Shuttle was compromise---need shuttle 2--- but it worked
← Shuttle loss will have serious impact to Nation
Shuttle was a design compromise–need shuttle 2–but it worked
Posted on January 7, 2012 by Bob
Dick Morris –
You're misreading me. I didn't say Rockwell had proposed a VTOL 2-stage version of the shuttle to NASA. As you note, it wasn't what NASA wanted, and as I noted, it couldn't have been built. VTO with Horizontal Land … that was doable, and if memory serves that's what everybody bidding on the shuttle contract proposed.
I wasn't there for that part of things. I started at Rockwell (still North American Rockwell, not yet Rockwell International, as I recall) in May of 1973. I was a college dropout, or maybe a flunkout, and even at the time it wasn't clear exactly which was the case. So I wasn't a "real engineer"; I was hired as an Engineering Aide II, an extra set of hands for a real engineer, to deal with scut work, and a week after I was hired I got shipped off to Tennessee for two weeks to help record the results from some wind tunnel tests at the Arnold Engineering Development Center — a little bit of the 21st century dropped down for some reason in the mid 20th century, in the midst of communities where most of the the technicians lived amongst farmers and people distilling whiskey and producing baseball bats from lengths of hickory…. Which was educational in a way, to learn I lived in such a country, never all one thing, but mixed up this way and that. And eventually I did become a Real Engineer, with a promotion to Member of the Technical Staff, a couple of years later when I was on the B-1 development program, and got a real college degree in Physics at West Coast University, which ain't famous but has served me well enough, along with GRE scores, to be considered for grad school, Which is another story.
My boss was a thin intense man named Mark Harthun. His boss was a genial hefty sort named John Lundgren, and I feel privileged in retrospect to havbe known them both. Harthun was a supervisor, he had two lead engineers and about a dozen other people, most of them new college grads. Lundgren had three groups, with about thirty people among them, and he was a manager — second rung on a ladder of 8 or 9 management rungs leading up to Vice President, if I remember correctly. Anyhow, we were the Aeroheating Group, and our job was to keep the shuttle from melting during ascent and reentry.
Chief problem on a VTOL system: re-entry and landing. You've got 25,000 fps of velocity to kill off, at a minimum, which takes a whole lot of fuel if you don't get aerodynamic braking. Which means wings or a wide fuselage, like a lifting body. Re-entry from the moon or planets, outside earth's orbit, gives you 35,000 fps or more velocity to deal with — twice the energy, twice the problem.
So Rockwell proposed a VTOHL system. The original design was a rwo-stage system: the booster was about the size of a football field, with a rather flattened oval cross-section; it had a two-man crew, if memory serves. The orbiter sat on the back of the booster, much as the real one did, but the proposal orbiter looked like an X-1, with small rectangular wings pretty much perpendicular to the fuselage at about 2/3?s the fuselage length. Think of a brick sitting on a one-foot ruler and you've got the idea. It was small, with maybe 15000 pounds of payload capability.
"And the air flow over it never goes turbulent," Joe Haney, one of the lead engineers said my first day, showing me a 6-inch wind tunnel model of the orbiter. "No matter what we test at, it doesn't go turbulent." Which didn't mean nearly as much then as it would later ….
What with this and that going on in Washington, some "improvements" were ordered. I think the flyback booster died first; the shuttle would be mounted to big expendable tanks instead — easier to design and cheaper to build, since a tank didn't need a human crew or engines or much else. This was supposed to be "temporary," and NASA would get around to building a real flyback booster as soon as funding was available. There was a war going on, and the government was running an unprecedented deficit, largely because of the cradle-to-grave socialism of the previous Democratic administration, which just about all the guys around me despised …. I'm sure you'll find this inconceivable.
Then NASA and the Air Force made an agreement, and the X1-like orbiter went away. The vehicle suddenly needed five times the crossrange (from 1250 miles to 6250 miles); it needed to carry 65,000 pounds of payload. Roughly speaking the cargo bay went from pickup truck sized to railroad car sized. So, bigger fuselage, wider and longer. Bigger wings to increase the vehicle's range — to slow down descent faster, to allow more steering during rentry. And bigger wings yet again, to deal with the orbiter's increasing size and weight…
And since the wings are larger, during reentry the boundary layer (the region where air goes from flowing "free" to not moving relative to the vehicle, and becomes hot as kinetic energy is transformed to thermal) is going to get thicker, and finally at last we did get turbulent flow on wind tunnels, implying we'd have turbulent flow in real flights….
Which was not good news, because a turbulent boundary layer is thinner than a laminar one. Air flow becomes chaotic. There's buffeting. And there's a lot more heat transfer to the aircraft structure. People write books about this stuff — the basic text on boundary layer theory was by a guy named Schlichting and for all I know it's still in print, there's not a lot of folks competing for fame in his corner of the churchyard.
Anyhow. The increased heating required bigger, tougher wings, which mweant more weight. It required some structural changes in the main fuselage, generally meaning more weight. All meaning changes in the vehicle's aerodynamics, meaning different return trajectories had to evaluated, and so on.
But of course the real fun thing that in was a tile-based Thermal Proctection System. People could write books about the TPS I rather think — really thick ones, like Russian novels, with — in the end –similar amounts of tragedy. Shuttle TPS was an absolutely insane kludge in my humble opinion == and because it took so large a workforce to maintain it was the single most important factor in destroying shuttle's economic rationale.
What would have been better? I always thought some sort of "boil-off" system to transfer cool air from the topside of shuttle to the bottom would have worked. But I wasn't asked, and long afterwards it strikes me a boil-off system would have needed some hefty internal ducting, which would have weighed a whole bunch. Shuttle just wasn't big enough for that. So, maybe some sort of liquid sodium heat transfer, as with nuclear reactors… but those transfer heat to water, rather than air. Better materials… well some titanium got substituted for aluminum, but basically the aeroheating would have overwhelmed the titanium as well. So the tile TPS won the day. (Remember, this was to be a _reusable_ system. The Mercury-Gemini-Apollo tactic of ablating heat shields wasn't open.) For what it's worth, Chuck Blumer and Serge Waiter, the high powered science guys in my group, never proposed any radical alternative in my hearing, though they did grumble about what had to work with.
Also, the TPS was fragile, about like chalk. "What's going to happen if a bird hits this?" I asked an older friend one day, after looking at a blueprint. The TPS over the front edge of the wing was about two inches thick at the most, and there was a gap of about a foot between it and the actual metal of the wing. "This'll break." Not a problem, he said to reassure me. The air flow around the wing would be thick enough and fast enough that anything a bird would be swept aside. "But even so," I said, persisting. "What if? What about a flock of birds." "For that, we keep our fingers crossed," he said wryly. And you might guess how well that worked out.
The TPS system was reasonably heavy. Not quite as bad as rock or steel, but the tiles weigh about as much as an equal volume of … Lava hand soap, I wrote originally, but that's too heavy, say that the tiles weigh about the same as dog poop, and you'll have a better idea … and they cover a great amount of area. Which increased weight.
I've said "increased weight" a lot, you may have noticed. And guess what, another improvement got ordained: A pair of nice bulky solid rocket boosters to be used on each shuttle flight. Two boosters collectively weighing 4.4 million pounds, which would generate 4.5 miilion pounds of thrust, and hopefully get this ungainly mother of a system off the launch pad at about the speed of a dying sparrow…
Well, that was the last of it. There were refinements to be made, of course. New trajectories to evaluate, new thicknesses of the TPS to consider, new structuraL changes in the fuselage, and so on. But that's how the basic system evolved. And it was all a horrible kludge as I well remember and you can now see. But it did work, well enough. Despite OMB's adding a couple years of delay to the program, despite losing about a third of the initial development budget, shuttle got built.
So yeah, it was seriously compromised, and it wasn't as rugged as we'dd all hoped and it killed 14 good men and women during its brief (138 flights!) lifetime. And the economics sucked.
So, should we not have built it? Should we have done The Honorable Thing, and walked away from the job? Or kept secret diaries to reveal the Waste & Corruption when it got bad enough to justify whistle blowing? Well, nobody did, and along the way those who hadn't been aware of how life departed from textbook examples got some education in real world engineering.
Engineering involves compromises. The world's complex, the customer's stupid, the money coming in isn't what was budgeted, and even your coworkers will make the occasional wee error. You had in mind making something glorious, and you have to settle. Other hand, you learn mistakes to watch out for, you learn tactics that do work, you get experience, and with it an idea of what kinds of problems are serious and what kind minor, and so on. Everything will be better when we get to version 2.0 of the software!
Most of us working on shuttle saw it as a temporary thing. No, it wasn't perfect, but it filled the time, it kept people busy till Washington loosened up and got back to funding space programs properly again. Which, in the worst world anyone could concieve of, would be the later 1970?s. Meanwhile, shuttle was a start toward the reusuable spacecraft of the future, far from perfect given the constraints, but a start. In five years or maybe ten, Shuttle 2 would take its place, and in the decades after, Shuttle 3 and Shuttle 4 and …. That's how it goes in the aerospace business, my friend.
And 40 years after the Space Shuttle started development, we are all still waiting for Shuttle 2.
And that's the "dream world" in which in which shuttle got built, shuttle got compromised, and shuttle went to hell and gone. Nobody got together and purposefully designed from the start exactly as it turned out. It was the best we could do at the time, with the technology and the money we had, and no one expected it to be perfect, or to last so long that anyone coming later would think that we had though it perfect.
Any questions?
Comment by mike shupp — January 3, 2012 @ 10:19 pm
Shuttle was a design compromise–need shuttle 2–but it worked
Posted on January 7, 2012 by Bob
Dick Morris –
You're misreading me. I didn't say Rockwell had proposed a VTOL 2-stage version of the shuttle to NASA. As you note, it wasn't what NASA wanted, and as I noted, it couldn't have been built. VTO with Horizontal Land … that was doable, and if memory serves that's what everybody bidding on the shuttle contract proposed.
I wasn't there for that part of things. I started at Rockwell (still North American Rockwell, not yet Rockwell International, as I recall) in May of 1973. I was a college dropout, or maybe a flunkout, and even at the time it wasn't clear exactly which was the case. So I wasn't a "real engineer"; I was hired as an Engineering Aide II, an extra set of hands for a real engineer, to deal with scut work, and a week after I was hired I got shipped off to Tennessee for two weeks to help record the results from some wind tunnel tests at the Arnold Engineering Development Center — a little bit of the 21st century dropped down for some reason in the mid 20th century, in the midst of communities where most of the the technicians lived amongst farmers and people distilling whiskey and producing baseball bats from lengths of hickory…. Which was educational in a way, to learn I lived in such a country, never all one thing, but mixed up this way and that. And eventually I did become a Real Engineer, with a promotion to Member of the Technical Staff, a couple of years later when I was on the B-1 development program, and got a real college degree in Physics at West Coast University, which ain't famous but has served me well enough, along with GRE scores, to be considered for grad school, Which is another story.
My boss was a thin intense man named Mark Harthun. His boss was a genial hefty sort named John Lundgren, and I feel privileged in retrospect to havbe known them both. Harthun was a supervisor, he had two lead engineers and about a dozen other people, most of them new college grads. Lundgren had three groups, with about thirty people among them, and he was a manager — second rung on a ladder of 8 or 9 management rungs leading up to Vice President, if I remember correctly. Anyhow, we were the Aeroheating Group, and our job was to keep the shuttle from melting during ascent and reentry.
Chief problem on a VTOL system: re-entry and landing. You've got 25,000 fps of velocity to kill off, at a minimum, which takes a whole lot of fuel if you don't get aerodynamic braking. Which means wings or a wide fuselage, like a lifting body. Re-entry from the moon or planets, outside earth's orbit, gives you 35,000 fps or more velocity to deal with — twice the energy, twice the problem.
So Rockwell proposed a VTOHL system. The original design was a rwo-stage system: the booster was about the size of a football field, with a rather flattened oval cross-section; it had a two-man crew, if memory serves. The orbiter sat on the back of the booster, much as the real one did, but the proposal orbiter looked like an X-1, with small rectangular wings pretty much perpendicular to the fuselage at about 2/3?s the fuselage length. Think of a brick sitting on a one-foot ruler and you've got the idea. It was small, with maybe 15000 pounds of payload capability.
"And the air flow over it never goes turbulent," Joe Haney, one of the lead engineers said my first day, showing me a 6-inch wind tunnel model of the orbiter. "No matter what we test at, it doesn't go turbulent." Which didn't mean nearly as much then as it would later ….
What with this and that going on in Washington, some "improvements" were ordered. I think the flyback booster died first; the shuttle would be mounted to big expendable tanks instead — easier to design and cheaper to build, since a tank didn't need a human crew or engines or much else. This was supposed to be "temporary," and NASA would get around to building a real flyback booster as soon as funding was available. There was a war going on, and the government was running an unprecedented deficit, largely because of the cradle-to-grave socialism of the previous Democratic administration, which just about all the guys around me despised …. I'm sure you'll find this inconceivable.
Then NASA and the Air Force made an agreement, and the X1-like orbiter went away. The vehicle suddenly needed five times the crossrange (from 1250 miles to 6250 miles); it needed to carry 65,000 pounds of payload. Roughly speaking the cargo bay went from pickup truck sized to railroad car sized. So, bigger fuselage, wider and longer. Bigger wings to increase the vehicle's range — to slow down descent faster, to allow more steering during rentry. And bigger wings yet again, to deal with the orbiter's increasing size and weight…
And since the wings are larger, during reentry the boundary layer (the region where air goes from flowing "free" to not moving relative to the vehicle, and becomes hot as kinetic energy is transformed to thermal) is going to get thicker, and finally at last we did get turbulent flow on wind tunnels, implying we'd have turbulent flow in real flights….
Which was not good news, because a turbulent boundary layer is thinner than a laminar one. Air flow becomes chaotic. There's buffeting. And there's a lot more heat transfer to the aircraft structure. People write books about this stuff — the basic text on boundary layer theory was by a guy named Schlichting and for all I know it's still in print, there's not a lot of folks competing for fame in his corner of the churchyard.
Anyhow. The increased heating required bigger, tougher wings, which mweant more weight. It required some structural changes in the main fuselage, generally meaning more weight. All meaning changes in the vehicle's aerodynamics, meaning different return trajectories had to evaluated, and so on.
But of course the real fun thing that in was a tile-based Thermal Proctection System. People could write books about the TPS I rather think — really thick ones, like Russian novels, with — in the end –similar amounts of tragedy. Shuttle TPS was an absolutely insane kludge in my humble opinion == and because it took so large a workforce to maintain it was the single most important factor in destroying shuttle's economic rationale.
What would have been better? I always thought some sort of "boil-off" system to transfer cool air from the topside of shuttle to the bottom would have worked. But I wasn't asked, and long afterwards it strikes me a boil-off system would have needed some hefty internal ducting, which would have weighed a whole bunch. Shuttle just wasn't big enough for that. So, maybe some sort of liquid sodium heat transfer, as with nuclear reactors… but those transfer heat to water, rather than air. Better materials… well some titanium got substituted for aluminum, but basically the aeroheating would have overwhelmed the titanium as well. So the tile TPS won the day. (Remember, this was to be a _reusable_ system. The Mercury-Gemini-Apollo tactic of ablating heat shields wasn't open.) For what it's worth, Chuck Blumer and Serge Waiter, the high powered science guys in my group, never proposed any radical alternative in my hearing, though they did grumble about what had to work with.
Also, the TPS was fragile, about like chalk. "What's going to happen if a bird hits this?" I asked an older friend one day, after looking at a blueprint. The TPS over the front edge of the wing was about two inches thick at the most, and there was a gap of about a foot between it and the actual metal of the wing. "This'll break." Not a problem, he said to reassure me. The air flow around the wing would be thick enough and fast enough that anything a bird would be swept aside. "But even so," I said, persisting. "What if? What about a flock of birds." "For that, we keep our fingers crossed," he said wryly. And you might guess how well that worked out.
The TPS system was reasonably heavy. Not quite as bad as rock or steel, but the tiles weigh about as much as an equal volume of … Lava hand soap, I wrote originally, but that's too heavy, say that the tiles weigh about the same as dog poop, and you'll have a better idea … and they cover a great amount of area. Which increased weight.
I've said "increased weight" a lot, you may have noticed. And guess what, another improvement got ordained: A pair of nice bulky solid rocket boosters to be used on each shuttle flight. Two boosters collectively weighing 4.4 million pounds, which would generate 4.5 miilion pounds of thrust, and hopefully get this ungainly mother of a system off the launch pad at about the speed of a dying sparrow…
Well, that was the last of it. There were refinements to be made, of course. New trajectories to evaluate, new thicknesses of the TPS to consider, new structuraL changes in the fuselage, and so on. But that's how the basic system evolved. And it was all a horrible kludge as I well remember and you can now see. But it did work, well enough. Despite OMB's adding a couple years of delay to the program, despite losing about a third of the initial development budget, shuttle got built.
So yeah, it was seriously compromised, and it wasn't as rugged as we'dd all hoped and it killed 14 good men and women during its brief (138 flights!) lifetime. And the economics sucked.
So, should we not have built it? Should we have done The Honorable Thing, and walked away from the job? Or kept secret diaries to reveal the Waste & Corruption when it got bad enough to justify whistle blowing? Well, nobody did, and along the way those who hadn't been aware of how life departed from textbook examples got some education in real world engineering.
Engineering involves compromises. The world's complex, the customer's stupid, the money coming in isn't what was budgeted, and even your coworkers will make the occasional wee error. You had in mind making something glorious, and you have to settle. Other hand, you learn mistakes to watch out for, you learn tactics that do work, you get experience, and with it an idea of what kinds of problems are serious and what kind minor, and so on. Everything will be better when we get to version 2.0 of the software!
Most of us working on shuttle saw it as a temporary thing. No, it wasn't perfect, but it filled the time, it kept people busy till Washington loosened up and got back to funding space programs properly again. Which, in the worst world anyone could concieve of, would be the later 1970?s. Meanwhile, shuttle was a start toward the reusuable spacecraft of the future, far from perfect given the constraints, but a start. In five years or maybe ten, Shuttle 2 would take its place, and in the decades after, Shuttle 3 and Shuttle 4 and …. That's how it goes in the aerospace business, my friend.
And 40 years after the Space Shuttle started development, we are all still waiting for Shuttle 2.
And that's the "dream world" in which in which shuttle got built, shuttle got compromised, and shuttle went to hell and gone. Nobody got together and purposefully designed from the start exactly as it turned out. It was the best we could do at the time, with the technology and the money we had, and no one expected it to be perfect, or to last so long that anyone coming later would think that we had though it perfect.
Any questions?
Comment by mike shupp — January 3, 2012 @ 10:19 pm
Sent from my iPad
No comments:
Post a Comment