From Now On, No Defeats, IV, 2: Leaning Towards France

I've got a great deal rattling around in my mind this week, and an unusually long writing period before the beginning of my graveyard shift tonight. So why not a bonus post, to finish off my discussion of the fighters of August?

This is not a fighter of August.

What was that again? A static view, thanks to Wikipedia:

9,240lbs dry weight, 2480hp from a Bristol Centaurus XVIIC 18cyl , 53.6L engine; almost twice  and rather more than twice the hp of the Spitfire I.

The Sea Fury is a pretty visible plane. It's on the air show and race circuit, in which people with money and a death wish fly 70 year old planes at excessive speeds far too close to the ground. If there's one thing that the Fury has to spare, it is excessive speed: the Wikipedia data box gives 460mph at 18,000ft.

I suppose that we have to thank the crazy people, even if we choose not to stand under them. Otherwise, we might not reflect at all on just what was done in the name of victory, 70 years ago, more-or-less today.



The reason I start with the Fury is that it isn't entirely unrelated to the August air war. Design of the Fury began in 1942, as Sidney Camm sat down to think about possible improvements of the Tempest, which would first fly in action in the spring of 1944, five months before the first Fury test flight. The Tempest was itself an improvement of the Typhoon, which had its first combat trials in 1941. Nor should I leave the impression that Camm was done with the Tempest. Development work on that plane continued to the end of the war. The Tempest VI was a near-contemporary to the Fury and a worthy rival. You will not, however, find any Tempest VIs flying today. That's sort of the difference between an 18 cylinder radial, with its heads poking out in all directions in a mechanic-friendly way, and a Napier Sabre, a water-cooled 24 cylinder double-inline 6.

This is the same engine that powered the Typhoon that was in service in 1942, contriving to drag 8,840lbs through the air at a maximum of 410mph in spite of having 2000hp to work with in the Sabre. In spite of plenty of warning,* Camm opted for a thick wing on the Typhoon in order to make room for all the stores for which the Air Ministry was asking. It was, ironically enough, the same mistake that Kurt Tank was making with the FW190 at the same time, and not entirely unrelated to the more-fixable mistakes being made at Republic as they developed their P-47, with its notorious early problems with compressibility blanking controls. (Slow-loading PDF.) Partially as a result, in August 1942, the Typhoon was mainly known as the plane that mysteriously disappeared a lot due to the back half falling off (or, later, just the tail). Camm "fixed" that the way we all learned to do with our Lego sets so long ago, by clamping on stiffeners on the outside, until aerodynamicists working in the difficult realm of transsonic physics figured out the elevator flutter problems.

So, instead, the plane of 1942 was the Spitfire IX, basically a lash-up of the Spitfire V fuselage with the new Merlin 61 two-speed, two-stage supercharger. This was, in some sense, inevitable. The Spitfire V was being beaten up over France by FW190s and over Darwin by whatever the Japanese were flying out of Timor. (Were they really Zeroes, or is this a case of the Japanese Army Air Force being overshadowed by the navy again?)

The deadly sense that enemy pilots had better materiel, the essence of a "Fokker Panic," made something like this inevitable. It probably didn't hurt that this allowed the "automotive industry style mass-production" plant at Castle Bromwich to continue making the same old fuselage. All the same, Vickers was able to switch to the Spitfire VIII instead, a better plane. So the RAF was still getting second-best to buck up the numbers.

Which isn't to dump on the Spitfire IX too much. After all, it was a very successful plane. By the  numbers. The problem was all this pouring-of-old-wine. Even the first generation of Merlin monoplanes had a bit of a problem with too much engine dragging too little plane around on the ground, and the intercooler between the two supercharger stages made the Merlin 61 nine inches longer than earlier engines. Nine inches is a lot of lever arm when you're building on the already excessive distance between a torquing engine and the rudder that is supposed to keep the plane from swinging about the runway. A key mission for aeroplane designers, often missed in practice,** is to make sure that our pilots are killed by the enemy in glorious battle, not by trees in landing and takeoff snafus.

So the Spitfire IX was a bit dangerous to fly. It's not like the lads at Kingston-upon-Thames could exactly point the finger! Okay, but back up for a second. Last time, I noticed that 32 year-old former apprentice carpenter Sidney Camm became chief designer at Hawker Kingston-upon-Thames in 1925. After cutting his teeth on the wooden, Rolls-Royce Condor-powered Hawker Horsley,  Camm  made his name with the metal-structured, Rolls-Royce Kestrel powered Hawker Hart.

No, wait, back up again, to 1918, or even 1911, to one of the first numbers of Flight ever published, and an advertisement for Isaacson's line of 5, 7, and 9 cylinder air-cooled radial engines giving 80--120hp for all of your amateur sporting aeroplane-building needs. This is how Great Britain entered the first war in the air: with  an air force flying French radial and even French water-cooled inline engines, because the British industry was basically all vapour ware. When people started to get upset about it, very upset indeed, attention went to airframes, not engines. The result was fortunate for Sopwith, as he placed design after design with the RFC and the RNAS, powered right up to the end of the war by rotaries, something of a dead-end in aeroengine development, although the concept keeps coming back on motorcycles.

Meanwhile, British engine makers (and, for some reason, shipbuilder and would-be integrated defence manufacturer Beardsmore) got their own versions of inline, water-cooled engines into service, making the name of Rolls-Royce, keeping the extraordinary engineers*** of Napier in the news for another generation. Rolls-Royce, Napier, Sunbeam and Beardsmore all delivered consistent power from smooth-running engines but the installations were too heavy for fighters, and the general thought was that water-cooled, like rotary, was wrong for planes in the long run. William Weir, the great technocrat, instead dropped vast amounts of money into projected radial engines like the ABC Dragonfly, and young Roy Fedden's brainstorm at Cosmos Engineering.

The result, as of 11 November 1918, was so much scrap metal. A lot of scrap metal. So much scrap metall, in fact, that many people were deterred from even entering the aeromotor market in the postwar period, because they expected to be flooded out by salvage engines made out of surplus spare parts assembled into less ambitious engines by the officially-sponsored salvage-and-disposal firm AEC, run by Frederick Handley-Page perhaps with predatory intent, if you believe Harald Penrose.

How did this happen? It happened because development takes time: it takes "failing forward." This is the lesson that policy planners persistently forget. If something is needed now, the time to produce it is now. That the time to start producing it was five years ago is not something that can be fixed without a time machine, but that isn't a problem if you just ignore reality. Well, the reality is that it took until 1928 for a metal-framed Sopwith-turned-Hawker plane to fly with a post-WWI engine. (That it happened to be a pressurised water-cooled engine makes this story a little more complicated, but I'll leave that as perhaps a commentary on the limits of the radial configuration, or of the people developing them.) Over a thousand Harts were made in a range of variants, and production ultimately spread throughout the industry, as the Air Ministry distributed contracts to keep good design firms going. Hawker made so much money off the project that was able to buy many of those design houses, keeping them going but subsuming their corporate identity under the Hawker-Siddeley name. (It is interesting given Kingston's surefootedness with planes that the  engine-design firm in its stable went into terminal senescence until saved by the transfer of the Metropolitan-Vickers jet turbine work.)

This is the Sidney Camm who came into the Typhoon contract only 10 years later. Again, I have to stress Camm's origins as a carpenter. The Hart could be a metal (structure) plane because Hawker had come up with its patent construction method, which used standardised lengths of market-supplied high grade steel pieces. Girders could be sawn to length, attached to each other with standard joins, and assembled, Mechano-like, into whatever structure the designer called for. It was a very direct replacement of wood with metal, and, in the Hart-Fury-Hind-Hurricane family, a very successful one. Camm was on a role.

Until the Typhoon, that is, but this isn't actually that surprising. The guts of the Typhoon showed some "Hawker method" ancestry, but only a little. Metal is not wood. The Hawker method had the advantage that it took standard lengths and cut them to size in a way that didn't affect their metallurgical properties. Practically any other way of handling metal, from grinding to pressing to extruding, does. Once you start producing hot metal, you need furnaces for heat treatment, at a huge increase in the cost of just running an aircraft factory.

Late in the war, there was a burst of mixed propaganda and interested journalistic coverage that allows me to do something towards underlining just how significant these changes were. The Air Ministry's Power Behind the Wings pamphlet tells me that in 1919, Bristol's engine design shop had a staff of 50 backing up Roy Fedden as he tried to turn the radial engine he'd dreamed up at Cosmos into the Jupiter. In 1945, with Fedden gone in an unfriendly industrial coup, the numbers were up to nearly 1000. More accurately, the Air Ministry tells me that Rolls-Royce was up to 1000, with another 1200 manning experimental machinery in the shops, making and testing new parts: failing forward.

This is the design side. At Kingston, the Sea Fury contract was being met in an integrated manufacturing facility that is fully documented in a special number of Aircraft Production that I would cite here in full if it weren't too dark to see right now, and if anyone cared, since, frankly, the details tend to bounce off. Somewhere there might be a skilled publicist who could extract illustrative anecdotes from the writeup. All I am going to say is that an amazing number of specialised machines are being used in an amazing number of steps to produce this aircraft. This is not only the kind of detail and complexity that was impossible in 1918. It's the very kind of work that has changed.

And there is research. Metallurgists tell us how a given piece of metal changes as it is being handled. Calculators, whose job title we never even managed to settle on before they were replaced by another kind of machine, are spinning circular slide rules and running IBM card-sorting machines specially rigged up to do detailed calculations, just  to tell us about air flow over the elevators. Behind that level of work is the effort of the people who design the cardsorter jury rigs and the "cleveland borers" and all of the other precision machine tools.

This is social transformation. The kind of industry that makes the Sea Fury just didn't exist in 1918. It was coming into existence, driven by the exigencies of failing forward, in 1942. It has its antecedents in other industries, particularly heavy powerplant manufacture, but it was not avaiable, could not be deployed in 1918. That was why Weir and the Air Ministry had to turn to a skeleton of skilled woodworkers whose most utilitarian previous application was boatmaking, but who were probably most employed making nice furniture. And this is how, by the steady and relentless drip of government money, in no great quantity in the interwar years, but in massive gobs during the war, that this tiny seed was turned into a massive design-and-production-complex that I can't even begin to talk about without using words like "precision" and "calculation," pointing inescapably forward through the changes that the war was wreaking on the electronics industry towards the computer age.

Failing forward: apart from having a plane that can shoot down a FW190 right now, it's not clear that anyone in 1942 has a clear picture of where they're going. Certainly not a picture of where they're going in 70 years or so. And yet they get there. It's sort of a miracle, and one that leads me to wave my hands at an "economy of knowledge," and imagine that some kind of efficient market place of skills is ensuring that there are operatives at the machines by arranging things so that they will learn what they'll need to make money in the postwar economy. It's in this respect, not combat effectiveness, that I'm retrospectively demanding new machines in novel configurations, rather than reheated Spitfire Vs in 1942. It may produce (relative) failures like the Typhoon, but it will lead to bigger payoffs in the end.

Well, there it is. Coherent as I have time to make it, I'm afraid.


*My all-too hasty workup of my notes for this all-too hasty post turns up the formerly ubiquitous, now all but forgotten C. N. H. Lock warning all-too clearly, on the basis of airscrew work, that thick sections lead to air flow divergence.
**Take a look at this series in Wikipedia: LaGG-1, LaGG-3, LaGG-5. These are the three major wartime upgrades of the Lavochkin-Gorbunov-Gudkoff plywood (very advanced plywood composites, not to take anything away from Soviet engineers) V-12 fighter. Check out the data boxs at the bottom of the articles. The LaGG-1 was good for 377mph at top speed, while the LaGG-7, of course, was much improved, at 405mph. So what about the LaGG-5? 355mph (at 5000 meters, of you were wondering). What happened? Well, the Finns captured 3 LaGG-5s, which wasn't unusual, and turned them over for evaluation. What was uncommon is that through all the filters of language and short-handedness at wartime editorial staffs, a published version of their evaluation managed to make into print, and then into translation and republication at Aircraft Engineering. A very slow-loading site has more information.
***If this is the sort of stuff you like, and  you haven't heard of the Deltic, check this out.

Now, I'm not saying that the Soviet engineers perhaps fibbed a little about the performance of the LaGG variants that weren't field tested by foreign, Axis engineers. Or, perhaps, it was the Axis guys who fibbed. What I am going to say is that if you check out the article, you discover this interesting tidbit: the landing speed of this aircraft with Vmax of 560km/h is 140km/h. This is a speed envelop well under 4-1, a death machine by the none-too-high standards of 1942, especially on rough landing grounds. But no-one is going to go out and check your accident records. They care about Vmax.