NACA Cowling?

[spicmart]

How effective were the NACA hood shapes in reducing drag of radial engines compared to inline engines? Were the late-war Japanese designs especially good at that?

 

[swampyankee]

How effective were the NACA hood shapes in reducing drag of radial engines compared to inline engines? Were the late-war Japanese designs especially good at that?

You mean NACA cowlings?

Very effective. A well-designed radial engine installation can have negative cooling drag, although this is quite unlikely (as it is for liquid-cooled engine radiators), but drag comparable to liquid-cooled engines is possible. It's just hard to get there.

 

[spicmart]

You mean NACA cowlings?

Very effective. A well-designed radial engine installation can have negative cooling drag, although this is quite unlikely (as it is for liquid-cooled engine radiators), but drag comparable to liquid-cooled engines is possible. It's just hard to get there.

Yes, NACA cowling. I wonder if it evolved during the war.
You know of an engine installation where negative cooling drag had been achieved?.

 

[swampyankee]

Yes, NACA cowling. I wonder if it evolved during the war.
You know of an engine installation where negative cooling drag had been achieved?.

It was developed in the late 1920s/early 1930s; NACA cowlings were used on, among other aircraft, the DC-1. I know of no aircraft where negative cooling drag was demonstrated.

 

[vikingBerserker]

They were very effective. On the Martin 123 which later became the Martin B-10, the Townsend Rings originally fitted were replaced with the NACA Cowlings and top speed jumped 10% and landing speed dropped 15% in the1932/1933 period.

The Townsend Rings themselves had improved the speed IIRC roughly the same amount vs none cowled engines

 

[Simon Thomas]

The effectiveness varied widely.
This is an interesting paper on the subject: The Cowling and Cooling of Radial Air-Cooled Aircraft Engines on JSTOR

 

[Shortround6]

The NACA cowling of the 1920s and first few years of 1930s was NOT, repeat NOT the NACA cowling of 1940 let alone the US cowlings of 1943/44.

The engine installations of the 1940s were a much more integrated system.

The Photo provided by vikingBerserker shows a better shaped and longer version of the Townsend ring.

However
Neither installation used ANY baffles between the cylinders.
Neither installation used any baffles between the cylinder heads and the cowl.
Neither installation used any means of adjusting the airflow through cowl to account for different speeds/flight conditions.
Neither installation used exhaust thrust 99% of the time.
Baffles between the cylinders started in the early 30s.
Adjustable exit flaps started about 1935 (Vaught SBU-1?)

Some aircraft got leading edges on cowls that were airfoil shaped and in some cases the exterior profile do match the interior profile.

It is an integrated system. If you want the lowest drag possible you have to arrange for the least amount of airflow through the cowl that will reliably cool the engine.
An illustration of this was the fact that the R-2800 C due to it's massive change in fin area (and fin material) on the cylinders and cylinder heads, needed 10% less cooling air flow than an R-2800 B series engine at the same power level. OR you can make 10% more power with the same air flow and not melt down the engine

Boeing had got their sums a little bit off on the B-29.

While a lot of the later US engines didn't reduce cooling drag like the P-51 a lot of them did get the exhaust thrust to a significant amount.
We have to be careful when trying to figure out what improvement did what.

 

[NTGray]

There is a LOT of information on NACA cowlings available online. It makes for interesting reading.

 

[swampyankee]

For piston-engined aircraft, the cooling system can easily be 25% of the parasitic drag. Both liquid- and air-cooled engines can, in theory, get negative cooling drag. While the Meredith effect gets a lot of mention, ejector exhausts are likely more effective in that regard.

 

[Shortround6]

ejector exhausts are likely more effective in that regard.

and that varies a lot. The US found that if they tried to use more than 3 cylinders on one pipe they lost most of the exhaust thrust.
The Navy R-2800s from mid war on had 6 exhaust pipes minimum. Not sure if they tried 8 pipes.

The P-36 is supposed to have had 22% more drag than the early P-40. The P & W P-40 with the two stage supercharged R-1830 is supposed to have had only 8% more drag that a P-40 (type not stated) and that was in 1943. It no only had exhaust thrust it looks like it was using the exhaust thrust to help move the air out of cowling, much like the FW 190.
Nothing I have seen breaks the drag down to figuring for the exhaust thrust. Which seems like a strange oversight.
However getting good figures on exhaust thrust is also hard as the exhaust thrust varies with speed and altitude even of the engine was making the same amount of exhaust.
Perhaps the drag of the engine installation doesn't jump around as much?

I would tend to doubt that anybody really got an air cooled engine actually give thrust using the cooling airflow. Just look at the size of the radiator duct on the P-51 and the amount of volume that it took up. Trying to slow the cooling air on a radial when you want it to slow down and speed it up when when you want it to speed up and trying to keep it somewhat orientated when you have more changes in direction in the airflow path just seems like it would be a lot harder.
When you get to something like the R-4360 you are trying to keep the engine from melting down, you can use brute force on the cooling air

Path for a single row of cylinders

 

[DarrenW]

The Navy R-2800s from mid war on had 6 exhaust pipes minimum. Not sure if they tried 8 pipes.

FWIW the F6F had a total of 10 exhaust pipes.There were 16 cylinders that shared a pipe and two which had their own pipe exclusively.

 

[Joe Broady]

The C-131 (and I assume the other members of the Convair 240 family) had an unusual cooling system for its R-2800 engines.

"Engine installation on the airplane differs from the usual installation in the use of augmentors extending from just forward of the engine firewall aft through the nacelle to the wing trailing edge. Five of the ten exhaust stack outlets of the engine extend to the forward end of each augmentor. Open space remains between the exhaust stacks and the mouth of the augmentor. When the engine is operating, exhaust gas is ejected from the stacks into the augmentors. The force of the exhaust creates a partial vacuum at the mouth of each augmentor, causing air that has entered the engine section around the propeller hub to rush into the augmentors. This action continuously draws cooling air across the engine. The pump action of the high velocity exhaust gases through the augmentors gives a jet thrust effect as the mixture is forced out of the aft ends of the augmentors. This thrust is considered to be sufficient at low airspeeds and high power settings to provide some positive thrust power. At moderately high airspeeds and lower power settings, the thrust is considered to be sufficient at least to nullify cooling drag. Perhaps the main advantage of the augmentor installation is in flow of cooling air over the engine. The cooling air is automatically proportioned to the power setting at all times due to the aspirating effect of the exhaust gases." (T.O. 1C-131A-1, Flight Manual USAF Series C-131A Aircraft, 1962)

To augment the augmentors there were four nacelle flaps which could be opened as required. Each nacelle flap was smaller than a conventional cowl flap.

 

[special ed]

A not close shot of VT-29 shows augmentor tubes AvPix Unlimited thread page 7 post 134. I don't know how to repost it here.

 

[Shortround6]

The Augmenter tubes and the shrouds/rear of the nacelle moved around a little bit on the various members of the Conair family

A lot of the DH Otters used a simpler system.

These systems were attempts to get more use out of the exhaust from the engine.

Since in WW II fighters the exhaust thrust and the exhaust power are not the same thing they were trying to do something that would benefit slower flying aircraft.

Exhaust thrust is simply mass of the exhaust X the veleocity of the exhaust gas at the end of the pipe/s.
However exhaust power is proportional to the thrust in relation to the speed of the aircraft. An F-15 sitting with it's brakes locked at the end of the runway is making a crap load of thrust but it is making no power. Once it starts rolling down the runway the power increases the fast it goes.

With the Augmenter tubes they were trying to convert the low mass, high veleocity "jet stream" into a higher mass, lower velocity jet stream that matched the aircraft speed better.
Getting the warm air out of the cowl gave them a head start and it helped move the cooling air through the cowling without using gear driven fan or other more complex solution.

It was ingenious doing about 3 things at once with no moving parts

 

[tommayer]

You mean NACA cowlings?

Very effective. A well-designed radial engine installation can have negative cooling drag, although this is quite unlikely (as it is for liquid-cooled engine radiators), but drag comparable to liquid-cooled engines is possible. It's just hard to get there.

Could you provide details of how a "normal" NACA cowl, say what was used on C-47s and C-45s, would have had to be modified to achieve zero cooling drag? Thanks.

 

[swampyankee]

Could you provide details of how a "normal" NACA cowl, say what was used on C-47s and C-45s, would have had to be modified to achieve zero cooling drag? Thanks.

It would be, largely, a matter of detail design, probably using ejector exhausts, which uses the otherwise wasted energy in the engine exhausts to pump air through the cowling. Zero cooling drag in any installation is difficult to impossible.

 

[Admiral Beez]

What of this double cowling? Armstrong Whitworth A.W.16.

Hawker Hoopee

 

[EverKing]

If you go to NASA Technical Reports Server (NTRS) and start a search for something as simple as "cowling" you'll get all sorts of interesting hits.

Some examples:

• Preliminary Investigation Directed Toward Improvement of the NACA Cowling (1942)
• Cooling Characteristics of a Pratt and Whitney R-2800 Engine Installed in an NACA Short-nose High-Inlet-Velocity Cowling (1944)
• The Effect of Streamlining the Afterbody of an NACA Cowling (1939)
• And many more!

What I found when referencing these original NACA reports is to check the references at the end of the report (end of the text portion, before the figures usually) and follow the rabbit-trail back. For example, Reference 1 in that 1939 report I linked above is "Full-Scale Tests of NAVA Cowlings," probably the most pertinent report for the OP question. I went back searched for a 1936-1940 date range and for the primary author of the report, "Theodorsen, Theodore" and found the cited reference, Technical Report 592 Here!

While it may not have direct comparison of Inline v Radial w/ NACA Cowlings, it does at least give you some idea of drag and airflow characteristics of the later which you can then compare to similar data published on liquid-cooled inline installations (you can probably search the NTRS for radiator installation reports as well, including ones on Meredith effect and I know for a fact there are reports on ejector exhausts and how to size/arrange them for best thrust).