Transterrestrial Musings

Comments

It seems to me that current spaceflight methods are like efforts in the late 1800's which tried to achieve winged flight with steam power.

If the limits of the technology were approached, it was theoretically possible, but difficult and not economical.

Any technology in which a "good" performance is one in which 90% of your to-orbit vehicle is fuel obviously has limits.

I don't think that we'll get the inexpensive, regular and safe access to space you're looking for until someone develops a better solution, just a the Wright brothers succeeded using an internal combustion engine, where previous steam efforts had, to be generous, met with limited success.

Posted by DocBrown at September 1, 2006 06:20 AM

Any technology in which a "good" performance is one in which 90% of your to-orbit vehicle is fuel obviously has limits.

It does, but we are far from them. The cost of propellant is less than one percent of the cost of launch, so mass fraction is not the reason that costs are high.

Posted by Rand Simberg at September 1, 2006 06:23 AM

It does, but we are far from them. The cost of propellant is less than one percent of the cost of launch, so mass fraction is not the reason that costs are high.

It's not the cost of propellant that is the "punisher" from the rocket equation--it's what a 90% mass fraction does to everything else on your spacecraft. Structures, engines, issues with reusability, etc. I agree with the first poster...we're stuck with the wrong technology. Problem is, there may never be a better one.

Posted by rocketEquation at September 1, 2006 06:49 AM

Nope, sorry.

Launch costs aren't high because of mass fraction. They're high because we throw the vehicle away, and have a low flight rate. It is not a technology problem--it's a financing and market problem.

Posted by Rand Simberg at September 1, 2006 07:15 AM

Launch costs ... [are] high because we throw the vehicle away, and have a low flight rate.

A high flight rate is absolutely necessary. Unlike Jon Goff though, I don't believe that reusability is a good solution.

Look at what it costs to inspect and refurbish the space shuttle main engines on every flight. It probably costs more than it would to simply build new ones each time - expendable SSMEs would allow production line methods could be brought to bear rather than the craftwork involved now.

But flight rates do indeed matter a lot. The development costs are amortized over the number of flights, so it would be the same whether there is a fleet of five doing 50 launches each or a fleet of 250 doing one launch each. With reusables (the fleet of five), you are stuck with a design that must be set at the beginning, and then you have to maintain those vehicles between bouts of the wear and tear of space travel. With expendables (the fleet of 250), one can iteratively improve the design, like the Russians have with Soyuz.

The secret to bringing costs down isn't to design the super-duperest spacecraft. The secret lies in bringing production line methods into the equation.

Either that, or change paradigm altogether, and go with a space elevator.

Posted by Ed Minchau at September 1, 2006 07:45 AM

Look at what it costs to inspect and refurbish the space shuttle main engines on every flight.

Which has nothing to do with what it would cost to operate a well-designed reusable engine. You cannot use a single data point (Shuttle) to draw general conclusions about reusable vehicles. That's the fallacy of hasty generalization.

Posted by Rand Simberg at September 1, 2006 07:48 AM

Rand, your comment about a "well-designed reusable engine" sounds like someone telling me that the space elevator is a perfect idea, we just need to work out some minor engineering details.

"You cannot use a single data point (Shuttle) to draw general conclusions about reusable vehicles."

So you're asserting that we should commit to reusables because we don't know how well they work?

Posted by DensityDuck at September 1, 2006 08:35 AM

...your comment about a "well-designed reusable engine" sounds like someone telling me that the space elevator is a perfect idea, we just need to work out some minor engineering details.

We don't have the materials for a space elevator. This is an absurd analogy. We do in fact have robust rocket engines (e.g., RL-10) that can be made quite operable and reusable. The problem with the SSME was that it had no margin, because the oversizing of the Shuttle drove it to extremes in performance requirements. I repeat, you cannot use a single data point to draw grand conclusions.

So you're asserting that we should commit to reusables because we don't know how well they work?

No, I'm asserting that we should be continuing to push on them because their potential is so huge, and we have no good reasons to think they won't work (a single flawed example doesn't constitute a good reason), and many examples of expendables over almost five decades have shown them to be intrinsically unreliable and expensive.

Posted by Rand Simberg at September 1, 2006 09:09 AM

"We don't have the materials for a space elevator. This is an absurd analogy."

Yes. Yes, it is an absurd analogy. That's the whole point. Is there something about modern life that just completely scours away people's ability to reach a logical conclusion through inductive reasoning?

It seems I need to spell this out for you. You say that we just need a "well-designed RLV engine", as though it were a trivial thing to create. Meanwhile, space-elevator advocates say that we just need to figure out how to felt solid diamond into a 10,000-km elastomer, as though that were a trivial thing to do. Are you seeing the similarity here?

"No, I'm asserting that we should be continuing to push on them because their potential is so huge..."

Again, this kind of rhetoric is heard just as loudly from the space-elevator group. "If this works, it will be so great! So we should keep working on it, instead of thinking about what else we might do!"

"...we have no good reasons to think they won't work (a single flawed example doesn't constitute a good reason)..."

Ahem. We are not complete morons. We are capable of doing the same design studies as anyone else, and those studies show that with current technology an RLV will not have as high a mass fraction as economical space access would require. The earlier commenter said it best--it didn't matter how smart people were in the 19th century, or how hard they worked, or how great airplanes would have been back then. It was not possible to build them and have them work.

"...many examples of expendables over almost five decades have shown them to be intrinsically unreliable and expensive."

Unreliable? By what standard? And how is an RLV going to be more reliable? It will, after all, be based on the same technology as existing expendables.

Posted by DensityDuck at September 1, 2006 11:41 AM

[Foolish and inappropriate condescension snipped]

You say that we just need a "well-designed RLV engine", as though it were a trivial thing to create.

No, I say it as though it's a possible thing to create. One has to spend some money on engine development, and do testing, but there are no fundamental barriers--we know how to do it. There is no similarity with space elevators, which is a fundamental strength of materials problem.

We are capable of doing the same design studies as anyone else, and those studies show that with current technology an RLV will not have as high a mass fraction as economical space access would require.

What studies are those? I've participated in a number of such studies, and have never seen such a result. In fact, there's no obvious and consistent relationship between cost per flight and mass fraction. It's at best a second-order factor.

Unreliable? By what standard?

By any reasonable standard of any other form of transportation. Only two nines of reliability is atrocious by any standard except the one that we've allowed ourselves to accept with expendable launch vehicles.

And how is an RLV going to be more reliable?

By flying a lot, wringing out the bugs, and not having infant mortality issues caused by the fact that every flight of an expendable is a first flight.

It will, after all, be based on the same technology as existing expendables.

Why do you persist in the mistaken belief that this is a technology problem, or that the low reliability of expendables is a technology problem? It is not.

Posted by Rand Simberg at September 1, 2006 11:51 AM

If the low reliability of expendables is not a technology problem, then what is the problem? And how is that problem not going to be applicable to RLV? And if you can't understand analogy, then why am I even talking to you?

Posted by DensityDuck at September 1, 2006 12:44 PM

If the low reliability of expendables is not a technology problem, then what is the problem?

The low reliability of expendables is a technology problem. One that will be solved by reusables, for reasons already explained.

And if you can't understand analogy, then why am I even talking to you?

I didn't say I can't understand analogy. In fact, it would seem to be you who has problems with it, since you make such poor ones.

I'm just wondering why I'm arguing about space technology (something I do for a living) with someone (who seems to know nothing about it) named "DensityDuck."

Posted by Rand Simberg at September 1, 2006 01:00 PM

> We don't have the materials for a space elevator. This is an absurd analogy. We do in fact
> have robust rocket engines (e.g., RL-10) that can be made quite operable and reusable

And XCOR has built engines that require much less maintenance than the RL-10.

Before declaring something theoretically impossible, posters should check to see if it's already been done.

Posted by Edward Wright at September 1, 2006 01:01 PM

> If the low reliability of expendables is not a technology problem, then what is the problem?

Learning-curve effects. *Every* vehicle has some reliability problems on its first flight.

> And how is that problem not going to be applicable to RLV

Because reusable vehicles will have a *second* flight. And a third. And a twenty-third.

With expendables, every flight is a test flight.

Posted by Edward Wright at September 1, 2006 01:04 PM

> The development costs are amortized over the number of flights, so it would be the same whether
> there is a fleet of five doing 50 launches each or a fleet of 250 doing one launch each.

According to whom? General Dynamics studied reusable and expendable rockets of similar performance (the X-15 and Atlas A). They found that reusable vehicles had 40% lower development costs.

Even if development costs were the same, you still have to consider *production* costs. To launch 250 ELVs a year, you have to *build* 250 ELVs a year. In the real world, no one can afford that.

The marginal cost of flying a reusable vehicle is propellant plus labor and maintenance. The marginal cost of flying an ELV is the manufacturing cost plus propellant plus labor and maintenance. (Yes, vehicles *do* require maintenance even before the first flight.)

> With reusables (the fleet of five), you are stuck with a design that must be set at the beginning, and
> then you have to maintain those vehicles between bouts of the wear and tear of space travel. With expendables
> (the fleet of 250), one can iteratively improve the design, like the Russians have with Soyuz.

What makes you think the new Soyuzes are greatly improved? Or that it's impossible to upgrade a reusable vehicle?

Today's B-52s are not the same as those built 50 years ago? The Basler turboprop conversion (DC-3) is not the same as the original. Both are vastly improved. In the aircraft world, this is done all the time.

Posted by Edward Wright at September 1, 2006 01:26 PM

We had well designed reusable engines over 40 years ago. The Apollo's F-1 was a hellaciously robust engine that would make an SSME cry with envy. The fact that they were thrown away with each flight was no comment on their extreme robustness.

There are several engines we currently discard that have more reusability potential than the SSME's.

Rand mentioned the RL-10. There are also the RD-170/180 series as well.

Comparing the SSME to an RL-10 is like comparing a NASCAR engine to the engine in a Honda Accord.

One is rebuilt everytime you run it 500 miles because it is meant to be run at redline. The other runs 150,000 to 250,000 miles without being rebuilt.

Yet both engines are far more similar than dissimilar.

Posted by Mike Puckett at September 1, 2006 03:27 PM

Even if development costs were the same, you still have to consider *production* costs. To launch 250 ELVs a year, you have to *build* 250 ELVs a year. In the real world, no one can afford that.

You're right. Nobody can afford that - not in the handcrafted way such things are done now. But when you've got production lines making them, all of a sudden it does become affordable.

Before Henry Ford, all cars were made by hand. After Henry Ford, average people could afford cars. We're still in the BHF era of space travel.

Posted by Ed Minchau at September 1, 2006 08:01 PM

You're right. Nobody can afford that - not in the handcrafted way such things are done now. But when you've got production lines making them, all of a sudden it does become affordable.

The Russians have been doing that for decades. Their rockets remain very expensive, and very unreliable.

Posted by Rand Simberg at September 1, 2006 08:23 PM

> Before Henry Ford, all cars were made by hand. After Henry Ford, average people could afford cars. We're
> still in the BHF era of space travel.

Nice rhetoric. Does it mean anything?

The machinery used to assemble rockets would blow Henry Ford's mind. The techniques used in rocket factories are similar to those used in aircraft factories.

If aircraft-like manufacturing techniques are what make space travel expensive, then they should make air travel almost equally expensive. But they don't.

Nor does anyone think it would be cheaper for Boeing to build a new 737 every time Southwest wants to fly from Chicago to New Orleans.

Even Henry Ford didn't suggest motorists throw their car away after every trip, just so he could build more automobiles.

Posted by Edward Wright at September 1, 2006 09:05 PM

Launch costs aren't high because of mass fraction. They're high because we throw the vehicle away, and have a low flight rate.

More the low flight rate than anything else. If we were churning out hundreds of RD-171's a year and miles of friction stir welded tanks we could get the cost of a 15 tonne launch down to about $25m.

Beyond that, yes, we'll need to make the vehicle reusable.

Posted by Chris Mann at September 1, 2006 10:51 PM

rocketEquation:

"I agree with the first poster...we're stuck with the wrong technology. Problem is, there may never be a better one."

Bullshit. A better technology was being developed in 1962, until politics killed it. One capable of getting a million tons to Mars in a couple of weeks. And now the name has been hijacked, possibly at least partially to bury the idea.

The name? Orion.

Posted by Fletcher Christian at September 2, 2006 05:18 AM

I'd like to see the 400m diameter, 8000000T Super Orion built.

Posted by Chris Mann at September 2, 2006 10:16 AM

There'd need to be a market for a reusable vehicle.

Posted by mz at September 3, 2006 01:14 PM

There is a humungous market for multiple reusable vehicles, if the cost per flight is low enough and the reliability high enough. Richard Branson is already starting to tap it.

Posted by Rand Simberg at September 3, 2006 01:30 PM

The Shuttle makes a poor case model as a reusable launch vehicle because in actuality it more resembles a rebuildable vehicle. It would have been fine for the engines to run as hard as they do and require such extensive rebuilding between launches. Only problem is the engine bay was designed with the mindset that the engines would be services every 20 flights not every flight. Those SSME pods are tight and constrictive for crews to work in effeciently. True engine interchangibility wasn't implemented in the design. This duality of concepts between the engines low tolerances and the orbiters poor engine pod accessibility I'm certan would contribute to higher maintenance costs. As well as being one of a long list of items that all make the Shuttle a poor case for a robust reusable vehicle design.

Posted by Josh Reiter at September 3, 2006 07:32 PM

Post a comment