Minimum clearance from air scoop to fuselage

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Eugene

Well-Known Member
Lifetime Supporter
Joined
May 26, 2017
Messages
1,713
Location
Merrill, Wisconsin, USA
Looks like if you design your air scoop correctly people will be copying your design for many years to come. Clearance from air scope to the fuselage seems to be different on every airplane. Wondering what kind of minimum clearance should we have?

Screen Shot 2021-06-30 at 09.28.40.pngac1624294322201866.jpgScreen Shot 2021-06-30 at 09.32.13.pngScreen Shot 2021-06-30 at 09.33.03.png
 

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,136
Location
Grand Junction, Colorado
The boundary layer gutter (the space between the fuselage and splitter plate) needs to be equal to or slightly greater than the thickness of the fuselage boundary layer. I'm not sure it's necessary unless you're flying fast enough to benefit from ram pressure. They did it on the latter model P-51s because the turbulence entering the oil cooler at high speed was causing an audible rumble and enough vibration that metal fatigue was a concern.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,336
Location
Saline Michigan
You need to reference a fluid mechanics text or other docs on how to calculate the boundary layer thickness at that point. Air will be flowing from the nose of the airplane all the way to the inlet. Once you know how thick that boundary layer is, make the bottom of the scoop a little higher above the rest of the fuselage.
 

Eugene

Well-Known Member
Lifetime Supporter
Joined
May 26, 2017
Messages
1,713
Location
Merrill, Wisconsin, USA
You need to reference a fluid mechanics text or other docs on how to calculate the boundary layer thickness at that point. Air will be flowing from the nose of the airplane all the way to the inlet. Once you know how thick that boundary layer is, make the bottom of the scoop a little higher above the rest of the fuselage.
Looking at this pictures of this 120 MPH Aircraft. Hard to tell exactly but if I have to guess opening of the air scoop is approximately 2 1/2 inches tall and about the same distance from fuselage.

Screen Shot 2021-06-30 at 20.57.34.pngScreen Shot 2021-06-30 at 20.55.38.png
 

Voidhawk9

Well-Known Member
HBA Supporter
Joined
Mar 26, 2012
Messages
638
Location
Timaru, NZ
You can go to the complexity of simming it or testing it somehow, or just use about 2" and call it good. That will probably be fairly close, and pressure-recovery shouldn't suffer much if you are off slightly, IMHO.
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
158
Location
Campinas, SP, Brazil
Very interesting paper, Mad MAC. One basic rule is undeniable: as the formula in the paper shows for both laminar and turbulent flows, the further aft it is the thicker the boundary layer becomes. However in the real world the boundary layer change from laminar to turbulent depends on quite a few factors: if it is a tractor or a pusher engine, the shape of the windshield, type of finishing, you name it. Sometimes a few vortex generators optimally placed here and there are enough to keep the boundary layer thin for longer.
I tend to agree with the suggested numbers - 1.5 to 2 inches. Another point is: do not make the inlet lips too sharp. Sharp lips may cause flow detachment inside the inlet and consequently loss of efficiency.
The best tool to optimize those things, in my view, for those mere mortals that neither have hi-end CFD software in their computer nor have a wind tunnel at home, is experimenting with wool strands.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,823
Location
US
The best tool to optimize those things, in my view, for those mere mortals that neither have hi-end CFD software in their computer nor have a wind tunnel at home, is experimenting with wool strands.
Yes, and the cheap, good light digital cameras available now really help. Eugene has apparently made good use of these.

From a big-picture, 100 KIAS perspective, will a splitter really make a difference in this case? Ingesting a bit of low energy air (that will soon slow dramatically along with the rest of the flow when it hits the radiator face) seems only slightly draggier than having to move the air laterally in the under-duct splitter cavity, and then have that air disrupt otherwise smooth flow on each side of the duct. Yes, if the radiator has a lot of pressure drop and requires a lot of static pressure to force the air through, then it will be important to keep the duct opening in an area of (only) high dynamic pressure. Otherwise, it would seem that a lighter and easier to fabricate (and modify, if needed) simple scoop might be just fine in this application.
 

Eugene

Well-Known Member
Lifetime Supporter
Joined
May 26, 2017
Messages
1,713
Location
Merrill, Wisconsin, USA
I tend to agree with the suggested numbers - 1.5 to 2 inches. Another point is: do not make the inlet lips too sharp. Sharp lips may cause flow detachment inside the inlet and consequently loss of efficiency.
Would this be a correct picture?

IMG_5942.jpeg
 

Norman

Well-Known Member
HBA Supporter
Joined
Nov 28, 2003
Messages
3,136
Location
Grand Junction, Colorado
No corners on the inside either. You can have flat walls that diverge from the free stream by 7 degrees but any more than that and there will probably be separation from the walls with recirculation (drag and loss of cooling) near the walls. If you want to get really anal about internal drag I'd use a model airplane airfoil for the whole cross section from the lip clear back to the face of the radiator.
 

Attachments

Eugene

Well-Known Member
Lifetime Supporter
Joined
May 26, 2017
Messages
1,713
Location
Merrill, Wisconsin, USA
No corners on the inside either. You can have flat walls that diverge from the free stream by 7 degrees but any more than that and there will probably be separation from the walls with recirculation (drag and loss of cooling) near the walls. If you want to get really anal about internal drag I'd use a model airplane airfoil for the whole cross section from the lip clear back to the face of the radiator.
Thank you! Makes perfect sense and corresponding what I am hearing from my Russian friend about not to increase 8° because after that you most likely going to see separation.
 

Mad MAC

Well-Known Member
Joined
Dec 9, 2004
Messages
795
Location
Hamilton New Zealand
Very interesting paper, Mad MAC. One basic rule is undeniable: as the formula in the paper shows for both laminar and turbulent flows, the further aft it is the thicker the boundary layer becomes. However in the real world the boundary layer change from laminar to turbulent depends on quite a few factors: if it is a tractor or a pusher engine, the shape of the windshield, type of finishing, you name it.
Yes in the real world one would only use the turbulent calculation as if it really did have much laminar flow, one would be running CFD anyway. It is only to give an indicative depth, I would be making sure there is a good margin over and above the calculated depth.
 

J Galt

Well-Known Member
Lifetime Supporter
Joined
Nov 29, 2019
Messages
65
Location
KMAN, Idaho
For your internal duct design, obtain a copy of Kuchemann and Weber, "Aerodynamics of propulsion". Some of the commonly believed internal duct shapes are incorrect, according to his text. In the book he gives an ordinate table for a streamline diffuser (internal duct) for various ratios of inlet area to radiator area.
Justin
 

FinnFlyer

Well-Known Member
Joined
Nov 19, 2019
Messages
60
Location
Bell, FL
I still need to understand the math, but inner radius can be smaller than outside radius. For an inlet designed for one angle of attack and optimized for airflow needed at one airspeed, it doesn't matter. But you'd design for full power at climbing AOA. Bigger outer radius lips then allows for excess airflow to spill around inlet with less turbulence (and thus less drag) at higher air speeds. For optimum pressure recovery (converting dynamic pressure in incoming air to static pressure just before the rad) the inlet needs to be trumpet shaped. You'd be surprised how much it needs to be pinched down near edges of rads.

Inlet lip:

1625754007626.png
Edit: As for duct (diffuser) shape, if the duct is long enough that won't have a change greater than 7 degreees (as in Norman's drawing), that may work. But typically the duct is shortened and you need a streamline (trumpet or horn shaped) duct.

Finn
 
Last edited:

rv6ejguy

Well-Known Member
HBA Supporter
Joined
Jun 26, 2012
Messages
4,566
Location
Calgary, Alberta, Canada
Just some food for thought. In practice, you don't need large radii on the inlet of the scoop and ingesting the boundary layer is of minimal concern as well. The fastest liquid cooled aircraft in the world- Voodoo, dispenses with the scupper completely. You need to slow the air anyway for pressure recovery at the rad face. Eliminating the scupper will result in lower frontal area.

voodoo-14.jpg

At climb speed the drag is pretty low so some separation makes little difference in drag, even if you have some. My tuft testing doesn't show any with 1/8" radii on my airplane in climb attitude.

For least drag and maximum cooling, be sure the duct is sealed to the heat exchanger face only. Don't thud air into the end tanks as so many people do. You'll cut cooling dag more by having a proper diffuser shape and a variable aperture on the duct exit to recover momentum.

The theory is great but instrumented flight testing is the true test.
 
Last edited:

Lendo

Well-Known Member
Joined
Feb 6, 2013
Messages
814
Location
Brisbane
rv6ejguy, do you have a variable outlet flap? I've always thought that was a good idea in controlling temps.
George
 

Eugene

Well-Known Member
Lifetime Supporter
Joined
May 26, 2017
Messages
1,713
Location
Merrill, Wisconsin, USA
How about something like this, Eugene?
I understand what you're trying to show me, but my real situation not corresponding to your picture at all. I will try to come up with a little sketch tonight. Oil cooler is position right behind water cooler. This installation was approved by Rotax and they actually like it this way because oil will be a little bit warmer than water. And this is what they want. But thank you for trying to straighten me out!!!

IMG_1740.jpegIMG_1732.jpegIMG_1733.jpegIMG_1735.JPGIMG_1739.jpegIMG_5787.jpeg
 
Top