Turboprop style cowling for water cooled auto conversion

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
Ive been mulling this around in my head for a while. Ive read almost all the watercooled topics on here. So if planning on a p51 style belly scoop for the radiator could one run a turbo prop looking engine cowling up front that has the large scoop looking inlet under the spinner for the intercooler. Run the big scoop instead of the side inlets next to the spinner.

Engine is ej25 subaru 4 cylinder with a gearbox. Going on pro composites vision. Im just trying to envision the cowling.
Or... stick with a conventional cowling and run ducts from a plenum behind the inlets to various areas. One being the intercooler, second being oil cooler, third a vent hose blowing air on the reduction drive and another bringing air into the cowling. And last air for the inlet air to turbo.

Figure the big opening under the spinner would feed a plenum instead of 2 seperate plenums on the side cheeks of the cowling.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,780
Location
Saline Michigan
Oh, I would go P-51 style belly scoop for radiator, then as slick a cowling as possible with a scoop with a splitter under the prop for induction, intercooler, oil cooler, and some general airflow. With the EJ25 setup, the front will not get real narrow across the cylinder heads, but it is narrower than a Lycoming, and will still look great.

The engine I am looking forward to is AeroMomentum's big inline four. Full on turbine looking with scoop for induction, intercooler, and oil cooler, then P-51 scoop on the belly for radiator. Zoom!
 

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
I figured the ej25 width would not be as streamlined as a turbine but was really wanting a slick looking cowling and way to cool the intercooler. Race 84 and rv6ej are the inspiration. And belly scoop is only logical answer for radiator.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
1,366
Location
Jackson
Do your CG calcs. Most auto conversions tend to be heavier than the typical Lyc they're replacing, but some aren't. Be sure that moving the rad aft of the cabin won't push the CG too far aft.

Charlie
 

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
Yes im paying attention to cg. Scott said i can have 300# firewall forward. But im asuming very short engine mount having it hug firewall. I should be 225hp and equivalent in wieght to an 0-390. But good thing is i have engine and gearbox before building plane so i can change things if cg doesnt play well.
 

Puggo1

Member
Joined
May 19, 2020
Messages
8
Hi,
Loving the threads here. Some very interesting topics. This is the only one I can contribute on.
Some years ago we designed a cowl for a Wilsch 3 cylinder diesel engine - see attached images. Like your thoughts, we wanted a sleek turbo look cowl around the two radiators for our RV9. The problem with the installation was the two radiators (water , intercooler) were stacked behind one another. During flight the airflow was choked and the cooling performance was marginal. Over time the radiators were moved and the aperture reduced for better flow through the intercooler only. The radiator was moved to the bulkhead and fed by a top slot with the air exiting under the fuselage.
A large square opening requires a longer divergent duct (max 9 degrees) to decelerate the airflow from flight speed to below 50mph (maximum radiator speed) . A shallow wide opening diverging in 2 dimensions is a better compromise to feed a radiator. Similarly a "smile" grin opening under the spinner would be a good means of feeding a radiator or intercooler in a firewall forward position.
A P51 style installation has its own issues but there have been many homebuilt aircraft with this. Cooling at low speed may require the assistance of an electric cooling fan mounted on the radiator. These radiator cores usually are derived from racing car open cores and not production cars to get the air flow rate. (think expensive).
Happy flying.
 

Attachments

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,150
Location
Calgary, Alberta, Canada
As discovered, the Wilsch setup performed poorly as I'd expect. Stacked cores create tremendous pressure drop and drag. The intercooler dumps heat into the rad resulting in lower Delta T. Simply a bad design.

Max divergent angle should be less than 7 degrees which is why it's so difficult to do a good job within the confines of a cowling. The airplane never has enough length to fit both the divergent and convergent parts of the duct required for low drag.

Properly done, the P51 type ventral setup requires no fan in ground operation. Mine works fine as does Russell Sherwood's. I am gathering photos, specs and video for a new video on rad design. This will describe in detail how it was done in my plane and Russell's.
 

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
^following yours to a t. Its like the bible on what works. Radiator definately going in belly scoop. Just need something for intercooler and oil cooler. Didnt plan on stacked. But just something about the sleek nose coming off the spinner that just draws your attention.
Im no engineer or savy in aero so didnt know the drag induced if did a single unlet under the spinner or if the side cheek inlets were less drag.
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,150
Location
Calgary, Alberta, Canada
The drag is due to momentum loss passing through the HX.

I think you could do a decent job for the oil and intercooler HXs feeding them from a Cheyenne style scoop. Split the air off to each one. Nice job on these. Exit the air at the edges of the cowling so you have clean, cool air for the ventral rad.

rv8scoop.jpg

rv8scoop2.jpg
 

berridos

Well-Known Member
Joined
Oct 10, 2009
Messages
1,129
Location
madrid
May i ask what Hw means?
The 7º rule considers a linear divergence of the tunnel as i understand it. Is there some rule for a progressivly widening scoop? Applies that 7º, the same for inlet and outlet sccop? After the rad the air heats and expands. What constructive rules deal with that expansions?
Thx
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,780
Location
Saline Michigan
May i ask what Hw means?
The 7º rule considers a linear divergence of the tunnel as i understand it. Is there some rule for a progressivly widening scoop? Applies that 7º, the same for inlet and outlet sccop? After the rad the air heats and expands. What constructive rules deal with that expansions?
Thx
HX is Heat Exchanger.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
1,366
Location
Jackson
beridos,

Probably the best known is the work of Kuchemann and Weber, Aerodynamics of Propulsion. Chapter 12 discusses cooling and diffusers, and the best version they found was called a 'streamline' diffuser. The chapter includes formulas for designing one; it looks like the profile curve of a trumpet bell. It's close, but not exactly, like an exponential horn; think old-school horn loudspeakers. I suspect that the difference is due to the 'loading' of the HX in the mouth of the 'horn'; that changes the 'impedance' of the air exiting the horn.

(probably a 4 year degree worth of ideas condensed into a paragraph....)

edit: The chapter also deals with exit ducts. While there are limits, they can typically be much shorter than inlet diffusers, and aren't as critical in shape/angle/etc.

Charlie
 
Last edited:

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,150
Location
Calgary, Alberta, Canada
May i ask what Hw means?
The 7º rule considers a linear divergence of the tunnel as i understand it. Is there some rule for a progressivly widening scoop? Applies that 7º, the same for inlet and outlet sccop? After the rad the air heats and expands. What constructive rules deal with that expansions?
Thx
The convergent section of the duct aft of the HX can use steeper closing angles as separation isn't much of concern here but my scoop has almost equal length on either side of the HX. Total duct length is 49 inches. You can stand an abrupt reduction in volume internally but this would likely give hard separation externally, making for higher drag. It would also give a less favorable vector to the exit air at near closed door angles used in cruise. The gentle curve aligns air better with the free stream with less turbulence. See how Voodoo (last photo) does it right with a very long, revised scoop using no inlet scupper (as the original P51 had). This is the fastest piston engined aircraft in the world.



8384.jpg 8949.jpg
8982.jpg voodoo2.jpg
 
Last edited:

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
The cheyenne style cowling and scoop is prety much exactly what i was thinking with it being streamlined and a single central opening.
Ill have to make some bows that go from the spinner to the firewall and play around with.
good to know there are proven versions of my thoughts and these arent just random useless ones.
Again no enginerr or aerodynamic specialist what is the effect of a raised hump on top the cowling? Say i have to put a raised hump on top to clear the intake manifold. Im making low profile one as low as i can but there still might have to be a bump .

This is one of those things i guess you can look at with theory and wont know till its flying and looking at tufts to see what the air is doing. Or at least ground testing and see what the prop does with the air.
 

slociviccoupe

Well-Known Member
Joined
Apr 1, 2011
Messages
366
Location
Palm Bay Fl.
Made a test spinner on cnc router (13" disk out of 1" foam) today and most likely not going to get desired look of the cowling with the subaru. It would take a very long prop extension to get the transition from firewall past the heads and to the spinner to look correct. Im probably stuck with a conventional cowling.
Will keep trying and see what i come up with.

Are there any shapes or designs of side cheek air inlets that are better than others. Same performance with less drag. Things to do and not to do. Like so many inches away from the side of the spinner.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
1,366
Location
Jackson
You can get bigger spinners than 13", if it's important enough to you.

The 'best' inlet shape is round, for the simple reason that it's the best skin area to inlet area using a circle. But. What are you feeding? It may be better to feed a rectangular heat exchanger with a rectangular inlet, simply because getting the transition from round to rectangular right (without flow separation) isn't simple. A boundary layer splitter is often good, but some research (see the CAFE Foundation work) says that if the inlet is in a high pressure area (like under/just aft of the spinner), the duct can ingest the boundary layer without issue.

The shape of the inlet 'lip' is the subject of entire research papers; it 'depends'... on what type of diffuser you're using. Internal diffusion with perfect consumption of all the air hitting the inlet wants a relatively sharper lip with smaller external dimensions, while external diffusion wants a 'fatter' lip so the excess air that spills around the opening can turn the corner easier.

Internal is the most efficient, but is rarely (likely never) done perfectly. External diffusion is easier, the duct is less critical, and the inlet can be oversized quite a bit for low speed/high load cooling, and an exit flap can be used to limit excess flow through the system at higher speeds. Net result can be close to a perfect internal diffuser, and a real-world external diffusion system will likely outperform a real-world internal diffusion system (like most of us will build).

Or so I read somewhere.

Charlie
edit: Ross makes a good point about prop blade root interference; many props don't have enough pitch near the root due to structural limitations, and it can affect cooling even with spitter-equipped inlets like the RV cowls.
 
Last edited:

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,150
Location
Calgary, Alberta, Canada
Round inlets are in vogue and some studies show they may be most efficient from an aero standpoint however it's hard to find round rads and intercoolers. Further from the spinner may be better as well on the ground as you have a better airfoil there from the prop than at the shank.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
14,345
Location
Memphis, TN
Lots of turboprops have cuffs on the blades to get air to the inlet. Easy if designing a composite blade, hard with a metal but doable. But it’s not going to be off the shelf.
 
Top