Chapter CXLIV: A Cooled Head in a Crisis Part I
Lord of Slower-than-real-time
- Dec 13, 2005
Chapter CXLIV: A Cooled Head in a Crisis Part I.
The Cooling Crisis is a convenient shorthand to describe the throng of concerns and long standing issues that all came to a head inside the Air Ministry in the Summer of 1937, before overflowing into the wider government. It must also be admitted that while the challenges of cooling high powered aero-engines were the notional cause of the crisis they were not the issues at the heart of the matter, if they were then the entire affair would probably never have become a crisis; Like any government department worth its salt the Air Ministry preferred to keep problems 'in house' and would doubtless have kept the matter from rising to the attention of the cabinet if that had been at all possible. That they failed to do so is due to the far reaching tendrils of the affair, dragging in the Foreign and Dominion Offices and touching on the ever contentious issues around industrial policy. The roots of the technical and political problems went back many years, but the specific trigger was the problems with the Arctic version of the Handley Page Hampden bomber. The standard Hampden in RAF service used Bristol Pegasus engines and was serving well, there was little wrong with the basic airframe, or at least little wrong that wasn't well known; it's nickname of 'The Flying Suitcase' was not exactly a sign of affection from the aircrew who had to squeeze into it. The Arctic version, which had been sold to Sweden and then selected by the Royal Canadian Air Force as an interim until the Wellington became available, came with Napier Dagger engines and was leading a far more troubled life. As one would guess from the chapter title the Dagger engine had severe problems with cooling, managing the impressive feat of over-heating on the ground then being over-cooled and seizing up while flying at altitude. While the full extent of the problem had been kept hidden from the Swedes and Canadians they had become aware there was an issue and were growing impatient with the lack of information. By the summer of 1937 the Foreign and Dominion Offices became involved as the queries had reached the diplomatic level, the Foreign Office was quick to warn of the strategic and diplomatic consequences should the Swedes drop the deal and instead turn to German suppliers, while the Dominion Office fretted about the Canadians turning to US engine suppliers or worse just buying an entirely American design, finally the Board of Trade were quick to complain about the prospect of losing either deal and the negative consequences for future sales to anybody else. None of this political pressure actually helped solve the problem, but it did ensure that the cabinet was taking a close interest in the Air Ministry's efforts, which was unfortunate for them given what was to follow.
The Napier H-24 Dagger VIII engine, as used in the Arctic Hampden, viewed from the supercharger end, the distinctive 'tall and narrow' profile of the H-block layout is apparent. Designed by Frank Halford the key feature of the Dagger was not it's air-cooling or the H-block layout but it's small power stroke and very fast spinning camshaft. Where other designers had sought extra engine power by making cylinders larger or just adding more of them, Halford had instead sought to increase revs, the Dagger running at 4,200 rpm for take-off where the Pegasus barely hit 2,500 rpm. The result was that that the 17L displacement Dagger produced the same power as the near 29L Pegasus, resulting in a more compact engine with considerably better fuel efficiency. The Achilles Heel of the Dagger was cooling, the earlier lower powered marks had just about managed reliable performance but managing the extra heat from the 1,000hp Mk.VIII would prove to be a far more difficult problem. As Napier specified ever larger ducts and intakes to try and stop the engine over-heating on the ground, they only made the over-cooling problem at altitude worse.
Technically speaking there was a very easy solution, take out the Dagger and put in the Pegasus engine from the RAF's version of the Hampden, just in case this proved necessary a number of engines from the Bristol production line were earmarked for the purpose. While the loss of range would be noticeable it was felt it wouldn't be enough for either customer to cancel the deal and the Pegasus itself was still being developed, so there was a small hope it would 'catch up' the difference by the time it came to deliver the final aircraft. It would however be an embarrassing to have to admit that the Dagger did not work as promised and failing to live up to the sales pitch would make future foreign sales trickier, hence the preferred solution was for Napier and Handley Page to properly fix the original design. This was where the real problems started, because the Air Ministry technical section was starting to have grave doubts that Napier had the capacity or even ability to do so. The high revving, H-block design concept was not a new idea, the Dagger was a development of the earlier and smaller Rapier engine and this was not a promising starting point. The Rapier had seen limited use, partly due to concerns over it's novelty but mostly because it proved to have terrible problems with cooling. To make matters worse for Napier they had inadvertently damaged their position with their planned follow up to the Dagger, provisionally called the Sabre, which they proposed should retain the same H-block layout and hi-revving concept but would be water cooled. Many in the Air Ministry saw that choice as an admission that either the Dagger's air-cooled approach was fundamentally flawed in some way or that Napier did not know what they were doing with air cooling. Neither option appealed, nor did the possibility that both answers were correct.
The Fairey Seafox prototype undergoing trials, note the 'tall and narrow' profile of the nose which indicates the presence of a H-block engine, in this case the Napier Rapier. Developed as a spotter-reconnaissance floatplane for the Royal Navy's light cruisers the Seafox was one of the last Fleet Air Arm aircraft to result from an Air Ministry specification and selection. As the Seafox was the only production aircraft to use the Rapier it fell to the FAA to deal with the problems of that engine, not least it's chronic cooling problems and the lack of power compared to what had been promised. While the Fairey airframe proved robust and manoeuvrable the Fifth Sea Lord began to suspect the FAA had been sold a pup with the engine, which to a large extent they had been. Strictly speaking there was no need for the Seafox at all, the FAA could just have procured more of their standard float planes, but the Air Ministry had wished to 'spread the work around' to Fairey, and especially to Napier, so had pushed the design through. The Admiralty vigorously indicating their unwillingness to receive sub-standard machinery just to help the industrial planning requirements of the Air Ministry would be one of the spicier strands of the Crisis.
While officialdom losing faith in their ability was a serious problem for Napier, who were almost completely dependent on military contracts for their survival, it was also a serious embarrassment for the Air Ministry, because Napier was an 'approved firm' and a member of the Ring. The Ring was the Air Ministry's scheme for maintaining capacity in the aircraft industry, after a shake-out in the immediate post-Great War period the industry had stabilised around a few key firms that the Ministry wished to see stay in the business of military aircraft and was prepared to support with orders and contracts to ensure that they did. The Ring can be seen as similar to the various Admiralty schemes to mothball and retain capacity for armour plate and large calibre gun construction, like their naval counterparts the Air Ministry sought to retain both design and production capacity and to keep research and development progressing even if it was not justifiable on purely commercial grounds. It is interesting to note at this point that the Army had no such scheme for tank or armoured vehicle manufacturers, it was the private sector efforts of Vickers and a few other key firms that kept that industry alive, a striking example of the defence priorities of successive governments.
The Air Ministry Ring of Approved Engine Manufacturers
- Armstrong Siddeley
- De Havilland
- Rolls Royce
(1) There was a separate Ring for airframe manufactures
(2) Some entities (Bristol, De Havilland and the Armstrong group) appeared on both lists, but significantly some did not.
Napier had earned their place in the Ring on the strength of the superlative Lion engine, a masterpiece of no-compromise design produced by one of the greatest engine designers of his generation, Arthur Rowledge. Boasting the bold choice of a W-12 configuration the Lion had incorporated countless advanced or innovative features and would find it's way into over 160 aircraft types, including dozens of RAF types, and would at various points power the Schneider Trophy winner and the World Land Speed and World Water Speed record holders. There were but two problems; firstly the Lion had been designed in 1917 so by the mid-1930s it was thoroughly obsolete and secondly Rowledge had long since left Napier to work for Rolls Royce, taking most of his design team with him. To be blunt it was dawning on the Air Ministry that Napier had designed a grand total of three engines since Rowledge had departed the firm in 1922 and none of them had actually worked properly. From a certain perspective this was not actually a problem, one of the justifications for the Ring was to provide the stability and support for engine manufacturers to develop new concepts without being afraid that failure would bankrupt them. As not every new idea would work out as hoped the Air Ministry would end up backing ideas that 'failed' and the manufacturer should not be punished for this. The counter-argument was that fifteen years support was surely enough time for Napier to have produced something valuable, even the incredibly complex sleeve valve project had only take Bristol five years to (mostly) get the hang of.
The racing power boat Miss Britain III being lowered into Poole Water prior to a race in 1933. At the heart of the aluminium clad hull sat a single Napier Lion VIID engine, the ultimate racing variant of the design and the last great hurrah of Sir Montague Napier before his death. Through supercharging and careful tuning the Napier engineers had coaxed 1,350hp out of the engine, a staggering figure when one considered the standard Lion in RAF service managed a shade under 500hp, though in fairness the engine could only manage such outputs briefly before needing an overhaul. While Miss Britain III would narrowly lose the Harmworth Trophy she would go on to win many races both in the UK and abroad, not least the prestigious Count Volpi Trophy in Venice, embarrassing Mussolini and the Regia Marina sponsored Italian entrants. These victories, and the many other triumphs for Lion powered craft on land, sea and air, had helped to mask the failure of Napier to successfully develop any sort of replacement engine. Alongside the troubled Dagger and Rapier they had also produced the mostly pointless Javelin, intended for the light aircraft market only a handful were sold as their intended customers chose the more powerful and more reliable Gypsy Six engine from De Havilland.
While this debate continued within the technical section of the Air Ministry the more urgent question of how to fix the Dagger was being considered. The Royal Aircraft Establishment at Farnborough were asked to look at it, but the Air Ministry decided to bring in some industry expertise as well. With Rolls Royce and Bristol considered too busy on other vital projects the obvious choice from the Ring was Armstrong Siddeley Motors (ASM), the firm had experience of air cooled engines and as part of the Hakwer-Siddeley grouping had a wise range of other expertise they could draw on. Unfortunately for the Air Ministry this approach did not help with the Dagger problem, instead it brought to the surface the problems that ASM were having with their new engine designs which, with a grim sense of inevitability, it soon emerged were related to cooling. As mentioned ASM were thought to be knowledgeable about this matter, they owed their place in the Ring not to a single great engine but a large family of good ones, the 'big cat' air cooled radials. So large and successful was that family of engines that the designers ran out of actual big cat names to use, for instance the 5 cylinder Mongoose engine gave good service in the RAF's training aircraft and shared many details and parts with it's larger brethren, but a Mongoose is at best a cat-like creature. Names were not the only thing that the 'big cat' engine family was running out of however, far more seriously the basic design was also reaching the limit of it's development by the early 1930s, in particular the lack of a central bearing between the two rows of the larger engine was proving a major bottleneck to further increases in either cylinder volume or RPM, the two main methods of increasing an engines power.
Recognising this the company had started developing it's next generation of engines, the dog family, which for our purposes began with the Hyena. The Hyena's 15 cylinders were arranged in three rows of five, three rows not being unusual enough the designers had then arranged the cylinders of each row inline behind the first row. Conventionally in a two row engine the two rows were staggered, to maximise the possible air flow and ensure the first row did not block air flow to the second row. This was impractical for a three row engine so ASM made a virtue of necessity and adopted the in line approach, vastly simplifying the valve and cam arrangements and relying on careful ducting and baffling to achieve the required cooling. While the Hyena itself encountered serious problems during testing the basic concept was promising enough to get an Air Ministry development contract for an improved version, the 21 cylinder Deerhound, which was projected to be capable of 1,500hp. Work had started in late 1935 and the Air Ministry had the engine tagged as an alternative/backup for the Bristol Hercules as part of their portfolio approach to engine development which required that any important engine programme should always have at least one backup. By mid 1937 the Deerhound prototypes were running and encountering problems with cooling, the airflow over the rear most row of cylinders was far too low and they were over-heating. It was at this point that the Air Ministry approached ASM asking for help with the Dagger issue, prompting the panic in certain parts of the Ministry to rise another few notches as it appeared that another member of the Ring was also unable to properly cool their engines. Proper investigation would eventually reveal a subtly different problem, namely a disconnect between the engineers at ASM (who, based on the problems encountered with Hyena, wanted the Deerhound to be water cooled) and the management (who wanted it to be air cooled, because they knew that was the configuration the Air Ministry would prefer if the Deerhound was to be backup for the equally air cooled Bristol Hercules). That the Air Ministry technical section had not flagged this up prior to the Deerhound being approved became another concern to add to the pile, but in the short term the Ministry still had the Dagger issue to solve so turned to the next member of the Ring, De Havilland.
The only application of the Armstrong Siddeley Hyena engine, the Armstrong Whitworth A.W.16 fighter, the distinctive 'air holes' in the nose line up with the cylinders behind and were intended to channel the cooling air to where it was needed. It must be noted that the Hyena was only fitted as an experiment to trial the engine, the standard hyena had conventional Panther engines. The A.W.16 failed to find favour with the RAF and instead found it's niche in China, the Hong Kong based Far Eastern Aviation Company selling 16 of them to the Kwangsi Air Force in the early 1930s, Kwangsi perhaps being more commonly known as the Guangxi Clique. Keen biologists may note that the Hyena, despite appearances and reputation, is technically a member of the cat family and is classified as Feliformia or cat-like. Some have suggested this is deliberate by ASM, the name chosen to represent the transition between cat and dog. While this is an elegant suggestion there is the slight flaw that ASM's first dog engine was actually the Mastiff, a heavily modified Tiger engine used to test various metallurgical and design concepts and so would have very much suited the name Hyena. A more relevant parallel is perhaps the one between ASM's understanding of taxonomy and their understanding of air-cooled engines, in both cases they had a decent understanding of the basics but were unaware of all the details and that occasionally led them into error.
De Havilland's place in the engine Ring was a slightly odd one at first glance, at this point their most powerful engine was the then brand new 425hp V-12 Gipsy King which developed enough power for the new Albatross Airliner and other commercial aircraft, but was deeply inadequate for any fighter or bomber with pretensions of being competitive. The explanation was that De Havilland's engines were widely used by the utility, communication and above all training aircraft of the RAF, roles were power was relatively unimportant but reliability, ease of maintenance and low cost were crucial. Naturally this meant De Havilland engines were all air cooled, that being the cheap and low maintenance option, so there was a certain logic in the Air Ministry turning to them. This was not the only reason however, it is apparent there was considerable concern about how competent the firms of the Ring actually were, with doubts circling around both Napier and Armstrong Siddeley, so bringing in De Havilland was also a chance to find out how wide spread the problem was. It is instructive to note that at no stage were doubts about Bristol or Rolls Royce expressed, Bristol's extensive track record of sales and licensing of air cooled engines being an eloquent enough defence of their skills, while the technical expertise of Rolls Royce was so taken for granted that no-one even thought to ask. As the De Havilland team got to work they essentially ignored the engine itself and focused on the installation, taking the Arctic Hampden trial aircraft and adjusting the exhaust and cooling outtakes on the wings. To the considerable surprise of the representatives from the Air Ministry these seemingly small tweaks worked, while the Dagger-engined Hampdens would never break records for reliability they would be good enough and that was all that was required. The explanation from De Havilland was quite simple, in their view cooling was not a problem of trying to force air into the front of the engine but about keeping the air pressure low at the exit, provided that was done correctly it would pull through the required volume of air. This implied that Napier, who had been focused on trying to control how much air went into the engine, had been looking at the wrong end of the problem for about a decade. Damning as that was on Napier, the Air Ministry realised that they had also failed to spot this issue despite it apparently being well known in the wider industry.
The Armstrong Whitworth A.W.38 Whitely shown in 'level' flight, it's distinctive nose-down angle being a feature of the design. The chief designer of the Whitley had been entirely unaware of the possibility of fitting flaps to the wings, so had instead just angled the entire wing down by 8º to achieve the same aerodynamic effect. While this did ensure the aircraft had a safe landing speed and could fit on the standard RAF runway, it did result in the aircraft flying with the same 8º down angle line of flight. While flaps were added to the design after initial testing the wings were never adjusted and remained angled on the production aircraft. This failure did not just result in a mildly ridiculous looking aircraft, but one that had considerably more drag than it should and so a lower top speed and shorter range. As the Air Ministry grappled with questions of how to share knowledge throughout the industry the example of the Whitely served as a reminder that similar issues afflicted the airframe design side of the industry.
While the nominal issue had been solved and the Arctic Hampden would go on to give entirely adequate service, it had left behind a string of problems that had managed to rise to the political level. While most of the cabinet were happy to move on once the panic had been solved, the Treasury and the Minister for Defence Co-ordination remained interested in aviation affairs. The Treasury were concerned that the Ring had been a waste of money, funding firms for not obvious benefit, and wanted to make sure there was some reform of the scheme to stop that happening again, Defence Co-ordination merely spotted a chance to expand their influence and leapt on it. As mentioned previously the Admiralty were no longer willing to accept 'bad' aircraft because the Air Ministry needed to keep the Ring in business, yet there clearly was a need to keep valuable defence industries supported during the lean times. Who better than an independent ministry, like the MoDC, to mediate the demands of the Air Ministry and Fleet Air Arm and oversee a co-ordinated policy? While the Air Ministry civil servants started a strong counter-attack to prevent this bureaucratic land grab, finding temporary allies in the Fleet Air Arm who had no desire to lose the control they had only just gained, the Air Board grappled with the actual problems thrown up by the Crisis. As they saw it there were three issues that needed dealing with;
- Which firms should be in the Ring
- How to properly disseminate knowledge and research around the industry
- How to stop the Admiralty disrupting the Ring and duplicating their efforts
A few thousand words on engine design and British aero-industry industrial policy. I considered just ploughing on through to what the Air Ministry will do about the problem, but I decided this had got long enough so you have a Part 2 to look forward to. You may, or may not, have noticed a lot more pictures in this one, it is a slight tweak to the format so any comment on that would be appreciated, even if it is just to say you hadn't noticed anything different.
The actual engine problems are all OTL, but the crisis was not. The Dagger engined Hampden (or Hereford as it was called in OTL) did not work very well, but as it was only ever in RAF service, and only ordered in case of a shortfall in Pegasus engine production (which never occurred), they just cancelled the order and converted them across to standard Hampdens. In Butterfly I think I have explained why that is not an option and so when various non-technical people start looking around they do notice a great many engine companies seemingly struggling with engine cooling and over-react, as the Air Ministry often did. They were a bit highly strung and prone to panics, which probably made it a tense place to work but overall they were probably one of the better inter-war ministries, certainly compared to say the French or German air departments. The Ring was of course OTL and there are a few questionable aircraft and indeed engine decisions they made that only make sense once you know they were trying to keep certain firms in business, the Seafox being just one example of their use of the Fleet Air Arm as a handy dumping ground for such aircraft. This is no longer an option and that will cause problems, as we shall see.
Napier are an odd bunch, a bit more on them in Part 2 as they fight to stay in the Ring, but they appear to have not really recovered from Rowledge leaving the firm. Because Rowledge really was that good, at Rolls Royce he would do Kestrel, Buzzard, the 'R' racing engine and a great deal of work on the Merlin, we are just around the time when you start to need teams to do engine work so it was not all him, but as started he did take his team with him when he left Napier so it was a hell of a blow. They had enough left to continue to tweak the Lion engine and it was a record breaking monster, but they struggled with a replacement. I quite like the dock photo with Miss Britain III in it and power boat racing / world water speed records seemed to excite the public imagination a lot more in the 30s than now.
So we come to Armstrong Siddeley Motors and our discussion on taxonomy. Hyenas are indeed part of the cat family, or cat-like feliformia should one wish to be precise. Morphologically (that is bodily structures and appearance) wise they do definitely look like dogs, but phylogenetically (that is the actual DNA and evolutionary relationship) they are much closer to cats, they also have a great many behaviours around parenting, marking territory and so on which are more feline-esque. As noted it would therefore be absolutely ideal for ASM to have used that name for their experimental transitionary engine, the one in between the 'big cat' and the 'dog' engines and had traits of both, but alas the opportunity was missed. One the actual engines, ASM did good but not amazing designs, the larger ones mostly ended up in Armstrong Whitworth and Avro designs (because all three companies were part of the same group), but they also provided a great many small low power engines for trainers, etc. The problems with Deerhound are OTL and best I can tell the engineers knew they had a problem, but management was telling the Air Ministry what the civil servants wanted to hear to get the contract. I suspect it was a corporate infighting issue as well, right around the time Deerhound got the development contract the bigger Hawker merger/takeover was happening, so a nice big contract win would help convince Hawker management to keep the engine division going. No proof of that, but it seems plausible to me. In any event it has now gone wrong and the Air Ministry are staring at them asking awkward questions (because the Air Ministry is also being asked awkward questions in Whitehall and wants to share the pain around).
De Havilland's fix for the Dagger is actually OTL, because in OTL the Dagger also ended up in the Hawker Hector biplane and didn't work their either. Here Churchill killed the biplanes, so the Hector never happened so wasn't mentioned. When De Havilland got asked to look at the Dagger they focused on the out-takes, pulling air out not trying to force it in, certainly that approach makes sense to me - if the way out is constricted then the airflow is always going to be limited. That also allows for the intakes to be made a bit smaller and so also help the over cooling at height issue (over cooling is when the engine oil drops below minimum temperature and becomes too thick to flow properly, so the engine is no longer properly lubricated). I don't think it's enough to make the Dagger a wonder engine, but it will make it good enough that the deals can go through, after all by definition most engines will be average.
We've mentioned the Whitley before and frankly it is a bit ridiculous that no-one thought to fix the wing at some point. That said it was ordered in a mad hurry, one of the first 'off the drawing board' orders the RAF made, they had committed to buying 80 around a year before the first prototype had even flown. I imagine the RAF looked at the time required to change the wing, re-run the flight trials, change the jigs in the factory, and realised that if they wanted a 'modern' heavy bomber in service quickly it was this or nothing. However the general point stands, someone should have tapped the designer on the shoulder and told him that flaps in wings were a thing he could do. What the Air Ministry needs to do is work out how to do that, because no-one on their side had noticed it either when they were looking at the drawings.