Cooling inlets & outlets and airflow direction: need the fluid dynamics experts

[Protocooler]

Hello,
there is a big discussion going on in the high performance forum about which way the aiflow is really going in a fast aiframe (likely from wing saddle towards the nose !!!!) and hence how we should cut out our cooling vents for true efficiency (we power fast electric gliders with 1000W+ most often)

We are not even sure wether inlets are really inlets and vice versa.
If some of you understand all that, please chime in the thread below and give us all a useful lesson! Please tell us were to cut what and which way round :)

http://www.rcgroups.com/forums/showthread.php?t=255485&page=5

thanks
Proto.

[jlk]

I always have used NACA ducts. As long as the fuse is still getting wider they should work well and they have, I believe the lowest drag. There is a formula out there for the shape and ramp pitch. They also work on exits, again as long as the fuse is getting narrower.

Jordan

I found this with a short look.

NACA duct design specs:
After wading through a bunch of documents and drawings on the NACA web site, I determined that an optimum NACA submerged inlet duct design should employ curved diverging ramp walls with a width to depth ratio between 3 and 5, and a ramp angle of between 5 and 7 degrees. The entrance lip at the rear of the duct should be a blunt airfoil leading edge shape; however, the edge formed by the duct and the surface elsewhere should remain crisp. The thicker the surface boundary layer in the area of the duct is, the less efficiently the duct will work. This means that if the surface boundary layer begins to detach from the fairing surface before it gets to the duct, you won't get as good ventilation as you would if the aerodynamics of your fairing were better in front of the duct.

[Sparky Paul]

The NACA duct..
It has been a miserable failure for feeding jet motors directly though.

[Protocooler]

Thanks,
interesting that the NACA inlets are located on the bottom, it seems bottom front of fuse is indeed a good overpressure area whereas the top would be more sucking air out...

1- Which lays the true question: is a NACA put on top a fuse, often midway through the nose, really useful at all (the air is expected to be turbulent in that area, neither sucking nor blowing in) ?

2- It seems the highest pressure area is right below the leading edge of wing, thus this would be ideal location for an inlet ? But which orientation ?

3- does the air go out from the nose cone (if there are openings in the firewall) as tends to prove the nosecone of a high speed DS soarer ejected in flight (underpressure created on nose, kind of sucking it off ) ?
Or will it go out by the tail or behind wing saddle ?

Common sense is not really applicable on aerodynamics, I don't listen what my mind says because it's probably wrong...

more insight much welcome !
...Proto

[Sparky Paul]

I believe this is the largest airplane which used NACA ducts for the engine air.
and a link... :)
http://prototypes.free.fr/yf93/yf93-1.htm
.
At model speeds, the outlet is generally 1.2 to 1.5 times the inlet area.. as a rule of thumb. No one has really investigated this.
Bill Wisniewski in the late '50s, early '60s flew control line speed models with NO cooling in or out, only the cylinder head exposed.

[kitfoxdrvr]

Common sense is not really applicable on aerodynamics, I don't listen what my mind says because it's probably wrong...

...Proto


Boy, am I learning that this is true!

[Salto]

Whether air goes in or out a duct will depend on one thing - the pressure difference between outside and inside the duct. If the pressure is higher outside, air will flow inwards. Even if a duct is placed in a high pressure area, air will not flow in if the pressure inside the fuse is higher. Therefore, to ensure a nose mounted duct will take air in, it is necessary to vent to fuse interior to some lower pressure area like the very back of the fuse.

The reported nose cone popping off is probably due to the fuse interior being "pumped" from some other high pressure leakage point.

I suspect the ducts are on the underside of the airliner in Sparky's pic because that is where the air was needed to feed the air-con or whatever, rather than because that was the area of highest pressure. At normal flying angles of attack the equivalent points on top of the fuse will have very similar pressure anyway.

For a model glider at speed the same would be true, ie ducts on top of the forward fuse will behave the same as ducts on the bottom. But when slowed for thermalling, the higher angle of attack of the fuse may favour the lower surface for duct location. But the key point here is that the fuse interior must not be fed from some other higher pressure leakage point, otherwise the flow will be reversed. So the outlet must be at a point of lower pressure like at the tail, or some other carefully considered point on the fuse. AND there must be no leaks from higher pressure points like at the wing saddle.

Proto, I think you'll find the flow all over the front of a streamlined fuse will still be laminar as the pressure gradient is still well and truly decreasing there. So a NACA style duct should work OK.

Graham.

[Sparky Paul]

Ducts on the bottom collect dirt and rocks.

[Julez]

I just put some beautiful NACA inlets at the nose of my robbe Arcus.
For the outlets, I had no better idea than to mount them backwards.
Is this ok, or is there any superior shape for outlets?

For the sizes, I just read Mark's and Jordan's postings:
http://www.rcgroups.com/forums/showthread.php?p=2705602#post2705602
I think it boils down to this:
At a given intake area, one achieves the biggest airflow with an outlet twice it's area.
At a given outlet area, one achieves the least drag with an inlet twice it's area.
Does that sound about right?

Cheers,

Julian

[nmasters]

is there any superior shape for outlets?


The outlet shape doesn’t really matter as long as it’s placed at a low pressure spot. On way to maximize the pressure drop though the compartment that I’ve never seen on a model is to put the exhaust pipe in the cooling air outlet thereby making an ejector. Google for “exhaust ejectors” and don’t ignore the antique railroad steam engine pages. There’s some real practical design info there

--Norm

[nmasters]

You can narrow that search with "Lempor Ejector"

[Gannet AME]

I just put some beautiful NACA inlets at the nose of my robbe Arcus.
For the outlets, I had no better idea than to mount them backwards.
Is this ok, or is there any superior shape for outlets?

A properly designed cooling air outlet the ramp gently turns the exiting air so that it leaves the cowling nearly parallel to the outside airflow. The ramp and the cowling wall should form a gently converging duct that accelerates the cooling air to nearly free-stream speed before ejecting it. See attachment.

[nmasters]

Oops! I forgot you guys are probably using electric motors. Ejector stacks can turn some of the wast heat of combustion engines into thrust but won't help you. It's possible to use the heat of an electric to produce some thrust but it's a much simpler design problem with an exhaust pipe. In either case the cooling air outlet should face downstream and you should make every effort to limit turbulence inside the engine compartment.

[KentuckyRanger]

I burned up the Motor in my Sapac Wilga 2000 because the Motor mount blocked all the Air inlets to the Motor and the intake vents did nothing more than circulate Air around the side of the Motor, not through it.

What I did was close off the intakes and cut the Cowl to where Air would come into the Motor compartment through the Motor instead of around it, then I made a Ramp to compress the incoming Air right into the Motor.
I also re designed the Motor mount after a crash that takes full advantage of the Air inlets to keep the Motor cool.

Here's some Pics, it might give you some ideas. It looks like crap and someday when I can get another Cowl I'm going to make it look allot better, but this works for now.
One thing to remember when integrating Air Ducts is to keep the turbulence down and always have an outlet so the in coming Air has a clean Path back out.
You can make a huge Air inlet, but without giving it a Path the Air pressure inside the Cowl increases, equalizes and simply goes around it thus canceling out any Air circulation.

AME's diagram is a perfect example of what I'm talking about.



http://i331.photobucket.com/albums/l454/FoamPlanes/MotorMount3Small.jpg

http://i331.photobucket.com/albums/l454/FoamPlanes/MotorMount6Small.jpg

http://i331.photobucket.com/albums/l454/FoamPlanes/Wilga2000CloseSmall.jpg

[Julez]

Thank you very much for your responses.

So it seems alright what I have done so far, but next time, I will skip the NACA- like tapering at the outlet.

Cheers,

Julian

[BMatthews]

Protocooler, you're right to assume that nothing is what it seems. Other than right at the nose airflow around an airframe can do some amazing things with localized areas of high pressure where you think there would be low pressure, turbulence that sort of fluffs things around without actually moving them and even places where you can have reverse flow.

Ideally you'd make a dummy model and fit it with a bunch of pressure test holes that connect to a bunch of small plastic tubes and run the whole lot through a wingtip that connects to a boom. Then attach it to a truck or car such that it's well away to the side or well in front of the vehicle and run it up and down a quiet road. Quiet is good so folks don't point and laugh..... Those tubes can then connect to a big bucket of colored water and by the amount the water is drawn up into the tubes or the amount that some tubes blow bubbles you can tell how much pressure or vacuum there is. For the bubble blowers if your bucket of Koolaid is deep enough you can extend the ends down lower until the bubbles stop. When they stop it's because the air pressure is matching the water pressure. The deeper you can go and still blow bubbles the more air pressure there is relative to the other pickup points.

The model for this can be pretty crude but the outside finish and shape has to mimic the final model design closely. But you can build it from some knot free construction wood in big slabs that you glue together and carve to shape. No need to worry about the wing loading on this one.... :D And results in many cases will only be applicable to that one design. Although with a few similar designs general trends will cross match.

Howzatt?

[BMatthews]

KentuckyRanger, I've got a model that I'm looking to move from brushed to brushless. Like your case I'm worried that the motor is going to be behind a big spinner and shielded by a big cowl. I'm looking at a small fan that is the same diameter as the motor casing that will collect and blow some air through the motor for cooling. It'll pick up air from within the cowl and it'll get into the bell using a gap between the motor and the spinner backplate. I'm hoping to try temperature measurements both with and without the fan housing. Should be interesting.

[Julez]

Bruce, you might want to try a "Turbospinner". :)

http://rcworldofplanes.net/ProdImages/dkt3623big.jpg

[BMatthews]

Bruce, you might want to try a "Turbospinner". :)

http://rcworldofplanes.net/ProdImages/dkt3623big.jpg

On some models I'd be all over that. But it would look out of place on this ol' girl..... :D

[Brandano]

you could also use a crude centrifugal blower to "suck" air through the motor. Modifying the backplate of the spinner might work for that too

[BMatthews]

you could also use a crude centrifugal blower to "suck" air through the motor. Modifying the backplate of the spinner might work for that too

Now why didn't I think of that..... thankyou.