Wednesday, August 2, 2017

Technical Appendix for June, 1947: Hermione Giffard Says, "Cool Your Jets!"

I feel very, very old to know that someone named "Hermione" is a postdoc at Utrecht with a very, very good University of Chicago Press book to her name. Dr. Giffard's Making Jet Engines in World War II isn't just an excellent treatment of the early days of jet production. It is also a meditation on the relationship between invention, development and production. In a neat little turn of phrase, she asks whether Allison's move to press forward with producing its own jet engine design might show that it is "Hard to do development without production," flipping what I would take to be still the controversial and out-there position that "It is hard to do production without development," that is, innovating-by-doing, on its head.

That said, I have a bit more on my plate here. Various Great Minds of the last age said that "God is dead," either hopefully or not, with the notion that it was time to prod the world into a post-Christian morality of some or another kind. Nietzsche and Beckett are my exemplars here, because Nietzsche's prophet is the overman, and who doesn't want to identify with an "overman," and because Waiting for Godot is so ludicrously appropriate.

Because we're waiting for Google.

Whatever you think of my potted literary criticism, at least it serves to frame my mad claim that technological innovation is dead.

The Aeroplane, some time in the middle of July, 1939. Yes, I should keep better notes. The top secret gas turbines are being unboxed by private industry in the far right hand corner, while the Mosquito is trying to get its nose in the door on the bottom left.

Dr. Giffard is not a terrifyingly pessimistic techno-futurist. (I shouldn't speak for her. It is quite possible that she'd be willing to gig it if you've got a tenure track post for her, but this is not the kind of thing that she says in her book.) She does want to de-centre the inventor-corporate research lab bipolarity in favour of a production-firm centred approach, but that turns out to be a bit much to do in a crunchy treatment of the way that the gas turbine aeroengine actually came to be. That is, she is very, very good on inventors, but the corporate labs tend to get laid by the wayside, which is too bad, in our modern age of Google Labs trying to capture Bell Lab's light in a bottle. (What light? What bottle?)

Giffard's is a different look at the way firms innovate. Her argument is that development and production occur together, and that it is meaningless to talk about "innovation" without customers. In the case of the jet engine, the customer is the state, and we have another box in the iterative innovation-development-production-development-production flow chart, the state arms of research and development. By this, one means the Royal Aircraft Establishment and the research agencies of the German Air Ministry, but the RAE has a more central role in her narrative. Indeed, it is far more central than in earlier explorations of the field. (Giffard would probably point to the fact that the previous business histories of the jet age have come out of the private sector, and roll her eyes at this point.)

Speaking of private business that got involved in gas turbine work at an early date. . . 

You may recognise this from last month as a schematic illustration of the Brown-Boveri "Comprex," a version of the "pressure wave supercharger," per Wikipedia. As that writeup is historically unsatisfactory, I'm pleased to report that following up on the further reading brings one to a very good paper [pdf], by Pezhnam Akbari and Razi Nalim, of Purdue, who confirm my suspicions that the Comprex was first fielded on the gas turbine locomotive Brown-Boveri was shopping around in May of 1939, when a paper on their early work in the field appeared in the Proc. Inst. Mech. Eng.  By this time, Brown Boveri had sold gas turbines both to blow Houdry process catalytic-cracking petroleum refineries and for electrical generation at a plant in Neuchatel. In the course of the article, Professor Meyer ever-so-innocently proposed aircraft propulsion as another field where the gas turbine might be applied. 

Everyone proceeded to ignore Brown Boveri, and have ever since. While the successor company remains a large, diversified and successful heavy engineering firm, it has never produced gas turbines for aerospace applications. On the contrary, it has had to countenance the intrusion of aeroengine firms such as Rolls Royce and Pratt and Whitney into the power generation business. (Although Pratt and Whitney have been taken over by Mitsubishi on a if-you-can't-beat-'em basis, so that's all good.)

Giffard has an explanation for that, and it's interesting and also somewhat revisionist. As Brown Boveri's babe-in-the-woods intervention suggests, nothing could be more natural than for steam-turbine manufacturers to look to gas turbines as the next horizon. The technologies have close parallels, and the steam turbine industry had a strongly innovative culture. Last week, in noting the passing of (Armiral) Reginald Bacon, I obliquely referred to Parson's Turbinia's legendary intervention at the 1897 Spithead Review. 
In the "Nikolai Tesla" school of the history of technology-

-Charles Parsons steamed  into the midst of the stately columns of battleships and armoured cruisers first class and steam rams and torpedo boat destroyers and proceeded to embarrass the Admiralty into ordering the novel technology by steaming rings around its fastest ships. That is not how things actually went down, but isn't it the kind of heroic inventor story that we like? It's therefore a little disappointing that Parsons and Co., still a going industrial concern, furiously innovating and inventing away, should not have had even a sniff of the gas turbine business --in marked contrast to British Thomson Houston, Metrovick, GE and Westinghouse, all of which were well launched into the field by the mid-Forties. Parsons is an exception, like Brown Boveri, it seems, and without even the excuse of being on the outside of the British military-industrial complex. 

When I say that Giffard is being revisionist, I mean that she takes issue with the idea that we should expect the steam turbine innovators to become gas turbine innovators. She takes this as a strand of Robert  Constant's Origins of the Turbojet Revolution, and I will not quarrel with her, it being some time since I've read Constant; more importantly, she points to General Arnold's decision to put the American steam turbine giants in charge of turbojet development. Although GE did go on to be a major player in the field, Giffard points out that it was its turbocharger division, and not GE's steam turbine division, that made a jet engine work. And with the exception of GE, all of the steam turbine companies exited the field, Westinghouse being the last out, in 1961. 

So, what happened? In looking at Brown Boveri specifically, I had always assumed that the answer would turn out to be something to do with weight control. Ship plant, never mind stationary powerplant, just is not subject to the same pressures to keep weight down; but that is going to turn out to be, at least, a very partial answer.

One version of the history of the jet aircraft features two inventors: Frank Whittle and Hans von Ohain. Of the two, Whittle was more prominent at an early date. Although neither Whittle nor Ohain's work led to a operational jet aircraft, narratives of invention rest on priority, not production. Since Whittle has the first jet aircraft patents, while Ohain was the first to run a jet turbine engine, they have priority. They are inventors. 

I have my doubts about the inventor's role in society, and I suspect that Giffard does, too. Ohain is fairly quickly discussed, since he actually seems to be missing some of the very particular gifts that a great inventor needs, such as a raging appetite for publicity. It might well be that he would never have become an inventor had he not linked up with Ernst Heinkel, and we pretty much know this Pappenheimer at a glance. He's your Howard Hughes type, and in Giffard's account of Heinkel Engine Works, his basic strategy from beginning to end was to stage one stunt after another to foce the RLM into funding the next. 

Inasmuch as producing stunts is actually good defence policy if not taken too far, the German Air Ministry got what it wanted; but Jumo, BMW and Daimler-Benz, the companies that actually carried the burden of developing the German jet engine and getting them into the war in time to defend Germany's skies (if the army had done its job) have never been very happy that people take Heinkel's bullshit seriously. 
Giffard calls the Me 262 (and, still more, the utterly insane Volksjaeger) an "ersatz" fighter. Which is completely to the point! It's an emergency technological expedient to deal with a military-technical emergency; and it would have worked, if the army could have kept the Allies west of the Channel and east of the Vistula. 

As for Ohain, he was promoted to the role of Whittle's shadow twin, and got to go to some nice conferences, so there's an upside to having had to put up having Ernst Henikel as your boss for six years. 

Frank Whittle and Power Jets, now there's a story. The outline is well known. Power Jet was created to design jets, in cooperation with the development assistance of British Thomson Houston. It was then given a factory to build them in; factory and design were then taken away by two later-stage partners, Rover and Rolls-Royce. As a result, Power Jet never built jets. It ended the war as a state-owned development institution, and then was taken over as a research laboratory. 

The key modification in Giffard's account is that we are no longer forced to choose between believing that BTH, Rover and Rolls-Royce were thieves and jealous of Whittle's genius; or whether Whittle was an incompetent asshole. (Both positions are egregiously libellous, but not without merit. Except for BTH, which seems to have been pretty much an innocent bystander.) Instead, the phasing reflects an eve-of-war/early-war phase of panic in which the Air Ministry expected to face a German bombing offensive that could only be stopped by jet fighters; and a later phase in which the air war could clearly be won by conventional aircraft (or, at least, in which it was clear that the bombing offensive required better conventional engines), and jet engine development became an issue of national pride and postwar sales. The shift from one stance to the other led to a change of emphasis from emergency engines to developmental ones. 

In its early days, when Power Jets had the full backing of the Air Ministry (revisionism time, again), the story is astonishingly  modern. Patents and patent rights are enormously important; Whittle is initially funded by a venture capital firm, O. T. Falk and Company. (The corporate history here may be a bit more complicated.) Whittle staffed his firm with academically-trained scientists and engineers, at a time when that was only beginning to become common, and even paid staff in stocks.

I could go on, but, of course, Whittle couldn't. His basic design was taken over and modified/simplified to produce both the de Havilland Goblin and a whole series of Rover/Rolls-Royce developmental engines, culminating in the Derwent and Nene.  It turns out that Whittle's design was too complicated, and all the engine designers who looked at it, even Frank Halford, who knew from complicated, switched Whittle's theoretically-superior reverse flow arrangement for direct flow. 
I may not have the clearest grasp of the technical aspects of the controversy.

This is a story that could be extended if we were to open the field to other inventors, individual or corporate. A. A. Griffith, the RAE scientist who was the motivating spirit behind RAE's on-and-off again efforts on axial turbines, is usually written out of the list of Great Inventors, perhaps for no better reason than that he worked for a government lab. Of course, he also does not have a working jet turbine engine to his credit, because his design was too complicated. The same may be said of Jack Northrop's Turbodyne, as an example of one of a surprising number of designs that do not play much of a role in the historiography. Jumo, which, uniquely, got an axial engine into service during the war, did so by sacrificing reliability and durability. 

The fact that the initial designs were too complicated shouldn't surprise anyone with a stereotype view of academic science. It's a bit more surprising to see a ludicrously complicated design emerging from within a firm, but that happened, too. I've already given away Giffard's explanation for the process of simplification, development, and refinement that finally turned designs that were sometimes too crude, sometimes too complicated, but always unready (because they weren't ready!) into working jet engines. It's a developmental process, one led by companies with an expertise in developing aircraft engines. That was in part due to the fact that there was far more complementary technology carry over than is sometimes acknowledged; but it also has to do with the fact that the aircraft engine companies were used to developing aircraft engines, and had mastered a process that was far more complicated than an inventor building a working prototype. 

So, a firm-centred narrative of innovation. That much is clear. But, more than that, a customer-centred one. In the end, it was the respective air ministries (War Department) that dictated the pace. The RLM needed to flee headlong into a technological future to escape a disastrous strategic situation. It did! Britain needed (at least after Barbarossa) to move into a postwar era in which it could market "high technology" in America and rebuild its economy. It did! America needed to midwife an engineering industry not previously exposed to European state-aided competition into the new era of the military-industrial age. And it did!

It's almost like government departments know what they're doing, sometimes. Or if that's too radical a conclusion for you, here is another, more disheartening one: There's no development, no innovation, no progress, if you will, without sales. Therefore, as GDP growth declines, so will technological innovation.
Still Waiting for You, Elon

Doom! Doom, I says!


  1. she asks whether Allison's move to press forward with producing its own jet engine design might show that it is "Hard to do development without production"

    This is very close to the resignation letter Whittle quotes in his memoir - "any organisation that claims to do research must also do development"

  2. It's the "No development without production" that's kind of the kicker here. If true, this greatly strengthens the link between deficient aggregate demand and low productivity growth in the modern era.

    Also, who else is amazed to see the Ministry of Supply replicate the Power Jets "Let's create a new company just to produce some developmental engines" model for Roy Fedden?