Meredith Effect and the P-51

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

Here's a pic I took of the radiator duct on a P51A (maybe an A-36?).

Location was Chino 2006.
 

Attachments

  • P51 parked Chino May 20 2006 k.jpg
    79.8 KB · Views: 329
  • P51 parked Chino May 20 2006 l.jpg
    84.6 KB · Views: 334
Seventy years later we still don't really know why the Mustang had less drag than the P-40 (Allison powered) or the Spitfire (Merlin powered). I find this amazing.
 
hrandy said:
...or the Spitfire...
Click to expand...
If you read the piece
the Spitfire never really capitalised on and exploited the air pump effect in the same way that the Mustang did, the Supermarine design team stuck with the symmetrical box radiators from the Mk IX all the way through to the end of the line for the Spitfire.
There doesn't seem to have been any attempt to separate boundary layer air from more stable air at the intake and the radiator flap didn't provide the same jet pipe effect that the Mustang achieved at the exhaust.
I personally think that it's significant that the Mustang carried out all of its cooling with one appliance (the ventral arrangement) while the Spitfire and many of its inline contemporaries used two (mounted under the wings, usually) where this would increase drag on the airframe.
 
hrandy said:
Seventy years later we still don't really know why the Mustang had less drag than the P-40 (Allison powered) or the Spitfire (Merlin powered). I find this amazing.
Click to expand...

We really do know.

The amazing thing about the approach (underwing radiator and scoop/cowling) used by NAA is that it was purchased from Curtis.

The modification from the P-51A to B was a major reduction in drag and high speed rumble and the P-51H used most of what was learned from the B/C/D design to reduce the drag even more.

The question we have been debating is not so much the form/parasite drag of the lower radiator cowl design but the actual benefits of the Meridith Effect - which are hard to quantify.
 
drgondog said:
The amazing thing about the approach (underwing radiator and scoop/cowling) used by NAA is that it was purchased from Curtis.
Click to expand...

G'day Bill.
Was that part of the $56,000 package? Being pre-Pearl Harbour, how is it that the AAF could apply pressure to a private firm?

 
Graeme said:
G'day Bill.
Was that part of the $56,000 package? Being pre-Pearl Harbour, how is it that the AAF could apply pressure to a private firm?
Click to expand...

Graeme - I will have to research here but remember the P-51 was built for Brits not USAAF. IIRC the data (and rights were secured from Curtiss by directive of USAAF so that NAA would have all necessary design details to build P-40 if the P-51 failed).

The legend is that the origin of the aft/under wing was contained in a P-36 Modification proposal delivered to USAAF in 1938, and the resulting data was incorporated in the 'complete package' delivered to NAA by Curtiss as 'part of the 'complete up to date data required to build the P-40.

The actual design was in no way tied to the 'P-36 preliminary design' but represented an available departure in concept from the Beard approach of the P-40... and the P-36 Data was sufficiently promising to lead NAA to that approach.

The contractual Price of the Mustang Mk I was $50,000 - the later price to USAAF for the P-51A was $56,000. I have no idea why the difference or the CLI/WBS changes that may have occurred between the US and Brit version - if any.

Your post also seems plausible and I can not refute the Kelsey version.
 
Colin1

Although I've always been impressed with the Mustangs' Meredith Effect radiator I have doubts that this is the complete explanation for several reasons.

In the 1990 edition of North American p-51 Mustang the author, Bill Gunston, states that "the Royal Aircraft Establishment (RAE) carried out careful tests and calculated the D.100 figure (the drag at an airspeed of 100 ft/sec [30.5 m/sec]) to be 50lb (22.7kg) compared to 65.5lb (29.7kg) for the new Spitfire V".

Gunston doesn't give much detail but because of the low speed I'm guessing these tests were on complete aircraft without propeller and no engine power in a large wind tunnel. Lee Atwood and others have stated that the "Meredith Effect" only became significant at speeds approaching 300 mph and of course with a hot radiator.

How do we account for the significantly lower drag of the Mustang compared to a Spitfire under a test regime that would minimize the Meredith Effect?

In late 1942 Roll-Royce began testing of several Mustangs the modified to take 60 series Merlins. They added charge cooling and oil cooling radiators just behind the propeller of the Mustang while retaining the original non boundary layer bypass radiator with some modifications for engine cooling. These aircraft produced speeds less than 10 mph less than the boundary layer bypass equipped production P-51Bs.

How do we account for the cobbled up Rolls-Royce experimental Mustangs being almost as fast as boundary layer bypass equipped production P-51Bs when roughly half the heat rejection of the experimental Mustangs went through very crude non Meredith Effect nose radiators?

Looking at published USAAF ratings, early Allison Mustangs with their non boundary layer bypass radiators seem to be thirty plus mph faster than P-40s with the same mark of Allison engine. Roll-Royce rated the Allison Mustangs about thirty plus mph faster than the Spitfire V at the same power level at the same altitude. The P-51B/Csand D/K s with the boundary layer bypass radiators seem to be about twenty five mph faster than similarly engined Spitfire VIIIs.

How do we account for boundary layer bypass equipped Mustangs seemingly losing some of their speed advantage over their contemporaries compared to non boundary layer bypass Mustangs?
 
Hi hrandi

Although I've always been impressed with the Mustangs' Meredith Effect radiator I have doubts that this is the complete explanation for several reasons.

The thread was presented as an insight into Meredith Effect and its implementation in the P-51, it doesn't state anywhere (that I can remember) that it was the sole reason why the Mustang was as good as it was. The P-51 air pump mechanism doesn't even completely overcome cooling drag but it does significantly reduce it.

In the 1990 edition of North American p-51 Mustang the author, Bill Gunston, states that "the Royal Aircraft Establishment (RAE) carried out careful tests and calculated the D.100 figure (the drag at an airspeed of 100 ft/sec [30.5 m/sec]) to be 50lb (22.7kg) compared to 65.5lb (29.7kg) for the new Spitfire V".

I've got no idea how to calculate a D.100 but back when the Spitfire was new, how did the RAE factor in the thrust recovered from cooling drag? Is there a paper that shows they've considered it (I doubt the actual mathematical analysis would be necessary)?

Gunston doesn't give much detail but because of the low speed I'm guessing these tests were on complete aircraft without propeller and no engine power in a large wind tunnel. Lee Atwood and others have stated that the "Meredith Effect" only became significant at speeds approaching 300 mph and of course with a hot radiator.

Atwood (and Meredith, and Junkers) were all absolutely correct and an air test at 100ft/s would not reveal anything significant about an air pump mechanism's performance. I'm not even sure what they would have learned about airframe drag at that speed.

How do we account for the significantly lower drag of the Mustang compared to a Spitfire under a test regime that would minimize the Meredith Effect?

The Spitfire was not exactly a 'dirty' airframe but it did have one or two aspects that could have been (easily) improved, the rake of the windscreen was steeper than in the Mustang and the adherence to boxy, underwing radiators with inefficient exhaust closing options was another. I don't know how aerodynamicists group various types of drag but the screen would have provided parasitic drag at higher speeds and the underwing radiators would not have recovered thrust as effectively as the Mustang's arrangement from cooling drag (maybe cooling drag is a form of parasitic drag, given that its effects are amplified with increases in level-flight speed).

In late 1942 Roll-Royce began testing of several Mustangs the modified to take 60 series Merlins. They added charge cooling and oil cooling radiators just behind the propeller of the Mustang while retaining the original non boundary layer bypass radiator with some modifications for engine cooling. These aircraft produced speeds less than 10 mph less than the boundary layer bypass equipped production P-51Bs.
How do we account for the cobbled up Rolls-Royce experimental Mustangs being almost as fast as boundary layer bypass equipped production P-51Bs when roughly half the heat rejection of the experimental Mustangs went through very crude non Meredith Effect nose radiators?

I wouldn't. Not without seeing alot more information concerning these tests, do you have a URL or other?

Looking at published USAAF ratings, early Allison Mustangs with their non boundary layer bypass radiators seem to be thirty plus mph faster than P-40s with the same mark of Allison engine. Roll-Royce rated the Allison Mustangs about thirty plus mph faster than the Spitfire V at the same power level at the same altitude. The P-51B/Csand D/K s with the boundary layer bypass radiators seem to be about twenty five mph faster than similarly engined Spitfire VIIIs.
How do we account for boundary layer bypass equipped Mustangs seemingly losing some of their speed advantage over their contemporaries compared to non boundary layer bypass Mustangs?

I would start by looking at the empty weights of the Mustang variants that you have just described: P-51A at 6,433lbs; P-51B at 6,985lbs and P-51D at 7,125lbs.
Normal take-off weights are then 8,600lbs, 9,800lbs and 10,100lbs respectively.
I have mentioned more than once on this forum that when the D variants began filtering down to front-line units they invariably went to flight commanders. They immediately noticed that their flights (still in their B and C variants) were some 3-5mph faster than they were in their Ds.
 
This would be the Mustang X
of which there were five conversions, I found this

By early Oct42 the first Mustang X conversion AL975 was completed and after ground running made a 30-minute flight on the 13Oct42 using a Merlin 65 driving a 10ft 9in diameter screw.
With the Merlin 65, the estimated maximum speed in FS gear was 427mph at 21,000ft and in MS gear 392mph at 9,600ft.
There then followed some experiment to see if an improvement in speed could be gained by modifying the intercooler radiator air exit louvres and the main radiator ducting; at the moment there was a shortfall of some 16mph below the estimated figure with these radiator configurations. All but 2mph was recovered by reducing the exit area of the main radiator by almost half. Also tried were changing the reduction gear ratio to 0.42:1 with the same Spitfire IX propeller, streamlining the intercooler radiator air exit louvres, fitting a larger intercooler radiator and deleting the fuel cooler, flattening the bottom section of the cowling slightly. The result was a small speed recovery and an improvement in the intercooler's effectiveness.
For the seventh flight, the 11ft 4in diameter screw was fitted and the intercooler radiator exit louvres on the starboard side were blanked off but there was no improvement in speed, there was even a small drop on the next flight with all the louvres open. It was thought that one reason for the speed shortfall might be that air passing through the intercooler radiator into the engine bay was being ejected through the cowling joints at right-angles to the flight path and to overcome this, a duct was fitted allowing the air to be piped directly from the radiator overboard, thus dispensing with the need for the louvres which were then blanked off. For the ninth flight, a tinned steel tank replaced the Mareng tank in the port wing which now made it possible to obtain performance figures in both MS and FS gears instead of MS gear only as before. On 08Nov42 a flight revealed a speed of 413mph in FS gear or 14mph short of the estimate and 390mph in MS gear.

Surprisingly, when the original 10ft 9in diameter screw was refitted, the speed in FS gear increased by 9mph to 422mph and flight tests confirmed that performance figures for this prop were contrary to expectation, being low below 26,000ft and high above this altitude.
For the next flights the intercooler radiator louvres were unblanked and it was found that the duct recently fitted was not responsible for the 9mph speed increase that occurred when the original screw was refitted.
Although there was still a speed shortfall, the rate of climb with the Merlin 65 proved to be better at low altitudes than the estimates and practically the same from 13,000ft to 34,000ft.
The original engine suffered bearing failure and on being replaced with another Merlin 65 attained the design speed of 427mph, once the main radiator front flap was fixed in a permanent position and sealed.

In Feb43 AL975 was fitted with an RM11SM Merlin 70 which had a max power output 50hp less than the Merlin 65, the same supercharger gear ratios as the Merlin 61 and a reduction gear ratio of 0.477:1. Another change was made following flight trials with Mustang X AM208, the addition of a dorsal fillet to the fin to improve directional stability; a large change in directional trim at high power and speeds had been experienced and there was a tendency to sideslip easily during manoeuvres. The dorsal fillet fitted to AL975 gave only a marginal improvement and so the fin chord was increased to give an extra 3sq ft of area and this did the trick.

The second Mustang X conversion was AM208 which made its first flight on 13Nov42 and was also fitted with a Merlin 65 driving a 10ft 9in diameter screw. It had the main radiator front flap permanently fixed and sealed as originally applied to an Allison-engined Mustang showing a 6-7mph speed increase as a result.
It was this mod that finally made it possible for a Merlin 65-powered Mustang to attain 427mph at 21,000ft in FS gear, vindicating the whole idea of fitting the Merlin. Performance trials with AM208 at A&AEE Boscombe Down culminated in Apr43 with max true airspeeds of 406mph at 10,000ft in MS gear and 433mph at 22,000ft in FS gear at the combat rating of 18lb/sq in boost.
The radiator front flap mod had resulted in a top speed increase of 11mph over AL975 without this feature. AM208 and the three subsequent conversions did not feature the slight bulge in the lower engine cowling that AL975 had in it original form. After the speed tests another Merlin 65 was fitted to AM208 in place of its original powerplant and climb trials were undertaken, these showed a rate of climb of 3,650ft/min at 7,500ft in MS gear and 2,840ft/min at 19,000ft in FS gear, the time to reach 20,000ft being 6.3 minutes.

The third conversion was AM203 first flown on 13Dec42 and fitted with a Merlin 65 driving an 11ft 4in diameter screw. AM203 was sprayed in a high-gloss finish to evaluate its effect on performance. It went to the Air Fighting Development Unit at Duxford on 23Dec42 and here achieved a max speed of 431mph at full throttle height of 21,800ft.
When the 10ft 9in screw was refitted, the same unexpected increase in speed was experienced, this time by +3mph in FS gear. The original 11ft 4in screw was refitted before the a/c was loaned to the USAAF for evaluation.
It returned to Hucknall on 22Mar43 to be resprayed in a matt finish and simultaneously had a dorsal fin fitted. There was found to be no difference in speeds produced by the two finishes, this was put down to the Mustang's already very smooth construction finish.

The fourth conversion AL963 flew with a Merlin 65 on 21Jan43. In Feb43 it went to Hucknall for the fitting of a dorsal fin. It was re-engined with a Merlin 65 Special with an SU Mk 2 fuel injector pump in place of the previous Bendix-Stromberg unit. The new engine's compression ratio was reduced from 6:1 to 5:1 with 100 octane fuel and the max permissible boost was increased to +23lbs/sq in for better low-altitude performance.
This engine was intended to give preliminary data for the coming RM14SM Merlin 100 series with higher capacity supercharging giving higher full-throttle heights and max boost of +25lbs/sq in. It gave AL963 a max speed of 370mph 1,000ft or 17mph faster than with the Merlin 65. At 25,000ft it gave 420mph, the same as the Merlin 65. Rate of climb also improved with the 65 Special by 900ft/min to 4,700ft/min from ground level up to 3,600ft and by 940ft/min to 3,980ft/min at 15,200ft.
After 45 3/4hrs flying time, the Special was replaced with a Merlin 66 to evaluate a new type of intercooler. This was moved from the nose to be incorporated with the main radiator just aft of the wings. This made it possible to reduce the size of the nose intake significantly, approximating that of the production P-51B. The new config caused a few problems with the efficiency of the main radiator and after a few hours flying time, AL963 reverted to the original config. By this time Rolls-Royce had agreed to a US design for a larger matrix incorporating both radiator functions and a slightly deeper ventral intake to accommodate it.
At some stage the dorsal fin was removed and replaced by leading-edge fin extensions that were also applied to AL975; AL963 completed 102 1/2hrs flying time before being stored on 06Jan44, from where it was SOC on 23Nov44.

The fifth conversion AM121 was the first to arrive at Hucknall but was kept on for performance calibration for 8 months. It first flew on 07Feb43 fitted with a Merlin 65. By this time most of the development flying had been done and AM121 was earmarked for the USAAF for evaluation. It went to AFDU at Duxford before being loaned to the 8 AF Fighter Command at Bovingdon. It was returned on 03May44 for fitting of broader chord fin similar to that of AL975 and on 11May44 it went to Rotol's Flight Test Department at Staverton to test the 0.020in thick Rotoloid protective coating for the blades in all-weather conditions; it flew 10 hours per day between 13 and 27May - remarkable for a fighter and equal to the best airliner utilisation figures of post-war years. Unfortunately it showed that large areas of Rotoloid coating had been shed from the blades as it was too thin in its current form.

On 27Jun44 AM121 returned to Hucknall where it was re-engined with another Merlin 65 and on 26Jul it went to 8 AF Fighter Command's Air Technical Section at Bovingdon again. It was finally broken down for spares in Aug44.

I fail to see what is so remarkable about the speed differential between the Mustang X and the production Merlin Mustangs; they are one and the same a/c with the only significant variation in aerodynamic form being the deeper chin arrangement on the Mustang X, this manifests itself on the a/c's performance in the speed differential demonstrated and without access to hard analytical data, would seem to be appropriate.
 

First - the P-51B was a lot faster than Allison equipped A and MkI.

Second - the Production P-51B was faster than the Mk I which received the first Merlin despite being heavier. The primary differences between the Brit Prototype and the P-51B-1 were;

a. Wing dropped 7 inches to accomodate lower cowl re-design
b. Merlin replaced Allison
c. Carburator cooling relocated from top of cowl to 'small chin'
d. Lower Radiator Cowl/Radiator, Oli Cooler and radiator exhaust shutter were re-designed
e. Armament arrangement changed in wing from one 1x.30/2x .50 in each wing to 4x.50 cal
f Hamilton Standard four blade prop with increase of 5" diameter over the Curtiss 3 blade

I will have to dig but the Thrust Line also changed slightly bewteen Merlin and Allison...as well as tail incidence (IIRC)
 
Also in comparing Allison engined Mustang vs the Warhawk, as far as speeds. I believe the P-39 was some 20mph faster than the Warhawk. I don't know how the cooling system on the Airacobra is classified. I have read little about it. I don't know (or think it is) a Meredeth effect design. It is just "cleaner" than that of the Warhawk.

I think this also leads to the fact that the overal design of the Mustang was cleaner, as was the Airacobra, as compared to the Warhawk. The Meredeth radiator on the Mustang may have just been "icing on the cake".
 
Of course we know -
The P-51 was carefully laid out - a very clean airframe without the extra lumps, bumps,
bulges and scoops that covered every other airframe. It had a very low frontal area,
and excellent streamlining. WHile the airfoil wasn't always achieving the goal of as much laminar
flow as wind tunnel tests of the airfoil shape, it waqs still a very low drag profile with very good
behavior in compressible flow.
The cooling system was generating enough thrust to pretty much cancel its drag. That's signifivant.
Hoerner's analysis of the Bf 109G shows the ccoling drag contribution to be 17% at Mach 0.55
That's about 300# of drag. Other World war 2 fighters carried a similar burden.

A word on the Meredith Effect in general, and how it applies to the Me 109 and Spifire intallations:
A belly radiator, or a flap on the exit side of hte radiator don't mean that there is any Meredith Effect.

In order to obtain the compression of the air ahead of the radiator needed to get any benefit (Both by cooling
through the radiator and as a ramjet) you need a very carefully profiled diffuser. (The front part of teh duct), fed clean (not boundary layer) air. In order to get a proper diffuser profile, the duct needs to be long enough, and hava a profile without sharp bends or sudden area changes. The underwing radiators on the Me 109 and Spitfire can't achieve this - the ducts are too short, and the change in duct cross section is too small.
The same holds true for the underfuselage ducts for the Huricane, the Ki-61, and various Soviet fighters.

In order to be able to use the energy of the heated, relatively high pressure air coming out of the radiator, you
also need a carefully profiled duct with a strictly controlled exit area. Neither the 109 or the Spit the expansion duct, or the exit area control. THe controllable flaps on those airplanes served only to control mass flow through the radiator.

It's not voodoo - just very careful engineering and execution in manufacturing.

--
Pete Stickney

.
 

All very true..

Also, the drop of the wing to accomodate the Merlin also gave the NAA engineers an opportunity to greatly reduce the drag - particularly the lower cowl geometry and location (vertically) to a.) minimize boundary layer feed and .b.) improve the inner plenum geometry in front of the radiator.

There was a significant improvement in just profile drag w/o consideration of the Meridith effect.

As was pointed out, although 'true' laminar flow expectations of moving separation back to ~ 40% where t/c max occurs, the P-51 wind did have less drag than the conventional t/c ~ 25-28% wings of the day. The imposed manufacturing surface quality was also a factor based on both the flush rivets combined with very good surface planes (i.e no ripples between ribs or spars).
 
syscom3 said:
How did the P-38 fare in this?

It should have the benefit of having the turbocharger exhaust act as a small jet engine.
Click to expand...
All inlines of the period enjoyed this phenomenon
whether they were turbocharged or supercharged, this was achieved with the shaping and canting rearwards of the exhaust stacks.

The wastegate of the P-38 does not appear to have been either shaped or canted (rearwards) in any deliberate attempt to dump heated gas into the airflow.

Edit: though the P-47 (notably, a radial) did implement a strategy for doing this
 
Last edited:
You must log in or register to reply here.

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)

Similar threads

BF-109’s and their engines
Replies
3
Views
1K
Laurelix
Any Sources on Sea Level max speeds for the Ki-43’s?
Replies
4
Views
505
taly01
Y
Tests on the La-5/7 series
Replies
3
Views
994
Trilisser
T
V
I found this an interesting report in a War Diary regarding the Corsair use in carrier aviation
Replies
9
Views
865
Escuadrilla Azul
E
XP-82/P-82B SAC manual info?
Replies
0
Views
702
BarnOwlLover
Menu
Log in
Register