing, takes very little time, and I don't need to get up out of my chair.

[rbogh901]

Originally Posted by Jack Olsen

The downside? Probably no meaningful or useful data.

The upside? It costs nothing, takes very little time, and I don't need to get up out of my chair.

I dunno, looks kinda useful to me. And, as stated, I can do it sitting on my arse.


All I need is a good silhouette then?