Tuesday, November 12, 2013

Postblogging October 1943: An Artificially Aged Technological Appendix

It turns out that I'm not the first person of our Modern Intertubes Era to be struck by the Bohn Aluminium and Brass Company's futuristic ads of the 1940s. Jim Edwards, of Business Insider, pointed out back in 2010 that "These Magnificent Paintings Of 'The Future' From 70 Years Ago Got Everything Hopelessly Wrong." That might be a little snarky. Lawnmowers are made of aluminum these days, and it's not the biggest mistake in the world to imagine them designed with an art deco touch back when art deco was still cool. 

Edwards should have substituted the aluminum high rise with button-adjustable room sizes for the lawnmower. Again, the vision is  not completely wrong, in that aluminum houses many North Americans today, and they are fairly modular and easily adjusted. As a young high school dropout at my employer pointed out the day that she was qualified as a cake decorator, "Now I can afford some class and get a double-wide." Which was not her exact phrasing, and, believe me, ex-con boyfriend, fake fingernails and all, she still had her tongue in cheek. No-one is so confused as to think that trailer parks are classy. As temporary housing, the old building codes here in British Columbia used to let you build them on the flood plains on the far side of the dykes. Hint hint, insert tornado joke here. Aluminum may have been cool once, but, like plastics, it has been "deglamorized."

After the jump, though, back with plane porn!

Mitsubishi A6M Zero, from Aviation-History.com.

The Mitsubishi A6M ("Mitsubishi Type Zero") is an otaku's dream. You wave in the direction of the Book of Five Rings and drop some quote about the invincibility of refined technique. Then you explain about how katanas are the ultimate badass sword on account of being made by supreme swordsmiths using a technique where they repeated hammer out a white hot metal bar into a long, flat rectangle, then folding it over into a billet, and then hammering it out again. You know how a piece of paper gets  harder and harder to tear as you fold it up? It's like that. The katana's steel has been folded over so many times that it is invincibly strong.

So if you have a culture that makes katanas for samurais, naturally it makes Zeros for naval pilots. (Who are also samurais.) Lighter than its big and cumbersome Western rivals (Wildcats smell like butter and have big noses!), the Zero answers to the precise control of the supreme, Asiatic martial-arts practitioner&etc.

Well, first, get over yourselves:

That's 1850hp of turbosupercharged, intake fan-cooled power. Oh, sure, the plant was unreliable and the turbosupecharger was never even installed, but it's not like the Japanese Navy failed to see the need.

So now that I've got some brisk exercise beating up a strawman with my wooden sword, I'll move on to the next part of the schtick, where I talk about wie es ist eigenttlich gewesen.  

In May of 1937, with the Nakajima A5N coming into service, the Imperial Japanese Navy issued a specification for its replacement. With the 4-5 year lead time typical of an aircraft development contract of the era, the resulting plane would appear in production quantities in 1941 and finish replacing the A5N in the next year. 

The legend is that the Japanese Navy's specification was unreasonably extreme. I am not at all sure that I see this. Typical of its era and requirement, the specification called for a short 39ft wingspan in order to fit onto an aircraft carrier. (The F4F managed a shade over 38ft, Brewster pushed it down to 35, and the  Royal Navy accepted the necessity of wingfolding and got 46ft from both Blackburn and Fairey).

Second, 310mph was required at 13,000ft. To put this in perspective, the Vickers Venom was shown off in public in the summer of 1936, had a legend speed of 312mph at 16,500ft. Only the most dewy-eyed Biggles fan would have imagined that a combatworthy Venom would make anything like this once properly tricked out with service impedimenta,  but, on the other hand, the Venom was clearly a failed specification. On the other hand again, the Zero was to be a carrier plane, and so could count on giving up some performance to contemporary land planes. 310mph was reasonable, and Wikipedia gives the Buffalo 321mph, no height specified. (If my aunt's old VW camper van could get a speeding ticket on a highway, then I take this number with the utmost seriousness that it deserves!)

Third, there was impedimenta, the amount to be carried. Since everyone was on about escort fighters in the summer of 1937, one might expect a very strict range requirement, implying significant fuel carriage as part of an excessive required service load. This is certainly the legend, but, again, I am not sure that I see it. 

While the escort fighter craze of 1937 usually led to twin-engine designs, the Japanese Navy seems to have had an attack of common sense and specified a single powerplant, the licensed version of the Gnome-et-Rhone 14N built by Nakajima, the Sakae, although here I may be misled by online splat sites, as it seems unlikely that the Navy would have forced designers to use a specific engine  

Nor was the range specification actually that demanding. It asked for two hours endurance at normal power At 80% of designed 950hp at 0.45 lb/hp hour this is a shade under 700lb of fuel, or about 120 US gallons, depending on air temperature. Six to eight hours at economical cruising power was also specified, but as one might ask an aeroengine plant to turn over at as low as 55% of rated full power in cruising mode, this would not have required any egregious loading with extended range tanks. (Is this the time to talk about "why didn't the Luftwaffe use drop tanks in the Battle of Britain?" No, this is probably not the time.) 

Certainly 120 gallons is a lot more than the Spitfire was asked to carry, but it is not a staggering load. Even the earliest P-51 was equipped to carry 170 gallons, and fuel is comparatively dense. There is not a great deal of room to spare in a single-engined high performance fighter, and, if you are making demanding use of it already, adding fuel tanks will push up parasitic drage quickly. If, on the other hand, you are making comparatively less demanding use of the structure, then there will be volume to spare for fuel. In the case of the A6M, the Navy was ready to compromise, specifying an armament of two machine guns and two cannons. This was still more than the prototype P-51, but, given the choice to use Japanese license-produced versions of the Lewis gun and the small Oelikon cannon, the requirement was again not demanding.

What the Wikipedia article, which may or may not carry over emphasis that a splatbook author may or may not have picked up from an actual designer, does emphasise is a requirement for a  complete radio set including a radio direction finder. This does not sound like very much, but it actually makes sense that this is the requirement at which designers would balk. Radios were clearly a challenge in the 1930s. Even German and American equipments fell somewhat short of British. 

But hold on there for a moment. The  incredibly supercilious and judgmental perspective to take here would be to point to Japan's relatively low national income and make some kind of slighting reference to natural British engineering superiority. It would be hard to pull that off in the 21st Century, when the national reputation of German and American aeronautical engineering is high, but surely not impossible.  Yet the "failures" I have mentioned have to do with the fact that the RAF began making and deploying aircraft VHF radio sets during the Battle of Britain. As I have pointed out before, the theoretical advantages of a VHF-band aircraft radio were obvious. The problem was that it would clearly take a great deal of money to develop them. The Americans chose not to spend the money to develop them, the Germans spent their money elsewhere. The Japanese. . . I am going to finish talking about the A6M. Hopefully I'll get to a place where we can see the implicit decisions about radio equipment for aircraft for what it was.

So the upshot is that Mitsubishi saw this specification as incredibly challenging, yet managed to fill it. As can be seen from the picture above, they did not exactly do so by pushing the limits of aerodynamic design. The retracting tailwheel is moderately impressive for its era, but the braced cockpit and distinctly old-fashioned engine cowling are workmanlike and no more. Yet the statistics of the plane are amazing: (Wikidump!)

General characteristics

  • Crew: 1
  • Length: 9.06 m (29 ft 9 in)
  • Height: 3.05 m (10 ft 0 in)
  • Wing area: 22.44 m² (241.5 ft²)
  • Loaded weight: 2,410 kg (5,313 lb)
  • Maximum speed: 533 km/h (287 kn, 331 mph) at 4,550 m (14,930 ft)
  • Range: 3,105 km (1,675 nmi, 1,929 mi)

That's 3,704lb, empty. It is not always easy to parse empty weights, never mind loaded, but the F4F-4 comes in at 5,895, the Bf109G-6 at 5893, the Spitfire Vb at 5,093. Even allowing for the fact that this particular iteration of the Zero was only sporting a 950hp engine, this is light, and this surely explains much.
So what does it explain? Famously, it ties back into a certain cliche about Asiatics being ... casual with life. Either they don't have feelings, like you or me, or samurais treat death as lighter than a feather, or some such. Anyway, samurai die like flies in service of their duty. Katanas are for samurais, and they are light and strong. So so are Zeros! In more prosaic terms, the Japanese authorities left armour plate, self-sealing fuel tanks and bulletproof glass of the Zero, just as their rivals tended to do in peacetime --and then did not add them once war made these measures of pilot protection more pressing. 
In fact, the Japanese got these into service as quickly as they could, in the A6M5c-2, following on improvements in the engine to keep the fighter up to par with its rivals. This is not to say that in critical phases of the war that Zeros were not springing sprightly through the air because they lacked the armour of their enemies, not even that this was not the main factor in their remarkably light airframe. Just that I have a point to make:

(Wikipedia again, emphasis mine, you can follow the citations at Wikipedia.)
Nakajima's team considered the new requirements unachievable and pulled out of the competition in January. Mitsubishi's chief designer, Jiro Horikoshi, felt that the requirements could be met, but only if the aircraft could be made as light as possible. Every possible weight-saving measure was incorporated into the design. Most of the aircraft was built of a new top-secret 7075 aluminium alloy developed by Sumitomo Metal Industries in 1936. Called Extra Super Duralumin (ESD), it was lighter and stronger than other alloys (e.g. 24S alloy) used at the time, but was more brittle and prone to corrosion[6] which was countered with an anti-corrosion coating applied after fabrication. No armor was provided for the pilot, engine or other critical points of the aircraft, and self-sealing fuel tanks, which were becoming common at the time, were not used. This made the Zero lighter, more maneuverable, and the longest range single engine fighter of WWII; which made it capable of searching out an enemy hundreds of miles away, bringing them to battle, then returning hundreds of miles back to its base or aircraft carrier. However, that trade in weight and construction also made it prone to catching fire and exploding when struck by enemy rounds.
Shorter: The Zero's remarkable lightness comes not from its conventional design or relatively backward equipment load. It is the result of using a more advanced alloy than its American and British rivals. Total Western superiority burn!

But first, another copy-paste dump, this time from my own material, gently corrected by J. Gilbert Kaufman:

Aluminum is the most common metallic element on Earth. Unfortunately, while aluminum is common, so is oxygen. The physical properties of aluminum, which is soft, lustrous, and ductile, come from its electron orbitals, which make Al-O a very low energy state from which the pure metal is difficult to reduce. Once a way was discovered to reduce the oxide in which it was normally found, that is, in the mid-1880s, its favourable strength-to-weight properties made it a popular industrial material. Incidentally, it is a useful addition to explosives due to its consequent high enthalpy of combustion. Industrial aluminum production is by electrical reduction, and uses a great deal of electrical current as well as, in WWII, the mineral cryolite. Pure aluminum is soft, and usually used in industry in alloy forms. The American Society of Metals recognizes 8 families of industrial aluminum designated by a prime number that indicates its primary alloying material and using a range of further indicators to designate the details of composition and working method: 1xxx is essentially pure aluminum; 2xx.x alloys it with copper; 3xxx with silicon, copper, and/or magnesium; 4xx.x with silicon; 5xx.x with magnesium; 7xxx with zinc, 8xxx with tin; and 9xxx with other elements; while 6xxx is an unused series. This system was not in use during WWII, and various national standards specifications survive today, notably in the United Kingdom, where the tradename “Hiduminium” and the RR specification reflect the semi-incestuous relationship of High Duty Alloys, Rolls-Royce and the National Physical Laboratory gave it the inside track in aluminum alloy development, but is useful to classify the alloys that were then in use. 

If you will click over to the Wiki article on hiduminium, you will see how hard it can be to reconcile traditional and modern designations. The standard alloy used in forged aircraft aluminum pistons during the era is now known as 5051, but hiduminium is there described as a family of nickel-containing alloys, which would not even be a family to the modern ASME. (We're negotiating with Sofia Vergara to play 7xxx in the TV adaptation.) “Duralumin,” the alloy that you will encounter in any discussion ofinterwar light alloy aeronautical materials, is close to modern 2017, but will be found in contemporary American nomenclature as 24S. But by mid-war, "24S" will more likely refer to 2024. German military aircraft captured in the first two years of war used an aluminum–copper alloy that is essentially 17S by the old standard designation, or 2027. By contrast the aluminum–magnesium alloys publicly revealed in the United Kingdom after the appearance of the Westland Lysander in the summer of 1938 were 5xxx series alloys similar to 5456 and 5083. 2xxx series metals are heat and “age” hardenable. Fighting my way clear of a thicket of terminology, this means that Al alloys have different responses to being heated up and suddenly cooled and to being held at elevated temperatures for extended periods. Both "harden" --I'll get to the scare quotes below-- the metal, but the latter does so in a way more congenial to post-hardening machining. These standard alloys were used for structural elements, for aircraft skins, including Alclad skins, and by British manufacturers for forged aeroengine pistons. 3xxx series had little, if any, aeronautical application until the end of the war when American manufacturers were able to work them into radiators with excellent results. Some 4xxx alloys are heat treatable, and their good flow characteristics made them particularly suitable for cast pieces such as some aeroengine pistons. 5xxx alloys are relatively strong and strain hardenable, with good toughness, and the low magnesium versions have good corrosion resistance. Strain hardening makes them difficult to extrude, requiring very powerful presse, but the basic similarity with steel allows application of the same kind of methods, particularly forging. 6xxx and 7xxx metals are heat treatable. 8xxx have various properties. 

7xxx, which you will by now recognise as the family that includes Sumitomo's ESD, is the strongest of all then-available alloys. That it is highly susceptible to corrosion is usually noticed. That it "strain toughens" is not. This means that it must be worked quickly after any heat treatment. It was supposed that its high stress corrosion characteristics ruled out is use in aeronautics. Developing pioneering German work, NPL scientists developed a 7xxx alloy in 1917–18, so called “E” alloy. The Germans marketed their own development in the 1930s, and NPL’s collaborators at Heavy Duty Alloys began to market an improved version of E alloy as Hiduminium RR77 in 1937. In the same year, Japanese scientists published an explanation for 7xxx’s high stress corrosion characteristics that implied a chemical solution, the cladding noted in the Wikipedia article. This work was evidently ignored in the west. I think next month's techblogging entry will suggest an explanation as to why.

It was not until the Mitsubishi A6M “Zero” was discovered to use a 7xxx metal structurally that the metal was used in western aircraft. Alcoa scientists patented 75s, or 7075 in May 1941 and announced it in 1943. Somewhat different from RR77, in Britain it received the designation RR78. Predictably, Alcoa and High Duty Alloy scientists chose to differ vigorously over which of the two was superior in various respects. RR77 was first used in late war undercarriages such as that of the Gloster Meteor, and ultimately replaced 2xxx and 5xxx extrusions and forged pieces in the British industry most notably in an evolutionary development of the Supermarine Spitfire, the Spitfire XVIII, while the Americans made a direct transition to the use of 7075 extrusions in the Douglas A-26 Invader and series continuations of various wartime airframes such as the Lockheed Hudson (A-20) and B-29 to the PV-2 Harpoon and B-50. Unfortunately, this led to something of a post-war recall fiasco as undercarriages began to fail from stress-corrosion, although I will not make too much about it as I have managed to misplace my note on the subject. It's in the SAE Proceedings for 1946, I think. Some year in there, anyway.[1]

I introduced this section with a driveby sneer at "otaku" who are so morally deficient as to know less about inorganic chemistry than I do. Low as that bar is, it is also entirely unfair, even by the standards of the way that I treat my strawmen around here. I just needed a way to get that analogy about a katana being strong in the way that a multiply-folded page is strong in here so that I could hold it up to the light and make fun of it. (See above.) The problem, of course, is that the analogy treats metal by analogy with any other "fabric." In the years since such analogies were the only way to understand the properties of metals*, we have had a  material sciences revolution. In a previous post, I talked about the understandable way in which the switch to metal in aircraft construction was something of a step backwards that, while quite satisfactorily motivated by manpower and maintenance considerations, allowed airframe shops to get by with reduced capital investments. Specifically, they did not have to invest in heat treatment equipment. They would have required ovens of some kind to pursue the plywood-plastic industry to where it was going (cheap houses in California), and they would need them, eventually, to take full advantage of new light alloys. But, in the context of the 1920s, metal-strip aircraft manufacture meant soldiering on with lathes, planers and bandsaws, only with a little more horsepower applied to work with aluminum, or a lot more to work with steel. In fact, from the machine shop perspective, aluminum was the lower technology solution. However it might be made, over there in the place where they make metal strips and ingots and stuff, Al was easier to work than steel!  

But to use light alloys, or steel, to its fullest, you need to do heat treatment, and that means that, finally, you have to reequip the shop. 

Here's where I get Isaac Asimov-y again. Quantum chemistry begins with the elemental table and all those orbitals up in the top right hand corner. We now understand materials as atoms locked into place by bonds formed by interlinks between these orbitals. The particular orbitals tell us about the energy required to make, form, and break those bonds, and the amount released when we do. They are an imperfect guide, in that no actually existing material consists of a perfect series of interlocked bonds, but rather of a system of dislocated bond-chains "doped" with impurities. The mechanical analogies that give us "strength," and "hardness" and "toughness" are thus ways of describing the particular family of bond-chain links that make up a particular physical phase of a given material (allotrope), and the actions of the impurities under displacing impulses. In short, "stress hardening," "stress corrosion" and the like can be seen as the migration of doping impurities. Using 7xxx as a high stress material is dangerous because impurities will migrate to the point of highest stress through the metal and congregate, creating a fault and a weak point. This is not a criticism of Sumitomo and Mitsubishi. The phenomena can be predicted, and, to a point, controlled, as they did. The whole point of the quantum chemical revolution in inorganic chemistry was to open up vistas of new light alloys for aeronautical work. 7xxx is not used very much in aircraft these days, I understand, but the first bit of progress away from it was to Mag-Thor, a material that makes a leap into the stress-corrosion dark seem positively sane in comparison, and beyond that we get to titanium, also a little nuts, albeit for different reasons. 

What we need to understand is only this: Mitsubishi's choice to use 7xxx was a daring, theoretically-informed, high-cost jump into the technological future, where factories had the equipment and organisation needed to use artificially-aged materials. American authorities famously recovered a flyable A6M in July of 1942.  The Wiki article captures the contemporary reaction as one of relief. The Zero was a paper tiger and the upcoming Hellcat could beat it with what might seem like the default American solution: MOAR POWER.

What is carefully not mentioned is the metallurgical analysis, which could easily have been done as early as Pearl Harbor.  The Hellcat took to the air with an enormous engine that ultimately flowed from the needs of civil aviation. It proved more than enough to master the A6M in flight, notwithstanding the state-directed technological innovation of the A6M airframe. As I have already suggested, we will begin to see next week how modern aluminum working methods were disseminated through American industry by the needs of military aviation once war had begun. Not to spoil my own material, but I have suggested that it is not just the material or the way that you work with it that matters here, but the way that the workplace is organised. This will be on the exam! 

So the difference here is this: for policy reasons upon which I will not dilate, America entered the Second World War with an underfunded armed forces and public research and development apparatus. The weapons that Americans used to win a world-shattering, logistically staggering world war against three ocean-separated G7 enemies simultaneously ultimately derived from the inadequate (for military purposes!) research and development that civil industry was willing to commit.

Japan, meanwhile, in spite of my qualification of it as "G7 level" was pulling its way out of the Third World. We know how successful that process was, but it seems to me that as long as we dismiss the A6M as something quaint and artisanal, a katana for the 20th Century, as it were, we fail to see the way in which the state can take the lead in research and development and push industry and society forward. That we continue to see the necessity and value of this only in the context of military affairs would be a tragedy were it not for the way that feedback loops lead back to industry. 

Americans defend the interwar underfunding of their armed forces in the light of lack of apparent threat. True. But in so far as apparent national security innnovation actually serves industrial (and social!) needs, I am going to suggest that the roots of failure lie elsewhere.

 I guess that I promised not to dilate on that subject, but, come to think of it, I have said a great deal about the roots of Roman failure in the state's increasing inability or unwillingness to tax. . . .  

[1]Kaufman, 10–16; 25–30, 32, 117; Frank King, “Aluminium I: 1900-1950,” Hist Met. 19 (1985) 39–43; William Doyle, “Aluminium 2: Hiduminium and RR Aluminium up to 1950,” Ibid., 44–7; Mac Young, “Aluminium 3: The Early History of Alloys,” Ibid., 48–56; and private communication from J. Gilbert Kaufman.
* Making it hard to believe that humanity ever learned to make steel when it had so little clue about what we were actually doing. Empirical knowledge for the win! 


  1. Awesome. And Mag-Thor is a gloriously crazy idea, high camp cold war technology. "There's only one, ah, issue. The entire airframe is highly radioactive, to say nothing of the floor in the machine shop. Which is precisely what we said to the union convenor."

    1. Come to think of it, not just radioactive but also inflammable...

  2. Pfft. Per Oakridge, average worker exposure is only 50 mrem/year. That's less than a chest x ray!(Assuming exposure to a 10kg Mag-Thor casting of 1.7% thorium at 0.5 meters distance for not more than 40 hours/year.)