Forecasting wartime spare parts levels is sort of like forecasting your requirements for rally driving unknown distances based on your regular work commute travel on good roads. The easiest way to up production numbers was to reduce the number of components set aside for spares.
The general rule was everyone was caught short the first times they encountered heavy combat and some made deliberate choices, the US had more warning time but had some trouble believing the numbers required (So did Churchill). Not only spare parts but the personnel to do the repairs, including moving the damaged aircraft as required. Going into largely unknown environments was another factor, like the tropics and flying at greater heights. The US published the two volume official history Global Logistics and Strategy looking at things from higher levels, it had more distances to cover during internal movements but the USN versus IJN plans aside not a lot about major expeditionary forces, the islands the navies were going to fight over did not have a lot of room and no one was forecasting how good you could make a wartime construction system. The merchant ships needed were not being built. The US had a much bigger civil than military aviation industry, which had different objectives and requirements, like reliability and economy versus absolute performance.
If anyone was building long range aircraft as standard requirement it was the IJN.
The US was heavier idea needs the usual qualifications, the B-17/24 ended up around the same weights as the Halifax/Lancaster, the US mediums were built on the assumption they would be intercepted and so went for survivability, the heavyweight P-47 started out about 600 pounds heavier than the Typhoon. Some of the extra weight is due to when the aircraft was designed, the later the more powerful the engines available, amongst other things those engines needed more airframe strength to stay attached, some early Merlin Mustangs did not have the engine bearer strength of the Spitfires, add more fuel as well. The Hellcat had a 2,000 HP engine but on specifications did not have much performance over the Zero with half the power, the weight went into other features, including pilot safety and good carrier landing characteristics.
The British could put forward the Stirling as a much more rugged design than the US heavies based on empty weights, how much of the P-40 weight is Curtiss or the way it is an update of an older design instead of then current US practice? The Empire State Building is over strength compared with later designs.
Always remembering there were people on the other side putting holes in your equipment. The Luftwaffe thought it took around 20x20mm or 2 to 3x30mm. hits to bring down a heavy bomber if no fire was started.
Given the USAAF reports its aircraft losses as 1/3 in USA, 1/3 non combat/war weary and 1/3 combat it is clear accident rates had a big impact on the maintenance system and those rates went up as average crew experience went down. RAF, end August 1939 75 out of 475 delivered Hurricanes had been Struck Off Charge, 27 out of 305 Spitfires, 38 out of 210 Gladiator I. The site
Martin's Aviation Pages lists British non training unit all causes aircraft casualties September 1939 to March 1940 as 584 losses and 897 damaged, by end August 1940 that was up to 2,794 lost and 3,585 damaged plus 233 abandoned.
The Narrow Margin by Wood and Dempster have a cumulative battle casualties table fr the RAF at home presumably for the war, to 1 July 1940 82 bombers category 2, 358 lost, 56 fighters category 2, 517 lost, other types 16 category 2, 104 lost, as of 31 October 1940 it was bombers 198/725, fighters 766/1,657, others 66/200.
According to the Ministry of Aircraft Production aircraft awaiting spares as a percentage of all aircraft within the RAF, end of month June 1941 to February 1942 was 8.5%, 9.7%, 10.2%, 9.9%, 9.4%, 8.2%, 8.4%, 8.0%, 7.0%. End June 1941 RAF Home Commands had 932 aircraft awaiting spares of which 611 were trainers, end February 1942 1,090 aircraft, 602 trainers. October 1941 to March 1942 spares requested in the month, around to over a third there was no stock and a similar number had over 10 month's stocks. Drill down and Hurricane spares were more plentiful than Spitfire, engine spares had under 15% no stock, propeller generally over 40% no stock, airframes varied between 13 and 48%, usually closer to the lower figure. We know any functioning repair system at a given date is going to have under repair and not under repair for <reason>, the question being what is an acceptable level of not under repair.
March 1942, aircraft storages, front line combat types, 212 aircraft needing engines (all but 2 Merlins), 479 needing propellers, 346 needing constant speed units, 193 (all Blenheim) needing turret, total 1,240, though some needed more than 1 component. January 1945, of the 8,949 aircraft and 512 airframes in storage units 1,292 had equipment deficiencies. (Top 5 were 121 Oxford I, 76 Anson, 71 Beaufighter TF.X, 60 Stirling IV and 45 Martinet.)
When introduced into RAF service the Merlin had an engine failure rate of 1 per 500 hours of flying, as of the December 1941 to May 1942 period it was 1 per 1,000 hours, the Sabre was 1 per 340 hours.
Some of the 15th Air Force 1945 monthly reports have Engine Change Reports, graphs of average running time per aircraft engine changed. generally speaking the R-1830 radials in the B-24 and the R-1820 radials in the B-17 had at least an average of 300 hours running when changed, few had 400, with the R-1820 ahead, most having 350 or more hours. Overhauled engines usually had around half the average hours of new engines. The V-1650 Merlins and the V-1710 Allisons usually averaged over 200 hours, few made it to 300 until March/April. Overhauled Allisons were around half the new engine hours, overhauled Merlins more like a quarter. In civil service the radials lasted several times longer between overhauls.
You find WWII history in some interesting places. Journal of the American Statistical Association, V 41 no. 234, June 1946, pages 190 to 203. "Actuarial Analysis of the operating life of B-29 aircraft engines", by O L Altman and C G Goor. Lots of obvious and not so obvious points.
1) R-3350's were in critical supply in 1944 and the first half of 1945. This is despite usually the number of R-3350 built to date being twice the number needed for B-29 built to date. Supply problems were easing around the end of the war.
2) Air transportation was used to fly R-3350's to the US for overhaul, at best 2 R-3350's would fit in a cargo plane.
3) Overhauled engines had around 10% lower "life" before the next overhaul.
4) The -23 was the carburettor and the -57 the fuel injected versions. The -23 was modified to improve reliability.
5) The statisticians noted the standard USAAF methodology for forecasting engine life was only suitable for a reasonably static population. Not where there was a steady arrival of more strength with new aircraft.
In terms of engine life, operations in India were the worst, since each combat sortie required three reasonably rapid climbs, India to China, China to Japan, China to India. Next came training in the US, finally the best were the units in the Marianas.
Expected life prior to first overhaul, early operations from India, 163 hours, -23 engines. Using modified -23 engines this had risen to 280 hours by February/March 1945 for aircraft operating from India and 304 hours from the Marianas. The figures for B-29s used in training were 221 hours and 310 hours versus the 163 and 280 hours figures above.
Operating from India a comparison between the modified and unmodified -23 engines showed 80% of the unmodified and 95.3% of the modified engines survived to over 100 hours, 33.9 of the unmodified and 81.5% of the modified engines survived to over 200 hours, 0.2% of the unmodified and 47.3% of the modified engines survived to over 300 hours.
In the Marianas, as of 20 November 1944 the average hours on each -23 engine removed was 91, by 20 January 1945 it was 151, as of 30 April it was 234. These figures include removals for engine model changes, modifications, accidents and battle damage. They are also under estimates of the normal engine lifetime because so many of the engines were new. The figures include new and overhauled engines, so it is either the number of hours since the engine was built for new engines or since overhaul for the overhauled engines.
Engine hours before removal as of 31 May was 259 hours, and 31 July 272 hours. These figures are for engines removed because of mechanical problems only. Even in July the steady number of new B-29s arriving drove down the average engine hours per removed engine.
A study as of 31 July 1945 noted in the Marianas the -23 engines 96.8% of new and 92.5% of overhauled logged more than 100 hours before replacement, 87.5% and 75.7% respectively logged over 200 hours, 62.7% and 43.4% logged over 300 hours, 19% and 8% logged over 400 hours, none logged over 500 hours.
As noted above the training schools in the US went through R-3350 engines quicker than the combat units in the Marianas, for example 57.9% (new) and 36.4% (overhauled) logged over 300 hours, but once this mark was passed the engines in the US held up more, so 24.6% and 10.4% logged over 400 hours, and 1.2% and 0.2% managed over 500 hours.
The fuel injected -57 engine had a higher time between overhauls, so in the above study 31.2% used in training logged over 400 hours, and 4.9% logged over 500 hours.
The USAAF Statistical Digest has the 20th Air Force flying between 10 and over 14 hour missions per the monthly averages, 621,975 hours of which 479,113 were on combat missions, at 200 hours per engine that is around 12,500 engines.
The US statistics on spares production indicate says airframe plants might not be making any, which makes some sense if they were assembly plants, or some plants were more dedicated to spares, it clearly varied, an example, Consolidated, San Diego in 1943, 25,656,000 hours making B-24, 3,185,000 making B-24 spares, Fort Worth no spares made for knock downs assembled, for other B-24 18,949,000 hours for assembly, 907,000 for spares.
The Spitfire and Bf109 undercarriage attached to the fuselage.
