Mosquito aerodynamics?

[ThomasP]

About the use (first use?) of the Meredith effect,

I cannot say authoritatively that the Hurricane was the first production aircraft to incorporate the Meredith effect, but Hawker engineers knowingly introduced aspects of the effect in the production cooling system, with the intent of reducing the cooling drag. Although they were aware of the potential in the Meredith effect (it had originated in the UK after all) and understood most of its theoretical finer points as to what would work, the Hawker design team did not pursue the maximum possible effect for practical reasons.

One reason was the cost of manufacturing what for the time would have been a fairly sophisticated assembly.

Another reason was available space and weight for the assembly. On first look this might not seem correct, after all the Hurricane was a fairly large airframe and the radiator was already in a similar location to that of the P-51, and it might be possible to rearrange the bits and pieces to maintain an acceptable CG. Keep in mind, however, that the Hurricane was intended to be an evolution of an older biplane design into a monoplane (similar to the evolution of the Wildcat), not a largely new design.

Still another reason was the expected top speed and sustained speeds of the new aircraft. The radiator system of an aircraft must be able to meet the sustained speed cooling requirement indefinitely, and the maximum power cooling requirements for various lesser periods (e.g. WEP for 5 min, Military for 15 min, Climb for 30 min, etc). Everything else being equal, the greater the speed, the greater the Meredith effect. In fact the Meredith effect becomes more efficient with square of the inrease in speed. The Hurricane II was effectively 100 mph slower than the P-51B, so if the exact same assembly as used on the P-51B ( at 430 mph) was used on the Hurricane II (at 330 mph) it would only be abut 56% as effective as on the P-51B.

Instead the Hurricane design team relied on a shorter but larger orifice inlet, with a boundary layer lip/notch, and a somewhat shorter exit chamber with a variable outlet area via a movable flap. The boundary layer lip/notch prevented air flow stagnation at the radiator inlet, the short inlet chamber combined with the larger orifice allowed enough air to slow down (via the pressure head effect) to efficiently absorb heat while passing through the radiator, the radiator heated the air up, and the short exit chamber with variable area outlet acted like an ejector nozzle depending on the aircraft speed, power setting, engine temperature, and flap position. In order to keep the controls simple, the pilot adjusted the flap depending on the engine temperature, not for maximum Meredith effect. The normal flap setting was used for maximum sustained speed which would correlate to a sustainable engine temperature, the more closed settings were used for higher speeds where higher air flow helped keep the engine cool temporarily, and the more open settings were for lower speeds at high power settings (while climbing for example) or while idling/taxying and during take-off .

If my information is correct the P-51B cooling system reduced the cooling HP loss by about 80%, or from about 400 HP to about 50 HP.

Using the P-51B system on the Hurricane II would have reduced the cooling HP loss by 45% (i.e. .56 x .80), or from about 260 to 143. This would have resulted in a speed increase of about 9 mph at 12,000 ft and about 13 mph at 19,000 ft. (i.e. if there was no reduction in cooling HP loss the top speeds would have been 9-13 mph less than they were in actual service.)

The actual system used on the Hurricane II only reduced the cooling HP loss by about 90 HP, or from 260 to 170, for a reduction of ~35%. This resulted in a speed increase of about 7 mph at 12,000 ft and about 10 mph at 19,000 ft.

So to sum up, for the loss of 2-3 mph top speed, the Hurricane II cooling system using a reduced effectiveness Meredith effect, saved an undetermined amount of time in development (check out the amount of time the NA team spent in wind tunnel and flight tests before they settled on a final design) and an undetermined amount in development cost, reduced the cooling system structure weight (probably) by about 50 lbs, and reduced the cost (probably) of the cooling system structure by about 50%.

 

[wuzak]

The Hawker Hurricane and Supermarine Spitfire were being designed when the effect was being studied by Meredith at the RAE.

The research was published in 1936 - after the Hurricane prototype flew, though information may have been given to them.

I'm not sure how much the Meredith effect was taken into account by Camm and Hawkers.

The Tornado prototype also had a belly radiator for initial flight testing, but it did not work too well, perhaps because of excessive turbulence. The radiator was moved to the chin position, which was also used for the Typhoon.

NAA had the benefit of 3 or 4 years to study the effect and put it into practice.

 

[davparlr]

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 always felt that leading edge radiators for the P-38 would have been much preferred over the four separate radiators used. Of course I wasn't making the trade-offs and I certainly don't know all the variables.

 

[tomo pauk]

I always felt that leading edge radiators for the P-38 would have been much preferred over the four separate radiators used. Of course I wasn't making the trade-offs and I certainly don't know all the variables.

NACA did all the calculations for you, and tested stuff in wind tunnel for a good measure.
NACA report
(they were also tweaking the pod & canopy to better combat compressibility, plus other tweaks to the wing)

Well worth the read: alternative P-38 with LE radiators and whatnot

 

[davparlr]

NACA did all the calculations for you, and tested stuff in wind tunnel for a good measure.
NACA report
(they were also tweaking the pod & canopy to better combat compressibility, plus other tweaks to the wing)

Well worth the read: alternative P-38 with LE radiators and whatnot

Thanks. It also may have made heating the cockpit easier. I think I read somewhere that one complaint was that the cockpit had a heating problem.

 

[Shortround6]

Thanks. It also may have made heating the cockpit easier. I think I read somewhere that one complaint was that the cockpit had a heating problem.

That could have solved (at least somewhat) by putting a generator on the right hand engine to handle the electrical load and putting an electric heater in the cockpit (or heated flying suit).
Might have saved a few planes too, as the Props were electrically operated as were the turbo regulators, aux fuel pumps, and oil cooler exir flaps in addition to the normal electrical stuff.
If you lost the left engine with the generator any long trip home became and exercise in electrical power management as all the electric needs had to be balanced against the power left in the battery.

 

[Shortround6]

Well worth the read: alternative P-38 with LE radiators and whatnot

It takes some careful reading as there are a few things that are misleading.

"The results are well known and documented: the P-38 in the ETO struggled the first 18 months of combat prompting 8th Fighter Command to pre-emptively phase them out in favor of the new P-51 as they became available at the end of 1943 and through the first half of 1944."

Bolding by me. The only P-38s in the ETO in 1942 and most of 1943 were a few photo recon planes. Before the P-38s that were there in late 1942 could fly more than few missions over the French coast they were all sent to North Africa to support the Torch invasion.
P-38s as fighters would not return to the ETO until Sept 1943 and went "operational" on Oct 15th 1943 with the 55th fighter group. Nov 3rd 1943 sees the 55th escort bombers Wilhelmshaven. P-51s (and P-38s) are used as escorts on a raid to Kiel On Dec 13th 1943.

So basically the P-38 as a bomber escort in the ETO went into service 6 weeks before the P-51.

Nowhere near 18 months.

 

[ThomasP]

Hey wuzak,

You are correct about the first flight and publish dates. But, Camm himself mentioned becoming aware of Meredith's findings during the development phase of the Hurricane, and subsequent incorporation of various aspects in the Hurricane's radiator cooling system.

For its first flight, the original radiator installation shroud(?) on the Hurricane prototype K.5083 was basically a carry-over from the Hawker Fury. It was basically just a u-shaped piece of sheet metal with somewhat convex shape to the bottom surface for strengthening, intended to protect the radiator from damage and not much else. It had no boundary layer lip/notch, a vertical front about 2" in front of the cooling array with thin protective bars across it, a vertical rear (with a larger orifice than the inlet) without the adjustable flap, and the total length of the shroud was only about 4-5" more than the radiator itself. If you look at the very early pictures of the prototype radiator this arrangement can be seen.

The radiator housing changed over the next 6-8 months, following the sequence listed below as far as I can tell from pictures and drawings of the time:

First it received a deeper inlet to about 8" and the protective bars went away.
Then the housing aft of the radiator was lengthened and shortly thereafter a short flap appeared on the bottom.
Then the inlet was deepened again to about 14", given a flattened oval shape, and a boundary layer lip/notch showed up.
Then the aft housing became tapered, reducing the size of the exit orifice, with a longer/larger flap (the original housing had a larger orifice at the rear than at the front).
Then the back section received more of a taper with the flap consequently moved farther back.
Then the side trailing edges were tapered into the fuselage.

The radiator itself was replaced with an improved design somewhere during the same period as above, prior to production.

When the Hurricane II came along the radiator was again changed, to a larger more efficient design, and the housing was deepened vertically with a more rectangular orifice, and slightly modified to keep the drag/cooling drag to a minimum. The radiator and housing were changed again with the advent of the Hurricane V, the radiator being significantly larger, and the housing deeper with a much larger inlet orifice.

I am new to posting stuff on forums so bear with me if the following attachments do not show up. I ran across these in the past and did not record the origin at the time, so not sure what to do about that.

 

[ThomasP]

Hey wuzak again,

I am not saying you are wrong about NA having 4-5 years to study the Meredith effect, but do you know if they actually did? Seriously, if you have read something specific that says so, please let me know, I would like to know for my own research purposes. Thanks.

 

[wuzak]

Hey wuzak again,

I am not saying you are wrong about NA having 4-5 years to study the Meredith effect, but do you know if they actually did? Seriously, if you have read something specific that says so, please let me know, I would like to know for my own research purposes. Thanks.

No, I have no direct knowledge of this.

But they did not guess the configuration and somehow get it right. Whether they had access to Meredith's research (probable, as it was published) or other research by NACA, or even by themselves, I could not tell you.

 

[ThomasP]

Hey wuzak,

Sorry for my vagueness, I did not mean to imply that NA guessed at the design work on the P-51. I just meant that I thought that their engineers really did not get going on the serious work needed to incorporate the Meredith effect into an airplane until they started on the P-51. I do not remember reading about any previous intent/attempt by NA to incorporate the technology in an airplane prior to th P-51, and I would be interested to find out if they had.

 

[Shortround6]

The Mustang may have been the first design they cut metal on that used a liquid cooled engine, I have no Idea what they had on paper.

American liquid cooled engines petered out with the last of the Curtiss conquerors in P-30s and the aviation scene in the US was dominated by air cooled radials for 4-6 years. The Allison was know to be in development along with the Army's hyper engine/s but the only engines most manufacturers could get their hands on were the radials.

This was one of the first NA aircraft (they took it over from Fleet (?) who took it over from Pilgrim division of Fairchild ( or the other way around?)

This had a lot more troubles than figuring out thrust from cooling.

First real NA design was father of the AT-6

There was also the XB-21, the 0-47 and the NA-40

predecessor of the B-25.

But no liquid cooled engines. If figuring out the Meredith effect was hard using a radiator then air cooled engines were a real pain in the rear.
Th e"principal" is the same but the methods are going to be a lot harder.

 

[tomo pauk]

Thing with Hurricane's radiator is that it was out of airframe by 100%, thus increasing frontal area as much as possible. Radiators on P-51 and many other fighters (Spitfire, Bf 109, Yaks etc) were burried within the airframe by a good deal, meaning increase of frontal area being smaller, and with it the drag.
LE radiators can be easily designed so they don't add any frontal area, as it was the case with, not only, Mosquito or Hornet.

 

[ThomasP]

Hey Shortround6 and tomo pauk,

"This had a lot more troubles than figuring out thrust from cooling."

Thanks for the NA history, Shortround6, and LOL! thanks for the laugh.

To tomo pauk's point, yeah, it it would have been nice if the radiator was semi-submerged, and in all seriousness maybe it could have been, but not without some serious redesign of the airframe?? The prototype design had full span trailing edge flaps, which was why the aft end of the radiator housing had the odd negative angle to it, and the front end of the housing was forced to end about where it did (on the production aircraft) due to the wheel wells being in the way. In order to incorporate the effect as far as they did they had to eliminate the center section of the flap to extend the aft part of the housing, and I do not know what would have to have been done to to the lower fuselage structure to semi-submerge the radiator, but I suspect that it would have been expensive (relatively).

As it was the potential extra 27 HP (if you got the maximum possible Meredith effect at the Hurricane's speeds) would have only added another 1-2 mph and/or made up for some weight gain and/or improved rate of climb by about 80 ft/min. Nothing to laugh at, but diminishing returns and all that.

It would have been interesting to see Camm's thin wing Hurricane with a more refined drag index. What else would Camm have changed if he choose to keep the basic design/construction? sigh.....

 

[tomo pauk]

...
It would have been interesting to see Camm's thin wing Hurricane with a more refined drag index. What else would Camm have changed if he choose to keep the basic design/construction? sigh.....

I'd go for beard radiator, proper exhaust stacks (not the draggy ones used historically) and injection carb = performance increase on time and on budget.
Granted, latest two suggestions aren't exactly airframe's designer's prerogatives.

 

[Shortround6]

The draggy exhaust may have been part of the "kit" needed for the night fighter requirement.

Note shields to help block exhaust glare from pilot's view.

The Hurricane and Spitfire were both supposed to be able to operate at night. Take-off, fly around and land without too many accidents.
Actually finding enemy aircraft (or even your own airfield) was another level of complication.

Some sort of flame dampening/suppression was needed to meet the requirement. If it cost a few MPH off the top speed, well, so be it.

 

[pbehn]

Hey wuzak again,

I am not saying you are wrong about NA having 4-5 years to study the Meredith effect, but do you know if they actually did? Seriously, if you have read something specific that says so, please let me know, I would like to know for my own research purposes. Thanks.

There was a pic posted somewhere showing the various incarnations of the P-51/Mustang/Apache series. This required a lot of wind tunnel work to get right. However, during a short period or time the Mustang/P51 went from being a small order of fighter planes for a foreign government to a vital part of the USA strategic offensive. Since the radiator design was vital to the range of the P-51 I doubt they had any trouble at all getting any wind tunnel time they wanted from 1942 onwards.

 

[pbehn]

The draggy exhaust may have been part of the "kit" needed for the night fighter requirement.
View attachment 533866

Note shields to help block exhaust glare from pilot's view.

The Hurricane and Spitfire were both supposed to be able to operate at night. Take-off, fly around and land without too many accidents.
Actually finding enemy aircraft (or even your own airfield) was another level of complication.

Some sort of flame dampening/suppression was needed to meet the requirement. If it cost a few MPH off the top speed, well, so be it.

From the Spitfire society web site …….It was soon discovered that simple changes to the ejector exhausts from simply blowing out to the side to being directed back would increase speed. The exhausts evolved from round outlets to fishtail in appearance which also had the bonus of reducing exhaust glare during night flying. These changes resulted in harnessing the exhaust gases provided an additional 10mph or 70 horsepower. The exhausts alongside forward facing intake ducts were used to heat the guns in the wing which were prone to stoppages at altitude as a result of the colder temperature, and superior to the earlier heating from the engine coolant radiator. During the Battle of Britain it was discovered that the Merlin engine would cut out when pursing Me109s in a high speed bunt dive due to fuel starvation in the float controlled carburettor. Initial solutions involved inverting the aircraft into the dive and also the fitting a restrictor in the fuel supply line and a diaphragm known as Miss Shilling's orifice, named after the female inventor (Beatrice Shilling) based at Farnborough at the Royal Aircraft Establishment. More permanent solutions involved moving the fuel outlet from the bottom of the carburettor to half way up and the use of fuel injection using a Stromberg pressure carburettor and finally an SU injection carburettor.

 

[Shortround6]

The rearward blowing exhausts were adopted fairly early, while photos like the 5th one in the series from post #68 can be found showing the extended exhausts with the slits ( I don't know the official name) most photos of Hurricanes in France, even with two blade wood props and fabric covered wings, show the rearward blowing exhausts although not yet fish tail.

 

[pbehn]

The "secret" to the P-51 radiator set is a combination of things, not one or two features by themselves.

View attachment 532536
1. You have the extended air scoop/boundary layer splitter. The turbulent/swirling air next to fuselage skin is diverted away from the radiator duct.
2. The air headed for the radiator is allowed/forced to expand considerably in long enough duct that major eddies/swirls are not introduced.
3. The expanded air is much lower velocity and as it passes through the radiator it creates less drag (drage being equal to the square of the speed)
3a. The slower moving air might pick up more heat (not sure about this)
4. The expanded and heated air is compressed and speeded up by the converging walls (or roof/floor) of the duct over enough distance that excess turbulence or resistance is not introduced (some probably is).
5. the exit opening is sized and adjusted inflight ot not only adjust to total airflow through the radiator but to try and make the exiting hot air have a velocity higher than the forward speed of the plane (this is where things get iffy as to whether it was actually making thrust)
6. the direction of the air (radiator exhaust stream) leaving the duct is pretty much in line with the direction of travel of the airplane.

Now lets compare the Mosquito radiator, simply because I am lazy and it is the first picture I found on the internet.
View attachment 532537

It certainly meets condition #1
It is expanding the air leading to the radiator as in condition #2, few neweddies introduced.
Is the air slowed down enough to be as low drag going through the radiator matrix? This may depend somewhat on the radiator matrix itself.
Has it picked up enough heat to materially expand the air (which is different that just cooling the engine)?
The Mosquito;s exit duct is shorter and somewhat more off center, a potential sources of drag/eddies.
and finally, even if the air leaving the duct is leaving at a higher speed than the aircraft is moving the direction of the just thrust is at more of an angle to the line of flight meaning some of the effort/thrust is wasted.

No I have not mentioned the difference in frontal area for each design and this are illustrative drawings and not blue prints with dimensions so perhaps too much should not be read into them.
I am not criticizing the Mosquito (the basic radiator design predates the Merlin power Mustang radiator setup) but just trying to show that there are large variety of factors that come into the situation and focusing on just a few may miss some of the others.

Does anyone know if the Mosquito leading edge intake reduced or increased drag with the radiator set up as is, and if the "Meredith effect" as far thrust goes was also combined lift, the outlet is pointing downwards as well as backwards?