Best USAAF fighter for escort missions over Japan

[GregP]

Oldskeptic,

You don't seem to have a good opinion of the P-47. Yes it was heavy. It wasn't any of the other things you described it as. In fact, the P-47 was one of the fastest, if not THE fastest production fighter of WWII in the P-47N model going 470 mph max. The guys who flew it loved it. It's toughness is well documented and it proved it's mettle many times over.

The P-38 was there early, with pilots that had NO training, and a couple of issues with the Allisons that were corrected as well as the fuel issues that was finally discovered and the jetting was corrected. The P-47, like all the other planes in the war, flew the missions it was assigned with the people in the state of training they had at the time.

I don't care HOW you cut it, the P-47 was not ineffective, delivering twenty times the bomb tonnage of the P-51's. Not sure where Drgondog is coming from about the P-51 being more effective at ground attack ... it wasn't. The P-51 DID shoot up a lot of planes on the ground, but they all flew the missions they were assigned.

P-47's flew 423,185 missions and dropped 113,963 short tons of bombs.

P-51's flew 213,873 missions and dropped 5,668 short tons of bombs.

Combat Air Support isn't shooting up enemy planes on the ground on airfields, it is support of the ground troops during ground combat with enemy ground forces. And when the P-47 strafed, it did so with eight 50's, not 6 like the P-51. The P-47 flew 34.6% of all USAAF fighter missions. The P-51 flew 17.5% of them. Which one sounds like it did more for the ground troops? Of course taking into account that the vast majority of the P-51 missions were bomber escort or fighter sweeps and not CAS.

 

[eagledad]

Gentlemen,

In 1950 the Rand Corporation did a study on American fighter aircraft and their vulnerability to anti aircraft fire. The study found:
29% of P-51's hit by AA were lost
26% of F4U's hit by AA were lost
25% of P-38's hit by AA were lost
25% of F6F's hit by AA were lost
22% of F4F/FM-2's hit by AA were lost
10% of P-47's hit by AA were lost.

(Source RM 402 Aircraft Vulnerability in WWII as presented in USAF Jabos in the MTO and ETO by William Wolf, page 53)

I am guessing that the study covers all theaters and all types of missions (from high altitude escort to strafing). However, I do not know how many aircraft were used in each sample, so I can't determine if a 3 or 4 point difference in percentages are statistically different. The study did mention that the distance the aircraft had to travel to get back to base would have an effect on the percentage lost.

Eagledad

 

[drgondog]

Here is an article that does not say x,y, or z is better but does say how it was done in one area.

Allied Strafing in World War II: A Cockpit View of Air to Ground Battle - William B. Colgan - Google Books


Beyond the P-51's vulnerable engine here is why the P-47 was a much better plane for ground attack and other things, from this site: The Republic P-47 "Thunderbolt"


You understand that claims without objective facts remain 'Claim' ??

In the entire history of military aviation, there has never been an airplane that could match the P-47 Thunderbolt for ruggedness and dependability. The pilots who flew it into combat called it "The Unbreakable" and "The plane that can do anything." They were not far from wrong.

Reflect on how hard an IL-2 was to kill..

P-47's often came back from combat shot full of holes, their wings and control surfaces in tatters. On one occasion a Thunderbolt pilot, Lieutenant Chetwood, hit a steel pole after strafing a train over Occupied France. The collision sliced four feet off one of his wings--yet he was able to fly back safely to his base in England.

The second site powered by 456FIS is so rich with factual errors that it is difficult to start.. but let's hit with one major question?
Where in hell does one pull combined air/ground credits in excess of "11,000"? The 8th AF was the only combat command to accurately record claims and credits. Period. The total number of air/ground credits for the P-47 in 8th AF was 1567 air and 740 ground. The total number of P-47 air victory credits (extracted laboriously from USAF 85) is 3661.

Do you suppose that some numbers in the long narrative are way, way off?

 

[eagledad]

Gentlemen,

Let us see what people on the receiving end thought about which Allied plane was best at close air support.

Q 5. TECHNICAL QUESTIONS.
a. Is there any difference in ground force reaction to P-47's, P-38's, P-51's, Spitfires and Typhoons?

A. 5. Difference in effect:

1. Lightning; Unpleasant for ground troops because of its quiet, almost noiseless approach, its maneuverability and speed, its heavy armament and bombs, exact aiming end hitting. Particularly effective against tanks. Direct or near hits destroy the tank,

2. Thunderbolt.} Armament and bombs; noisier, easier to identify from the ground, also very maneuverable and fast. Otherwise, same effect os Lightning, probably lighter bombs. (I, myself, was hit by Thunderbolts five times.) Direct or clcse bomb hits destroy the tank, weapons penetrate all tank armor except that of the "Koenigstiger".

3. Spitfires and Mustangs have not impressed me particulerly in their effect.

The above opinion was voiced by Generalleutnant Fritz Bayerlein, at the time of his capture Commending General of the SS Corps, but during most of the time covered by this report commander of the Panzer Lehr Division.

This information can be found at

Mediterranean Allied Air Forces. Air Surrender Documents

Eagledad

 

[RpR]

Do you suppose that some numbers in the long narrative are way, way off? [/B]

Off?
Probably.
How far off?
I do not know.

I listed that site partly because the gent speaking gave his opinion of what he liked best and why, BUT actually more for the information on the P-47 M which is not often spoken of.

 

[parsifal]

Gentlemen,

In 1950 the Rand Corporation did a study on American fighter aircraft and their vulnerability to anti aircraft fire. The study found:
29% of P-51's hit by AA were lost
26% of F4U's hit by AA were lost
25% of P-38's hit by AA were lost
25% of F6F's hit by AA were lost
22% of F4F/FM-2's hit by AA were lost
10% of P-47's hit by AA were lost.

(Source RM 402 Aircraft Vulnerability in WWII as presented in USAF Jabos in the MTO and ETO by William Wolf, page 53)

I am guessing that the study covers all theaters and all types of missions (from high altitude escort to strafing). However, I do not know how many aircraft were used in each sample, so I can't determine if a 3 or 4 point difference in percentages are statistically different. The study did mention that the distance the aircraft had to travel to get back to base would have an effect on the percentage lost.

Eagledad

fuinally someone produces some numbers to give us a framework to use. I still would like to see or know how this survey was conducted. How did they reach these conclusions?

 

[OldSkeptic]

Oldskeptic,
You don't seem to have a good opinion of the P-47....

No I think the P-47 was a pretty good plane for what it was designed to do. It's weaknesses was its cost, complexity and fuel consumption (roughly the same as a twin engined P-38). As I keep saying at 30,000ft it was a hot ship indeed. But when you got down low then its performance dropped and things like it's weight really started to work against it. Remember a book (somewhere in my pile) when P-47s first came to the UK and they had (at least at one airfield) had to ban mock dogfights below 10,000ft because they lost too many planes and pilots.

And this was an issue, a manoeuvre at 30,000ft that lost you 10,000ft was one thing, doing it at 10,000ft was another. Mushing was a real issue where at pull out you could lose thousands of feet, hence the use of relatively shallow dive angles and bombing techniques.

It was the mainstay of the escorts until they had enough Mustangs to replace them, where they went to the 9th (and 19th etc) for CAS. So they had the plane and production lines, replaced it for escort duty as fast as they could and put them into CAS, it was (and the P-38) all the US had, so they used it. The Mustangs by that time were largely used for escort and air superiority work with opportunistic strafing and bombing thrown in.

And all credit to the pilots, they did the best they could with it and did well despite its low level shortcomings.

The P-38 was another thing, large, expensive, buggy, complex to maintain, complex to fly with a crippling mach limit. Even with the later models it was not thing to be in at 30,000ft over Germany. 20,000ft and less in the Med or the Pacific it gave a good account of itself, but a front line fighter against top opposition it was not.

As I said before, you have 2 main viable design options for your CAS plane. Small fast and manoeuvrable that relies on surprise for the attack and survival or a flying tank that can take a hammering. Anything else is a 'compromise' design, like the P-47, Typhoon, Fw-190, P-38 or Corsair. I'd argue that the 190 and Corsair (later models) were fundamentally better 'compromises' than the others, and of that five the P-47 and P-38 were the poorest of the bunch (the P-38's mach problems were so acute that even low down there were fairly severe dive restrictions on it).

At least the Tiffie had speed low down and could pull better dive angles with a better pull out (ie higher G). It had a very good cruising speed, which was much more important than max speed for that sort of work.... But it did require a lot more maintenance than any of the others except maybe the P-38 (big black mark, hence the main reason why I put it in 3rd place).

So it was lucky the Jug was so tough (I don't dispute that at all) but it had to be because it's poorer low level performance (which meant operational restrictions) meant that it was more likely to take a hammering than than some of the other faster and/or more manoeuvrable aircraft that were used in that role.

 

[RpR]

So it was lucky the Jug was so tough (I don't dispute that at all) but it had to be because it's poorer low level performance (which meant operational restrictions) meant that it was more likely to take a hammering than than some of the other faster and/or more manoeuvrable aircraft that were used in that role.

I do not know just where or why you say what you say as noting pasted here, or that I have ever read anywhere supports your opinion.

The only restriction for the Jug was a slightly higher altitude because as you said it could mush, so the pilots were told if they were not doing a two-tier sweep to maintain x altitude before diving.
In a two-tier sweep there would be the low altitude aircraft who attacked the ground and a second tier to watch for opposing fighters who often did not attack the fighters but made sure the boys down low were able to avoid them because if they went after the fighters, the ground attack aircraft would lose their eyes in the skies.

 

[drgondog]

I don't care HOW you cut it, the P-47 was not ineffective, delivering twenty times the bomb tonnage of the P-51's. Not sure where Drgondog is coming from about the P-51 being more effective at ground attack ... it wasn't. The P-51 DID shoot up a lot of planes on the ground, but they all flew the missions they were assigned.

Greg - my comments about the statistical comparisons between P-47, P-38 and P-51 'effectiveness' were solely related to the 8th 8F Mission and focused on the statistics available for strafing airfields. I invited the reader to review the data absent pre-conceived notions regarding 'toughness'. Specifically I stated that "I am Not saying the P-51 is better at CAS'

P-47's flew 423,185 missions and dropped 113,963 short tons of bombs.

P-51's flew 213,873 missions and dropped 5,668 short tons of bombs.

Totally irrelevant to my discussion. The post that started the discussion was the one in which the Poster declared that the F4U and AD were Far less susceptible to ground fire - and I responded by saying that I doubt that any reliable set of data and analysis would yield a large disparity between those aircraft re: flak losses in Korea

Combat Air Support isn't shooting up enemy planes on the ground on airfields, it is support of the ground troops during ground combat with enemy ground forces. And when the P-47 strafed, it did so with eight 50's, not 6 like the P-51. The P-47 flew 34.6% of all USAAF fighter missions. The P-51 flew 17.5% of them. Which one sounds like it did more for the ground troops? Of course taking into account that the vast majority of the P-51 missions were bomber escort or fighter sweeps and not CAS.

Greg if you want to take me to task regarding my comments, a.) read carefully what I said, and b.) respond to the facts that I presented. To refresh your memory tell me why the much vaunted toughness of the P-47 failed to yield lower loss per aircraft destroyed in the airfield strafing mission. (please don't wander into tonnage of bombs dropped - my comments had Zero to do with CAS in the ETO or Korea).

I would like to see the Rand Report, Eagledad, to understand their source data as well as the assumptions pool used to extract the statistics - Note: I do not disagree with them, they somewhat support my point earlier regarding not much difference between F4U vs P-51 susceptibility to flak (29% to 26%), but the devil in making the comparisons is aligning the threat environment to the observable and provable facts. Thanks for posting the study.

 

[drgondog]

Off?
Probably.
How far off?
I do not know.

I listed that site partly because the gent speaking gave his opinion of what he liked best and why, BUT actually more for the information on the P-47 M which is not often spoken of.

I listed that site partly because the gent speaking gave his opinion of what he liked best and why, BUT actually more for the information on the P-47 M which is not often spoken of.[/QUOTE]

My challenge to you was to source evidence that the difference in vulnerability to flak between P-51 and the AD/F4U was materially higher and you came back with the PR piece on the P-47.

Addressing the PR claim of "11,000+" combined air/ground victory credits is fairly simple. The USAF Study 85 examination yields either 3661 or 3634 depending on the gap of P-51 conversion timing for various 8th AF fighter groups but lets take 3661 as the higher number for this discussion.

Of the 3661, the ETO represents 2658 of the 3661, all other theatres comprising the 1003 remaining air victory credits.

In the ETO, only the 8th AF documented ground victory credits and also the only one that parsed and reviewed them as an assigned command authority. The P-47 destroyed 740 air to ground according to the 8th AF VCB. The 9th AF P-47 victory credits may be found in squadron histories - or may not - but in any case there was no review and command authority at 9th (or 12th or 15th or 5th or 13th etc) which bothered with ' official tallies of P-47 air to ground victory credits' for the entire AAF during WWII.

So, the official reviewed and accounted for P-47 victory credits is 3661 air and 740 ground for a total 4401..

That leaves us (the P-47 Public Relations essay you linked us to) to 'find 6600 air to ground victories'. Where do we need to look for that quantity? 9th AF where the 36th, 358th, 362n, 365th, 366th, 404th, 405th, etc Fighter Groups were primarily tasked for CAS at or slightly behind MLR against armor and rail and bridge targets? 12th AF? with similar mission tasking?

PTO? CBI? both with exceptionally low P-47 distribution among combat aircraft compared to ETO.

Back to the only question I posed regarding possible conclusions one could make for the common tasking regarding actual losses while strafing LW airfields compared to the number of reviewed and credited aircraft destroyed on the ground.

The Mustang destroyed 50% more per German aircraft lost to airfield flak, per aircraft lost and 200% more than the P-38.

Why?

 

[RpR]

The Mustang destroyed 50% more per German aircraft lost to airfield flak, per aircraft lost and 200% more than the P-38.

Why?

Do not know.

You are trying to reduce what happened in WWII to number crunching which is good for statisticians but compared to people who flew the missions says little.
If you are trying to reduce effectiveness of ground attack to aircraft destroyed on the ground, then you are leaving out every other item destroyed or killed by ground attack.
Your parameters are too narrow to give an accurate look at anything; if one goes by GregP's numbers that says far more about which is best than item p destroyed on the ground.

 

[drgondog]

Do not know.

You are trying to reduce what happened in WWII to number crunching which is good for statisticians but compared to people who flew the missions says little.

Trot out your Facts - rather than Your Opinions based on what other people say or feel. Do your own research. Your comment that I am 'trying to reduce what happened in WWII" is just silly and candidly, offensive.

If you are trying to reduce effectiveness of ground attack to aircraft destroyed on the ground, then you are leaving out every other item destroyed or killed by ground attack.

What I did is throw out a set of facts that make you question your simplified conclusions regarding the vaunted 'much greater toughness of the P-47' and asked you to comment of the facts presented to you -

Your parameters are too narrow to give an accurate look at anything;

Parameters: isolate strafing attacks on airfields in France and Germany by US fighters in the ETO; Collect the data by type of aircraft in context of primary objective targets destroyed: Parse the data and aggregate in categories of "Numbers of aircraft destroyed versus US aircraft losses due to the attack" OK lets for the moment stipulate that the parameters are insufficient to draw conclusions? Your turn - pick the parameters you wish to parse to derive relative 'survivability' or toughness' and trot them out.

if one goes by GregP's numbers that says far more about which is best than item p destroyed on the ground.

Could you re-read what you just wrote and explain what you are trying to say relative to the questions I posed for you?

And then go back to the post in which you stated that the F4U and AD were far more survivable in Korea than the Mustang - and I asked you to trot out the data? You keep trying to get away from the root of the discussion and you offer nothing but platitudes.

 

[eagledad]

Parsifal and Drgondog,

Like you, I also would like to see the actual study. Please note, that I did mention that the data presented was found in another book, "USAF Jabos in the MTO and ETO" by William Wolf. I have been to the Rand website and found that for 30 dollars, one can have a copy of the report. At this time, I need my cash for other items.

As for the Mustang v F4U vulnerability in CAS in Korea, I would like to point out that Vought came up with a specialized version of the Corsair for CAS, the AU-1. If the F4U was so invulnerable to AA in the CAS role, why would Vought go to the trouble of an extra design, which included, if I remember correctly, additional armored plates?

Just some thoughts.

Eagledad

 

[parsifal]

I happen to think at this stage that a radial offers some theoretical advantages to flak damage, but Im not convinced that it makes a meanigful or substantial difference in the real world. But those figures are what they are, and they suggest differently.

Maybe thats as good as its going to get

 

[RpR]

Could you re-read what you just wrote and explain what you are trying to say relative to the questions I posed for you?

And then go back to the post in which you stated that the F4U and AD were far more survivable in Korea than the Mustang - and I asked you to trot out the data? You keep trying to get away from the root of the discussion and you offer nothing but platitudes.

You are trying to reduce what works and does not work in a war where thorough records do not exist down to book-keepers numbers, that does not work.

R. Strange McNamara did that in 'Nam and got a lot of good people killed and crippled because HE HAD the NUMBERS.

I will take the words of a vet of what did and does not work anytime over a chart that ignores all the oddities of war that do not fit into a numbers chart.
A leading officer in Korea would have preferred to have the P-47 based on his experience in WWII; that along with pilots who have said the Jug could take more punishment than the Mustang is all the proof I need.

As you are a number cruncher, do you have a number that shows the number of Mustangs that did not return for unknown reasons compared to a Thunderbolts that did not return for unknown reasons?
This is a serious question.

Parsifal:
The simple fact that an Air-Cooled Radial cannot lose its entire cooling capacity all at once as any liquid cooled engine can, makes world of difference at base value.
That difference may be the difference between getting an aircraft that is junk back to base with its pilot, or the pilot having to bail-out and be captured.
Loss of a pilot is far more important than the loss of an airplane.

 

[drgondog]

There is distinct odor of gaseous emissions in the air.. it is easier to exit stage right than joust with it.. anybody have a match?

 

[GrauGeist]

 

[OldSkeptic]

The problem with getting the numbers together is that you really have to go down to the sortie level. No point in comparing plane X vs Plane Y when they had totally different mixes of tasks.
This can be done with incredible amount of research if the data is there, but it is not easy. Took decades of work by some very dedicated people to get the BoB sorted out.

My analysis of the V1 campaign showed me how difficult it was and that was really simple in comparison. I had good data (collected and sorted out by many others) that I used, but even so it took weeks to pull it all together ... and I do this sort of stuff for a living.

So when you compare kill/loss ratios, or ground kills you have to really do a lot of work and some of it will be impossible.

One issue is over claiming, that applies to air-ground stuff just as much as it did air-air (almost certainly far worse). So only by matching records can you make some sense out of it. But that is largely impossible for air-ground stuff since the records are non existent or incomplete (by comparison the air-air ones for all sides are models of accuracy.. and they are full of holes).

So you might say that 'claims' showed (eg) a Corsair killed so many of X on the ground. But (a) was it true (b) where and when did it happen?

OR studies, some taken right after things like the Battle or Mortein, showed that air-ground over claiming was rife for very understandable reasons (not saying the pilots, except for possibly a very few were lying). But with fractions of second to see what happened when your bombs/rockets hit and all the smoke and dust came up... I mean accuracy of claims was never going to be good. Two (3/4/etc) pilots all hitting the same thing and all claiming them. 4 tanks claimed when there was only 1 (maybe). Plus it was far harder to sort out what was damaged on the ground compared to air-air stuff (and that was hard enough). That truck/tank/etc you hit and claimed as destroyed was actually patched up and running again in an hour....

So unless someone is prepared to spend the decades on it to filter and match claims vs hard German data on losses, then I treat all these 'statistics' as very 'iffy' (and I work as a statistician).

Plus there is the system of rewards and humans, being humans, tend to put themselves in the best light. The US system of allowing ground kills to count towards awards was probably not a good idea, since people will (again with no malice intended) will interpret things to put them in the best light. I have long suspected that the massive numbers of reports of US 'blue on blue' attacks was caused, in part, because of their award system. Some pilots becoming a bit desperate and pushing the envelope and not being careful enough.

The Luftwaffe had similar issues caused by its reward system. Since so much has been done on the BoB we do know that some of the 'throat chasers' in that battle did tell a few fibs. As the Luftwaffe skills went down and they faced planes (like the Mustang and Typhoon) that were more similar in profile to their own ones I wonder how many 'blue on blue' ones they had. Galland tells of how he was nearly shot down by being mistaken for a Hurricane....

I met a guy in Germany (worked for my father) who flew 109s, shot down 3 times ... twice by his own side.....

The RAF, by no means was exempt from this, but were probably a bit better in air-air stuff, perhaps the 'Battle of Barking Creek' very early on helped them to reduce it in the future. In the daylight with Spits their profile was so unique that is was harder to make a mistake. But I am sure that the RAF had many of its own 'blue on blue' as well, maybe a bit less than the US or the Luftwaffe because the ethos and training was different. But I have never read of in the Normandy and post-Normandy period where the US claimed that the RAF shot down any US planes .. while that period was rife with reports of RAF planes being shot down by the USAAF.

So any 'claims' data I treat with huge amounts of skepticism unless they have been verified by careful cross research. And a lot of them I simply do not believe.

Give an example, on the air-ground stuff the Typhoons and P-47s get all the credit, but I have read various things that showed the Spits (IX/XVI) in the 2nd TAF actually killed a lot more MET.. right .. maybe ..maybe not. But you cannot ignore the role of the Spits and Allison Mustangs in the British air ground stuff.

 

[drgondog]

The problem with getting the numbers together is that you really have to go down to the sortie level. No point in comparing plane X vs Plane Y when they had totally different mixes of tasks.
This can be done with incredible amount of research if the data is there, but it is not easy. Took decades of work by some very dedicated people to get the BoB sorted out.

My analysis of the V1 campaign showed me how difficult it was and that was really simple in comparison. I had good data (collected and sorted out by many others) that I used, but even so it took weeks to pull it all together ... and I do this sort of stuff for a living.

So when you compare kill/loss ratios, or ground kills you have to really do a lot of work and some of it will be impossible.

One issue is over claiming, that applies to air-ground stuff just as much as it did air-air (almost certainly far worse). So only by matching records can you make some sense out of it. But that is largely impossible for air-ground stuff since the records are non existent or incomplete (by comparison the air-air ones for all sides are models of accuracy.. and they are full of holes).

Certainly a factor which makes any degree of 'award' precision impossible for individual claims. The works by many individuals (Dr. Prien is an example) examining and research individual squadron records which still exist as well as 'the collective tied together by internet' has added greatly to this body of research folks for the aggregate claims.

Give an example, on the air-ground stuff the Typhoons and P-47s get all the credit, but I have read various things that showed the Spits (IX/XVI) in the 2nd TAF actually killed a lot more MET.. right .. maybe ..maybe not. But you cannot ignore the role of the Spits and Allison Mustangs in the British air ground stuff.

The only comment that I disagree with (mildly) is the 'impossible' nature of validating ground destruction. The 8th AF Intelligence Officers collaborated with the pilots in methodical ways to sort chaos into tangible 'belief's'. The pilots, as a briefed process began to assign one of the high cover pilots to turn on gun camera film and take pictures of the field they strafed - for those missions that had planned high cover. Burning aircraft were the only targets permitted as 'destroyed'

Overclaims existed as a matter of course for many specific airfield scores but more on an individual basis rather than aggregate for large 'bags' primarily.

The intelligence officers broke the daily claims into locations of the claimed aircraft to assist in sorting out the individual gun camera film and tried to match it with the film of the airfield. The primary purpose was to attempt to sort out 'shared' claims into a.) assignment of different pilots shooting as same aircraft which at one time in the process would burn or blow up, and by virtue of the 'mapping' process, be assigned to an X' on the master location drawing by the IO's. That document and the gun camera film would travel, along with the local pilot preliminary award in the Mission Summary Report, to 8th AF Victory Credit Board for review and final disposition.

The mere fact that causing a fire was the minimum requirement for a locally approved Destroyed claim was very serious reason for belief of that aircraft was not likely return to service resulted in high probability of an observed aircraft to truly be destroyed or >60% damaged which was LW criteria for scrapping it..

It wasn't perfect but there were, ONTH, many aircraft that were damaged very badly but did not burn for lack of fuel. Me 262s fit in that category most of all.

Specifically if the target didn't burn or blow up it would get only a Damaged claim so 8th AF rules for air to ground victory credits were sound if not perfect.

I was distracted after my last post while looking for a match to clear the air, but the answer to the question 'Do I have the data for the number of Mustangs and Thunderbolts which did not return for 'unknown' reasons?' is yes as I examined every one of the MACR's at NARA and HRC which later went to Form3'. As a 'number cruncher' I carefully sorted out the 'unknowns' into 'last seen strafing' with the LW KUJ reports (yes I read and recorded what they stated when available (mostly complete) to note what the reporting LW report said about time and location. I have a spreadsheet which carries the 'unknown' category into multiple probability categories for 'likely cause'

 

[Jenisch]

BTW, the IL-2 also had an inline engine. That plane was famous for holding battle damage. Was it's radiator much different than that of the P-51?