liquid cooling info

[airguy]

While doing some research on the cooling system used by the Galloping Ghost P51 racer at Reno, I ran across this interesting piece about designing radiators for liquid-cooled aircraft engines. It has some good info and some good literature references, I thought you guys might find it interesting.

http://protonet.org/doc/Liquid_cooling_5_LTR.pdf

 

[TSwezey]

Great info and he probably put that liquid cooled stuff in my brain back when I was one of his students in 1983 and 1984.

 

[elippse]

There were also articles in several issues of Contact! magazine. Check out their back issues on their site, contactmagazine.com, such as #45 which devotes 8 pages by Hans Mayer to the duct designs on many American, English, and German WWII planes. Also #62 on radiators, and #53 on the Mustang cooling by J.L. Atwood, former president of North American. Contact! has a wealth of info on liquid-cooled engines, including augmented cooling, a real treasure-trove of info!

 

[JDanno]

Todd,
You an Aggie?
My Dad graduated there in 1958 with Phd. Physics. It was still male only then. Dan

Great info and he probably put that liquid cooled stuff in my brain back when I was one of his students in 1983 and 1984.

 

[rv6ejguy]

The protonet article is one of the best I've read on cooling and it takes a few reads for it all to sink in. This was published in Contact! many years ago. If you apply the knowledge here on rad placement and duct shape, chances are things will cool well.

Hoerner's book is probably the best published on this subject with actual testing.

The Hans Mayer articles draw some erroneous conclusions and have some questionable information in my view although some of the information is useful. The high diffusion angle depicted in his drawings would doubtless cause serious separation within the duct even with a guide vane.

 

[pierre smith]

Todd, my late boss asked about vortex generators...

...... The high diffusion angle depicted in his drawings would doubtless cause serious separation within the duct even with a guide vane.

....inside the upper chamber shelf, leading to the radiator on his P-51. Would this not be an appropriate place for VG's?

Thanks,

 

[TSwezey]

Todd,
You an Aggie?
My Dad graduated there in 1958 with Phd. Physics. It was still male only then. Dan

Class of '86, Corps of Cadets, BS Aerospace Engineering

 

[TSwezey]

....inside the upper chamber shelf, leading to the radiator on his P-51. Would this not be an appropriate place for VG's?

Thanks,

Pierre I don't have a clue about any of that but Ross might know. Not knowing the flow it would probably best be tested in a wind tunnel.

 

[rv6ejguy]

Guide vanes are more effective in gently turning the air within a duct without turbulence. Studies after the war found most under wing submerged rad setups had serious flow separation and consequently higher drag and less cooling.

High divergent angles are usually bad but we are limited by practical duct lengths in most aircraft to diverge airflow gently- yet another compromise.

With regards to Galloping Ghost at Reno with the boil off/ loss type cooling system eliminating the rad, this has been tried before many years ago and it made no difference in speed. The modified P51 rad ducts like Strega and Voodoo have been highly refined so that they offer no net drag and possibly even some net thrust with the high speeds and spray bar water being injected.

Despite the less than good press on the P40 rad setup, tests during WW2 with ventral rads like the P51 had always showed a clear loss in speed over the under engine setup on this airframe. There are many factors in play here.

 

[SHIPCHIEF]

I had read somewhere that Curtiss tried a P-51 type radiator on the P-40, but it did not work for them.
P-40's certainly were fast in a straight line or a dive. I have seen one fly with a P-51 at an airshow. It was during the pull-out with a turn that the P-40 lost energy while the P-51 retained more. I don't think that had much to do with the cooling layout, but rather the wing's span loading and airfoil efficiency in a high lift situation.
But back to the cooling; When I began to realize that mounting the cooling system aft of the wing spar required carrying the weight of all that tubing and the coolant therein, I went back to the chin radiator.
As for a boil off cooling system, you are back to the excess weight problem; how can you carry enough total loss coolant to offset the weight penalty of a radiator over the course of any kind of trip?
Add to that, can you dependably get coolant at any airport where you might land?
What about coolant chemicals? plain water leads to corrosion inside your expensive engine, so total loss expensive chemicals is out too.
When you look at the whole airplane, Simple and Light are the best way to go. Never forget that.
Alternative engines still have to measure up to the power output, power to weight ratio, fuel consumption, usability and relaibility of traditional aircraft engines.
Lately I've been trying to understand "Plume Drag". That's the drag caused by the air exiting your aircraft, from leaks around the canopy, rudder cables, engine cooling air etc. Perhaps the Galloping Ghost's cooling system suffers from that instead of traditional radiator form drag? Sometimes you trade one kind of problem for another. That's OK if you're looking for a challenge.
I'm not building a alternative engine because I think Lycoming makes a bad engine.

 

[TSwezey]

You really need a wind tunnel to fully test everything. Aerodynamics is a bunch of theory perfected by real world testing. Todays super computers can almost simulate real world conditions but I don't think anybody has the money to do it for RV testing. I know my cowl was like having a parachute up front. I had way too much cooling which is better than not having enough.

 

[rv6ejguy]

I had read somewhere that Curtiss tried a P-51 type radiator on the P-40, but it did not work for them.
P-40's certainly were fast in a straight line or a dive. I have seen one fly with a P-51 at an airshow. It was during the pull-out with a turn that the P-40 lost energy while the P-51 retained more. I don't think that had much to do with the cooling layout, but rather the wing's span loading and airfoil efficiency in a high lift situation.
But back to the cooling; When I began to realize that mounting the cooling system aft of the wing spar required carrying the weight of all that tubing and the coolant therein, I went back to the chin radiator.
As for a boil off cooling system, you are back to the excess weight problem; how can you carry enough total loss coolant to offset the weight penalty of a radiator over the course of any kind of trip?
Add to that, can you dependably get coolant at any airport where you might land?
What about coolant chemicals? plain water leads to corrosion inside your expensive engine, so total loss expensive chemicals is out too.
When you look at the whole airplane, Simple and Light are the best way to go. Never forget that.
Alternative engines still have to measure up to the power output, power to weight ratio, fuel consumption, usability and relaibility of traditional aircraft engines.
Lately I've been trying to understand "Plume Drag". That's the drag caused by the air exiting your aircraft, from leaks around the canopy, rudder cables, engine cooling air etc. Perhaps the Galloping Ghost's cooling system suffers from that instead of traditional radiator form drag? Sometimes you trade one kind of problem for another. That's OK if you're looking for a challenge.
I'm not building a alternative engine because I think Lycoming makes a bad engine.

The Allison engines in the P40 was well down on power compared to the Merlins in the P51s and the 2 stage, 2 speed superchargers made a vast difference in altitude performance. The ventral radiators tried on the P51 showed a loss of over 20 mph both times they were tried. Curtiss returned to the chin rad setups for production aircraft. The point being here is than well designed chin setups can be low drag and this configuration is more suitable for RVs.

Boil off systems are only practical for race aircraft but most times they have been tried, no net gain in speed was evident.

Most of the clipped winged P51s have had a good portion of the air leaks plugged to reduce plume drag, control surfaces being the main culprit. Any time you have air exiting at a tangent to the free stream, especially at below free stream velocity, you get separation and turbulence. This would not directly affect Galloping Ghost as the radiator duct has been completely removed and faired over and only a tiny aft facing vent exit for the steam exists.

 

[SHIPCHIEF]

Ross;
So what you are saying, is that the Galloping Ghost' coolant exhaust can be pressurized steam ejected rearward thru a nozzle, for negative drag Wow, that sounds like thrust I didn't think of it like that when I wrote my reply.
Still might not work for a 400 mile cross country flight to Death Valley.
I just looked up the Saturation table for steam. 300 degree F jacket water would be 67 PSI steam pressure. Minus some pressure drop to get it back to the nozzle, but still, that WOULD be thrust. Can a Merlin survive with 300F coolant temp?

 

[airguy]

Ross;
So what you are saying, is that the Galloping Ghost' coolant exhaust can be pressurized steam ejected rearward thru a nozzle, for negative drag Wow, that sounds like thrust I didn't think of it like that when I wrote my reply.
Still might not work for a 400 mile cross country flight to Death Valley.
I just looked up the Saturation table for steam. 300 degree F jacket water would be 67 PSI steam pressure. Minus some pressure drop to get it back to the nozzle, but still, that WOULD be thrust. Can a Merlin survive with 300F coolant temp?

I believe they are running a pressurized coolant loop, with a heat exchanger in a boiler to cool the coolant and send it back to the engine. This limits your steam temp (and pressure) considerably, I would be surprised if they run more than atmospheric pressure for the steam discharge.

 

[rv6ejguy]

Any thrust would be minimal due to the low mass ejected. This is in contrast to the highly developed, non standard, scoops used on Voodoo, Strega and Dago Red where rad air plus spray bar water mass is heated, converged and ejected at maximum velocity parallel to the free stream.

 

[TSwezey]

Any thrust would be minimal due to the low mass ejected. This is in contrast to the highly developed, non standard, scoops used on Voodoo, Strega and Dago Red where rad air plus spray bar water mass is converged and ejected at maximum velocity parallel to the free stream.

It would be practically non-existent!

 

[SHIPCHIEF]

I read the link at the head of this thread, and sat back a few days to think of my own cooling system.
My radiator is in the chin and has no upper lip. I'm presuming that the airflow under the prop is laminar because the airflow presents directly to the lower cowl face.
The lower lip of the scoop is not raked back much, as a low subsonic speed would make a very blunt bow wave I hope to get some minimal pressure recovery across the opening. But I worry about flow separation inside the lower lip at low airspeed with steep climb, right when it's needed most with the engine at climb power. The angle of the lower cowl face plus the divergent lower scoop wall would be a pretty sharp change of direction for the incoming air in this condition.
I'm using a fairly thin Griffin radiator to assure good airflow at low recovery pressure, and the scoop opening is fairly large, about 26% of the radiator face area. But the length of duct is too short to make any meaningful divergence. I mostly read as much as I could, then used TLAR from there....After my first design would not cool during ground runs.
Now I have effective water and oil cooling during ground runs to 25% power. I have not tested at higher power yet. I wonder how high % of power should it cool before 'brake release'? We have all read of first flights terminated after a single circuit with a hot engine, and I don't want to go there.
I can accept large drag at high speed initially if good cooling is the result. Eventually reducing any excess cooling drag as time goes on.

 

[Rotary10-RV]

Some things are some things aren't (true)

The Allison engines in the P40 was well down on power compared to the Merlins in the P51s and the 2 stage, 2 speed superchargers made a vast difference in altitude performance. The ventral radiators tried on the P51 showed a loss of over 20 mph both times they were tried. Curtiss returned to the chin rad setups for production aircraft. The point being here is than well designed chin setups can be low drag and this configuration is more suitable for RVs.

Boil off systems are only practical for race aircraft but most times they have been tried, no net gain in speed was evident.

Most of the clipped winged P51s have had a good portion of the air leaks plugged to reduce plume drag, control surfaces being the main culprit. Any time you have air exiting at a tangent to the free stream, especially at below free stream velocity, you get separation and turbulence. This would not directly affect Galloping Ghost as the radiator duct has been completely removed and faired over and only a tiny aft facing vent exit for the steam exists.

Ross carrying water to boil off also increases HP required to maintain level flight. The amount of extra weight carried can slow the plane considerably. They were probably hoping that they would be lighter when it really counts, at the finish!
With regard the P-51, P-40 layouts there is simply no question every NACA and War Report paper lists the underbelly scoop with a boundrylayer seperator as lower drag than the P-40 style chin radiator. The P39 might be considered to have a "chin" radiator is is under the engine.
The Allisons produced excellent power at low altitudes, in fact the Allison powered P-51 prototype was faster than the Merlin, AT SEA LEVEL unsupercharged. The requirement was to fly faster and higher than a B-17 at altitude, where the Merlin two stage supercharged system was totally superior. Comparing the layouts has too may variables. Even when willing to go to the trouble of a ring radiator like some of the FW designs you often create more drag when rejoining the exhaust air in a low pressure area. Unless you have a wind tunnel your going to need to adhere to good basic designs. The most common thread I have seen on all the papers I've read is that "getting the bad air out" is the most critical part of a duct design. The most often ignored feature on homebuilts is that having a clear area behind the radiator is as important as having a good duct leading to one. This is rarely possible in inside the cowl designs.
Bill Jepson

 

[rv6ejguy]

Ross carrying water to boil off also increases HP required to maintain level flight. The amount of extra weight carried can slow the plane considerably. They were probably hoping that they would be lighter when it really counts, at the finish!
With regard the P-51, P-40 layouts there is simply no question every NACA and War Report paper lists the underbelly scoop with a boundrylayer seperator as lower drag than the P-40 style chin radiator. The P39 might be considered to have a "chin" radiator is is under the engine.
The Allisons produced excellent power at low altitudes, in fact the Allison powered P-51 prototype was faster than the Merlin, AT SEA LEVEL unsupercharged. The requirement was to fly faster and higher than a B-17 at altitude, where the Merlin two stage supercharged system was totally superior. Comparing the layouts has too may variables. Even when willing to go to the trouble of a ring radiator like some of the FW designs you often create more drag when rejoining the exhaust air in a low pressure area. Unless you have a wind tunnel your going to need to adhere to good basic designs. The most common thread I have seen on all the papers I've read is that "getting the bad air out" is the most critical part of a duct design. The most often ignored feature on homebuilts is that having a clear area behind the radiator is as important as having a good duct leading to one. This is rarely possible in inside the cowl designs.
Bill Jepson

All the Unlimiteds carry vast amounts of spray bar water anyway (over 1000 lbs in some cases) so this is all even and induced drag varies inversely with speed so the weight does not make a huge difference at 500 mph.

Actually Curtiss tested ventral rads on 2 types of P40s and they were noticeably inferior to the chin setup when tested for top speed. Lots of people don't believe this but I've got two sources stating this. It's all in the details and sometimes looks are deceiving.

Strega, Voodoo and Dago omit the boundary layer separator. I don't think anyone knows more than these folks about liquid cooled installations as they have done some 500+ mph laps at Reno.

 

[airguy]

Strega, Voodoo and Dago omit the boundary layer separator. I don't think anyone knows more than these folks about liquid cooled installations as they have done some 500+ mph laps at Reno.

Except that those same racers generally have the ability to make more power to overcome drag at will - but they are pretty much against the wall as far as dissipating heat. They are (generally speaking) willing to add drag for better cooling because they know they can generate sufficient HP to overcome that drag. Engine temps are the limitation they are up against now. It's been a few years now since I've even suspected that Strega was giving all she had, they only run these birds hard enough to maintain the desired finishing order unless it's a passing situation in the last lap.

Galloping Ghost is an unknown - she can generate the power, there is no doubt about that. "Unofficial" speeds of 550mph have been reported during testing at Minden, UT on a simulated Reno course in prep for Reno 2010, and then of course we didn't get to see her run flat out in the Gold on Sunday due to high winds, so nobody really knows (except Jimmy Leeward and crew) how she'll handle it and what she'll do. I talked to a few of the crew about the cooling system and was told that they use one wing tank for race fuel and one for boil-off water, presumably both are very low at the end of the race for minimum weight.

 

[skylor]

...induced drag varies inversely with speed so the weight does not make a huge difference at 500 mph.

This would be true in level flight, but since these guys have to turn as well (typically 4-6 G's) induced drag due to weight does have some significance since induced drag itself is increased by load factor SQUARED...

Skylor

 

[rv6ejguy]

This would be true in level flight, but since these guys have to turn as well (typically 4-6 G's) induced drag due to weight does have some significance since induced drag itself is increased by load factor SQUARED...

Skylor

My point was that all Unlimiteds already carry many hundreds of pounds of spray bar water and ADI and you have to do the same to be competitive so the playing field is level for everyone. Speeds are not significantly different from second lap to last lap so I still contend that weight is not a big factor here.

 

[rv6ejguy]

Except that those same racers generally have the ability to make more power to overcome drag at will - but they are pretty much against the wall as far as dissipating heat. They are (generally speaking) willing to add drag for better cooling because they know they can generate sufficient HP to overcome that drag. Engine temps are the limitation they are up against now. It's been a few years now since I've even suspected that Strega was giving all she had, they only run these birds hard enough to maintain the desired finishing order unless it's a passing situation in the last lap.

Galloping Ghost is an unknown - she can generate the power, there is no doubt about that. "Unofficial" speeds of 550mph have been reported during testing at Minden, UT on a simulated Reno course in prep for Reno 2010, and then of course we didn't get to see her run flat out in the Gold on Sunday due to high winds, so nobody really knows (except Jimmy Leeward and crew) how she'll handle it and what she'll do. I talked to a few of the crew about the cooling system and was told that they use one wing tank for race fuel and one for boil-off water, presumably both are very low at the end of the race for minimum weight.

I think any of the 3 fast P51s mentioned can top 540 straight and level for short periods. Strega and Dago have both run laps of over 500 mph at Reno so they are doing well over 500 mph on the straights.

Engine mechanical strength is the worry in Unlimited Racing not water temps so much. My point about the lack of rad scuppers on the the fast '51s is that they would not do it if it increased drag. Flat plate area is critical at these speeds so the lower profile scoop with suitable mods does the job for them.

When the boil off systems are proven to be superior with some race wins over Strega and the likes, I'll buy it. 20 years of incremental refinement on Strega will be hard to beat as many have found out.

Longevity is mainly a function of manifold pressure and duration. You run minimum MAP to win the race because these engines are grenades with the pins pulled. It is unlikely any would survive the full 145+ inches for the entire race duration.

It is already common on the other aircraft to have one wing tank dedicated to spray bar water and one ammo bay dedicated to ADI. I believe Dago carries 900 pounds of spray bar water and 345 lbs. of ADI for the race.

Getting back to the whole boil off concept for RVs, it is not an option for cross country flight, strictly useful only for very short duration races or record attempts.