The P-38 with a P-51 type mounting for the radiators and coolers.

[Draconis]

Thinking about the OTL development of the P-38 the biggest change was from the earlier P-38 versions to the P-38J. This is the most visibly obvious change as the look of the front of the engines nacelles changes from the more streamlined shape to one more resembling the P-40s' nose.


Glacier Girl, a P-38F a provides a good view of the front of her engine nacelles.

And here is a later model, either a P-38J or L. Showing the change in the look of the nacelles.

These changes to the the P-38 were required to accommodate the more effective intercooler needed by the more powerful versions of the Allison engines with its new turbocharger. The old leading edge intercooler design wasn't adequate. The new chin location for the intercooler, packed in with the oil coolers worked but the wider shape did introduce more drag. I think a different approach to where to locate the oil coolers and prestone radiators could have been examined and developed in late 1942, early 1943. The new intercooler could still be able to fitted in about the same place in the nose.


Here are two pictures of P-51s showing their radiator housings from different angles. This helps
provide an idea of the width and length required for this installation. The P-51 radiator housing design
was able to utilize the Meredith effect.

Instead of the OTL changes the Lockheed engineers did for the P-38 I think a more advantageous though radical design change would see the fitting on a P-51 type radiator housing to each P-38 engine nacelle.

These would hold both the prestone coolers and the oil radiators. Pretty much a close copy of the P-51s' design in shape and size it would be able to use the Meredith effect as well reducing drag. The front of the engine nacelles would also be much slimmer thereby reducing drag. There would only be one opening needed for the intercooller intake slightly offset from lower centre of the front of the nacelle.

What about the main landing gear? Well, why does the placement of the radiator housing have to be directly under the centre of the underside of the nacelle same as the P-51? I think there is room to fit the housing on the lower inner side of each nacelle between the landing gear and the underside of the wing. Placed low enough to allow the flaps to be opened.

So try to visualize a P-51 style scoop and housing on the inner side of each engine nacelle starting about halfway back of wing chord. The benefits are the improved over all streamlining of the P-38 with the smaller nacelle frontal cross section, the removal of the draggy radiators from the tail booms and the advantage of using the Meredith effect. Also the weight of the oil coolers and radiators are moved closer the the planes' centre of gravity. And there is a reduction in piping for the prestone radiators. This all would require a larger redesign effort than the OTL changes but it would provide a better, faster airplane.

Perhaps while they are at this Lockheed can hasten the addition of the anti-compressibility tuck dive recovery flaps into this new production model. It will still need those too.

 

[HB of CJ]

Excellent question and thank you. My uneducated guess here would be mounting the bigger turbos and inter cooling plumbing would have been too difficult to hang it all under the engine. Also aerodynamics come to mind. Such under front mounting MAY have slowed down the aircraft? The designers had room behind the engines. They had little room just below. Finally the production versions of the P38 were "good enough". There was a nasty world war going on.

However I do wonder at times if the P38 might have had more earlier success in Europe with the dive brakes and turbo compound Allisions? Also a three gun version with all 20 mm cannons.

I am misspelling Allision.

 

[PorkNuggets]

There are two big problems as I see it. First is that there just isn't much room back there. The P-38's booms and gondola are substantially narrower than the P-51's fuselage and much of that space is occupied as is. You can't really mount it on the boom itself without impinging on turbos and/or landing gear, and you can't really put them under the central wing assembly without giving up a lot fuel and or payload capacity because the radiators (along with thier pluming and duct-work) will be occupying space historically occupied by fuel tanks and the scoop will be obstructing your two primary hardpoints.

Second, and possibly more important though harder to characterize is the matter of center of gravity vs center of lift/drag. That's a lot of weight that you're proposing to shift around. It's going to have an effect.

As it stands I think EverKing's NACA Modified P-38 is likely the best anyone can hope short of a full ground-up redesign.

Edited: Typo

 

[marathag]

There are two big problems as I see it. First is that there just isn't much room

Really is only room in the Chin area. You might be able to do some better NACA tested ducting up there for some Meredith effect. Much of that is the exhaust design as the inlet.

 

[Draconis]

Excellent question and thank you. My uneducated guess here would be mounting the bigger turbos and inter cooling plumbing would have been too difficult to hang it all under the engine. Also aerodynamics come to mind. Such under front mounting MAY have slowed down the aircraft? The designers had room behind the engines. They had little room just below. Finally the production versions of the P38 were "good enough". There was a nasty world war going on.

I'm not suggesting relocating the turbochargers. And the new intercooler location is where they moved it to in OTL.
The engine coolant (prestone) radiators and the oil coolers are what is being moved. Packed together into a housing nearly identical the P-51s'. BTW, this radiator relocation could have been done on the earlier P-38 versions too still keeping the original leading edge, soon to become inadequate, intercooler location. I just thought that the redesign from the P-38H to the chinny P-38J was a good time to implement these other changes.

 

[Draconis]

There are two big problems as I see it. First is that there just isn't much room back there. The P-38's booms and gondola are substantially narrower than the P-51's fuselage and much of that space is occupied as is. You can't really mount it on the boom itself without impinging on turbos and/or landing gear, and you can't really put them under the central wing assembly without giving up a lot fuel and or payload capacity because the radiators (along with thier pluming and duct-work) will be occupying space historically occupied by fuel tanks and the scoop will be obstructing your two primary hardpoints.

Second, and possibly more important though harder to characterize is the matter of center of gravity vs center of lift/drag. That's a lot of weight that you're proposing to shift around. It's going to have an effect.

As it stands I think EverKing's NACA Modified P-38 is likely the best anyone can hope short of a full ground-up redesign.

All excellent points but I think they can be managed. And yes, I agree that EverKing's NACA Lightning is the best and quite plausible P-38 ATL story. My idea is more of a how to further improve the non-NACA OTL Lightning at a later date.

Room to mount the radiator and oil cooler is the biggest concern. But most of the assembly would be fitted outside of the nacelles. Look at the P-51s' housing.

If you visualize the P-51s' fuselage contour from the front of the scoop to the rear exit it's clear that all of the oil cooler and most of the radiator are located outside of the fuselage. I think the small bit of radiator and ducting that would project into the side of the P-38s' nacelles could be accommodated. And that's if the Lockheed designers don't simply fit everything on the outside of the nacelles. It's not that big a change from the P-51 fit. And there are already coolant lines running to the existing rear mounted radiators. Not a big job to shorten them for the new housing.

The coolant radiators and the weight of their mountings are moved forward from their tailboom location causing the C of G to be moved forward. The oil coolers and their mountings are moved rearward form their nose mounted position in the engine nacelles thereby moving the C of G rearward. Do these balance each other out completely? Hard to say precisely but I think it can be managed. This is one of those rare situations in aircraft design when you get two improvemment at once for one change. Having as much weight as possible located as close as possible to the C of G paradoxically improves both maneuverability and stability. The key reason that gave the P-39 its excellent handling qualities was the weighty engine located in the centre of everything.

There is no impact on the fuel tanks. They are located in the wings and are not affected by the radiator housings mounted on the side of the engine nacelles. There are no fuel tanks in OTL P-38s nacelles.

The inner wing hard points could be affected by the housings depending on just exactly how large the housings are and how large the items carried on the hard points would be. Perhaps no change is required. Or maybe the hard points will need to be extended lower a few inches and/or relocated a few inches toward the gondola. Doesn't appear to be any show stoppers here.

 

[PorkNuggets]

I feel like you're overestimating the size of the P-38 relative to the P-51. The P-38's booms and pilot's gondola are all approx 38 inches wide. There are 96 inches between the centerline of the gondola and the centerline of each boom. This gives a length of roughly 58 inches for the inboard wings. The scoop and radiator assembly on a P-51 is approx. 44 inches wide. Meaning that a wing mounted scoop would leave less than 7 - 14 inches for ordnance depending on it's exact position. That's not much, and it certainly isn't enough for the 150 and 300 gal. drop tanks that were carried historically. Likewise even if the most of the P-51's radiator resides "outside the fuselage", the picture you linked makes it look like there's a good 5 - 10 inches of it that does not. As such I'd say my observation above still holds. If you mount it on the nacelle it's likely to impinge on the landing gear and/or turbos, and it can't be mounted on the wing without substantially impacting fuel capacity.

Finally, while it's true that having as much weight as possible concentrated in the center of gravity enhances maneuverability by reducing moments of inertia and thus the control forces (torque) required to effect a change in attitude, it also tends to produce an oscillatory effect unless the center of lift/drag is also well aft of the CoG. This is why we put streamers on kites.

 

[EverKing]

Love the idea and I have some thoughts but they will have to wait a couple days before I am back at a proper computer to look it over in detail.

 

[Draconis]

snip

Finally, while it's true that having as much weight as possible concentrated in the center of gravity enhances maneuverability by reducing moments of inertia and thus the control forces (torque) required to effect a change in attitude, it also tends to produce an oscillatory effect unless the center of lift/drag is also well aft of the CoG. This is why we put streamers on kites.

It seems we have differing views on the engineering difficulties to be faced here. As I don't know exactly what the dimensions and sizing would be it's hard to quantify how difficult this modification would prove. But I don't see it as being insurmountable or even particularly difficult and I also think the benefits would outweigh the costs.

I wouldn't try to draw too close a comparison between the aerodynamics of a kite and an airplane. The kite needs a large amount of passive stability because there is no active controlling to deal with the kites changing attitudes and the motions of the air. Airplanes with their stabilizers, dihedral and 3-axis control surfaces and pilots to operate them require only a minimal amount of stability. How much inherent stability depending on whether it's a fighter or a cargo plane.

 

[Draconis]

Love the idea and I have some thoughts but they will have to wait a couple days before I am back at a proper computer to look it over in detail.

Thanks. I'm looking forward to your take on this. It's all based on my realization that there is no engineering or aerodynamic reason that the P-51 type radiator installation has to be located only on the belly of an airplane. Considering the main landing gear placement on the P-38 I decided the inner side of the nacelles would serve.

Another possibility is locating the housing on the top of the nacelles between the wing roots. The rads would certainly be better protected from ground fire there. This would of course require a major redesign to relocate the turbosuperchargers but I always thought they should have been located closer to the engines. This redesign would admittedly be a real handful for Lockheed and unlikely.

I'm sticking to my OP idea of the rad housings fitted to the side of the nacelles at the time of the new intercooler changes.

 

[riggerrob]

WI you mount new radiators on the outboard side of tail booms?
This will eliminate several inference problems.

 

[Draconis]

WI you mount new radiators on the outboard side of tail booms? This will eliminate several inference problems.

If the radiator housings are located on the outer side of the engine nacelles this will place them directly in front of the turbocharger air intake scoops. And it's even more of a tighter squeeze between the landing gear doors and the extended flaps. Here is another picture of the beautiful Glacier Girl with her flaps fully extended for landing.

The advantages for placing the radiator housings on the inner side are it's better to place the weight as close to the centre of gravity as possible. Including along the longitudinal as well as the lateral axis. Also this puts the oil coolers and Prestone radiators in a better protected spot. As they are then partially shielded from bullets and flak shrapnel.

The problem with either location are the extended flaps as can be seen by the picture. I think that is more of a difficulty than the location of the inner hardpoints. Trying to squeeze the housing between the inner side gear doors and still have the exit port below the fully extended flaps is challenging.

One possible work around is to change the main landing gear doors from the two door clamshell design to a single larger door hinged on the outer side. Instead of the door opening toward the housing it just needs to clear the side as it opens and closes. That would provide a few more inches of room that will permit the radiator housing to be mounted a little lower. Would that be sufficient? I wish I had the design savvy to be more than guesstimating.

 

[EverKing]

OK, here's a quick back-of-napkin analysis of the idea.

The TL/DR version is that there isn't enough room in the booms aft of the main gear housing to allow full radiators in a P-51 style installation package. To do so would require a complete redesign of the booms.

Here's the proof (sorry about the low resolution, these are the best I had available which accurately show the installation):

By looking at this these drawings, you can see that the radiators are considerably larger/wider than the existing booms. To accommodate a similar cooling capacity you will need to maintain a roughly similar area of the radiators (OK, you may be able to go slightly smaller--more in lines with the early P-38D through G radiators--if you design the ducting more efficiently) which means you will need to still have some radiator bulge on either side of the boom. Even then, the central structure of the boom will likely interfere with proper and efficient airflow so you'd still have to use the two-radiator per engine arrangement in something similar to this (very rough sketch):

Even that may not gain you much in terms of improved aerodynamics, although proper engineering may be able to get a little reduction in Net Drag at high speed due to Meredith effect thrust. I doubt such an installation will ever be able to gain Net Zero drag for the installation, much less a Net Thrust.

Hanging the ductwork on the inboard boom only could work if you redesign the central support structure of the boom, but with the P-38's 23016 inner-wing section you may get a complete loss of effective airflow at high speeds (Mach 0.65+, certainly by the onset of compressibility over the inner-wing at Mach 0.68) due to the separation of airflow in the entire area between the booms.

The better solution, if you are dead set on a Mustang-style radiator installation, would be a full redesign of the booms to allow a single radiator per engine with the ducting on the bottom side of the boom. This would likely end up looking something like the P-82/F-82 with a P-38 center gondola instead of full cockpits in the two fuselages. I can't fathom the weight gains from this or any other stability issues with the added weight aft of the CoG.

Anyway, those are my quick thoughts on this. A worthy pursuit but one that I fear is out of reach without a complete redesign of the airplane.

 

[Draconis]

The P-51 radiator housing is just over 8 feet long. A bit more than 2 feet wide. And at the deepest about 3 feet where the radiator is. The depth would vary depending on how much of the radiator and ducting could be buried into the nacelle. I have made the assumption that the same size and number of coolers and radiators, one of each, that is sufficient to cool the Merlin V-1640 will suffice for the Allison V-1710. Especially if the intake is moved closer to the propellors.

It is a real problem trying to figure out how best to fit the housing on the side of the nacelles under the wings. Hard to say what effect it would have on the airflow under the wing. Particularily at high angles of attack.

There is just enough room to put the housing on the top of each nacelle. It's a little over 9 feet from the base of the spinner to the forward edge of the top of the turbocharger. That location would keep the weight close to the C of G. This spot would also have the intake scoop just barely a foot back from the propellors. However it would require some changes to how the exhaust pipes are run to the turbocharger. Plus the redesign of the blast tube and turbocharger cooling inlet piping. Even so this maybe a better approach depending on the plumbing difficulty.

Might look a little odd though.

 

[b0ned0me]

Wouldn’t it be best to just go full P3851 and also delete the Allison+Turbo in favour of Merlin+supercharger? Then you can encroach on the volume used by the turbo to sort out your cooling thing installation.

 

[Draconis]

Wouldn’t it be best to just go full P3851 and also delete the Allison+Turbo in favour of Merlin+supercharger? Then you can encroach on the volume used by the turbo to sort out your cooling thing installation.

That's a whole different ballgame. It raises the old Merlin versus Allison debate which is not what I'm trying to examine here. Nor do I want to reconsider an entire redesign right from the beginning of the P-38 program in my thread. In late 1938 into 1939 I don't think Kelly Johnson and his design team at Lockheed would have included using the principles of the Meredith Effect in their new design. I don't think they would have known about it at the time.

When the P-51 was being developed in mid-1940 the designers at North American had learnt about the Meredith Effect. Possibly being informed by the British experiences with the Spitfire. But that knowledge wasn't available to
Lockheed 2 years earlier.

What I'm attempting to do is implement the use of a more modern (1940s instead of 1930s) designed cooling system utilizing what was known by Lockheed by 1942. And the OTL P-38s' intercooler relocation in late 1942 seems like a good point to do this. However the engineering and aerodynamic considerations are challenging.

 

[b0ned0me]

What I'm attempting to do is implement the use of a more modern (1940s instead of 1930s) designed cooling system utilizing what was known by Lockheed by 1942. And the OTL P-38s' intercooler relocation in late 1942 seems like a good point to do this. However the engineering and aerodynamic considerations are challenging.

But that’s exactly my point. As soon as you start looking into optimising the cooling system you run into trouble because the plane was quite literally designed around two V-12s with turbos. It’s almost impossible to get a P-51 style cooling arrangement without getting rid of the turbos, and since the V-1710 never had a competitive supercharger setup you then end up going to the 1650 as an expedient solution.

1710 to 1650 isn’t quite plug and play but not a huge engineering challenge. Losing the turbos means a lot of weight and space shuffling but so does the new cooling system plus there are lots of other bits that can be moved. So a lot of engineering effort and much calculating but it seems like it should be feasible, might yield a bit of additional aerodynamic cleanup at the same time, and might not be any harder than trying to shoehorn Meredith radiators into volume already full of turbo, ducting etc.

 

[Draconis]

I wouldn't want to replace the Allison V-1710 and its turbocharger. There are benefits to having a turbosupercharged
engine though it did took a while to work the bugs out with the OTL P-38.

And I'm not sure swapping engines would help that much with the P-38. You can't put the radiator housing on the belly of the nacelles because of the landing gear. And I think that trying to fit the housing on either the outer or inner side of the nacelles is unsound because of the disturbance to the airflow under the wings especially with the flaps extended. Putting the housing on top of the nacelles between the nose and the turbocharger is a lousy location too. Trying to relocate the exhaust piping and ducting is too difficult. Having a three foot high hump on either side of the gondola will restrict the pilots' side views also.

I think my idea of trying to install a P-51 style radiator housing would not have merited the effort required and would've been rejected. In OTL they would not interrupt production to change over to paddle-bladed propellors which would have been a much easier change. So I don't think Lockheed would have attempted the rad housing modification. I wonder if the OTL new design P-38J Prestone twin radiator housings on the tailbooms could have been redesigned to better utilize the Meredith effect? That seems like a much simpler project.

@b0ned0me Here is a link to a discussion on the merits of turbocharging versus supercharging. It includes an interesting comparision between the R-2800 horsepower ratings at increasing altitudes of the supercharged F4U and the turbocharged P-47. I think the RR Merlin could have benefited from a turbocharger for some planes. For example Lancaster bomber crews needed all the altitude that they could get.

https://ww2aircraft.net/forum/threads/supercharger-vs-turbocharger.31451/

 

[Draconis]

@EverKing It looks like the NACA modifications you describe in your TL still stand as the most plausible way to improve the P-38. Especially at the very beginning of large scale production for the war. I still wonder if the NACA mods would have obviated the need for dive recovery flaps. I recall you're of the opinion they would still be needed.

 

[EverKing]

I think my idea of trying to install a P-51 style radiator housing would not have merited the effort required and would've been rejected. In OTL they would not interrupt production to change over to paddle-bladed propellors which would have been a much easier change. So I don't think Lockheed would have attempted the rad housing modification. I wonder if the OTL new design P-38J Prestone twin radiator housings on the tailbooms could have been redesigned to better utilize the Meredith effect? That seems like a much simpler project.

That's pretty much the same conclusion I reached in my previous post. I really don't know how much gain you could expect from the rework or whether it would be worth the effort. It all depends on how effectively the Lockheed engineers could redesign the "hourglass assembly" and the "coolant radiator scoop" (Items #3 & #9 in the first image of my previous post) to produce effective Meredith thrust. I showed a rough sketch of this concept at the end of my post.

It looks like the NACA modifications you describe in your TL still stand as the most plausible way to improve the P-38. Especially at the very beginning of large scale production for the war. I still wonder if the NACA mods would have obviated the need for dive recovery flaps. I recall you're of the opinion they would still be needed.

Thank you. I had played around with all sorts of possible P-38 upgrades and none of the them seemed to really do the trick until I discovered the NACA report on Compressibility in the YP-38 which inspired my (on-going at a glacial pace) timeline.

As for the dive flaps I think they would likely still be adopted. NACA estimated about a 0.08 - 0.1 Mach improvement in the delay of compressibility iirc. This would put the NACA P-38's dive limits in 0.76 - 0.78 Mach range. Some fine tuning of the air flow at important junctures and I think they could safely massage it to 0.8 Mach. But at 30,000 feet with the massive acceleration of the two turbocharged V-1710s and 15,000-18,000 lbs of airplane I think they would still quickly be hitting these numbers in even short power-on dives. OTL P-38 could approach its Mach limit in level flight above 30,000 feet so I don't think an aerodynamically cleaner P-38 wouldn't also be nearing its limit, increased though it may be. I think adding Dive Recovery Flaps would still be desirable to give a little more cushion for safe dives from altitude and to keep up / catch up with some of the fast diving opponents. Remember, Republic even added Dive Recovery Flaps to later versions of the P-47 and many (most?) of the early Jets also adopted them--even the OTL dual power Convair XP-81 was designed with DRFs from the start.