Mosquito aerodynamics?

[llemon]

Thanks again.

I posed this question a couple of years ago. But as some more knowledge seems to have accumulated since I ask again.

Which radiator installation for inlines offer the least drag and the most thrust respectively?
First place must be the Mustang-style one which uses the Meredith-effect well. It has a sufficient expansion chamber aft of the radiator for the warm air which exits through the aft outlet. I wonder how efficient the ventral cooler installations on other fighters were, for example the Yak-3 and the Italian series 5 fighters. Those seem to lack a big enough chamber.
Second (at least I think) is the annular/drum installation found on late-war German aircraft. Though it looked draggy it was surprisingly aerodynamic. Maybe because it was part of the fuselage? Here louvers are used as air outlet. No idea about thrust effects.
Late in/after the war the British experimented with an air outlet installation which was movable fore and aft and thus regulating the amount of air outlet.
By renouncing the use of louvers, which enlarged the area exposed to airflow, the drag should not increase.
Third would be leading edge radiator, which I intuitively would rate as less draggy draggy than the annular/drum one. But according to a report it isn't. I wonder why as the inlet is not mounted outside of the main body.
Fourth had to be underwing radiators which gave the Me 109 and especially the Spitfire a massive drag penalty.

As far as engine cowling aerodynamics are concerned it seems that the Japanese fighters were the most advanced. If one looks at them they seem more refined and curved. The J2M Raiden was more aerodynamic than a predecessor, the A6M Zero I think, even though it had a larger fuselage cross section diameter. It was achieved by a carefully shaped engine cowling.

If you look at the wing shapes of Japanese and Russian planes, they have less blunt wing leading edge than the German planes and the thickest part of the wing seems to be more far aft, almost of laminar shape in some cases. Can one deduce that, like in the case of the Mossie, that these airfoil are more aerodynamic than the German ones?

I've read a paper that argues that a 109 type installation is superior to the 51. This is because the expansion area in front of the radiator is small enough that separation isnt an issue, duct area is minimized, and it can hit the needed expansion ratio with a minimum of length. It is interesting that the spitful had a very 109 like radiator instead of a 51 style belly scoop. Of course boundary layer ingestion is the biggest issue and its hard to see how such a radiator installation can overcome it even with a slot.

The late war 152/209 radiator installation are a drum type, which despite the looks isnt really similar to the annular installations used earlier. By sticking the radiator horizontally you can get a good expansion ration while avoiding separation and do it in a very short duct. Blohm actually had a very similar arrangement with the 155s radiators which look more like 51 scoops but in actuality have the radiator nearly horizontal.

The issue with leading edge radiators is spilliage spoiling the airfoil over the wing. Even if you dont get spilliage the slot for the intake will help ruin the airflow. Making the exit slot work with a ramp is also a bit of an issue.

In hindsight the long nose allison looks like the perfect candidate for drum style radiator, but that is getting off topic.

Yaks and Las used a Clark-yh profile, around 14% at root iirc. Most Japanese aircraft were using NACA derived airfoils, often 23000s. Tightly cowled engines like the J2 were tried elsewhere - expecially by Curtiss. They are good on paper but getting them to work in practice is hard, and the raiden had lots of cooling problems for most of its life.

For the era the Bf 109 was introduced, it's wing was very modern and thin.
Wings of 109F and 109E probably differed only at tip and where the radiators were installed, so most of existing tooling for the wing can still be used in order to maintain production rates.

Willy Messerschimtt love the NACA 2R airfoil. Its a kinda strange thing that so many messerschmitt aircraft used it. I don't know of any other manufacturer who employed it as a wing airfoil. The 2R was originally intended for props as far as I can tell. Finding info on it is much harder than other NACA series airfoils but in fairness I havent tried that hard.

The Fs wing originally was a straight up clipped E series aerodynamically. They found that the reduction in span had too much of an adverse effect on handling so they added some elliptical tips. This was done because the F wing was a substantial structural redesign and it was easier from a tooling/produciton standpoint to make the tips bigger rather than extending the spar and going back to an E range span.

 

[tomo pauk]

I've read a paper that argues that a 109 type installation is superior to the 51. This is because the expansion area in front of the radiator is small enough that separation isnt an issue, duct area is minimized, and it can hit the needed expansion ratio with a minimum of length. It is interesting that the spitful had a very 109 like radiator instead of a 51 style belly scoop. Of course boundary layer ingestion is the biggest issue and its hard to see how such a radiator installation can overcome it even with a slot.

The paper you've mentioned is IMO fishy. As you've noted, ingestion of boundary layer is a problem. Bf 109 have had 3 duct openings, for it's two coolant radiators and one oil cooler, vs. single opening on P-51.
Spiteful took a lot of fuselage structure from Spitfire, so unless a major redesign of fuselage is done it will probably not be possible to have P-51-style radiator on Spiteful.

The issue with leading edge radiators is spilliage spoiling the airfoil over the wing. Even if you dont get spilliage the slot for the intake will help ruin the airflow. Making the exit slot work with a ramp is also a bit of an issue.

Leading-edge radiators were suggested by NACA for the (Y)P-38, to be installed in the extended (chord increased by 20%) section of the wing inboard the engines.
Seems like LE radiators worked on Mosquito, Hornet and Tempest I.

Yaks and Las used a Clark-yh profile, around 14% at root iirc.

LaGG-1/3 and La5/7 used NACA 23016 at root.

 

[pbehn]

It is interesting that the spitful had a very 109 like radiator instead of a 51 style belly scoop. Of course boundary layer ingestion is the biggest issue and its hard to see how such a radiator installation can overcome it even with a slot.

.

The Supermarine Spiteful had a Supermarine like radiator developed by Supermarine, from the Supermarine Spitfire.

 

[llemon]

The paper you've mentioned is IMO fishy. As you've noted, ingestion of boundary layer is a problem. Bf 109 have had 3 duct openings, for it's two coolant radiators and one oil cooler, vs. single opening on P-51.

One advantage to the 109 style radiator is the wetted area from the ducts is much less than a belly scoop. Also the internal duct area is less despite having two radiator - this is because the expansion part of the duct is dependent on the height of the rad. And given that it is a much shorter rad in 109 style the duct can be shorter.

But none of this really matters since there is no good way to get rid of the boundary layer. Although you can get a sense of the logic behind such an installation.

Leading-edge radiators were suggested by NACA for the (Y)P-38, to be installed in the extended (chord increased by 20%) section of the wing inboard the engines.
Seems like LE radiators worked on Mosquito, Hornet and Tempest I.

I wasn't implying that leading edge radiators dont work. Just that they probably arent as good of an option as 51 style belly scoop or annular.

LaGG-1/3 and La5/7 used NACA 23016 at root.

Got them mixed up with the mig. I knew that two of the 3 major newish soviet fighters used clark-yh.

The Supermarine Spiteful had a Supermarine like radiator developed by Supermarine, from the Supermarine Spitfire.

The spitfire started off with one deep underwing radiator. Then two. And finally ends up with two very shallow and wide under wing radiators in the spitful.

I wasn't implying anything about supermarine copying the Germans. Just that the spitfuls rad arrangment is much more like a 109F than a MKII. What that means exactly, were they aping the germans or did they arrive at the same solution for similar reasons, would have to be discovered by digging around in archives.

 

[Shortround6]

We can look at the drawings all we want.
We know what they were trying to achieve.

Without test results of different set ups that give pressure in the ducts at various places, speeds of the airflow through the duct at various locations in the duct, and/or photography of smoke streams through the ducts/radiators we are just guessing.

The Mustang radiator is generally acknowledged as one of the best of WW II, I listed a bunch of factors. I have no idea if one or more of those factors was actually working as they thought/hoped or how close to the top of the heap any one factor was compared to all the other set ups. All we know is that the total combination was one of the best, if not the best.
One or more factors could have been below par if the others made up for.

As for the idea that smaller ducts have less wetted area and thus less friction/drag. I am not sure that is right. I don't know about air, I know a bit about water (but it is incompressible)

but here is a friction loss chart for fire hose.

Pick a flow like 500gpm and look at the 2 1/2, 3 and 3 1/2 diameter hose. Going from the 2 1/2 to 3 1/2 hose increased the "wetted" area by 40% but the force needed to push the water through the hose dropped to less than 20% , The water was moving "slower" and for a total volume of water, less was in contact with hose wall at any given time.

Somebody who has studied air flow could very well correct me.

At times on the training ground (depending on which training officer we had) we would put pressure gauges mounted on short lengths of pipe into a layout of several hundred feet and we would also measure the pressure of the water exiting nozzle. Most of the time we got the expected result, sometimes we did not.

Please note that most of the time doubling the flow in a given size hose quadruples the friction loss which is what we would expect from the square law.

 

[pbehn]

The spitfire started off with one deep underwing radiator. Then two. And finally ends up with two very shallow and wide under wing radiators in the spitful.

I wasn't implying anything about supermarine copying the Germans. Just that the spitfuls rad arrangment is much more like a 109F than a MKII. What that means exactly, were they aping the germans or did they arrive at the same solution for similar reasons, would have to be discovered by digging around in archives.

How many did the Bf 109 start out with? The radiators for oil and water cooling were always below the spitfires wings. A radiator for the intercooler was also added later and the oil cooler made bigger to cope with increased temperature/work load. There was a simple reason that the radiators were not low and slim on a Spitfire, it needed a complete re design of the wing as you can see below.

 

[Milosh]

What was the total radiator area size (coolant and oil) for the Spit IX and 109G?

 

[Zipper730]

From "The Mosquito Manual", I know there were two sections involved. The first was a Piercy section (Norman Augustus Victor Piercy) and the second was the RAF 34.

So, basically it combined the RAF 34 and Piercey sections, with the crest at 40% T/C?

It does have some traits I can see in later laminar flow foils -- particularly the trailing edge starting to develop a barely visible cusp.

I'm confused about what you wrote regarding T/C of the airfoil -- was it 13 or 15%?

 

[spicmart]

The late war 152/209 radiator installation are a drum type, which despite the looks isnt really similar to the annular installations used earlier. By sticking the radiator horizontally you can get a good expansion ration while avoiding separation and do it in a very short duct. Blohm actually had a very similar arrangement with the 155s radiators which look more like 51 scoops but in actuality have the radiator nearly horizontal.

This is a known drawing showing annular and drum radiator! Can you specify why the drum would be better other than offering more area? I don't quite understand the separation issue and it doesn't offer more expansion space than the annular one.

Here is the report of the British tests that tomo posted in another thread before. It says that frontal radiators are less draggy than leading edge radiators.

tempest | 1946 | 1441 | Flight Archive


I don't know really know about wing profiles and how they are named or classified.
Can you tell me some sources where I can educate myself?

 

[pbehn]

So, basically it combined the RAF 34 and Piercey sections, with the crest at 40% T/C?
It does have some traits I can see in later laminar flow foils -- particularly the trailing edge starting to develop a barely visible cusp.

I'm confused about what you wrote regarding T/C of the airfoil -- was it 13 or 15%?

I didn't write it, I found it on a blog, cant really tell who wrote it. As I can see it is a combination of Piercy and RAF 34 with some of d Havillands own ideas/experience. From what I can see people were learning all the time, it was more "laminar flow than previous and less laminar flow than later but non were actually laminar flow anyway.

 

[spicmart]

The issue with leading edge radiators is spilliage spoiling the airfoil over the wing. Even if you dont get spilliage the slot for the intake will help ruin the airflow. Making the exit slot work with a ramp is also a bit of an issue.

In hindsight the long nose allison looks like the perfect candidate for drum style radiator, but that is getting off topic.

Yaks and Las used a Clark-yh profile, around 14% at root iirc. Most Japanese aircraft were using NACA derived airfoils, often 23000s. Tightly cowled engines like the J2 were tried elsewhere - expecially by Curtiss. They are good on paper but getting them to work in practice is hard, and the raiden had lots of cooling problems for most of


Willy Messerschimtt love the NACA 2R airfoil. Its a kinda strange thing that so many messerschmitt aircraft used it. I don't know of any other manufacturer who employed it as a wing airfoil. The 2R was originally intended for props as far as I can tell. Finding info on it is much harder than other NACA series airfoils but in fairness I havent tried that hard.

What do you mean with spillage?

Can you elaborate on the Russian and Japanese wings?
What do the 14% mean drag wise?

Why do you include the Allison?

The 2R isn't the most aerodynamic of wings? For one striving for ever more speed Willy would have been better advised to have looked for at at least used others.

Sorry, the many and direct questions. Not to be in polite, just curious.

 

[wuzak]

We can look at the drawings all we want.
We know what they were trying to achieve.

Without test results of different set ups that give pressure in the ducts at various places, speeds of the airflow through the duct at various locations in the duct, and/or photography of smoke streams through the ducts/radiators we are just guessing.

The Mustang radiator is generally acknowledged as one of the best of WW II, I listed a bunch of factors. I have no idea if one or more of those factors was actually working as they thought/hoped or how close to the top of the heap any one factor was compared to all the other set ups. All we know is that the total combination was one of the best, if not the best.
One or more factors could have been below par if the others made up for.

As for the idea that smaller ducts have less wetted area and thus less friction/drag. I am not sure that is right. I don't know about air, I know a bit about water (but it is incompressible)

I think the wetted area comment referred to the shape and size of the duct's effect on the external air flow, not the duct's internal flow.

The P-51's radiator was one of the best of WW2, no doubt, but it was far from perfect.

The expansion duct was not symmetrical, with the roof being quite steep, which may have led to flow separation.


In the case of the wide, low radiator ducts, as used on the Spiteful, there was actually boundary layer thickening on the roof of the duct. The boundary layer air is not moving very quickly relative to the aircraft, so this reduced the effectiveness of the radiator. I can't recall exactly how much of the radiator the boundary layer covered, except that it was around 1/8 the depth of the radiator. Which is significant. This was from tests done of the production style Spiteful radiators.

The best solution for dealing with the boundary layer was to provide a duct, which exhausted to the top of the wing before the radiator. Obviously this was not the best solution for the aircraft, as it disturbed the wing.

Another solution was to provide a duct the entire length of the radiator duct, gong above the radiator.

 

[pbehn]

What was the total radiator area size (coolant and oil) for the Spit IX and 109G?

My point was that the two planes are different. The Bf-109 always had the oil cooler under the engine, even when the water cooling was moved to the wings. The Spitfire always had oil and water cooling in the wings and added the intercooler. The Spitfires radiators would have been wider and flatter but there wasn't space for that because of the undercarriage.

 

[Shortround6]

Radiators (and the airflow through them) have to be sized to the power the engine is developing in the cylinders and the amount of airflow going through the radiator at the appropriate altitudes.
Like a DB605A making 1355PS at about 18800ft to which you have to add the friction in the engine + the power needed to drive the supercharger which is compressing the the Air to about 6lbs of boost.
The V-1659-3 in an P-51B made 1330hp at 23,300ft to which you need to add the friction and the power to drive the supercharger which is compressing thinner air to 18lb boost and is using thinner (less dense) air going through the radiator to do the cooling. You also have to know if the engines in question put a similar amount of heat into the oil and the coolant.
The Allisons and Merlin used in the P-40 did not and the Merlin P-40s need a different radiator and oil cooler it order to cool properly.

Just measuring radiator size leaves out several considerations/factors.

 

[pbehn]

Just measuring radiator size leaves out several considerations/factors.

I always thought a radiator was based on volume. The surface area of the radiator and airflow is what governs its worth, if the frontal area is reduced it must be made deeper which is a bit of a bind because they work best with a large frontal area and small depth, just what an aircraft designer doesn't want.

 

[Shortround6]

You are correct if the other conditions don't change, You can't use the same radiator for a big V-8 towing a trailer even if it works fine on a low powered six cylinder engine.

For our aircraft the air at 23,000ft is about 87% as dense as the air at 19,000ft so to get rid of the same amount of heat at 19.000ft your radiator only needs to be 87% as big assuming you have the same volume of air going through it per minute. Yes the air is colder at 23,000ft but not enough to make up for the density difference.
Two stage Merlin is probably making more power in the cylinders in order to drive the two stage supercharger even if power to prop isn't that different so the radiator has to sized to the total cooling load.

 

[pbehn]

You are correct if the other conditions don't change, You can't use the same radiator for a big V-8 towing a trailer even if it works fine on a low powered six cylinder engine.

For our aircraft the air at 23,000ft is about 87% as dense as the air at 19,000ft so to get rid of the same amount of heat at 19.000ft your radiator only needs to be 87% as big assuming you have the same volume of air going through it per minute. Yes the air is colder at 23,000ft but not enough to make up for the density difference.
Two stage Merlin is probably making more power in the cylinders in order to drive the two stage supercharger even if power to prop isn't that different so the radiator has to sized to the total cooling load.

Those are all problems faced by Supermarine in cooling the Merlin and later Griffon in the Spitfire from when it was first designed to the end of the war, in an airframe that was originally laid out for evaporative cooling. With the place in the wing where the radiator was placed hemmed in by the undercarriage only a radiator scoop with a bigger frontal and cooling area was possible, without stopping production for a completely new wing.

 

[Shortround6]

That may very well be.
But that is not what I am arguing/

I am arguing that trying to compare the 109 to the Spitfire by comparing the size of their radiators isn't going to tell us much unless we know the heat load they are trying to get rid of and I am guessing that the 109 was usually trying to get rid of less heat and/or doing it in higher density air that needed smaller radiators.

 

[pbehn]

That may very well be.
But that is not what I am arguing/

I am arguing that trying to compare the 109 to the Spitfire by comparing the size of their radiators isn't going to tell us much unless we know the heat load they are trying to get rid of and I am guessing that the 109 was usually trying to get rid of less heat and/or doing it in higher density air that needed smaller radiators.

That was my point S/R when it was suggested that the Spitfire ended up with a configuration like the Bf-109.

 

[nuuumannn]

Just going back to the original question, the Mosquito was an extraordinarily clean design and finish, despite its physical size, as can be seen here.

TV959 16

TV959 02

Both the Spitfire and Bf 109's radiators sizes and configurations were determined by their location. This is a Spit XIV radiator and you can see how the size has increased to cope with the installation of the Griffon.

Radiator

The Bf 109's was similarly located, but in later, post Friedrich models became more sophisticated.

Radiator location

Radiator