About Propellers Other Products Articles Bookshop Gallery Links Contact Supercool
 The Trouble with NACA-4digit airfoil sections



Propeller Dynamics

Essential reading for model aircraft contest fliers. This is the only book on the market explaining propeller theory in non-mathematical terms. A rattling good read, I know, I wrote it.

Read more...

More on cooling airplane engines  
from  Graham White

 

Hi Supercool. 

How about the following to add to your Ducted Cooling Piece

You are one of the few people who have taken a close look at cooling a high performance model airplane engine. As you correctly stated, significant gains are achievable through a carefully designed cooling system. Starting out with the basic premise that heat is energy and if this energy can be recovered in some fashion, it's equivalent to increasing the power of the engine which finally equates to a faster model. 
 

Miss Trinidad

This is R-2000 powered Yak 3 with a Pete Law designed ADI system. Note the attempt at creating exhaust pumping effect through the cowl. Two outlets with seven jet stacks on each side.

North American Aviation spent countless hours in the Cal. Tech wind tunnel tweaking the shape of the P-51 'Dog House' in order to maximize the Meredith effect. There must be dozens of books written on the P-51 but the one I'd recommend is the one written by Robert Gruenhagen; Mustang. The Story of the P51 (Genesis Press). Most airplane books tend to be what I call 'units and markings books', i.e.,. not much in the way of technical detail, engineering aspects or development but lots of details on what ½' stencil markings went where. In Gruenhagens' book, numerous photos are shown depicting the various shapes and designs NAA engineers developed for the dog house.

 

Bristol Hercules engine cut-away drawing

The point of the forgoing is to stress the fact that just because a huge amount of heat energy is dissipated to the atmosphere it does not necessarily imply that it's easy to capture. A good example of this is the Hawker Hurricane. Similar placement of the cooling system as the P-51 but absolutely no Meredith effect. With an air cooled radial, things get a little trickier due to the confines of the cowl. However, a carefully designed cowl can, at the minimum, reduce cooling drag. To pull this off, as you correctly stated, take a look at a top unlimited Reno Racer such as Furias, a modified Sea fury powered by a Pratt & Whitney R-4360-63A or Rare Bear a modified F8F Bearcat powered by a 'bitsa' Wright R-3350. 
 

Bristol Hercules cylinder

In the case of Furias, an annular gap of 2' is used and this has to take care of the cooling requirements of a huge twenty-eight cylinder, four row radial. So how's this pulled off? Fortunately, these ultimate hot rods have the expertise of the world's best aerodynamicists at their disposal. The late Bruce Boland, designer of the late lamented Tsunami, was one of these experts who gave freely of his time and expertise to assist the racers. Pete Law, another Lockheed Martin thermo dynamics expert also assists with these projects.
 

Critical Mass

Note the after body. Highly modified Sea Fury powered by a 3350 made up form the best components.

 Getting back to Furias, the trick, as you mentioned in your article, is to slow down the incoming air. This is achieved via a classic convergent/divergent duct. In other words, a larger diameter spinner directs the air through the aforementioned 2' annular gap and from there, an aft fairing that covers the nose case and magnetos terminates a the base of the first row of cylinders. The aft faring also tapers down as it extends to the cylinders. The inside of the cowl receives similar treatment, everything being flowed for ideal expansion and pressure increase. Even these measures are not sufficient to cool an R-4360 so seven spray nozzles are built into the leading edge of the cowl nose bowl. At race powers, water is injected into each of the seven cooling plenums that make up the 4360 cooling system. 
 

Hi-Tech Fosters

This is race #38 Precious Metal powered by a Griffon 58 with a 74 blower and Bendix PR-100 carb.

In other words, raw water is squirted on to the hot (500 degree head temps) cylinders. Additionally, copious amounts of ADI (anti detonation injection) fluid is injected directly into the engine. ADI fluid is made up of a 50/50 mix of water and methanol. Again, my good friend Pete Law is responsible for 90% of the ADI systems used by the Unlimited class racers. 
 

Sea Fury oil cooler

Note the nice entry to the cooler core. It's a new cooler that uses fin and tube technology rather than the stock matrix cooler. Also note the Pete Law designed spray bar.

Yet another source of heat rejection, obviously not applicable to a two-stroke is oil cooling. Pete has designed a number of 'boiler' systems whereby the oil cooler, typically a bundle of copper tubes which allows air to flow through the tubes and oil to flow around the exterior of the tubes. By immersing the oil cooler in a bath of water and allowing the hot oil to boil off the water, cooling drag from the oil cooler is eliminated. 
 

Rare Bear afterbody

Note the fairing over the nose case that extends to the base of the front row of cylinders. This shape creates a nice convergent/divergent flow path for the cooling air.

 

Rare Bear ram scoop

Notes the two weld beads that support the flow splitters to direct ram air into the Bendix PR-58 carburetor.

Another Pete Law innovation is the use of spray bars. In other words, the oil cooler, which is located in a duct and exposed to the air stream, has water sprayed on it from a spray bar system. Typically, the spray bars are made from aircraft fire extinguisher tubing. Spray bars are also used on coolant radiators for P-51s. It's a little known fact that between ADI fluid, spray bar fluid and fuel, Dago Red, the worlds' fastest P-51, will consume over 1,000 pounds of fluid in a 15 minute race. That's right, ½ ton of fluid in 15minutes. Of course, a 4360 powered racer such as Furias will consume considerably more. Lets take a look at engine design.

 

The attached graph is one Pete Law generated for a presentation to racers . They were astounded at the temperature rises shown on this chart. And this temp. rise is not just for induction air, it's applicable to all air taken on board. That's why it is key to take on the minimum amount air for cooling requirements ® or any other requirements. A good example is a radial engine installation. Many folks, including yours truly at one time, thought that the drag caused by opening cowl flap on a radial installation was due to the parasitic drag of the flaps sticking out in the breeze. As it turns out, not so. The vast majority of the drag is induced drag due to delta MV component, the parasitic drag is inconsequential. Pete spent 2 hours on the phone with me yesterday explaining how all this works. It was tough sledding (for me) because Pete has an IQ in the stratosphere so when having a conversation with him you'd better be prepared for some heavy duty laws of physics. The foregoing also helps explain why spray bars are so effective, they minimize the mass air flow requirements for cooling thus contributing to a remarkable reduction in drag. This may be effective on models ® but maybe not. Dunno; but it's worth a shot.

Air cooled radials were hampered by inadequate cooling in the early years until luminaries such as Sam Heron figured out the best way to manufacture an air cooled cylinder. His method was surprisingly similar to that employed by Cox. Screw the entire cylinder head onto the cylinder. The only difference being the fact Cox reversed the gender of the treads, i.e., they used a male thread on the head and a female thread on the cylinder ® other than that they were identical in concept. I have often wondered if Cox got the idea when they first designed the Thimble Drome in the late '40s from full size practice. 
 

Attached is remarkable piece of work that Pete Law did. Obviously not applicable to models but nevertheless neat stuff.

Pete was instrumental in Darryl Greenamayer's record breaking attempts with the Lockheed F-104 that got the world's air speed and altitude records. So I've attached a graph that shows ram temp rise for that aircraft. The F-104 info. Is a the lower left, the rest of the chart shows SR-71 performance figures ® Pete was the chief thermo guy on that project. This stuff is all now de-classified but Pete told me that Air Force pilots cruised at a 'conservative' Mach 3 on missions..!!! Pete also developed a water injection system for both the F-104 and the SR-71 in order to reduce ram temp. rises. It was never fitted to the SR-71, they figured the Russians couldn't catch it even without water. But one could just imagine the instant vaporization of the water as it hit 800 degree air.

After Heron established the basic concept of how to design an air cooled cylinder, the next hurdle was to increase the number and depth of cooling fins. By the early 1940s it was apparent that the limitations of casting technology had been reached, but horsepower was now restricted because heat could not be rejected fast enough. The answer was actually developed in the late '20s, early '30s by Roy Fedden who was responsible for all the mighty Bristol radials. The solution was to forge the cylinder head out of a solid aluminum billet and then cooling fins could be machined in as closely and deep as necessary. By the end of the big high horsepower radial era, cylinders on engines such as the R-4360 were absolute jewels. Finning was so close and deep, they actually looked fragile, which in fact they were. 
 

Michael Brown's ram scoop

Michael Brown picked up an additional 10in.Hg. manifold pressure or about 400 ® 500 horsepower with this new (for 2002) ram scoop on his 3350 powered Sea Fury. Note the bead of weld as the scoop transitions through 90 degrees. This is the flow splitter inside the scoop. He finished 2nd. In the final Gold Race.

However, it was the only way to get the maximum surface area exposed to the cooling air stream which is a nice segway into the next aspect of cooling air cooled engines. When cooling air enters a cowl, it's important to ensure that every ounce of that air is put to good use by cooling something. No use in just having a cylinder stuck in the breeze and hope that some air will do its job. To this end, Pratt & Whitney, Bristol and all the other manufacturers of large air cooled radials spent hundreds of design hours perfecting baffles that would capture cooling air as it entered the cowl and ensured that it was forced around the cylinder and through the fins. Bristol, in particular, had a significant challenge ensuring that its sleeve valve engines received sufficient cooling air down into the depths of the junk head. Once the air had passed through the cylinder it had to be dumped over board. But even here, careful design of the outlet directed cooling air through a fairing colloquially known as a dish pan. Mass air flow through the cowl was controlled via a ring of flaps called cowl flaps. The greater the mass air flow through the cowl, the greater the drag. So it was incumbent upon the flight engineer to ensure that the ideal cylinder head temperature was maintained; too high and the engine would be cooked, too low and excessive drag was created. 
 

Furias

Furias powered by an R-4360-63A with a -59 nose case. Runs 68in. Hg. at 3,100rpm. Quite conservative for a racer. Laps around 420.

 

Furias Spray nozzles

This shows the 2' annular gap and how the spray nozzles are arranged around the nose bowl. Hard to imagine that such a narrow gap can feed enough cooling air to a massive 28 cylinder 4,360 cubic aircraft engine.

Exhaust is another good source of energy. There are several ways to utilize exhaust energy, most of which would be inapplicable to a model but worth looking at just the same: (i) turbosupercharging, (ii) jet stacks, in othe words utilize the thrust which is typically 400 pounds for a Merlin, (iii) augmenter system and (iv) turbo compounding whereby the exhaust energy is used to drive a gas turbine which feeds power back to the engine. 
 

Strega's propeller blade

Much of the foregoing has absolutely no application or use in model applications with one or two exceptions. First off, the interior shape of the cowl is just as important, if not more so, than the exterior. So it would be a good idea to design a convergent/divergent shape into the front half of the cowl. Next, cooling air that enters the cowl should all be put to good use doing what it's supposed to do; carry away heat rejected through the cooling fins. And once the air has picked up the rejected heat, the exit duct should, again, be given a lot of thought. This includes the shape and the exit area. It may be possible to design an augmenter system whereby the exhaust pulses are used to 'pump' cooling air through the cowl. Convair designed a twin engined commuter aircraft powered by a pair of Pratt & Whitney R-2800s in the late 1940s. The eighteen exhaust stacks were utilized to pump cooling air through the cowl. Exhaust and cooling air were mixed and dumped overboard over the trailing edge of the wing. It was claimed that 500 pounds of thrust and 20mph resulted from this innovation. Cast cooling fins are not the optimal design. Machining them in from solid billets offers a lot more flexibility to the extent more and deeper cooling fins can be used and thus reduce cooling drag. Remember, the less air that enters the cowl, the less drag results.
 

Rolls Royce power

The attached pic. shows my R-R Griffon (in the bed of my truck) and the Merlin. As I mentioned, I'm in the process of returning to Florida so I've temporarily parked those two engines in a local museum.

Another aspect that hasn't been touched upon is that of induction ram recovery. With a front induction engine, a carefully designed ram duct for the induction system could gain 1 to 2 in.Hg. at 150mph. This is a significant amount. Again, look at a high performance WWII fighter for good ram air induction systems. The opening, which should be normal to the air flow, must be sized correctly; too large and air spills out of the ram air duct and creates drag. Too little and full ram recovery is not possible. If, for arguments sake, the induction venturi is at 45 degrees, a duct that 'bends' the air from horizontal to 45 degrees is required. And it may even need a flow splitter within the duct to further assist the ram effect. For a rear induction engine a similar duct can be used and again, the duct should discharge into the venturi, preferably well sealed. Only with this system the air needs to bend through 180 degrees.
 

1.5" annular gap on Sea Fury

This is Nelson Ezell's Sea Fury powered by a Wright R-3350-26W. The annular gap is 1-1/2 inches and it cools just fine. Unfortunately, I couldn't get a picture of the after body but many of the 3350 powered Sea Furys' use a 1-1/2' or 2' gap. Of course, a lot of heat is rejected through the oil coolers which have spray bar assist.

For now, we won't go into reduction gearing but it does not have to be difficult or bulky. The Farmen epycyclic system offers the most compact system that would easily fit within the confines of a speed pan.
 

P51 Mustang cooling duct

Authors' background. Recently joined NASS. In the late 50s and early 60s I did all classes of tam racing. Currently I collect full size engines and restore them to running condition. Presently have (2) P& W R-4360s. R-R Merlin, R-R Griffon, P&W R-2800 and Continental IV-1430. Authored two books on aircraft engines: 'Allied Aircraft Piston Engines of World War II' and R-2800, Pratt & Whitney's Dependable Masterpiece'. Both books are available through the publisher www.SAE.org


Author of numerous magazine articles on automotive and aviation subjects.

I'm also the executive editor of 'Torque Meter' the journal for the Aircraft Engine Historical Society. www.enginehistory.org

Very truly yours, Graham White.

 Back to top