Tuesday, July 24, 2012

From Now On, No Defeats, IV: The Fighters of August, I

It's time to start talking about the aeroplanes of August.

In the tradition of this post:

The comparison I'm making here is a bit strained. Unlike the Citro├źn four wheel drive, the Nash Rambler was not an import. It wasn't even designed in Britain or with British input, unlike the Nash-Healey or Metropolitan. It was just inspired by European approaches to carmaking. Which, given the difference between the driving experience on the two continents, helps explain why AMC isn't with us anymore. Sorry, George, but if you can't fit a Baby Boom's worth of kids into the car and drive them to Disneyland in something approaching comfort, you're just not going to win the market in the long run. High-speed, low displacement engines mounted in weight controlled chassis are for performance and fuel economy on roads with frequent speed changes, not comfort and reliability on long 55mph runs.

Which, naturally, brings me to Fighter Command's "Rhubarbs" and "Circusses," the lean towards France that began in 1941. This doesn't really belong in the province of "no more defeats," given their mediocre results. (See whining, here, 282ff.)  The problem was that the Luftwaffe was doing decisive stuff in Russia and the Middle East, and there was this huge air force in Britain that the Luftwaffe was declining to fight. So that air force started flying to France and looking for Germans to fight. 

It did not go well. The secret of air forces is that they are projections of industrial bases. They have to be, because they use themselves up. A few sedate training engines of the World War era delivered multi-thousand-hour runs, but the fighting engines lasted a much shorter time, and no-one could guarantee that they wouldn't stop working over mid-Channel, as opposed to being condemned on the bench.  The more an air force flies, even without an enemy gracing it with its presence, the more planes will be lost, and the more pilots will die. 

That being said, the GAF did come up and fight, occasionally, when it thought that it had reason. In February of 1942, the Air Defence of Great Britain got egg on its face when it failed to respond in time to a major German air mobilisation. Next month, seventy years ago this 19 August, Combined Operations Command will provide a pretext for an air battle, and Fighter Command will go into action again. The point will not be to make the Dieppe Raid more successful. The Dieppe Raid is just a raid. The fact that it happens is most of its rationale. It's meant to  hold German air assets in France. It's part of a global battle for air superiority. 

So let's talk about fighters and the battle for air superiority.

When you look at the triumphalist statistics of aircraft production, you might easily take air campaigns to be an industrial war of attrition.

That's the problem with the whole "numbers that are bigger than other numbers" school of writing about stuff. If you've ever read your way into the details of steel production, swimming way over your head in technicalities that only make sense to people who've actually done it, you will have encountered discussions of "pours" and "heads" and "ingots." In order to keep the conversation brief, and avoid displaying how much I don't know about the subject, I will propose that this language means that steel is poured out of the buckets that it is melted in into, uhm, ingot-thingies. The last bit of the pour is the "head," and, for various reasons, the steel in the "head" isn't worth proceeding with. So, at this stage, the steelmakers take the "head" and recycle it, putting it back in the bucket and melting it all over again. Obviously, this is quite wasteful, so the steelmakers aim to minimise the size of this "head" with various sorts of  technical-doing-thingie-stuff. 

Again, I'm trying to avoid being more technical, because, if I am, I'll just be hilariously wrong. But here's what I do understand: the worse your technique is, the larger will be this head. And so you will recycle more steel, and so, just based on the total number of ingots produced, your factory's statistical steel production will go up! This actually happened in World War II. It's one of the reasons why American steel production reached such awesome totals. And while the vast majority of that steel was actually produced and went to market and was turned into excellent munitions, some of it was turned into unfortunate boondoggles, like the 5%* of Liberty Ships that would have been condemned at the end of the war had they not gone into the National Strategic Reserve instead. (Good ass-covering, dudes!) 

Which is most certainly not to single out American producers of one particular product. As the heat of summer of 1942 seems to bake the ambition out of everyone and bring the flood of war news to a trickle, British aircraft production plants are producing Hawker Hurricane IIs and Supermarine Spitfire Vs, while German plants are producing Messerschmitt Bf109Fs, and the pilots are going, like, "Gee, thanks. I guess."

Because these are fighters, designed specifically to wage the battle for air superiority. The pilots need to know that they're flying the best planes available. If their weapons are second-rate, they're going to get the idea that they're being sent to their deaths. This was what was called in World War I a "Fokker Panic." Lif is, of course, unfair, so this is a tribute to the ascendancy of the Albatros fighters from September 1916 to June, 1917.

The First World War is a good example of the importance of perceived technical advantage to pilots, underlined by the extremely rapid turnover of fighter designs at the front. One firm, T. O. M. Sopwith's operation at Brooklands, managed to deliver three generations of fighters in a little more than two years: the Pup; the famous Camel; and the unjustly obscure Snipe. The extreme rapidity of design turnover at Sopwith's is pretty typical, and indicative of the fact that no-one really had any idea what they were doing. Aircraft manufacturers and engine manufacturers were basically both "failing forward" in their search for viable aerodynamics and reliable, lightweight power. Then, on top of that, to combine the two streams into a useful aircraft that was then sent aloft in the hands of a pilot who was too ill-trained to understand just how badly trained he was?** The First World War was crazy. No-one knew what they were doing, and that meant they were prey to people who talked like they knew what they were doing. I'd say more about that, but my good buddy Brett Holman already has a tag on his blog for Noel Pemberton-Billing, probably the paragon of the breed. 

The point of "failing forward" is that no-one understood how to make a good fighting aeroplane. The most that you could ask from each new essay from a given firm was that it would fail in a new way. The same might be said of a new aeroengine, except that the commitment of time and effort in developing a new aeroengine was far greater than the investment in a new plane. People sometimes elided over this distinction, but they usually had their reason.

For example, when Ernst Heinkel (and later the Messerschmitt firm) claimed that they had new fighter jets almost ready for service in 1943 (or 1941, or 1939), they sort of elided over the fact that the engines weren't ready, and couldn't possibly be ready before 1944. Given the amount and credibility of the literature that has taken on these claims over the year, beginning with the 1946 Journal of the Royal Aeronautical Society "Jet Turbine" symposium of 1945, one might well wonder why people still make these claims. The answer, of course, is that we don't know the technical issues well, and this makes us susceptible to being led astray by the irresponsible and the grifter. The more pressure there is on a technology, the darker the cloud of obfuscation grows. There's a lot to learn here.

So, you're going to say, "What makes you think you're so smart? I've seen your transcripts. I know just how weak your understanding of that math and physics stuff is, Lund." And, true, I admit it. My assumption here is that people who do it a lot know how it works. That is, I'm betting that failing forward makes you better at stuff eventually. So, it seems to me, we really need to look at the situation through the eyes of applied experience. In the summer of 1942, the pilots who were underwhelmed by the Bf109G and Spitfire V were being given a look into the Hawker Typhoon and the Focke-Wulf 190.

That's a lot to chew, so this blog post is certainly not going to extract the meat from this discussion. But I can set the stage. Firstly, it's not obvious, but the Hawker Typhoon is the direct descendant of the Sopwith Camel. At the end of the war, the British imposed an "excess profits" tax on the armaments industry based on items sold, and Sopwith had sold thousands of planes. Soon enough, the firm of Sopwith Aeroplanes was no more. The official version is that it had gone "bankrupt" due to the Excess Profits tax, but, immediately that it folded, the CEO of Sopwith, T O M Sopwith reappeared as the principle of H. G. Hawker Engineering, incorporated 1920, and in the process inheriting Sopwith's contracts.

You know what? This is not the kind of thing that gets by the fellows down at the Treasury. I probably shouldn't suggest that there was the government tacitly cooperated in an outrageously obvious corporate shellgame without evidence, but....

Oops. Wait. I just did. Anyway, while I'm not going to suggest that the Treasury was fooled, aviation historians could easily go astray. Sopwith was the owner and proprietor of both firms, but seems to have taken a back seat in the direction of Hawker. The way the story has been framed has always led me to suspect that Sopwith might have been a man of independent means, but the tradition is that we avoid saying that sort of thing in engineering-industry-related-autobiography.*** The titular Harry Hawker was an Australian fighter ace and test pilot who was dead in a crash by the end of 1921.  So when we talk about the  man who designed the aircraft that the old Sopwith firm built after it was renamed after Harry Hawker, we are talking about Sidney Camm, the young (born 1893) carpenter's apprentice who came into the industry in 1915 (if I'm remembering Penrose correctly, and I probably am not), and was hired on at Hawker as "senior draftsman" in 1923, and became Chief Designer in 1925 at the ripe old age of 32.

I don't know if aircraft designers call it a "Festschrift," but Camm got one when he retired. I suspect that it may be the same racket, given that the editor somehow gets an authorial credit. (Not that I'm pointing a finger of accusation at anyone I know. No sir, not me. Does having monographs on your C.V. count for aircraft designers, too?) Anyway, it's not entirely unfair in this case, because John Fozard is also the author of a very interesting article in the collection, “Jubilees in Design and Development.” It sticks in my mind, fifteen years after I read it, because of Fozard's blase approach to aircraft design. Far from the melancholy story of failure, underachievement, and agonising delay that you get in other versions, Fozard plots out a graph showing that aircraft development has always been a direct function of the number of planes built. Yawn. At Kingston-on-Thames, we just think that aircraft design is so boring. You would, too, when you've done so many.

Well, that's probably not the approach to take when you're talking about the Hawker Typhoon, because it was  dog. You put the best engine in the best chassis that you can, and you get the fastest machine that you can get, and, somehow, it's still not good enough. I wonder if Camm had a son who contemplated running for Prime Minister?

But it's a dog that we can learn from! That's what "failing forward" is all about! It's time to dissect failure and learn. Or, rather, it will be when  I come back to this subject, though, because I seem to have somehow run out of time in the day again. 

*Hand counted here,  since the facts are not always clearly articulated in the literature. Did you know that one of the economies made on the Liberty ships was the omission of firefighting equipment in the engine rooms? Guess what was the main cause of the very large number of accidents that took most of the surviving Liberty ships out of service in the 1960s. Cost cutting can have consequences.

**In his long, digressive, and gossipy account of the pioneering days of aviation, Harald Penrose (first of four volumes) at one point observes just how quickly ideas about what made an experienced pilot "experienced" changed. By the 1930s, there were pilots with thousands of hours in the air. These were often men like Charles Lindbergh and Air Vice Marshal Bennett, and their experience still told. In contrast, "experienced" pilots in 1918 often had less than 100  hours in the air. They didn't know what they didn't know. Which is probably a good thing, because if they actually understood what they were doing, they would have refused to go up in those crates. Seriously. The Royal Naval Air Service was the first air arm to work out a method of training pilots, and the work started in December, 1916.

***Fortunately, we now have Google, and, thus, easy access to this. So, yes, T O M Sopwith was a child of privilege, although considering that he gave Britain the Camel, the Hurricane, and the Harrier, I think we can say that he earned his privilege anew. Enjoy the link, by the way: it really is a slice of upper-class "civil engineering" in the golden age of the underdraining and navigating of old England.


  1. Are you sure you mean the Typhoon, not the RR Vulture-powered Tornado?

  2. No, the Typhoon was a dog, too. Just to explain, I'm saying that coming from an iconoclastic position on the whole "fighter bomber" thing. The Tiffie was an awesome fighter bomber, but the reason that there were fighter bombers is that there lots of fighters around that couldn't hack the air superiority role.

    So what do you do with them? You use them as ersatz light bombers. Real light bombers have two engines, because this reduces crew losses from mechanical casualties and a bombardier, who greatly increases your per sortie bomb hit rate. At the very least, as with the Red Air Force's Shturmoviks, they have gunners to reduce combat casualties.

    SE, S/S fighter bombers have none of that. Sorry, Fighter Mafia, but it's true.

  3. Well...Blenheim < pretty much anything. Obviously Mossies all round would be better, but that's asking for a pony.

  4. Britain First! Service deliveries of Rothermere's folly began on 10 March, 1937. It's basically a contemporary of the He-111 and much earlier than the Hampden --the original "fighter bomber," according to Handley Page prewar advertising. (Okay, I lied. They called the Hampden a 'bomber-fighter.')

    I tend to think of the Blenheim as a plane that earned a bad reputation in its last years of service. Fast bombers don't have to be faster than fighters. They have to be fast enough to narrow the interception envelopment to the point where few fighters are in a position to catch them. The Blenheim had a good start at that. Remember that it was introduced to the world as a "300mph bomber" by the Yugoslav air force at the same air show where the Yugoslavs showed off the High Speed (Hawker) Fury.

    Now, this isn't the place to get into the world of propeller plane speed claims. Suffice it to say that they all manage to be both technically accurate in some narrow sense and wildly misleading in others. The Blenheim looked a great deal faster in 1937 than it did in 1941, and so did the High Speed Fury--the aircraft that was supposed to be intercepting it.

    So times change. The real question is why the Blenheim's operational life was extended past that of the Fury, and that comes back to the Shadow Scheme, which had built up a considerable head of steam behind its conventional, poppet-valved Bristol Mercury radial engine. Had it been possible to bring the Taurus and Hercules into production in decent numbers at an early date, the Blenheim would be another historical footnote like the Wellesley, presumably replaced by Beauforts and Bothas and Beaufighters.

    Whether more Beauforts and Bothas would have been a god idea is another question. The counterfactual here is that the Taurus was never developed very far (and the Botha wasn't actually Taurus-powered). This comes back to the question of the tradeoffs between investing in production of low-tech machines like the Mercury versus investing in the productive capacity that will make it possible for the shadow industry to make more technically-demanding machines like the Taurus.

    Just to follow this tangent a little further, the big problem with the Taurus was the sleeve valve, which was very metallurgically demanding. Yet it was precisely more demanding metallurgy that lay in store for the aeroengine industry as the jet turbine came in, and which has broader implications across a wide range of industries. Being able to make and machine the more refractory alloys required in both sleeve valves and jet turbine blades means being able to make everything from better tanks to more productive refineries.

    Which brings me back to the incredible closed recursive loops that we can see in action in 1920s/1930s defence research and development. We start with Ricardo and the Aeronautical Research Council pushing for specific solutions to the problem of predetonation and end up with new alloys being used in jet engines that completely sidestep the problem, as well as in a broad range of entirely novel applications.