Reliability of WW2 fighters.

[AL Schlageter]

Originally Posted by Crumpp

Yes it is and do you understand what that means?

I don't think you do if your intent is to call into question the fact it is a representation of maintenance hours.

Do you?

[renrich]

I think the notion that air cooled engines are more reliable than liquid cooled is based somewhat that they are more resistant to battle damage. That is the reason the US Navy ruled out liquid engines prior to WW2.

[davparlr]

Originally Posted by Crumpp

You can see by examing this document that the R-2800 offered no advantage in number of maintenance hours over any other aircraft engine.

http://afhra.maxwell.af.mil/aafsd/aa...t114and115.pdf

I am not sure this tells us anything about the availability of aircraft at the squadron level but only tells us how long it takes to overhall the engine at depot. I am not too knowledgeable but I suspect that depot overhall manhours starts with the engine out of the aircraft and probably does not include the cooling system on the liquid cooled engines. As such, the hours are similar for work accomplished, 11.8 hr./cylinder for the V-1710 to 13.4 hr/cyl for the R-2800 (average, 1945).

The real importance to warplane availability is things like mean time between failure (MTBF), mean time between scheduled maintenance (MTBSM) (like engine overhall), mean time between unscheduled maintenance (MTBUM), mean time to repair (MTTR), etc. These are what planners must use to determine how many of 100 aircraft assigned will be available for combat at any given time. Improvements in these numbers at the weapon system (aircraft) level are force multipliers.

Since we don't have these numbers for the engine systems, predicting weapons system reliability is difficult. However, as a rule of thumb, these numbers are proportional to complexity. As such, when you calculate in the liquid cooling system reliability, it seems the air cooled engine should have fewer failures, thus more reliability. Also, it must be noted that the 18 cylinder R-2800 engine in itself is pretty complex (note that the overhall number for the R-2800 is higher than the smaller cylindered stablemates).

[Crumpp]

but only tells us how long it takes to overhall the engine

That is a very good indicator of maintenance guys. Why do you think Bill, another aircraft owner, thought it was such a good find?

Here is why it is a very good indicator of maintenance requirements:

There are several levels of "major overhaul".

Service limit - Parts are checked to ensure they fall within service limits.
This overhaul has little chance of reaching TBO as the standard is the minimum service limits. If a part is allowed a service limit of .010 and it is at .009 when checked the part continues life. It does not need to be replaced in a service limit overhaul. These are the least expensive overhaul because generally, there is less disassembly involved. Parts are replaced as needed.

New Limit Overhaul - Parts are checked to ensure they fall within new part tolerances. Engine time is zeroed and reset. This is the overhaul the US Military most commonly does as it has it's own mechanics to perform the work. The engine logbook continues life. Parts are replaced as needed

Re-manufactured limits - Performed at the factory by the factory being the major difference between a "New Limit" major overhaul and "re-manufactured Limits". Engine time is zeroed and reset. Factory issues new warranty and new logbooks. Parts are replaced as needed.

The common theme with all of the overhauls is, "Parts are replaced as needed.".

Therefore we can examine the document and see that in 1943, the R-2800 required considerably more time to overhaul than it did in 1945. Why? The engine became more reliable. The parts were made stronger or the points of stress reduced and the overhauls did not need as many parts replaced.

Partial disassembly and checking tolerances is one thing. Complete disassembly and reinstallation of a new part is quite another.

You can now see that the R-2800 developed into a reliable aircraft engine and that all aircraft engines are very similar in their maintenance.

All the best,

Crumpp

[Crumpp]

Understand too that crank tolerences on most aircooled engines can be checked by removing the barrels while leaving the corncob on the mounts.

Liquid cooled inlines require removal of the engine and disassembly of the case to check the crank. So pick your poison, both engines take about the same amount of time to overhaul. Do you want more barrels to overhaul or do you want to remove the engine?

All the best,

Crumpp

[drgondog]

Originally Posted by Crumpp

That is a very good indicator of maintenance guys. Why do you think Bill, another aircraft owner, thought it was such a good find?

Here is why it is a very good indicator of maintenance requirements:

There are several levels of "major overhaul".

Service limit - Parts are checked to ensure they fall within service limits.
This overhaul has little chance of reaching TBO as the standard is the minimum service limits. If a part is allowed a service limit of .010 and it is at .009 when checked the part continues life. It does not need to be replaced in a service limit overhaul. These are the least expensive overhaul because generally, there is less disassembly involved. Parts are replaced as needed.

New Limit Overhaul - Parts are checked to ensure they falls within new part tolerances. Engine time is zeroed and reset. This is the overhaul the US Military most commonly does as it has it's own mechanics to perform the work. The engine logbook continues life. Parts are replaced as needed

Re-manufactured limits - Performed at the factory by the factory being the major difference between a "New Limit" major overhaul and "re-manufactured Limits". Engine time is zeroed and reset. Factory issues new warranty and new logbooks. Parts are replaced as needed.

The common theme with all of the overhauls is, "Parts are replaced as needed.".

Therefore we can examine the document and see that in 1943, the R-2800 required considerably more time to overhaul than it did in 1945. Why? The engine became more reliable. The parts were made stronger or the points of stress reduced and the overhauls did not need as many parts replaced.

Partial disassembly and checking tolerances is one thing. Complete disassembly and reinstallation of a new part is quite another.

You can now see that the R-2800 developed into a reliable aircraft engine and that all aircraft engines are very similar in their maintenance.

All the best,

Crumpp

Gene - slightly off on a tangent but still in the strike zone. While I no longer fly or own my own ship 'the dreaded engine overhaul' to New Limit OR Re-manufactured Limit was by far the worst sound - then and now.

I have a close friend with both a Gulfstream IV and Challenger at Hayward who is having turbine blade replacements for a cool Million and a Half on the Challenger - approximately 20% of the total value of the entire ship.

Has anyone priced out a Re-manufacured Limit $$ for either the R2800 or 1650-7 lately?

This is why Mustangs get sold.. The previous owner got one 25 years ago, flew it very little but one day the dreaded day comes.

Of course the stuff that Crump and his guys do on rebuilding an Fw 190 ranging from inspection of spars (and repair/re-manufacture) of 60 year old combat airframes (and engine parts) is unbelieveably expensive and doesn't even fit in the discussion.

[Crumpp]

This is why Mustangs get sold..

You are so right, Bill. The cost of airplanes is staggering.

> The following is a shameless plea for donations!

We live off donations and are non-profit. One of the reasons I am out posting on these types of boards is to help keep up the interest in WWII aircraft and get the word out on the WWII Aircraft Foundation.

Buy some Christmas gifts from our shop, make a donation, or become a member. We need your support and only through that support can these airplanes continue to grace blue skies.

All of you are more than welcome to visit the Museum and the Restoration Shop.

WWII Fighter Aircraft Foundation

All the best,

Crumpp

[Crumpp]

Hi davparlr,

There is no mystery to solve. Failure predictions give us the maintenance time schedule. This schedule is posted in the appropriate pubs on the type. Keep in mind too, that as an aircraft, our margins are much smaller than we would find on any other form of transportation.

The real importance to warplane availability is things like mean time between failure (MTBF),

Time Between Overhauls is a function of time between failures. Most aircraft engines have a TBO of around 2000hrs. For example, both the R-2800 and the BMW801D2 both had 2000hrs TBO's.

mean time between scheduled maintenance (MTBSM) (like engine overhall),

All engines have a recommended regular maintained schedule. This schedule is dictated by probable failures of subcomponents. These schedules are very similar in scope and frequency by the fact physics limits our design choices.

The schedule is posted in the engines manuals.

mean time between unscheduled maintenance (MTBUM),

Impossible to quantify and has little bearing IMHO. Everyone works to reduce this portion. Given a reasonable passage of time, both operators and design teams will reduce this to the point that once again, all aircraft engines are similar.

All the best,

Crumpp

[davparlr]

Originally Posted by Crumpp

Hi davparlr,

There is no mystery to solve. Failure predictions give us the maintenance time schedule. This schedule is posted in the appropriate pubs on the type. Keep in mind too, that as an aircraft, our margins are much smaller than we would find on any other form of transportation.

I still do not think that rebuild time tells us anything about reliability of the engine. In a way it corresponds in that a more complex engine typically takes more time to rebuild and a also has a lower MTBF. However, an engine may be very difficult to rebuild but be built to such standards as to be very reliable. In fact, in designing equipment and installation on modern aircraft, this is a criteria. If the part is difficult to repair or is hard to get to, reliability must be built into the component. Still rebuild time is high but reliability is up.

Impossible to quantify and has little bearing IMHO.

Actually, operationally this is of major concern since all aircraft related aborts are due to unscheduled maintenance (if it could have been scheduled, it would have been). As you mentioned, operationally, this can indeed be quanitified over time and incorporated in aircraft availability planning and corrective action implemented.

On newly developed engines, it is a large impact, but, as time goes by, fixing these are part of the maturing process and equipment get more reliable. For comparing reliability of engines, only mature designs should considered and MTBUM should indeed have little bearing.

[Crumpp]

On newly developed engines, it is a large impact, but, as time goes by, fixing these are part of the maturing process and equipment get more reliable. For comparing reliability of engines, only mature designs should considered and MTBUM should indeed have little bearing.

Over a fixed and finite time period such as determining the reliability of WWII aircraft engines it is impossible to only consider mature designs.

Hence my statement that MTBUM has little bearing in the context of comparing WWII engines.

In a way it corresponds in that a more complex engine typically takes more time to rebuild and a also has a lower MTBF. However, an engine may be very difficult to rebuild but be built to such standards as to be very reliable.

An engine with a 1500hour TBO has a shorter MTBF than an engine with a 2000hour TBO.

A 2000hour TBO is a 2000hour TBO whether it is on a 4 cylinder engine or a 28 cylinder radial. FYI a Lycoming IO-360 4 Cylinder air-cooled engine, PW R-2800 series, and the BMW801 series all have a TBO of 2000hours. The MTBF of these engines is comparable and the TBO is a function of that MTBF.

Certainly the more complex engines designer worked to increase the MTBF and that workload was higher than the team that worked on a simpler engine.

The complexity of the engine is irrelevant if both engines last 2000hours and we are looking to compare reliability.

If we want to compare reliability as a function of complexity then your point is valid. The R-2800 required much more development time than the Lycoming IO-360.

If the engine we are comparing are reliable and shows little or consistent wear, then the Overhaul Man-Hours will remain consistent as well.

If the engine is not reliable, then we will see a large reduction in overhaul man-hours as that reliability is increased by the designers as well as end user’s input.

All the best,

Crumpp

[davparlr]

Originally Posted by Crumpp

Over a fixed and finite time period such as determining the reliability of WWII aircraft engines it is impossible to only consider mature designs.

Hence my statement that MTBUM has little bearing in the context of comparing WWII engines.

No problem here. This is what I agreed.

An engine with a 1500hour TBO has a shorter MTBF than an engine with a 2000hour TBO.

A 2000hour TBO is a 2000hour TBO whether it is on a 4 cylinder engine or a 28 cylinder radial. FYI a Lycoming IO-360 4 Cylinder air-cooled engine, PW R-2800 series, and the BMW801 series all have a TBO of 2000hours. The MTBF of these engines is comparable and the TBO is a function of that MTBF.

No problem here.

Certainly the more complex engines designer worked to increase the MTBF and that workload was higher than the team that worked on a simpler engine.

The complexity of the engine is irrelevant if both engines last 2000hours and we are looking to compare reliability.

If we want to compare reliability as a function of complexity then your point is valid. The R-2800 required much more development time than the Lycoming IO-360.

If the engine we are comparing are reliable and shows little or consistent wear, then the Overhaul Man-Hours will remain consistent as well.

All of this is fine.

If the engine is not reliable, then we will see a large reduction in overhaul man-hours as that reliability is increased by the designers as well as end user’s input.

I am not sure of what this says. But this is my main point.

Say that both the R-2800 and the IO-360 (hypothethically, since I do not know what the TBO of either is), has identical TBO, say 2000 hrs. This would imply that both have similar reliability. However, I am quite sure that the manhours to rebuild is much higher for the highly complex R-2800 than the manhours for the simpler IO-360. In other words, manhours to rebuild is a good reflection on complexity, but a poor reflection of reliability.

Even similar types of engines can have the same reliability but one can be more maintainability friendly and thus have less maintenance manhours to rebuild. I am sure you have seen designs in aircraft and cars that were not maintenance friendly, and, some that were.

[Crumpp]

(hypothethically, since I do not know what the TBO of either is),

The TBO for both engines is 2000hours, davparir.

However, I am quite sure that the manhours to rebuild is much higher for the highly complex R-2800 than the manhours for the simpler IO-360.

True. The much simpler IO-360 requires less man-hours to overhaul.

In other words, manhours to rebuild is a poor reflection of reliability.

I have edited your sentence. This is not a true statement. Manhours is a good reflection of reliability.

I have explained why already. The problem is in how you are viewing the data I believe and what you expect to see from the data.

You cannot look at the specifics but rather examine the trends. The data is excellent for trend development because it represents the extent of wear under normal usage of the engine as a function of it's MTBF. It's not tainted by combat or prototypical testing.

To use this data for example:

If both our engines are reliable, then the difference in man-hours will remain constant. If one engine is not reliable, then the trend will be our average man-hours decreases as our engine becomes more reliable during it’s developmental lifecycle. This is a function of both the end user’s and design teams working to overcome problems and improve the engine. The reliable power plant will remain generally fixed in average man-hours to overhaul. It’s already reliable within the constraints of physics.

Understand?

Now when you get to a certain level, the complexity of the engine is academic and it becomes a matter of just picking your poison. There is little to choose in the complexity of a V-1650 and R-2800.

We can examine the data and see that the R-2800 experienced a ~286% reduction in overhaul man-hours during the war and the V-1650 experienced a ~170% reduction in overhaul man-hours.

That tells me that the R-2800 started out as considerably less reliable a power plant than the V-1650 series.

We can also conclude by examining the overhaul man-hours that by the end of the war, the R-2800 had developed into at least an equally reliable power plant as the V-1650. It probably reached the extent of its reliability within the constraints of physics.

All the best,

Crumpp

[pbfoot]

I may be off track here but from what I've gleaned from various sources is that after 200hours of combat flying they rotated the Spits back to OTUs or moved them to less strenuous duties

[davparlr]

Originally Posted by Crumpp

You cannot look at the specifics but rather examine the trends. The data is excellent for trend development because it represents the extent of wear under normal usage of the engine as a function of it's MTBF. It's not tainted by combat or prototypical testing.


Understand?

Crumpp

Aaaah, I see!

Let me make an editorial change to your editorial change to my comment that I think correctly reflects your position,

"Manhours trends are a good reflection of reliability."

I agree with this! Good observation.

[drgondog]

Originally Posted by pbfoot

I may be off track here but from what I've gleaned from various sources is that after 200hours of combat flying they rotated the Spits back to OTUs or moved them to less strenuous duties

That does suprise me. The Mustangs in the 8th AF generally trended to 500 airframe hours before being declared War Weary and assigned to OTU or in Group Clobber Colleges. From memory the highest time P-51B I found for that category in the 355th was 640 hours and it flew steadily from March 1944 through October 1944 before declared WW.

The A/B/C/D/K Mustangs were designed to 8g limit and 12g ultimate which I believe was higher than the Spits. The P-51H was designed to 7 1/2 and 11.25 to reflect more to Brit standards to conserve weight.

I suspect the lower limit load envelope may have something to do with the lower airframe hours?