Chapter CLX: A Season for Decisions Part 2.
It is a useful approximation that the process of producing a new aircraft for the RAF ran as follows; operational requirement (what the aircraft had to do); specification (the 'hard numbers' the aircraft had to achieve in terms of size, speed, altitude and so on); competitive tender (the various manufacturers producing designs and mockups); prototype (construction of prototypes of the chosen design or designs); trials (the prototypes being tested to see if they met the specification and what properties they had) and production (order being issued to actually build the selected aircraft). If there were problems with the prototype(s) this could be lengthened by a Development stage where an improved prototype was produced and tested, while if things were urgent the entire prototype stage could be cut and the design 'ordered off the drawing board' on the basis of the tender design, the first production aircraft effectively serving as the prototype and trials aircraft. For our current purposes the two most relevant phases are the production of the requirement and specification documents and the selection for prototyping decision. The importance of the former should be obvious, the operational requirement and the specification were the documents from which everything else flowed, even a 'perfect' design could be an in service failure if it was trying to do the wrong thing. The end of tender stage decision point was also important as it was the point at which the Air Ministry had to pick the best design(s) to order prototypes of or in extreme cases issue a production order. The accusation is sometimes made that if the if the final aircraft was not a success then the Air Ministry had picked the wrong design at this stage, though this does raise the less frequently engaged with question of what best design actually means. Was a design that stuck closely to the specification requirements better or worse than one which excelled in one thing by sacrificing it's performance in other areas? The unsatisfactory answer was generally 'it depends', not just on the specific trade-offs made but on wider strategic thinking, the politician climate and the Air Staff's views on what they though should be possible. It must also be emphasised that even if all the tender submissions had broadly similar projected performance it was not a matter of just picking the one with the biggest numbers. The tender design would have to be turned from paper into metal and it was not unknown for some loss of performance to occur in this process, sometimes due to over-optimistic projections in an effort to win the tender but more often due to design team inexperience or unexpected problems with a new or unfamiliar technology; the Air Ministry often required designers to push the limits of what was possible and sometimes the limits pushed back. The final area of consideration was workloads and the Air Ministry's wider plans, if a company was already committed to other projects that were of greater importance then that could rule out their bid even if it was technically the best. It should be no surprise this was an especially contentious area, requiring as it did agreement on both the 'best' designs in the various specifications and on the relative importance of the various requirements. For completeness is should be acknowledged there were any number of unofficial and indeed officially discouraged factors at work, most notably individual views on the relative merits or demerits of the companies bidding. Given the large number of factors to consider, and to try and dilute the previously mentioned individual prejudices, the decision on which tender to proceed with was made at a formal design review conference. A mix of civilian and military technical specialists would be gathered from the Royal Aircraft Establishment (RAE), Air Staff and Air Ministry to consider the tenders from several perspectives, as was often the case with such gatherings the informal chats during the light refreshments were at least as important as the official agenda items. With an understanding of why and how the tenders were being assessed, let us turn to the actual designs.
We begin in a manner the Air Staff would approve of by looking at the bombers. After the shocks of the summer air defence exercises and the Lindemann Committee's first findings on offensive technology and tactics a reasonable first question could have been if any of the designs should be proceeded with as they had been specified to out-dated thinking. In reality this was never even discussed, on practical level scrapping the designs and starting with a new operational requirement would mean 18 months to 2 years worth of delay, the bulk of that time in the manufacturers producing brand new tender designs. With the comparative rarity of a defence friendly Chancellor in the Treasury the senior ranks in the Air Staff and indeed Air Ministry wished to take full advantage and get squadrons ordered; despite the war emergency orders and a great deal of effort to rush aircraft into service much of Bomber Command was still in biplanes. From the technical perspective it should also be said that many of the recommendations training, navigation, bomb sights etc had very little to do with the actual aircraft or could be easily incorporated by allowing some provision for extra equipment. The exception was bomb size, when area effect had been king the expected bomb load had been large numbers of 500lbs whereas the new approach was looking at fewer but larger bombs to maximise impact and blast. While this emphasis on size not quantity was arguably a change to the specifications, certainly a number of designers would complain they would have done things differently had they known, military aircraft were essentially a monopsony (the government was the only customer) so there was not a great deal the complainers could do. It is sometimes suggested that speed was another change, the importance of a fast cruise to minimise the reaction time available to the enemy air defences was certainly one lesson coming out of the Chain Home exercises. This misses the point that the Air Staff had long wanted faster bombers, indeed they had declared speed to be the most vital defence on a bomber since the early 1930s. The issue was they didn't think it was sufficient as they assumed that bombers would get attacked by enemy fighters, if not on the way to the target then certainly on the way back, at which point they would need defensive guns. In their typical false precision fashion the Air Staff decided on the 80/20 policy, deciding that 80% of the defence of a bomber was it's speed and 20% defensive guns, but that the 20% was vital. That there was almost no basis for these figures beyond assertion, not indeed any way to use them to quantitatively assess design trade-offs and choices, was irrelevant as it was mostly used as a justification to try and kill of the internal debate about unarmed bombers, a subject we shall return to later. For our present purposes the key issue was that the speed and defensive armament were in tension, more specifically speed and turrets were in tension because turrets were almost uniformly terrible for an aircraft's aerodynamics. It was well known that the single easiest way to improve the speed of almost any bomber was to remove it's turrets and smooth over the resulting hole, the resulting reduction in parasitic drag would substantially improve both maximum speed and best economic cruising speed. The Air Staff were well aware of this but as mentioned had no quantitative way of assessing the trade off, the impact on speed (and to an extent fuel consumption and range) of a turret could be worked out to a reasonable accuracy, but the benefits of a turret were entirely situational, depending on how an enemy chose to attack, if the friendly bomber was in formation and so on. It should also be noted that, fittingly for an organisation so concerned with destroying the enemy will to win, much effort was expended on thinking about the impact of turrets on bomber crew morale and if being able to 'fight back' against attacking fighters from all angles would improve their morale and so performance. The net result of all this was a certain amount of indecision, while it was agreed that a nose and tail turret were the absolute minimum necessary for any bomber beyond that there was a tendency for turrets to be added, deleted, moved and generally tinkered with during testing and indeed operational service. The full details can doubtless be found in those books which exhaustively focus on the history of a single type, the relevant point is that the trade off existed and was significant factor in both design and assessment.
A Bristol Blenheim Mk.I in flight during a training exercise, the large glass cage like object on the top just aft of the wings is the dorsal gun turret. While not the subject of any the deliberations the Blenheim was a good example of the performance cost of a turret; the original civilian Type 142 had achieved 307mph but by the time the military version had entered service as the Blenheim Mk.I the top speed was barely 280mph. While there were many adjustments necessary to militarise a plane, the turret alone cost around 15mph and that was in it's 'aerodynamic' retracted position. When fully deployed and actually able to fire the speed impact was even more significant. In exchange for this loss of speed the early Blenheim crews gained the dubious benefit of a single Vickers 'K' .303" machine gun, which perhaps explain why an unarmed bomber was not seen as a particularly radical option.
There were two bomber specifications about which a decision had to be made and a third which was playing the role of the spectre at the feast. Of the two main specifications B.12/36 was officially for a heavy bomber while P.13/36 was a general purpose medium bomber for worldwide use. The latter description was seen as something of an incantation, a ritual phrase the Air Staff used to show the Ministry and other interested groups that they were ticking boxes other than just strategic bombing. This was somewhat undermined by the Air Staff regularly referring to P.13/36 as a 'heavy medium' or 'medium heavy' bomber, indicating that they still viewed the strategic role as the key capability of the design, though in fairness they had included a requirement for the design to be able to carry and launch two torpedoes giving it some multi-role ability. The P.13/36 was specified as twin engined both as a nod to it's 'worldwide use' requirement (fewer engines meant less maintenance and so a small ground crew) and on cost grounds as engines were generally the most expensive single item on an aircraft. Cost was important because the design was intended to equip the majority of the bomber squadrons and so form the bulk of the future main bombing force. The specification had been written when the speed faction were in ascendance so only the minimum nose and tail turrets were specified, though the more experienced designers had wisely made provision for additional turrets to be added. The review process was relatively straightforward, while eight firms had responded to the tender most could rapidly be dismissed. As outlined above this was not always on technical grounds, as an example the Bristol submission was rated highly both on design and practicality, but the company was heavily committed elsewhere and so was deemed unlikely to be able to deliver without sacrifice elsewhere. As a telling example of the position of the Air Ministry and the wider government at the time one of those projects was the civilian Bristol Blackpool airliner; beating back the commercial challenge of US airline manufacturers was rated a comparable priority with rearmament and Bomber Command modernisation. At the other extreme there was the Hawker submission, as it was expected that a company would respond to an Air Ministry request for tender they had put a design in, however as their main focus was the fighter specification it had very much been a bare minimum token effort. In between there was the Fairey submission, a streamlined and well laid out bomber but which was submitted as using the Fairey P.24 Monarch engine, after all the engine rows of the summer the only choice that would have more annoyed the Air Ministry would have been an imported foreign engine, unsurprisingly this was rapidly ruled out. The preferred option was the Avro 679 and this was duly selected for prototyping, this is sometimes seen as a surprising choice given the firm's previous effort the Avro Anson which had been marked for replacement shortly after entering service due to being slow and too short ranged. In fact the Anson was a mark in Avro's favour, because while it's performance was poor it was what Avro had promised in their tender, that said performance was inadequate for the role was the fault of the Air Staff and the specification not the designer who had provided what was asked for. The only point of debate was the 'second string', the backup design that should be pursued in case of issues with the first choice. This was the point where the third specification made it's presence felt and began to influence the decision making process.
The project looming over the discussion was B.1/35, an older specification which most manufacturers had pulled out of as they (and much of the Air Staff) felt it had been superseded by later specification. Only Vickers had persevered and were working on a design that was essentially a larger brother to the Wellington, sharing the same geodesic construction and twin engined layout while being bigger and heavier. The common heritage with the Wellington had been enough to ensure considerable Dominion interest in a 'large Wellington' and this, along, with a reluctance to abandon the work that had been put in so far had been enough to keep the project going. The significant chance had been the switch from Bristol Hercules engines to Rolls Royce Vultures, the increase in engine power being necessary to keep the projected speed acceptable in the face of the ever growing weight of the design. While this switch further increased it's attractiveness to the RCAF and RAAF contingents (swapping 'exotic' Bristol sleeve valves for an line Rolls Royce unit made domestic production, or at least local engineering support, far more feasible) it also further increased the similarities with the favoured P.13/36 designs which also used Vultures. There was therefore a strong argument that B.1/35, which had become politically uncancellable due to the Dominion interest, could serve as the 'second string' design, freeing up money and design capacity for other projects. If nothing else the Air Ministry would need a good argument as to why it needed three very similar sized prototypes, all with the same engines and all intended for the same mission. The argument that emerged was a risk based one, in line with the usual portfolio approach, the need for a non-Vulture engined option to cover the possibility of problems or delays with that engine. At this point there were no major concerns about Vulture, some teething issues had been identified but after the re-organisation of the summer more resources were being deployed and the Air Ministry's Resident Technical Officers in Rolls Royce were confident in the basic design layout, however it was not unknown for issues to occur on seemingly solid engines as the Cooling Crisis had demonstrated. As this was the future mainstay of Bomber Command some insurance was considered wise just in case, so attention turned to the available options. The second place design had been the Handley Page HP.56 and that submission included an alternate engine, sadly it was the Alvis-Napier Sabre which was at an even earlier development stage than Vulture not to mention the ongoing disruption caused by the Alvis-Napier merger, making it an even less certain option. The logical solution was that if two engines were insufficient, then four could be provided, a change which 'only' required new longer and thicker wings and a great deal of internal re-arrangement around controls, fuel lines and similar. The initial preference was for four Rolls Royce Merlins, on the basis of an Air Staff study on expected bomber performance and the fact the Merlin was a proven established engine at this point. Handley Page however managed to get that switched to Bristol Hercules, on the basis of their much stronger experience with fitting and cowling air cooled radials and because the extra power would be needed to offset the increased drag from the larger aircraft. An interesting point would be that the torpedo and catapult requirements were removed at the same time so Handley Page could concentrate on the design of what was in many respects a new aircraft. That a specification that called for a twin engined medium bomber with torpedo capability would result in a design for a four engined effectively heavy bomber getting approved for prototyping did attract a degree of comment, however the reasoning was sound and at least at this stage the Air Staff did have a large amount of latitude to exercise their professional judgement. Once there were actually flying prototypes to look at there would be more pointed questions from the politicians to answer.
The Supermarine drawing office in their newly refurbished Woolston works in Southampton. The 1930s were a transition time for aircraft design, twenty years earlier the teams were often less than ten people and much of the design process was not codified but relied upon the experience and engineering judgement of the chief designer while twenty years later and teams were hundreds or even thousands strong, the first commercially useful computers were available and the firmer theoretical understanding of aerodynamics and stress meant design was far less of a black art. The late 30s had design teams in the dozen, had some analytical methods to assess stress and from wind tunnel testing a solid but far from complete understanding of aerodynamics. As a result much of the chief designers contribution in this period was marshalling and combining the work of the team rather than detailing every part; a vital job but not an irreplaceable one, or so the Air Ministry hoped.
A naive observer might expect that as the P.13/36 specification had resulted in a four engined bomber being ordered for prototype this might impact the assessment of the B.12/36 specification, which was also for a four engined heavy bomber. This sort of thinking cut no ice with the Air Staff who had the reasonable defence that the two specifications were for substantially different aircraft, not least the fact that B.12/36 required double the bomb load. The resulting aircraft were intended to serve in a small number of squadrons who's wartime role would be to hit targets either at long range or that required a very large bomb load be dropped; the specified overloaded launch capacity was 14,000lb of bombs, a shade over 6 tons. A more meaningful comparison is perhaps the fact that a fully fuelled and loaded Bristol Blenheim light bomber only weighed 14,500lb, however measured the required bomb carrying capacity was truly immense. The designs also all had a ventral or dorsal turret in addition to two at the nose and tail, being intended for distant targets the Air Staff expected the bombers to be engaged more often and so need more turrets for defence. All of this came at a cost of course, the target cruising speed was almost 50mph slower than it's twin engined counterparts due to the extra weight and the increased drag from the larger size and the extra turrets. The discussions on this specification were more involved and it took several rounds of conference before a preferred design was agreed upon. At the technical level the Supermarine design was favoured, it was exceptionally "clean" aerodynamically and had much of the bomb load stored in the wings, allowing for a smaller fuselage and so less drag, all of which combined to make it exceptionally fast for such a large aircraft. The issue therefore was not the design but the company, not only were Supermarine were heavily committed to the Spitfire and another project but just as seriously their chief designer R J Mitchell had recently died, casting doubt on their ability to convert the paper design into a working aircraft with the stated performance. However the promise of the design, along with an understanding that Vickers would support Supermarine during the project and that all of Supermarine's non-Spitfire work be cancelled, was enough to see the design taken forward to prototyping. The Merlin engined Type 318 variant was tentatively selected as the Air Staff wanted to spread their bets on engine choice, though Supermarine were warned they would likely be held to their claim that engine type could be 'easily' changed. The other project was the Stingray, the latest upgrade to the venerable Seal/Sea Gull/Walrus line of amphibious biplanes that served as everything from gunnery spotters for cruisers to search and rescue aircraft. Naturally the Royal Navy objected and the Committee for Imperial Defence also made their views clear as the type had been slated for use by Australian and New Zealand on their cruisers, so the project was not cancelled but transferred to a different firm. A positive spin would see this as evidence of more joined up thinking between the services, though the Air Staff instead viewed it as confirming their worst fears about non-RAF types interfering with 'their industry' instead of just being grateful for the aircraft they were provided with.
With a degree of uncertainty looming over the preferred design the choice of second string backup attracted considerable attention. As with the medium bomber many of the tendered designs were quickly dismissed and need not detain us, however one failed contender is worth briefly mentioning as the reasoning is enlightening about the views of the Air Staff and Ministry at the time. The Armstrong Whitworth "Super Whitely" was rejected very early on and the company told to focus their efforts on the getting the Sabretooth Tiger engines into the standard Whitley, with the added instruction to finally sort out the wing incidence issue with the design while they were re-stressing the design for the 'new' engine loads. Interestingly as with Bristol their civil aviation projects were also cited as a reason not to give them the work, as we shall see when we look at the state of civil aviation in a later chapter they were heavily involved in another high profile airliner. The actual choice for second string came down to Shorts vs Boulton Paul and technically there was no clear winner, both designs needed some more refinement and both firms had made unfortunate engine choices; since the specification had been written less than a year earlier the Napier Dagger and Rolls Royce Kestrel had changed from acceptable options for a frontline aircraft to being preferred for export, trainer use or various non-aircraft purposes. As both would need re-designing for new engines the choice came down to industrial and workload factors, how busy the firms were and their capacity to produce the prototype and then the final design. Shorts were initially favoured, they had experience in large four engined aircraft and through their Short & Harland joint venture had a pool of available design staff they could share the workload with. Yet it was Boulton Paul who were selected, machinations elsewhere meant that Shorts were required to focus on other projects while Boulton Paul would find themselves short on work and in need of a project to keep their workforce busy. That said the Boulton Paul design was not a bad design picked solely due to workshare concerns, despite their lack of large monoplane experience the design was assessed as far easier to manufacture than the Shorts effort and the rear turret was particularly admired for it's low drag layout. The point is that had decisions in other specifications gone the other way Boulton Paul who have been focused on other projects while the Shorts design would be selected because it was the lower risk option from a company experienced in four engine monoplane design and construction; the stated reasoning of the tender design conferences should always be read with some caution, while wider industry and strategic concerns often drove their decisions they were reluctant to ever admit to this, if nothing else telling a company they had only got the job for workshare reasons was seen as a bad for morale at the firm in question and an invitation for yet more unwelcome questioning from those in government who disliked the entire Ring approach.
A Bristol Bombay bomber transport prototype on a trials flight, the under fuselage ordnance racks for the bomber part of it's role clearly visible. The need for a replacement for the biplane Vickers Vincents and Valentinas had risen up the priority list after the high profile role of No.216 squadron during the early stages of the Abyssinian War and the dramatic trans-continental 'Smuts Flight'. While in very heavy used by the RAF commands in the Middle East and India the bomber transport type had not been highly valued by the Air Staff who would have preferred more bomber. But with the Air Ministry, Committee of Imperial Defence and various 'air minded' MPs keen on the type and a higher priority given to the Aerial Policing mission, the Bombay was pushed forward for production. With Bristol heavily committed elsewhere the job was given to the Shorts & Harland factory in Belfast, they were also tasked with a number of modifications to the aircraft using the design resources Shorts had intended to use for the heavy bomber prototype production.
Before we leave the heavy and medium bombers behind us there is the matter of the B.17/37 specification which while never issued still managed to cause a great deal of heated debate inside the Air Staff. With the Lindemann Committee still working away on it's first recommendations report and the 'Ideal Bomber' project to convert those into operational requirements not even started, in theory the Air Staff should have been waiting before issuing any new specifications. In practice there was a great deal of jostling to get various pet projects started to create some 'facts in the air' or even just designs to bolster the relevant arguments. At one extreme came the faction that was basically happy with the current approach save for the lack of defensive firepower, recognising more turrets probably wasn't an option they wanted more lethal turrets, specifically 20mm cannon armed turrets instead of 0.303" machine guns. Given the large weight of said cannons and their ammunition supply they could not just be 'dropped in' to replace standard turrets, instead new aircraft would have to be designed around them. That a practical 20mm turret did not exist at this point, so there was nothing to design around, was deemed a minor issue and something for the manufacturers to sort out in their tender designs. At the other extreme the unarmed bomber faction were pushing for a high speed, high altitude bomber that could not be intercepted and so would not need weapons. In between was to be found those parts of the Air Staff that had to interact and deal with politicians, their main concern was a bomber that had a Far Eastern scale operational range, or at least something close, to show the RAF was responding to political concerns and so deserved a lot more funding. The compromise was the B.17/37 specification which attempted to do all of those things and yet was surprisingly coherent, at least on paper. On the defensive side the starting point was the RAE's belief that a remote controlled turret was probably technically feasible or very close to it, this was relevant because if gun turrets didn't have to fit a gunner in they could be a lot more aerodynamic, especially if the guns were concentrated in a tail turret. On the range issue the specification took it's cue from the trend towards accuracy and traded away a great deal of bomb load for fuel as well as specifying 'overload' launch to maximise the launch weight. Finally the cruising altitude was set at over 35,000ft, which required the aircraft to be pressurised, the cruising speed was to be at least 300mph while the range was set at being able to haul a mere 4,000lb of bombs 4,500 miles, which would put Taiwan in plausible range of a Malay based bomber. While none of the factions were pleased it ticked enough boxes for them to at least accept it, there was just one issue; the design was hideously impractical. While the technologies required existed they had only been used individually on highly specialised experimental aircraft or in the laboratory, certainly they had never been used all together while in the -50ºC air temperature of the expected altitude. While the exercise had comprehensively proven that some compromise would be required in the next generation of bombers, and prompted a number of research programmes to be urgently started, the Air Council was forced to step in and put a stop to the arguments. They instructing all sides to wait until the 'Ideal Bomber' report on what was both desirable and possible was issued later the following year, this was done in the full and certain knowledge that the instruction would just delay the argument as there was little hope of the report being universally accepted without a fight. Having dealt with the larger bombers we must move onto their lighter brethren, an area which in addition to all the technical and industrial considerations of the heavies also had politics, strategy and diplomacy to deal with.
---
Notes:
Slightly longer than I had hoped, a different author would probably have cut off the entire first section but I left it in so everyone could understand how RAF aircraft procurement works because that is the sort of story this is.
The first three specifications are all entirely historic, however the selected aircraft are not for I hope solid enough reasons. If you know your British WW2 bombers then I hope you are nodding appreciatively and can enjoy speculating on the changes, if you have neglected your studies in that area I hope it was still an enjoyable read as I tried to steer clear of a confusing and distracting dump of numbers and figures. R J Mitchell's death is OTL, the Supermarine B.12/36 bomber was his last major design and the Air Ministry's concern about his absence is also historic. I would say this period is around the end of the 'Great Man' of design/engineering, before this point you could say one person designed that aircraft or engine, but around this point it becomes one person designed a key element and post-WW2 it was all mostly huge design teams. That said there is still space for my 'Great Idiot of History' theory as while one person cannot single handedly produce a war changing design, one idiot can still make really stupid decisions - see the He-119 Kurtie was discussing at the top of the page.
The 17/37 specification existed and was nothing like what was described, but it got abandoned for a similar reason - the RAF was unsure what a good bomber looked liked. The 'Ideal Bomber' study was an actual project of the Air Staff but the OTL version was aimed on producing a Standard Bomber, merging the Heavy and Medium categories and incorporating all the newest technologies, etc. In Butterfly it will be more concerned about producing a bomber that can survive in a world of RDF/radar and then be capable of finding and hitting precision targets, so all the listed often contradictory ideas about speed, altitude, armament will have to be balanced. Oh and it has to have a much longer range so the RAF can be taken seriously in chats about offensive war in the Far East.
I love the Bomber Transport idea not least because it seemed to actually work, it's another thing I think Past Pip should have covered in a bit more detail when talking about 216 Squadron and the early days of the Abyssinian War. But then if he'd done that we'd never have made any progress at all so perhaps it is for the best it was glossed over.
I'm hopeful the next one will be shorter as so much has been explained here and the Air Ministry used the same procurement process for everything, but I've thought that before so we shall have to see.
