-
#120
by
Downix
on 18 Apr, 2011 04:20
-
To bring this back to the RL-10, to manufacture an RL-10 requires a lot more setup than a mask and die change. Many of the tools used for it are customized, and the skills to operate it similarly custom. You cannot take, say, the tools needed to make the primary LOX turbine impeller and swap them easily to manufacture something else.
-
#121
by
jimgagnon
on 18 Apr, 2011 05:10
-
To bring this back to the RL-10, to manufacture an RL-10 requires a lot more setup than a mask and die change. Many of the tools used for it are customized, and the skills to operate it similarly custom. You cannot take, say, the tools needed to make the primary LOX turbine impeller and swap them easily to manufacture something else.
I agree that the comparison with semiconductor manufacturing has some value. Both aerospace and chip fab are close to the edge of what's physically possible with today's technology. However, the value lies in the differences: mem fab is actually a good one. Chip makers are loath to throw anything away; the more waste in a process, the more the waste in materials but more importantly the lost time in making the garbage in the first place. So modern memory chips have a testing procedure that, when it detects portions of a chip that are just bad, can blow fuses so that part of the chip is inaccessible. Now you have a good if not somewhat smaller chip, which you can still sell. I'm no LOX turbine impeller machinist, but it's my impression that in aerospace if you see a defect in a component, you guys melt it down and try again, with the resulting extra cost of labor and materials, not to mention what it may do to a schedule.
I hate to keep bringing up SpaceX but it's not a mistake that Musk's other company builds a car. With the demands of performance, emissions and the price of fuel, automobiles are a high tech, mass produced product. As an example let's assume that you could get the car and rocket engine guys in the same room, and the two agree on an impeller design that worked well in both a LOX pump and a turbocharger. So, when mass producing these impellers the rocket guys get the best of the best, the zero defect parts and the car guys get parts at are good enough for the demands of a car. No waste in labor or materials, and mass production quantities the rocket guys can only dream of.
If a common design isn't possible, the engineers could at least agree to a common fabrication plant, metal choice, toolsets and operator skills so that the same line can be used to create the impellers both teams need. Save money that way, and makes the car parts better in the process.
Aerospace components have to be of the highest quality; you can't call out for a RMA at Max Q 11km up because a component fails. However, greater commonality with mass production can not only reduce component cost but help with some of the issues of low production assemblies.
-
#122
by
Downix
on 18 Apr, 2011 05:23
-
To bring this back to the RL-10, to manufacture an RL-10 requires a lot more setup than a mask and die change. Many of the tools used for it are customized, and the skills to operate it similarly custom. You cannot take, say, the tools needed to make the primary LOX turbine impeller and swap them easily to manufacture something else.
I agree that the comparison with semiconductor manufacturing has some value. Both aerospace and chip fab are close to the edge of what's physically possible with today's technology. However, the value lies in the differences: mem fab is actually a good one. Chip makers are loath to throw anything away; the more waste in a process, the more the waste in materials but more importantly the lost time in making the garbage in the first place. So modern memory chips have a testing procedure that, when it detects portions of a chip that are just bad, can blow fuses so that part of the chip is inaccessible. Now you have a good if not somewhat smaller chip, which you can still sell. I'm no LOX turbine impeller machinist, but it's my impression that in aerospace if you see a defect in a component, you guys melt it down and try again, with the resulting extra cost of labor and materials, not to mention what it may do to a schedule.
To discuss semiconductors a bit, we had a major problem with one of our supplied components. We learned how it failed, and we created a pair of chips which fixed the issue, an external solution added after the fact. Of course, this would not work on a rocket engine, can't just add another turbomachine assembly when the primary LOX supply line looses pressure 50 seconds into a flight.
-
#123
by
Robotbeat
on 18 Apr, 2011 05:37
-
To bring this back to the RL-10, to manufacture an RL-10 requires a lot more setup than a mask and die change. Many of the tools used for it are customized, and the skills to operate it similarly custom. You cannot take, say, the tools needed to make the primary LOX turbine impeller and swap them easily to manufacture something else.
It's not impossible, though. Modern CNC machine tools are getting more flexible, not less. You can print regeneratively cooled rocket engines in 3D, for instance. The same machine can (and does) make all sorts of other things in 3D (during the same day or even shift...). This is the direction manufacturing is going, more flexibility with less cost in changing the tooling (less need to change the tooling). So, even less set-up than in the silicon fab industry.
Whatever replaces the RL-10 will use less specialized manufacturing equipment. I think that many of the very small aerospace start-ups don't have their own big machine shop; they order parts digitally from a machinist (perhaps in a totally different part of the country), and they receive their rocket engine via fedex or whathaveyou. This sort of horizontal integration needs to become more common. Flightweight components with the right material compatibility are very expensive right now if you go the traditional aerospace route. The smaller companies end up looking for short-cuts... parts used in other industries that happen to work good enough for rocket engines, though they often have to make their own solution to the low-cost, flightweight, material compatibility optimization problem. Many of them are starting to offer these solutions to other small aerospace companies. This sort of horizontal integration needs to happen more often, not less. That XCor is offering their experience in medium-sized (initially soft-)cryogenic propellant pumps as part of a cooperative venture with ULA is evidence of this.
XCor has actually a lot of experience testing engines repeatedly, and is willing to sell engines or even whole vehicles to other companies to operate, instead of having the vehicle manufacturer be also the operator. In spite of the seeming success of SpaceX's vertical integration, one sign of a healthy market is horizontal integration, IMHO. If you try to horizontally integrate in an unhealthy market, you tend to go out of business. That's why Orbital is using foreign parts when they're doing horizontal integration but SpaceX can do it domestically with vertical integration: the liquid rocket engine market in the US has been stagnant. Elon had initially talked about doing horizontal integration when he started SpaceX, but now are the poster boy for vertical integration because of the fact that it's just SO expensive to do horizontal integration in the domestic US rocket market. These small guys like Armadillo Aerospace and XCor are doing all they can to improve it, and there are signs they are succeeding, though there's a long, long ways to go.
-
#124
by
Downix
on 18 Apr, 2011 05:59
-
Actually was just discussing this the other day. You're quite right, with the rise of 3D printing, additive manufacturing, the whole dynamic of the industry is turned on its head. It's not there yet, but it's getting there very rapidly. I have worked with 3D printers before, and the potential is absolutely mind boggling.
-
#125
by
MP99
on 18 Apr, 2011 07:28
-
Go read an interview from Jim Maser where he admitted the RL10 price quote was obscenely high to prove a point to the government on the dire state of the liuid engine industry. As much as the vox populi amateurs are trying to make something nefarious out of this, it's all quite open. PWR is as peeved about the state of the industry as you are.
PWR has only known of the end of the SSME / overhead sustainment contract since the FY11 Obama budget was announced in February 2010. Prior to that it was planned by both NASA and PWR as a follow-on to the SSME contract that would develop J2X and the liquid booster engine for some HLV.
Since no-one has linked these:-
http://www.reuters.com/article/2011/03/07/us-unitedtechnologies-space-idUSTRE72663L20110307http://www.spacenews.com/civil/110204-engine-costs-drive-atlas5-prices.htmlLots of downsizing planned and going on...
Jim Maser, president of Pratt & Whitney Rocketdyne, said his company already had plans to close half its office and factory space over the next three years, and might have to lay off hundreds of employees unless NASA mapped out a shuttle successor plan within the next four to eight months.
The United Technologies Corp unit also winnowed its list of suppliers from 600 to 200 over the past year, Maser told reporters, underscoring his commitment to cutting costs and making space systems more affordable.
cheers, Martin
-
#126
by
Bernie Roehl
on 18 Apr, 2011 11:09
-
http://www.spacenews.com/civil/110204-engine-costs-drive-atlas5-prices.html
Thanks for posting that, Martin.
Some interesting quotes:
prices for the Russian-built RD-180, a liquid oxygen/kerosene engine that powers the Atlas 5’s main stage, are also on the rise. The RD-180, the single most expensive Atlas 5 component, is built by NPO Energomash of Khimki, near Moscow.
[...]
Sowers said the Russians are simply charging what the customer is willing to pay. “The Russians have learned over the years the wonders of capitalism, and have become really good at letting the market drive their prices up,” he said.
Yes, yes indeed. Not that we have any capitalists here as well. No, of course not. That couldn't possibly be the reason for the recent price increases.
between 1999 and 2010, the cost to launch an Atlas 5 rocket was between $100 million and $125 million. Under NLS 2, which NASA signed with ULA and other rocket manufacturers in September, the Atlas 5 cost range grew to between $102 million and $334 million per mission.
$334 per mission!Norman cautioned there are risk factors that could drive Atlas 5 prices even higher than the upper range of the NLS 2 prices [...] A failure to reach agreement, Norman said, could raise NASA’s Atlas 5 costs by as much as $140 million per mission as ULA’s overhead costs are swept into individual launch contracts.
So we're talking what, up to
$474 million for an Atlas V launch?
It also appears that ULA is making the launch vehicle heavier, therefore reducing its performance:
We’ve lost about 400 to 500 kilograms of performance in terms of what is being offered.
*sigh*
-
#127
by
Bernie Roehl
on 18 Apr, 2011 11:48
-
Modern CNC machine tools are getting more flexible, not less. You can print regeneratively cooled rocket engines in 3D, for instance. The same machine can make all sorts of other things in 3D. This is the direction manufacturing is going, more flexibility with less cost in changing the tooling.
Whatever replaces the RL-10 will use less specialized manufacturing equipment.
Agreed. Of course, an engine would have to be designed from the outset to take advantage of these new manufacturing techniques. You couldn't simply print up an RL-10 from existing blueprints, for example. (I know you know this already, Robotbeat -- I'm just mentioning it in case anyone thinks that would be a possibility).
Of course, the RL-10 was designed half a century ago. Given the recent tenfold increase in pricing, it appears its days are numbered. ULA is looking at alternatives (though not quickly enough to remain competitive, I suspect), and they're the only customer for the engine.
-
#128
by
Jim
on 18 Apr, 2011 11:59
-
the two agree on an impeller design that worked well in both a LOX pump and a turbocharger.
If a common design isn't possible, the engineers could at least agree to a common fabrication plant, metal choice, toolsets and operator skills so that the same line can be used to create the impellers both teams need. Save money that way, and makes the car parts better in the process.
Not feasible. There are many requirements that are incompatible.
-
#129
by
Chris Bergin
on 18 Apr, 2011 13:05
-
Incorrect (i love this auto-anti profanity filter)
So you're confrontational, distruptive and admit to using profanity. Classy.
If people want to post on this site, you're going to be civil, or I'm going to kick you the incorrect off this site.
-
#130
by
edkyle99
on 18 Apr, 2011 14:12
-
http://www.spacenews.com/civil/110204-engine-costs-drive-atlas5-prices.html
$334 per mission!
EELV costs have long been much higher than commonly believed. Note that the article says that the Launch Capabilities Contract covers another $140 million per launch that is not usually quoted as a launch cost! In other words, even the most basic Atlas V launch has always cost well more than $200 million when the true program costs of the launch complex and factories, etc, were included. Delta IV too.
- Ed Kyle
-
#131
by
Antares
on 18 Apr, 2011 17:42
-
No. That's only since the ELC contract has been in place, and the ELC contract is with the DoD not NASA. The share of the ELC for non-DoD customers is added to the price of the launch contract or task order, be it another government customer or a commercial customer.
-
#132
by
edkyle99
on 18 Apr, 2011 21:14
-
No. That's only since the ELC contract has been in place, and the ELC contract is with the DoD not NASA. The share of the ELC for non-DoD customers is added to the price of the launch contract or task order, be it another government customer or a commercial customer.
ELC has been in place since, what, 2005? Prior to that the costs were eaten by the contractors, presumably, but the costs still existed.
It doesn't matter if DoD pays or NASA pays or if the contractor writes it off. The end result is still the same. It is money spent to launch rockets - and the grand total is a lot more than the usually-quoted sums.
- Ed Kyle
-
#133
by
Jim
on 18 Apr, 2011 23:34
-
No. That's only since the ELC contract has been in place, and the ELC contract is with the DoD not NASA. The share of the ELC for non-DoD customers is added to the price of the launch contract or task order, be it another government customer or a commercial customer.
ELC has been in place since, what, 2005? Prior to that the costs were eaten by the contractors, presumably, but the costs still existed.
It doesn't matter if DoD pays or NASA pays or if the contractor writes it off. The end result is still the same. It is money spent to launch rockets - and the grand total is a lot more than the usually-quoted sums.
- Ed Kyle
No, when NASA or commercial customer quotes a price it includes ELC.
-
#134
by
Robotbeat
on 19 Apr, 2011 02:13
-
One more significant player in the field of hydrolox engines:
http://forum.nasaspaceflight.com/index.php?topic=10685.msg726156#msg726156
Blue Origin also proposes to speed development of its Reusable Booster System through accelerated testing of its 100,000 lbf liquid oxygen/liquid hydrogen (LOx/LH2) engine. Development of Blue Origin's restartable, deep-throttle engine is pacing the entire orbital RBS program. Under CCDev 2, Blue Origin proposes to test the full-scale thrust chamber at NASA's Stennis Space Center and, optionally, perform development testing of the engine's fuel and oxidizer turbopumps.
Blue Origin requests $10,400,000 in NASA funding for the RBS Engine Risk Reduction Project with the possibility of an additional $3,000,000 for optional milestones. Partnering with NASA will not only accelerate the Reusable Booster System; it will also speed development of a low-cost LOx/LH2 engine suitable for a variety of other upper stage applications and deep-throttling exploration missions.
They won the $10,400,000, and it should get them through pressure-fed testing of the full-scale thrust chamber assembly (TCA) by May, 2012.
It's been known for a while that Blue Origin sucked up a lot of hydrolox propulsion experts, and this is the fruit. 100,000lbf hydrolox deeply throttleable, low-cost engine (supposed to be reusable, too). PWR, watch out!
-
#135
by
Antares
on 19 Apr, 2011 02:16
-
Hmmm, I wonder if Musk and Mueller would swallow their pride and use someone else's engine.... (jeopardy music).... No, I think they'd buy the company instead.
-
#136
by
Namechange User
on 19 Apr, 2011 02:17
-
What? LOX/LH2? Stupid people. Didn't they get the buzz memo?
-
#137
by
Robotbeat
on 19 Apr, 2011 02:45
-
What? LOX/LH2? Stupid people. Didn't they get the buzz memo?
It's quite true that the "New Space" companies (with the notable exception of Blue Origin) have, until quite recently, been pretty prejudiced against hydrogen as a rocket fuel. They are "warming up" to hydrogen, now, as evidenced by XCor and ULA's joint project and our Jon Goff's Altius Space Machines and ULA joint project (Jon Goff has also done a few papers extolling the virtues of a hydrolox-based depot).
-
#138
by
Nate_Trost
on 19 Apr, 2011 13:17
-
Hmmm, I wonder if Musk and Mueller would swallow their pride and use someone else's engine.... (jeopardy music).... No, I think they'd buy the company instead.
Bezos net worth is $18 billion. If there was an improbable buyout scenario, it would be that he'd buy SpaceX if Musk ran into trouble, not that he'd sell Blue Origin.
-
#139
by
Antares
on 19 Apr, 2011 13:48
-
Musk will never sell SpaceX. The only way he leaves is if he goes public and somehow loses a majority of shares. But the subject of this thread is not SpaceX and Blue Origin.