I know there are other engineers on this board...time to ask for some help, as I'm a computer engineer and not an aeronautical engineer. What would I need in a wind tunnel to be able to accurately test a model of a C3? I'm looking at the Cd, lift generated and drag generated in order to minimize drag and lift (may require downforce).

Could someone with knowledge of these things point me in the right direction for a DIY tunnel (like this one: http://www.ceeo.tufts.edu/graphics/windtunnel/) and/or computer modeling software?


Thanks. :cool:

 

Check the archives. Try the restoration section. Ask Julia. This was an interesting topic several months ago.

 

You might want to set up something similar to the wright brothers. Here's something I found on the net:

[url}http://wright.nasa.gov/airplane/shortw.html[/url]

Basically, a C3 is the worst airfoil that could be conceived by an aerodynamic engineer. Airflow is random and non-laminar, and the non-linear shape of the body produces a lot of boundary layer seperation.

What were you thinking about doing? I've determined that a wing would only add more drag, and make the car more unstable. I believe the only improvement to the car is to keep it clean and waxed to minimize the boundary layer (and the undesireable effects when it peels off), and use some sheet metal, fiberglass sheets, or other rigid material to create a lower portion of an airfoil underneath the car...or possibly make it flat to the ground to minimize drag underneath (though this is risky cause it will also create a vacuum to pull the car down...which is desireable, however due to the upper body surfaces the amount of vacuum will be unstable). Another option is to remove the spare tire and mold the underneath of the car with fiberglass, sheet metal, etc. to produce a better shaped airfoil that has a decent negative angle of attack (this is what the C5's and C6 does with it's geometry and body shape).

As for a wind tunnel...if you plan on staying under 200 mph then you just need rubber bands to hold up a test airfoil, a household fan, and some aluminum A/C ducting. It's also advisable to have a six pack of beer and lots of patience. :)

 

Yes...we had a great thread on this a couple of months ago but C3s CoD are the worst then improved over its fourteen years. Began at about .52 and then finally settled in about .44. In contrast, C4 was .34 while C5 poked through at .29!

I believe C2s were worst in terms of lift and takeoff...the big blocks kept em on the ground...drove Zora nuts!

 

i'm very interested in this. if you come up with a really good solution, please let me know. i have also intended to create a tunnel aft of the diff since my spare tire carrier is removed as well as a splitter for the front.
this may sound stupid, but hey, i did it when i was 16. i designed a car that i thought was really nice looking and aerodynamically pretty clean. i molded it out of plaster type stuff and carved out the design. to test it, i got a hair dryer and a piece of string (like kite string) and tied it to the grill inside the dryer. i also had a 1/12 model of an '82 to compare it to. granted, this is by no means the most technically sophisticated experiment but pretty darned neat for me. anyway, the string showed me exactly how the air broke over the surfaces of both the '82 model and my design (which was better by the way ;) ). i think in the long run, it helped me "see" the air over car designs. it's not really accurate but that experiment showed me how the air moves over certain angles and i still use that experience today.
as for the c-3 design, what i've seen that needs addressing is definately the under car shape, the ride height, the front end shape (especially any pre '80 car) and the cowl/windshield area. these are just the areas that can reasonably be worked on. to have an 'aerodynamic' c-3, just buy a c-4 :p:

 

For the wind tunnel to give you any data you would have to measure the pressure drop across the model to calculate out the cod or col. That would take a manometer of some sort and you would have to know the velocity of the air you are pushing across the model, I would think that it could be done fairly easy but I wouldn't gaurantee that your results would be consistent or accurate. Airflow is pretty tough to measure especially when it is turbulent. Althought I am probably not an expert in the matter.

 

Hi,

(This is gonna be a long post):

I just graduated (Dec 20, 2002 was my last final) as a Mechanical Engineer. Studying sub-sonic flow can be complicated so getting extremely accurate results from a homemade windtunnel may be difficult... but it is possible to get somewhat reasonable estimates of Cd, Cl etc.. You have to really understand the basics of fluid dynamics, ie. what is and what causes drag (or lift). Here is a quick lesson on lift and drag, (I hope I'm not insulting anyone's intelligence and apologize in advance if you already know this stuff).

The fluid forces acting on a body can be divided up into 2 parts: lift and drag.

Drag is the force exerted on the body by a fluid moving in the direction of the free stream flow. Drag depends on the shape/surface conditions (smoothness, area etc..) and also properties of the flow (velocity, laminar or turbulent flow etc..). Drag can be further split up into two types: 1) Skin friction drag and 2) Pressure drag. Skin friction drag is due to shear stresses in the boundary layer on the body surface ie. the fluid has to move around something and "touches" the body although for subsonic flow, the flow properties tend to change further upstream of the body almost as if the fluid anticipates a change in direction. Pressure drag is what causes the majority of the drag. This is caused by flow reversal due to an adverse pressure gradient. This happens due to seperation of the flow from the body which generates a wake (a mess of turbulent flow) behind the body....bascially almost a vacuum that wants to suck the body "back".

Lift forces are primarily caused by a pressure gradient between the top and bottom portions of the body. The pressure gradient is caused by the flow changing velocity and thus increasing or decreasing the local pressure on portions of the body.

The "general" equations for the coefficients of lift and drag on a aerofoil (derived from dimensional analysis) are:

Cl = 2*(Lift force)/[(free stream velocity squared of the fluid)*(platform area of body)*(density of the fluid)]

Cd is the same as above but replace Lift force with drag force.

To be able to estimate the Cl and Cd, you will need to know the velocity of the flow, the density of the fluid (can be estimated from the idea gas law) and the forces.

I have to look through my old "Lab manual" to see what we used to measure the actual flow velocity and forces but it is not as easy as it seems. The "lego" thing on the link you provided looks like it will work but is not that easy to build. Like someone suggested, the easiest way may be to measure the pressure differences and then the velocity and plugging all that into the above equations. Making sure the flow is smooth before it hits the body is also something to take remember.

I've said too much and will apologize again if you already knew this.

Mike
(Jr. Mechanical Engineer looking for a job :D )


[Modified by 71stingray454, 6:30 PM 12/31/2002]

 

What he said! :jester :jester :jester

 

You're going to have problems getting a model to be accurate. Besides the technical problems of measuring everything, you have to worry about whether the airflow over a model is even remotely similar to that over the real thing. Something called the Reynold's number is usually used to give an indication of whether two flows will be similar.

Its defined to be R=vL/u where v=velocity, L=characteristic object size, u=kinematic viscosity. So if you use normal air in both cases, u will be the same. A 1/12 scale model would therefore require v to increase a factor of 12 to keep the same R. So 100 mph real car speed would require a 1200 mph wind tunnel :eek: (well that wouldn't work either, but that's the idea).

The Reynold's number isn't that sensitive an indicator, but a household fan probably doesn't get much more than 10-20mph, and a factor of 100 would make a difference. You'd need a way to manipulate the viscosity and/or get much higher flow speed.

 

:iagree:

You are right. It is possible to simulate high speed air flow around a body by using water as your fluid at a much slower speed since the Reynolds number can be written as a function of the density of the fluid (since the kinematic viscosity is a function of density) ie:
Re=density*velocity*length/(dynamic viscosity)

Here's an example comparing something with a length of one meter in air with a model 1/12th the size in water.

Thus using air as the fluid, a length if 1 meter, a speed of say 100km/h, (27.8m/s), density of 1.23kg/m^3 and kinematic viscosity of 1.79E-5, the Re is approximately 1908905. Now keeping the same Re but instead using water as the fluid, with density of 999kg/m^3 and a model of 1/12 scale, the velocity of the water required would be approximately 26m/s (94km/h) instead of the 1200mph (or 200km/h). Getting water to flow smoothly up to that speed ain't easy either.

Designing a windtunnel is almost a science on its own.

Mike

 

Clutchdust,
I actually did the yarn test on my 71 vert a couple of years back. I taped 6" pieces of yarn all over the Vette and then drove it at speeds varying from 45 to 90 while a friend took pics from an adjacent car.
With all the negative things I had read about the C3's poor aerodynamics, I expected to see the yarn demonstrating turbulence at numerous points. However, I saw just the opposite - smooth flow just about everywhere.
I expect it would be more turbulent at much higher speeds but I don't drive that fast.

 

You guys have covered just about everything that I recall from Dr Foss' Fluid Mechanics classes. Reynolds is in the discussion now, so everything is good.

Carry on! :D

 

Ok, two questions from the recesses of my mind. (I do have 3 credit hours of Aero...thanks to the core curriculum at my alma mater, but thanks for the quality discussion so far)


First: Speaking of water tanks...IIRC, dye is used in the tanks to analyze flow...same as smoke in some wind tunnel experiments to visualize the flow...say I mounted a smoke emitting device at the front of the car to see the smoke as it passes over/under/around the car...would that be helpful for seeing where to design in the "negative lift" underside?




Second: The Ferrari 360 with its nifty body venting generates "negative lift" unlike, say, a Porsche that generates lift at high speeds. I guess it is a lot like the F-1 cars...would it be possible to design "aero-tunnels" in the car to cancel out the natural lift of the body?


Y'all rock. :smash: :cheers: :cool:


[Modified by rainman69, 10:59 PM 1/1/2003]

 

current discussion level


:confused: <----my head. where's the aspirin?

 

New question(s): What's the effective area for the car? Is it frontal area or is it a surface area like when looking at a 3-D airfoil?

Is the .52 a 2-D or 3-D drag coeficient?

So many questions...

 

The effective area used is the frontal area. Surface area as distinct from frontal area is in some sense "absorbed" into the drag coefficient in the equation. Skin friction isn't a very large effect.

 

[QUOTE]First: Speaking of water tanks...IIRC, dye is used in the tanks to analyze flow...same as smoke in some wind tunnel experiments to visualize the flow...say I mounted a smoke emitting device at the front of the car to see the smoke as it passes over/under/around the car...would that be helpful for seeing where to design in the "negative lift" underside?[\QUOTE]

Yes, that would be useful. You could try smoothing out areas of turbulence etc. Overall try to keep as smooth a flow as possible. That reduces drag coefficient. Downforce is a lot more complicated, and would I think require a real analysis. Wings and spoilers are still the best way to get downforce for us amateurs (unfortunately).

I suppose it could be possible to design in F1-style diffusers, but do you really want to chop up your car that much for what's probably not going to be a huge effect at reasonable speeds?

 

http://www.smokemup.com/auto_math/hp...resistance.php

Hm...I like this. I just wonder how "accurate" the 1968 L88 data is. Although, it might not be too far off seeing as a (stock) 1968 L88 did 210+ at Bonneville in 1979.

 

Interesting discussion here, but my old '72 with SB ~400 hp and overdrive did 150 indicated mph with the top down one evening....car was stable as hell, and pulling like mad....when I came unglued upon my glasses blowing off.....lifting did feel like an anchor was dropped out back though....

at that point, I"m not TOOOO overly concerned as the car has more ***** than I do....

GENE

 

TTT


Anyone have comment on my ideas on getting the car to "hunker down" at high speeds?

Right now the plan is to redesign the front air-dam and to design some sort of baffle that replaces the spare tire carrier. :smash: