Roof Rack "Wind Tunnel": A Naive Idea?

[autogyro46]

I've not yet seen any reference to this, and it would seem so obvious that there must be something wrong with it.

However, it occurred to me that a 1/5 scale model affixed to a "ground plane" mounting on the roof of a passenger car tow vehicle, could achieve Reynold's numbers close to 1 X 10E6 at normal highway speeds, making it possible to provide reasonable esimation of CD, CL, etc. of a full scale version (chord ~ 4.25m)

One would need to confirm that the flow over the roof was reasonably laminar, ( which generally seems to be the case)and perhaps it might be advisable to enclose the model in a tube of sufficiently large dimensions to to avoid boundary effects, perhaps even adding a honeycomb flow straightener up front.

What am I missing here?

 

[ivymike]

you can do the same thing in the comfort of a lab, if you buy a big fan?

 

[autogyro46]

Well, yeah...
But that becomes a big project in a hurry if you want decent Reynold's numbers. Not to mention that you have to characterize, rebuild, tweak, repeat...

I've included a link for a design that is at least 1/5 too small.
The cost scales something like the 3rd power of the linear dimensions, without even throwing in the cost of instrumentation.

It strikes me that one could get at least a fair qualitative idea of what the limitations of a design would be ( perhaps using something like openfoam CFD for more quantitative analysis) before scaling up , and renting time at a real (ie professional) tunnel

Thanks.

 

[patprimmer]

Where would you drive this.

How would you accurately control speed and account for naturally occurring wind.

What power supply do you need for the instruments and can you safely provide that in a moving car.

Wind over the roof of a car is not traveling parallel to the ground.

The stability of the car may be severely compromised.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[GregLocock]

We used it to establish the pressure profile over our solar car in 92.

Results were good and helpful.

I analysed the results, didn't take the data so I don't know how hairy the drive process was.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[autogyro46]

Thanks to all:

Pat-
The expected dimensions of the model are somewhere around 100cmL X 40cmW X 20cmH, and the shape fairly aerodynamic, so I wouldn't imagine that it would be worse than the stuff normally carried on roof racks.

Pneumatic instrumentation is fairly lightweight,(thanks Honeywell) and although I've yet to select the loadcells , they should come in under 1kg. (I design medical instrumentation for a living, BTW) I'm thinking of Bluetoothing the data. (which we do a lot)

Greg-
First, I don't know if you're near Queensland, but my best wishes, hopes and sympathy to you and family, friends, colleages, etc.
I'm delighted to hear you've already done this. It would seem the most direct way for a first pass.

 

[Peter7307]

Autogyro,
One point which may complicate matters is isolating the signals from the test model from those generated by the motion of the carry vehicle to give accurate data.

Never done it so I am not sure how much of a problem it may be.

Interesting thought though.

Peter.

 

[GregLocock]

Instrument the groundplane as well, to get around that.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[autogyro46]

Greg:
I was about to say the same thing... This approach, by the way, seems to be a natural for the "elevated ground plane" approach suggested by Katz. One might still need to fiddle with it to eliminate the remaining boundary layer effect, but it suggests that a few well placed pressure tranducers on the plane would be helpful.

Thanks

 

[tffy]

Yeah, it's been done before. The same guy, Burt Rutan, whose company did the stuff in the picture a few years ago (in developing SpaceShipOne) used his car with an airplane scale model on top in developing his first few aircraft. No pictures of that I could easily find online.

http://www.scaled.com/images/uploads/gallery/spaceshipone/ground_general/wind_tunnel_800.jpg

-Gene

 

[bradleyelwood]

autogyro

Yes it will work. How well it will work depends upon how good the airflow is and little things like where you can find a straight smooth section of road and of course you need to do the "runs" when there is absolutely no wind. Instrumentation may be a challenge. If you can settle for approximate results, just calibrate the speedometer. One word of caution is that air flow over the top of a curved car may not be smooth. I suggest you do your own investigation of air flow over the vehicle using a pitot tube and manometer. For crude results, just use a Magnehelic gage and look for changes of air flow as a function of distance from the vehicle. If you want to get fancy, build a "boom" which supports the model in front of the vehicle.

As for flow straighteners, wind tunnels use them for all the velocity profile problems with "flow inside a duct", elbows, fan velocity profiles, etc. You probably will not need anything for flow staighteners:

I have used the technique with good results on models for helicopter suspended loads and have had good correlations. Of course I was not looking for 1 per cent accuracy.

Try it and let us know how it works.