Radial vs liquid cooled engines

Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules

29
0
Feb 22, 2012
Was any engine better suited for combat during WW2 than the other? I know radials offered greater durability, but did they offer any other advantages? And did liquid cooled engines have any inherent advantages over radials?
 
Define durability?

They may have been more resistant to combat damage due to no cooling system but as far as engine life goes ( hours before/between suggested overhaul) those were all over the place for both types depending on particular engine model, country, type of service ( fighter, bomber, transport), conditions of service ( grass field in England, North African sand, Russian steppe dust, Pacific coral rock dust) and how hard the engine was "pushed". WEP settings or water injection/MW-50.

Liquid cooled engines had less drag for a good part of the war. V-12s can have a gun firing through the prop hub. Liquid cooled engines (again, depending on engine) could usually tolerate more boost.
 
Radial engines have the following advantages:
1. Because a radial engine cannot be deprived of its coolant (air) it is much more resistant to battle damage than an LC engine, which can be put out of action by a single hit to any part of the liquid cooling system.
2. Because all cylinders of a radial engine are exposed to similar air flow the engine tends to maintain a universal operating temperature. In contrast the different cylinders of an inline or V liquid-cooled engine receive less progressively less benefit as the coolant moves through the block and heats up.
3. An radial engine can be installed and removed much more easily than an LC engine as it is not hooked up to a separate cooling system. It is also easier to service as the cylinder heads are simple to access and the radial is effectively restricted to push-rod activated two-valve heads (this last is a disadvantage also – see below – but it does make the service intervals longer )
The liquid cooled engine has the following advantages:
1. An LC engine does not need to expose all cylinders to the airflow, which greatly reduces drag compared to a radial.
2. Water is a much denser medium than air and will therefore remove much more heat from an engine for a given volume. This means that a liquid cooling system can be made to control engine heat at operating levels where an air cooled engine would not be able to cope.
3. Because the inline or V configurations enabled by liquid cooling have cylinders aligned in banks, a single camshaft and chain can be used to actuate valves in multiple cylinder heads. This opens up the option of multi-valve heads, increasing the efficiency with which the cylinder imports fuel mix and exports exhaust gasses, leading in turn to fuel savings and/or increased power. As the radial's cylinder heads are isolated from each other, the same kind system for a radial engine would require separate cam chains and camshafts for each individual cylinder head, a scarily complex proposition.
 
There are some generalisations in that.

Not all air cooled engines were radials. And not all radials were air cooled. Granted, the major production radials were all air cooled and only a few liquid cooled radial prototypes were made.

Radial engines have the following advantages:
1. Because a radial engine cannot be deprived of its coolant (air) it is much more resistant to battle damage than an LC engine, which can be put out of action by a single hit to any part of the liquid cooling system.

Radial engines are also far more likely to have lose things like cylinder heads. They are also larger targets. Air can be deprived from the cooling system on an air cooled engine.


Radial engines have the following advantages:
2. Because all cylinders of a radial engine are exposed to similar air flow the engine tends to maintain a universal operating temperature. In contrast the different cylinders of an inline or V liquid-cooled engine receive less progressively less benefit as the coolant moves through the block and heats up.

I wouldn't say that is necessarily the case. Rear cylinders in multi-row radials have a harder time, for instance.

As for liquid cooled engines, it depends where the coolant enters and exits, and the coolant path. They were not all the same.


Radial engines have the following advantages:
3. An radial engine can be installed and removed much more easily than an LC engine as it is not hooked up to a separate cooling system. It is also easier to service as the cylinder heads are simple to access and the radial is effectively restricted to push-rod activated two-valve heads (this last is a disadvantage also – see below – but it does make the service intervals longer )

That is a sweeping generalisation. Liquid cooled engines with annular radiators could be changed in the same amount of time as the equivalent air cooled QEC. But you're right, it can take a lot of time to disconnect two hoses, particularly if they have quick disconnect fittings.

Ai rcooled engines were not restricted to 2 OHV. Some Bristol radials had pushrod operated OHV, before they went sleeve valve. The Armstrong Whitworth Deerhound had 3 or 4v per cylinder (can't recall) operated by single overhead camshafts.


The liquid cooled engine has the following advantages:
1. An LC engine does not need to expose all cylinders to the airflow, which greatly reduces drag compared to a radial.

Assuming we are talking V-12s, the LC engine has much less frontal area, which partially explains drag.


Radial engines have the following advantages:
2. Water is a much denser medium than air and will therefore remove much more heat from an engine for a given volume. This means that a liquid cooling system can be made to control engine heat at operating levels where an air cooled engine would not be able to cope.

It also allows tighter fits in piston/bore, reducing tthe amount of oil burned in the engine. Liquid cooling allows better cooling control, IMO.


3. Because the inline or V configurations enabled by liquid cooling have cylinders aligned in banks, a single camshaft and chain can be used to actuate valves in multiple cylinder heads. This opens up the option of multi-valve heads, increasing the efficiency with which the cylinder imports fuel mix and exports exhaust gasses, leading in turn to fuel savings and/or increased power. As the radial's cylinder heads are isolated from each other, the same kind system for a radial engine would require separate cam chains and camshafts for each individual cylinder head, a scarily complex proposition.

No chains involved in (most) aero engines. Most V-12s used geared shaft drive to the cams. Far mor reliable than belts.

As mentioned before, the AS Deerhound had overhead cams driven by geared shafts. So to the Lycoming XR-7755.

Fuel/air mix distribution to cylinders was a problem for both V-12 and radial engines during WW2, and had more to do with the intake side of things rather than the number of valves.
 
There are some generalisations in that.

Not all air cooled engines were radials. And not all radials were air cooled. Granted, the major production radials were all air cooled and only a few liquid cooled radial prototypes were made.



Radial engines are also far more likely to have lose things like cylinder heads. They are also larger targets. Air can be deprived from the cooling system on an air cooled engine.




I wouldn't say that is necessarily the case. Rear cylinders in multi-row radials have a harder time, for instance.

As for liquid cooled engines, it depends where the coolant enters and exits, and the coolant path. They were not all the same.




That is a sweeping generalisation. Liquid cooled engines with annular radiators could be changed in the same amount of time as the equivalent air cooled QEC. But you're right, it can take a lot of time to disconnect two hoses, particularly if they have quick disconnect fittings.

Ai rcooled engines were not restricted to 2 OHV. Some Bristol radials had pushrod operated OHV, before they went sleeve valve. The Armstrong Whitworth Deerhound had 3 or 4v per cylinder (can't recall) operated by single overhead camshafts.




Assuming we are talking V-12s, the LC engine has much less frontal area, which partially explains drag.




It also allows tighter fits in piston/bore, reducing tthe amount of oil burned in the engine. Liquid cooling allows better cooling control, IMO.




No chains involved in (most) aero engines. Most V-12s used geared shaft drive to the cams. Far mor reliable than belts.

As mentioned before, the AS Deerhound had overhead cams driven by geared shafts. So to the Lycoming XR-7755.

Fuel/air mix distribution to cylinders was a problem for both V-12 and radial engines during WW2, and had more to do with the intake side of things rather than the number of valves.

Hi Wuzak,
I think we are at risk of repeating our discussion in the fighter bomber thread, though we seem to be finding a bit more common ground.
I'm sorry but no, you can't deprive an air-cooled engine of air. Either the flaps were open when hit, in which case they might get locked open, or they were closed when hit, in which case the plane was flying with them closed and will continue to do so with them locked closed. In contrast, once a liquid cooled engine lost its coolant there was ZERO cooling and the thing would seize up in short order. I've heard of plenty of LC aircraft overheating and going down after being hit in the radiator but I can't recall any where an air-cooled plane seized up due to damaged cowl flaps. And how likely is the cowl mechanism of an air-cooled motor to be hit compared to a liquid cooling system that encompassed the entire top end of the motor and had a radiator hanging out in the breeze?
Re the cylinder heads of a radial more exposed to being shot off, radial engine fighters can and did sustain such damage and keep on operating. The same damage to a LC engine would result in rapid loss of coolant and inevitable engine failure
How was my observation that radial engines are easy to remove a 'sweeping generalisation'? You mentioned LC engines with annular radiators as being just as easy to install and uninstall but how many combat aircraft were thus equipped? The Ju88 and Fw190D/Ta152 sure, but the vast majority of LC engine aircraft have had radiators attached to the airframe
Regarding multi valve heads on radial engines I said the radials were EFECTIVELY limited to two valves per cylinder, as were all the widely used radials of WWII I can think of. I didn't say it was impossible to make a multi valve radial, it was just complex and impractical. How many Armstrong Deerhounds were produced? Ten, I think. And the Lycoming had a production run of two, both of which were extremely unreliable.
To put it into perspective, a V-12 A Rolls Royce Merlin had two drive shafts driving two cams, one for each bank of six cylinders. A twelve cylinder radial with bevel-drive cam-actuated valves would require twelve drive shafts operating twelve cams – six times as many moving parts as the Merlin for the for the same result. Did any wartime production radial manage this?
Finally, if the number of valves in a head was of little relevance to the breathing of the engine, why did designers bother with multivalve heads at all? Why do they now?
 
Last edited:
I'm sorry but no, you can't deprive an air-cooled engine of air. Either the flaps were open when hit, in which case they might get locked open, or they were closed when hit, in which case the plane was flying with them closed and will continue to do so with them locked closed.

No, you can't deprive an air cooled engine of air. You could deprive it of the required mass flow of air. It isn't just the cowl flaps - the cowl at teh entry could be damaged and thus deprive the engine of air.


In contrast, once a liquid cooled engine lost its coolant there was ZERO cooling and the thing would seize up in short order. I've heard of plenty of LC aircraft overheating and going down after being hit in the radiator but I can't recall any where an air-cooled plane seized up due to damaged cowl flaps.

B-29s certainly had issues overheating at takeoff when the cowl flaps were closed.


Re the cylinder heads of a radial more exposed to being shot off, radial engine fighters can and did sustain such damage and keep on operating. The same damage to a LC engine would result in rapid loss of coolant and inevitable engine failure

You provided one example of an air cooled engine losing a cylinder/head and surviving, and I provided one example of a liquid cooled engine that got its pilot home after being shot up.

I still would say that a radial with a cylinder/head missing is unlikely to make it back to base.


How was my observation that radial engines are easy to remove a 'sweeping generalisation'? You mentioned LC engines with annular radiators as being just as easy to install and uninstall but how many combat aircraft were thus equipped? The Ju88 and Fw190D/Ta152 sure, but the vast majority of LC engine aircraft have had radiators attached to the airframe

What about radials with turbos? How much easier was it to remove an R-2800 from a P-47 than a Merlin from a Spitfire? I mean, disconnecting two water hoses must have been a real chore (4 if it was a 2 stage with intercooler).


Regarding multi valve heads on radial engines I said the radials were EFECTIVELY limited to two valves per cylinder, as were all the widely used radials of WWII I can think of. I didn't say it was impossible to make a multi valve radial, it was just complex and impractical. How many Armstrong Deerhounds were produced? Ten, I think. And the Lycoming had a production run of two, both of which were extremely unreliable.

Only a handful of Deerhounds, and they had their problems too. Interestingly, Armstrong Whitworth had wanted to make the Deerhound as a liquid cooled engine, but the Air Ministry wanted an air-cooled competitor for Bristol.

Plenty of 4v Pegasus and Mercuries were built before and during the war.


To put it into perspective, a V-12 A Rolls Royce Merlin had two drive shafts driving two cams, one for each bank of six cylinders. A twelve cylinder radial with driven bevel drive cam actuated valves would require twelve drive shafts operating twelve cams – six times as many moving parts as theforo the for the same result. Did any wartime production radial manage this?

The Deerhound had 21 cylinders, yet only 7 camshafts and associated drives. The Wolfhound was to have 28 cylinders, and yets still only 7 camshafts and associated drives. The (liquid cooled) Wright R-2160 Tornado had 42 cylinders and 7 camshafts and associated drives.


Finally, if the number of valves in a head was of little relevance to the breathing of the engine, why did designers bother with multivalve heads at all?

That isn't what I said. I said, or was trying to say, that more issues lay with the distribution of the air/fuel mix into the cylinders - for both radials and in-lines.

There are advantages to 4 valves over 2 valves, other than breathing - such as being lighter and not requiring as heavy a spring to control it.
 
No, you can't deprive an air cooled engine of air. You could deprive it of the required mass flow of air. It isn't just the cowl flaps - the cowl at teh entry could be damaged and thus deprive the engine of air.




B-29s certainly had issues overheating at takeoff when the cowl flaps were closed.




You provided one example of an air cooled engine losing a cylinder/head and surviving, and I provided one example of a liquid cooled engine that got its pilot home after being shot up.

I still would say that a radial with a cylinder/head missing is unlikely to make it back to base.




What about radials with turbos? How much easier was it to remove an R-2800 from a P-47 than a Merlin from a Spitfire? I mean, disconnecting two water hoses must have been a real chore (4 if it was a 2 stage with intercooler).




Only a handful of Deerhounds, and they had their problems too. Interestingly, Armstrong Whitworth had wanted to make the Deerhound as a liquid cooled engine, but the Air Ministry wanted an air-cooled competitor for Bristol.

Plenty of 4v Pegasus and Mercuries were built before and during the war.




The Deerhound had 21 cylinders, yet only 7 camshafts and associated drives. The Wolfhound was to have 28 cylinders, and yets still only 7 camshafts and associated drives. The (liquid cooled) Wright R-2160 Tornado had 42 cylinders and 7 camshafts and associated drives.




That isn't what I said. I said, or was trying to say, that more issues lay with the distribution of the air/fuel mix into the cylinders - for both radials and in-lines.

There are advantages to 4 valves over 2 valves, other than breathing - such as being lighter and not requiring as heavy a spring to control it.

I refuse to surrender!
1. Shooting up the cowl of an air cooled plane seems more likely to increase airflow, rather than decrease it. Lots of holes in anything tend to let stuff in, not keep it out. Sorry, I just can't see how a hail of machine gun fire is going to seal up the cowling over a motor.
2. If the only extra issue in removing a liquid cooled engine undoing a couple of pipes, I've been buying the wrong cars. Or maybe I should give up on routinely flushing radiators, replacing hoses, hooking the whole thing up again and mixing and replacing the coolant. Never worked on a turbo – don't they come out with the engine in one piece?
3. The relevance B-29s with closed flaps overheating on take-off to the idea that gunfire could close the flaps on an air-cooled aircraft and thus cripple it in combat is pretty tenuous
4. I provided an example of an aircooled aircraft that flew across the Channel with a cylinder missing. You provided an example of a liquid cooled aircraft that got 'shot up' and made it back to base. Was its cooling system punctured? Did it make it back across the Channel?
5. True, Bristol did produce a heap of multi-valve radials during the war. They were used to power such brutes as the Fairey Swordfish and Gloster Gladiator. The top end of the performance envelope belonged to multi valve Liquid cooled designs and two valve air-cooled radials
6. "The Deerhound had 21 cylinders, yet only 7 camshafts and associated drives. The Wolfhound was to have 28 cylinders, and yet still only 7 camshafts and associated drives. The (liquid cooled) Wright R-2160 Tornado had 42 cylinders and 7 camshafts and associated drives" - Yep, and neither of them got past the prototype stage…
7. Yes, there are advantages to multi-valve engines over two valve engines aside from breathing – but better breathing means more power and power was at the top of the list for a combat aero-engine
 
2. If the only extra issue in removing a liquid cooled engine undoing a couple of pipes, I've been buying the wrong cars. Or maybe I should give up on routinely flushing radiators, replacing hoses, hooking the whole thing up again and mixing and replacing the coolant. Never worked on a turbo – don't they come out with the engine in one piece?

Not the turbos of WW2. The P-47 turbo, for instance, was in the rear fuselage.

There were a few Daimler Benz prototypes with turbos that would be removed with the engine.


3. The relevance B-29s with closed flaps overheating on take-off to the idea that gunfire could close the flaps on an air-cooled aircraft and thus cripple it in combat is pretty tenuous

It does show that cooling can be quite borderline on air-cooled engines.


4. I provided an example of an aircooled aircraft that flew across the Channel with a cylinder missing. You provided an example of a liquid cooled aircraft that got 'shot up' and made it back to base. Was its cooling system punctured? Did it make it back across the Channel?

Well, no, because it was in North Africa.


5. True, Bristol did produce a heap of multi-valve radials during the war. They were used to power such brutes as the Fairey Swordfish and Gloster Gladiator. The top end of the performance envelope belonged to multi valve Liquid cooled designs and two valve air-cooled radials

Bristol would have made big 4 valve radials had they not moved into sleeve valves. Persues, Hercules and Centaurus would likely have been 4 valve engines had Fedden not persued sleeve valves.



6. "The Deerhound had 21 cylinders, yet only 7 camshafts and associated drives. The Wolfhound was to have 28 cylinders, and yet still only 7 camshafts and associated drives. The (liquid cooled) Wright R-2160 Tornado had 42 cylinders and 7 camshafts and associated drives" - Yep, and neither of them got past the prototype stage…

Yes, that is correct. Just pointing outthat you need not use a camshaft for each cylinder.


7. Yes, there are advantages to multi-valve engines over two valve engines aside from breathing – but better breathing means more power and power was at the top of the list for a combat aero-engine

All true.
 
Last edited:
Just remembered another couple of examples of LC QEC modules - one production, one pprototype.

Rolls-Royce developed a Merlin QEC for the Beaufighter. This was used to convert the Manchester into the Lancaster.

GM's Fisher Body Division developed a QEC for the V-3420 to be used as a replacement for the R-3350 if needed. The QEC contained all coolers and the (experimental) turbo. This was tested in the XB-19 and then used in the XB-39 (without the turbo).
 
On the whole I maintain the relative advantages and disadvantages of the air-cooled v liquid cooled engines stand up; resistence to battle damage, and ease of amintenance for the air cooled radial against lower drag and the easier application of multi-valve heads for the V or inline liquid cooled motor. Its interesting that at the comencement of WWII the opinion of most designers of land based fighters was heaviy in favour of the liquid cooled engine, but subsequent experience demonstrated that the death of the air cooled radial had been greatly exaggerated. The radial remained a popular choice for carrier based aircraft, thanks to doubt to its ease of amaintenance and durability, and the arrival of the hugely influentual Fw190 in Europe demonstrated that a carefully installed radial could still trump the liquid-cooled fighters. Some excellent aircfaft also came about when airframes originally intended to house liquid cooled V-s had radials boilted on to see what would happen. The Japanese Ki-100 was derived from the Ki 61 in this way, as was the Lavochkin La-5 from the underwhelming La-3. Even the great Sydney Cramm was equivocal about mating the Hawker Tempest to the Bristol Centaurus, but that version his fighter was clearly superior to its Napier-Sabre equiped predecessor.
As to which was 'better', the air-cooled radial or inline/V liquid cool engine, utimately the arrival of the jet and turbo-prop removed any chance of the question being answered definitively. They both seem to be going well in civilian use today, and only now there is one more option - the Wankel Rotary. Wonder what they might have done with that during WWII?
 
During the History of the piston engine the air cooled and liquid cooled "sides" traded top honors several times. A lot of it had to do with drag. While the early installations of both engines were high drag, the liquid cooled engine's lower frontal area and ease of cowling was balanced somewhat by the rather inefficient radiators of the day. Also the early liquid (meaning water) systems tended to be leaky even without combat damage. AS the air-cooled engines (meaning radials, high powered air cooled v-12s or H 16/24 cylinder engines were a minority and had a few problems) got better installation, like Townsend rings and early NACA cowlings the liquid cooled camp countered by going to Glycol and/or pressure cooling with higher boiling points which allowed for smaller/lighter radiators. the P-36 with it's aircooled engine had 22% more drag than an early P-40 did. But as has been noted above the air-cooled engine installations got better, if a bit more finicky, until there was little to chose between a liquid cooled installation and an air-cooled one. Since, however, the balance changed during the course of WW II one statement cannot cover the entire war.
 
During the Fighter Conference in 1944, there were pilots from Britain, manufacturers, AAF and USN. THere was a poll taken of the preferred engine for combat. The R2800 won handily over the Merlin, the only two engines that got many votes. Lindberg chose a radial, air cooled engine for the Altlantic trip because it was lighter and more reliable.
 
if we can simply put it, it whould be radial give strengt, inline give speed in comparelse.
 
Not necessarily true. Me-109s had a radiator in each wing. They could be individually isolated in case of battle damage. As long as one wing remains undamaged you can limp home.

Annular radiators are another way to mitigate damage to liquid cooling systems. Hitting an annular radiator means hitting the prop and/or engine. Even without a radiator leak such damage is likely to seize the engine.
 
Not necessarily true. Me-109s had a radiator in each wing. They could be individually isolated in case of battle damage. As long as one wing remains undamaged you can limp home.

"They could be individually isolated in case of battle damage"

The COULD be but usually were not, only certain models had the isolation valves and not even all of some of those models. Ground crews went to some lengths to salvage the valves from crashed/wrecked aircraft to fit to those aircraft that were not built with them.
 
On the whole I maintain the relative advantages and disadvantages of the air-cooled v liquid cooled engines stand up; resistence to battle damage, and ease of amintenance for the air cooled radial against lower drag and the easier application of multi-valve heads for the V or inline liquid cooled motor. Its interesting that at the comencement of WWII the opinion of most designers of land based fighters was heaviy in favour of the liquid cooled engine, but subsequent experience demonstrated that the death of the air cooled radial had been greatly exaggerated. The radial remained a popular choice for carrier based aircraft, thanks to doubt to its ease of amaintenance and durability, and the arrival of the hugely influentual Fw190 in Europe demonstrated that a carefully installed radial could still trump the liquid-cooled fighters. Some excellent aircfaft also came about when airframes originally intended to house liquid cooled V-s had radials boilted on to see what would happen. The Japanese Ki-100 was derived from the Ki 61 in this way, as was the Lavochkin La-5 from the underwhelming La-3. Even the great Sydney Cramm was equivocal about mating the Hawker Tempest to the Bristol Centaurus, but that version his fighter was clearly superior to its Napier-Sabre equiped predecessor.
As to which was 'better', the air-cooled radial or inline/V liquid cool engine, utimately the arrival of the jet and turbo-prop removed any chance of the question being answered definitively. They both seem to be going well in civilian use today, and only now there is one more option - the Wankel Rotary. Wonder what they might have done with that during WWII?

I agree that the liquid cooled engine has another system to maintain and be susceptible to battle damage, so it is more vulnerable to battle damage. I, however, believe that the difference in vulnerability is overexaggerated. It is done in two ways - exaggerating the vulnerability of liquid cooled engines to battle damage ("just one shot from a small calibre gun will do the trick"), and exaggerating the ability of air cooled engines to survive battle damage (using one or two examples of a particular engine having survived after a cylinder was shot off).

Ease of mainetnance is a more difficult sell - because in both categories some are easier to maintain than others.

Also, in what way was the Tempest II superior to the Tempest V?
 
My suspicion is that most members of this forum are not old enough to remember when cooling problems were a major cause of breakdowns in cars with liquid cooled engines, which was most of the cars back in the late forties and early fifties when I started driving. That may be the reason that most don't seem to think that liquid cooled aero engines in WW2 were more problematic than air cooled ones. I can't relate how many times in various cars I have had to stop with steam coming out from under the hood and search for water to put in the radiator. It was mostly because the cooling system had a small leak because the car or truck was elderly( three or four years old) not because a bullet had gone through the radiator or a hose.

Further, we have debated often that the oil coolers in the Corsair made it more or less more vulnerable, especially to ground fire, than the Hellcat. When you look at the size and location of those oil coolers, compared to the cooling systems of the various liquid cooled engines and some say the Corsair was practically a death trap and on the other hand say liquid cooled engines are no more vulnerable than air cooled ones, I am nonplussed and amused.
 

Users who are viewing this thread

Back