Monday, February 10, 2014

Postblogging January 1943: The Most Inevitable Technical Appendix Of All

*


Remember those housing start numbers from last time time. The reason I went and looked them up is that I made the mistake of skimming The Economist from the summer of 1947, and was about ready to slit my wrists in the bath. (Fortunately,  my CAD 1000/month one bedroom in Kitsilano doesn't have a bathtub. So yay for high rents through artificial scarcity.) Here's a Trotskyite, of all things, giving us the low down. The fact that The Economist was not known for its Trotskyite sympathies did not prevent its "American Survey" feature from confidently announcing through the whole of the summer of 1947 that by the time that next week's number came out, the bubble would have burst, and that stagflation would be upon us. And they  hadn't even invented "stagflation" yet! The author would just go, oh, there's something fundamentally wrong with the American economy, it's about to crash. Oh, and by the way, the cost of living is out of control! Recession+inflation, just around the corner!

So, yeah, the inevitable technical appendix. It's about jets, of course. 

The trope is that when an old scientist-technician-guy sits down to write his memoirs, there has got to be a bit where the Olds tell him that whatever he's working on will never actually work. There's places and times where I can see that conversation happening, but when you hear it in the context of aviation in the late 1930s, it's ....weird, is what it is. Computers are about the freshest technological progress there is, and I'm 49, and I do not remember a time when there were not computers --right down to predictions of self-driving cars** and artificial intelligence taking our jobs. If I were a 49 year old project manager in 1939, I would very distinctly remember the day that the first heavier-than-air plane flew. And I'm telling Frank Whittle/Hans Ohain that jets will never work? Seriously?

And, yet, jets did not work in 1939--45. Sure, Me 262s and even Gloster Meteors shot some stuff down. As with most of the technological developments of World War II that we notice and care about, the turbojet revolution was mostly in the pipeline when the war ended. 

There's probably some law of human nature here. Plenty of technologies were developed, field-tested and used during the war, but they tend to get the "But what have you done for me lately?" Treatment. Oh, sure, penicillin saved millions of lives, and the technology for mass producing it led directly to the massive postwar orange juice concentrate industry (seriously), apparently via the first experiments with Tang, but does anyone even remember that it was the production of penicillin, and not its "invention," that was the issue? On the other hand, if the technology was still on the way, its proponents had a fairly obvious reason for talking it up: it needed all the help it could get to win funding in the newly constrained postwar environment. All you have to do is look at the helicopter. The folk memory of 1944 is that everyone was promising that we would all be commuting in helicopters real soon.*** There were plenty of articles that said that, of course, but there were also plenty of "Let's be realistic about helicopters" articles. As it happens, it was the semi-trailer, reefer and shipping container that transformed postwar logistics, and while the shipping container has hired a publicist, the rest of it is all lost history.

So what would you have said if you were writing a "Let's be serious about jets" article in February of 1944?


Well, the first thing that you would say, of course, is that Ernest Heinkel is a great big fraud. If you were German. And that's not theory. Ralf Schabel actually quotes the CEO of Daimler-Benz to that effect.  Heinkel claimed that his He280 jet fighter could have been in service much earlier if the Air Ministry had just supported it, and this occasioned the tart reply that it was not the plane, but the engine that mattered, and engine developments have timelines. Of course, Heinkel was trying to develop his own line of jet engines on the side, so it gets a bit more complicated. (Wiki articles! Oh, the wonders of our age.) Frankly, I do not think that that makes Heinkel less of an asshole. He had basically already tried that with the He177, an aircraft designed, admittedly to Air Ministry specifications, to give him an excuse to play with and appropriate a DB engine. The fact that Richard Fairey tried the same stunt twice (1,2) and that Geoffrey de Havilland made it work, or more accurately, Frank Halford, does not make Heinkel any less of an asshole. It shows that assholes often succeed in business. (Actually, the "asshole axis" might need some development here. It certainly explains the Mamba/ASX/Sapphire. We're just left wondering where the Handley-Page and Fedden jets are.)

It goes Jaguar-Tiger-Deerhound-ASX-Sapphire-Austin-Mini. There's a biography by Martin Nutland, now, and I  lifted this photo from the promotional website.
 So these are guys you should probably not trust to give the fullest and fairest account of their lives. Having dwelt on it for far too long already, I still cannot help repeating my belief that the most interesting part of the story could be the Armstrong-Siddeley connection, and it is a very good example of buried history.

Fortunately, the British mechanical engineering sector was like about twice as many starving dogs as there was room for in their pound, all trying to get a last, desperate bite of the last bloody fragments of Merlin development when the keeper wandered up to the edge of the pit, staggering under the weight of a massive, bloody joint labelled "jet turbine." So if you're looking for the science and engineering of actual jet engines, you do not have to go very far to look. Open a number of the Proceedings of the Institute of Mechanical Engineers or the Journal of the Royal Aeronautical Society, and it'll be there. 

Not that anyone's looking. It's all, "OMG, Hitler lost the war by making the Me262 be a bomber!"

Here's the thing: a jet turbine is really simple. Air goes in a hole, is injected with fuel, and sucked through a turbine. Compression leads to adiabatic heating, and blam! the reaction product "jets" out the back and pushes the plane along. I would say that the Devil is in the details, but then I would look at the design of the Napier Sabre, and I would laugh and laugh and laugh. It's possible that the thingies they're putting on F-35s and whatnot are more complicated --more detailed-- than the Sabre, but certainly no jet engine of the 1940s/50s was. 

Put more exactly, we confront several related problems:

i) That fluid flow across the turbine blades receive net acceleration. There is no "compression stall," in which the power produces by reducing jet fuel in air is not wasted. Actually, though, every jet of the 1940s was threatened by compression stall at low speed. More than anything else, this explains why the first jet fighters were not ready for prime time in 1945.

ii) That enough power is imparted to the reaction mass to be worthwhile. This is a little complicated, though. The Me262's engine was not particularly efficient, but because jets are more efficient at high speed, it still went faster than any propeller-driven aircraft.

iii) That gas impinging on the turbine blades not be so hot as to cause metallurgical failure. This is, again, a little complicated in that the turbine blades are being required to spin around at very, very high speeds. They are thus already at risk of entering a structural failure mode through "creep," and hot gasses flowing over the blades clearly does not help. Early German jet engines had turbine blades made out of milled aluminum, and, later on, the production models embraced low-alloy steels, as Germany as short of alloying materials. This made German jet turbines both short lived and inefficient, really through no fault of German engineering. (Although because Corelli Barnett, I am still tempted to mock.) The Whittle-Rolls-Royce line of development in Britain focussed on nickel-based alloys of the kind that the British industry had just got through developing in an attempt to make Roy Fedden's sleeve valve engines work in the real world.  There is no way that the Germans could have afforded that! On the other hand, alloy steels will work. 

Fortunately, the solution to all of these problems is at hand, because the problem is (almost) exactly like that of the steam turbine, already solved! No wonder that A. A. Griffith, the Principal Science Officer at the Air Ministry Laboratory, spent most of the 1920s and all of the 1930s puttering around with jet turbines. He does not get the attention that he deserves, first because he was working on turboprops rather than jets, and second because he started with the  more complicated version of the jet engine with the greater potential, the "axial flow" concept, in which additional increments of compression are gained by putting more turbine "stages" one behind the other. Frank Whittle and, initially, Hans Ohain, instead just made the blades bigger, relying on greater power being imparted by an expanded single stage. The tip area of the blades were going faster, thus pushing harder --"centrifugal flow." 

Everyone thought that Griffith was right, except for Whittle, and even though the Meteor initially flew with a centrifugal engine, people kept thinking that Whittle had backed the wrong horse, short as well as long term, until a United Nations pilot flying over the Yalu one bright day noticed that a speck on his windshield was getting closer, really, really fast. How embarrassing.

As Jakob Whitfeld notes, in his now-available-online PhD thesis on the Metrovick jet engine programme, the engineers coming into the jet problem had extensive experience with exactly these problems from internal combustion propeller engines. The supercharger problem was like the turbine design problem, while metallurgy was fairly obviously involved in improving both kinds of engines.

Unfortunately, this familiarity was misleading, at best. Jakob's blow-by-blow gives a good sense of the way in which experimental turbine sections kept returning disappointing results. The better known Whittle unit was, meanwhile, hanging fire, somehow unable to deliver the expected power --as were parallel German projects.

So what was the problem?

Here's a sketch from J. Reeman's "The Turbine for the Simple Jet Engine," published in a War Emergency number of the Proc. Inst. Mech. Eng. and based, if Google can be trusted, on bench tests of turbine blade efficiencies done in 1941, although the paper itself is postwar.


Oh, for the days when engineers could draw. As you an see, a great deal of attention  has been paid to designing the airfoil profile of the turbine blades here. Reeman is very proud of them, and a little latetr he is going to point out that while wartime British turbine blades achieved 87% efficiency, German blades were at 75%. That is something to be proud of, but what follows definitely is not.


That would be math right there. Or, rather, "math." Reeman is a dab hand with the drafting pencil, but he clearly has not a clue how to deal with gas flows across surfaces where unit velocities vary. As, in fact, they do when you're busy imparting energy to them. To be fair, if he wrote this as a partial differential equation, as he would need to do to be accurate, he would just end up with an equation that he could not solve, so there is some justification for the classic engineering view of PDEs as a mystery invented by mathematicians to torture undergraduates,

At this point, I could go on about how science and engineering at the end of the war was bumping up against the need for computers to solve the problems they had made for themselves at the same time that they were engaged in inventing computers. Today, though, I will not. No dead horses, today.As you might have guessed from the sense of ennui that I communicated at the head of this post, I am feeling in a bit of a "been there, done that" mood. The fact is that, without computers to solve specific problems of transsonic gas flow, the only way that you are going to get to working jet engines is by building lots of prototypes and trying them out. 

You need to throw money at them. Unexpectedly, I come back to my digression at the top. You want new technology? Spend money. You don't? Well, don't cut research and development budgets while idly dreaming of the miracles to be delivered by inventors, entrepeneurs, and the genius of the private sector, real soon now.

**"In 1960, Ohio State University'Communication and Control Systems Laboratory launched a project to develop driverless cars which were activated by electronic devices imbedded in the roadway. Head of the project, Dr. Robert L. Cosgriff, claimed in 1966 that the system could be ready for installation on a public road in 15 years." That's further away than commercial fusion power! Thanks, Wikipedia. It looks as though Dr. Cosgriff was on about this for something like eight years, and then found a safe landing studying traffic light timers, 
***Also, "Jeepmarine."

No comments:

Post a Comment