The "Real Aerospace" Thread

[Mjolnir]

hi hi

Well, when you put it like that, it sounds basic and easy.

It's SpaceX. They aren't pursuing cost reduction via the use of ultra-complex supertech, that's Reaction Engines (and pretty much every other spaceplane and SSTO project that's ever existed). They like "basic and easy" wherever they can get it.

Scaling up powered landings to the first stage of an orbital launch system (and not even a small one) is far from trivial...the powered reentry with the thruster firing into a hypersonic airstream in particular is something nobody's ever tried before...but the basic technology required has existed for decades. That doesn't make it less of a big deal that someone's now actually doing it...the launch industry's been stuck in a rut, doing little other than maintaining the status quo ever since the Shuttle replaced the Saturn V. SpaceX has already achieved a huge reduction in launch costs, and first stage reuse will significantly improve their ability to do so. They're aiming for an eventual order of magnitude reduction with a future fully-reusable system...that's going to make a big difference.

[icekatze]

hi hi

I imagine it is easy to reduce costs when one doesn't need to have a large research and development department for things like designing new rocket motors. Space X is doing some cool things, and I am trying to give them some credit where it is due here, but they have managed to get a lot of cost reductions by getting free research and development from other sources. Like on the development of their rocket motor.

But time will be the ultimate judge of whether or not this works. The space shuttle was supposed to be a reusable cost saving measure. When they planned it out, they figured on the costs of maintenance, of replacing various parts and thought it would be cost effective, but it turned out to be an order of magnitude more costly than they anticipated.

If Space X is able to significantly undercut their competition with a rocket that is comparable to their competition, then I expect they should eventually come to dominate the market, and none of what I say will make much of a difference, but for now out of 130 some launches worldwide this year, they launched fewer than 10.

[Siber]

One thing that amazes me about a lot of SpaceX re-usability detractors(not aimed at anyone in this thread, just a general comment), is that their payload prices are extremely competitive already, and that's with rockets weighed down by equipment and complexity designed for reuse but sold as expendable. If for some reason reuse is a total bust they could strip out the legs, fins, and I imagine some engine gimbals capability, and launch using all their fuel instead of reserving some for boostback and landing, and have an even cheaper launcher with even larger payload.

[Absalom]

hi hi

I imagine it is easy to reduce costs when one doesn't need to have a large research and development department for things like designing new rocket motors. Space X is doing some cool things, and I am trying to give them some credit where it is due here, but they have managed to get a lot of cost reductions by getting free research and development from other sources. Like on the development of their rocket motor.

And yet, if they can do this, so can anyone else currently designing a launch system. It's not as if NASA, for example, actually needs to redesign the motors every time it designs a new rocket.

But time will be the ultimate judge of whether or not this works. The space shuttle was supposed to be a reusable cost saving measure. When they planned it out, they figured on the costs of maintenance, of replacing various parts and thought it would be cost effective, but it turned out to be an order of magnitude more costly than they anticipated.

It would really be preferable to compare it with something that didn't suffer from such decisive feature creep. The original mission profile for the space shuttle could have been accomplished by a horizontal launch & landing system based on an almost entirely conventional airframe, but the addition of polar orbiting for military launches apparently forced a flyback capability equivalent to the circumference of the Earth. The choice is understandable: NASA was afraid that Nixon or one of his successors would cancel the shuttle like they cancelled Apollo. None-the-less, it makes the shuttle an orange compared to most other launch system's apple.

[icekatze]

hi hi

I could be wrong about this, but to my knowledge, NASA has not been sharing their rocket motor blueprints with the Russians or the Chinese. There are some very real concerns with launching rockets with a large number of small motors.

I mean, I am really hoping that low cost space launches will become a reality, but there are still a lot of hurdles left to go. NASA tried to get the Shuttle's main engines reusable for 55 flights, and after years and huge amounts of money, they still weren't able to achieve that goal with all parts of the engine.

“The most fundamental breakthrough is a rapidly and fully reusable rocket. It doesn't help if it's partially reusable.”

- Elon Musk

[Mjolnir]

hi hi

I could be wrong about this, but to my knowledge, NASA has not been sharing their rocket motor blueprints with the Russians or the Chinese. There are some very real concerns with launching rockets with a large number of small motors.

I mean, I am really hoping that low cost space launches will become a reality, but there are still a lot of hurdles left to go. NASA tried to get the Shuttle's main engines reusable for 55 flights, and after years and huge amounts of money, they still weren't able to achieve that goal with all parts of the engine.

“The most fundamental breakthrough is a rapidly and fully reusable rocket. It doesn't help if it's partially reusable.”

- Elon Musk

SpaceX puts a great deal of money into R&D. They aren't just flying NASA-designed hardware with a SpaceX logo slapped on the outside. They've had access to a great deal of NASA data and expertise, but Merlin and the rest of the Falcon 9 are their own design, and they are doing things that even NASA doesn't have any experience with, such as the 9-engine cluster and supersonic retrofire followed by flyback and landing of an entire first stage booster, and even groundbreaking research into adapting CFD simulations for acceleration with GPU hardware. Even the PICA-X heat shield material they use on the Dragon has been heavily modified for manufacturability, cost reduction, and improved reusability.

The SSMEs were extremely complex high-pressure hydrogen-fueled engines, and as with the rest of the Shuttle (and its replacement, the SLS), there was no drive to make them more economical. The Merlins perform multiple test fires (and some perform multiple firings in each flight) without any refurbishing required. It isn't some theoretical possibility, it's part of normal operations and an important capability of their upper stage. At minimum, those can be taken off and put on another stage if the rest of the stage is irreparably damaged...which is unlikely given the degree of function required to successfully land.

And yes, Musk wants full reuse. First stage reuse with 90% of the engines is not an insignificant part of that. They're no longer pursuing second stage reuse for Falcon 9 due to the payload limitations, but they plan for it with their next launch system...which uses an engine design that should be fundamentally easier to reuse than the Merlin, due to lower temperatures in the full-flow turbopumps and lack of coking from the methane fuel.

[icekatze]

hi hi

Hold up a second. Merlin is most definitely not Space X's own design, or at least, it is as much their design as an AJ26 is not a Russian motor with some tweaks and aerojet's name slapped on. Merlin is a Fastrac MC-1 rocket motor with a different turbo pump and some tweaks. (Fastrac was developed by NASA.)

[Mjolnir]

hi hi

Hold up a second. Merlin is most definitely not Space X's own design, or at least, it is as much their design as an AJ26 is not a Russian motor with some tweaks and aerojet's name slapped on. Merlin is a Fastrac MC-1 rocket motor with a different turbo pump and some tweaks. (Fastrac was developed by NASA.)

Don't be ridiculous. The Fastrac was a much smaller engine than the Merlins and was never developed beyond a test-stand breadboard engine. The Merlin 1A was at most a new and much larger engine design that used the Fastrac design as a reference, and with the various upgrades, the Merlin 1D isn't even the same engine as the Merlin 1A. The Merlins have gone through at least two regeneratively cooled combustion chamber/nozzle designs, and the Fastrac isn't even a regeneratively cooled engine...the only remaining similarity is that they use the same thermodynamic cycle. It is nothing remotely like the AJ26 rebranding of NK-33s hauled out of some warehouse where a Soviet bureaucrat stashed them.

SpaceX is not achieving their cost reductions by flying NASA experimental testbeds.

[icekatze]

hi hi

Actually, the Fastrac was heavier than the Merlins, but whatever. Tweaking the nozzle was cool and all, but it is still an iteration on past work. I look forward to seeing if they can succeed in making recovery cost effective, they've got one step out of the way, I'll be sure to follow them along the next few.

[Arioch]

The SpaceX webcast of the launch and recovery is fun, with a huge audience of employees going berserk.

[Mjolnir]

hi hi

Actually, the Fastrac was heavier than the Merlins, but whatever. Tweaking the nozzle was cool and all, but it is still an iteration on past work. I look forward to seeing if they can succeed in making recovery cost effective, they've got one step out of the way, I'll be sure to follow them along the next few.

Fastrac was indeed nearly twice the mass, at 910 kg compared to 470 kg, but nobody measures rocket engine size in mass. The Fastrac only had 284 kN of vacuum thrust, compared to 825 kN for the latest up-rated Merlin 1D...the Merlin is nearly 3 times larger. It would take 25 Fastrac engines to match the sea level thrust of a Falcon 9. The T/W ratio of the Merlin 1D is 180...Fastrac was 32. Yes, Fastrac had an utterly abysmal T/W ratio, while the Merlin sets records.

Fastrac had an ablatively-cooled combustion chamber and expendable carbon fiber nozzle, Merlin has a regeneratively cooled combustion chamber and a radiatively cooled niobium alloy expansion nozzle.

Fastrac was fixed thrust, set prior to launch, the avionics were limited to opening and closing valves and firing an expendable ignition cartridge. The Merlin 1D can throttle down to at least 70% (possibly more with the recent updates to the 1D), quickly and precisely enough to allow landing of the first stage. Merlin is also capable of relighting multiple times in a single flight.

Merlin has a vacuum variant with even higher thrust and far deeper throttling, down to 39%. There'd be little point in even trying to put a Fastrac on an upper stage.

SpaceX also makes heavy use of 3D printing to reduce complexity and manufacturing costs and improve performance, including parts of the Merlin engines...a technology that didn't even exist when Fastrac was canceled.

They didn't "tweak the nozzle", they developed a series of several new engines, and several launch vehicles using them. At best, they used the Fastrac design as a reference while designing the Merlin 1A...they immediately went well beyond it. The 1C and 1D aren't even entirely the same kind of engine, they both are regenerative cooling designs and each takes a very different approach to combustion chamber construction, while Fastrac used a thick layer of phenolic composite ablative material.

As for the notion that SpaceX could cobble together a few over-budget, behind-schedule, canceled NASA projects into a medium-heavy launch vehicle that operates with a small fraction of the launch costs of the competition...really? They're reducing costs by blindly copying designs from an organization that has consistently failed miserably at reducing costs? Do you honestly not see how absurd an idea that is?

[Arioch]

Do you honestly not see how absurd an idea that is?

There was really no need to add this insulting comment at the end of your argument.

[icekatze]

hi hi

Like I said, they upgraded the turbo pump to have a higher flow rate, something that was within the design's capabilities, though never utilized in testing. Regenerative cooling is certainly nothing new (late 1800s, early 1900s), you put tubes around the engine bell and pump propellant through it, that's hardly a revolutionary design change. Not to mention the fact that the Merlin gets its excellent thrust to weight ratio at the expense of specific impulse.

And after speaking with people who work on space launches, I will continue to maintain that the engine was only one of many examples of places where Space X got free assistance from NASA. Which is fine for them, and honestly the US space industry could use the cooperation.

a small fraction of the launch costs of the competition

I mean, I guess qualitative size is open to interpretation, but I personally think that 13402 dollars per kilogram (Falcon 9) is not a small fraction of 13725.5 dollars per kilogram (Long March 3B) to GTO. The Atlas V is actually better, costing less than 11494.25 dollars per kilogram to GTO. (The ULA also gets plenty of free assistance.)

Space X is a fine company, and I've been trying to give them credit where it is due this entire time, but I don't think they are the second coming of space Jesus. If and when they manage to prove that they have a reliable, reusable launch system that reduces costs, then I will happily give them credit for that as well.

[Grayhome]

[Mjolnir]

Like I said, they upgraded the turbo pump to have a higher flow rate, something that was within the design's capabilities, though never utilized in testing. Regenerative cooling is certainly nothing new (late 1800s, early 1900s), you put tubes around the engine bell and pump propellant through it, that's hardly a revolutionary design change. Not to mention the fact that the Merlin gets its excellent thrust to weight ratio at the expense of specific impulse.

No, it's not "like you said" at all. These are not even close to being the same engine with a different turbopump and "tubes around the engine bell" (which is not what they did and which completely misrepresents the engineering issues involved).

As for launch costs, I don't know where you got them, but your Atlas V numbers are way off. With the current no-reuse Falcon 9 prices, for GTO, Falcon 9 is $12600/kg, which is about 1/3 of the ~$35000/kg an Atlas V would cost. At your $11494.25/kg figure, an Atlas V 551 (the highest-capability Atlas V that has launched) would only cost about $100 million, and Tory Bruno has given $164 million as the minimum cost for their lowest-end configuration (which would be the 401, with 4750 kg to GTO).

For LEO: Falcon 9 gets $4650/kg, Atlas V gets $13182/kg. About 2.8 times cheaper there.

Atlas V is simply not competitive. Of the launchers with comparable capability, only the Delta IV is more expensive. It can't even compete against the ESA launchers, let alone the Russians and Chinese. Not even ULA claims otherwise, Tory Bruno has specifically stated that they are targeting much lower prices with the Vulcan in order to compete (aiming for a target that is still higher than the no-reuse Falcon 9, but at least comparable), but they are years away from launching it. The Chinese are the only ones who have come close to Falcon 9 launch costs, and they have heavy state support.

[icekatze]

hi hi

I got the numbers from Former ULA CEO Michael Gass, and the ULA's 2014 launch contract. 164 million is the cost for a low end Delta rocket, not a low end Atlas rocket.

It is also important to note that when Space X, or anyone, does launches for the US government, there is another roughly $30 million per launch in government oversight that is not present during commercial launches.

[Absalom]

hi hi

Like I said, they upgraded the turbo pump to have a higher flow rate, something that was within the design's capabilities, though never utilized in testing. Regenerative cooling is certainly nothing new (late 1800s, early 1900s), you put tubes around the engine bell and pump propellant through it, that's hardly a revolutionary design change.

This is like saying that the Fastrak is a minor revision of the Apollo DPS. On top of that, even if you went with feature set over actual physical design (and you should almost always go with actual physical design), the current Merlins wouldn't be Fastraks (or as far as I can see, any other pintle-injector design) because none of the others had regenerative cooling. This is not a minor change, given that it means the rocket engine is not designed to engage in auto-canibalism.

If your NASA friend thinks that the Fastrak and Merlin are the same engine just because they both use a pintle injector, then they certainly don't belong in the actual launcher branch of NASA due to having a personality unsuited to the job. That kind of obsesively-invative outlook really only belongs in experimental science, not applied science, since applied is about doing things, instead of preparing for doing things.

Not to mention the fact that the Merlin gets its excellent thrust to weight ratio at the expense of specific impulse.

As long as it can achieve orbital insertion of the payload, this is perfectly reasonable. You'll never get good Isp from a chemical rocket because pure chemical rockets just can't achieve the enery / mass ratios needed, so you're better off focusing on ion engines & such as soon as you get your orbital characteristics to a state that allows them to achieve the job. I won't say that higher Isp chemical rockets are inherently bad, but if that cuts into your reusability quotient (just as an example), then you need to ditch the extra Isp, because the rockets currently cost much more than the fuel, so if you cut rocket costs by half (which a single reuse of a lower stage probably won't quite achieve) then you might get around 40% cost reductions. Fuel, meanwhile, commonly costs ~10%. Thus, reusability is more important than Isp while your reuses are few (or none), because each reuse makes more of an impact.

Space X is a fine company, and I've been trying to give them credit where it is due this entire time, but I don't think they are the second coming of space Jesus. If and when they manage to prove that they have a reliable, reusable launch system that reduces costs, then I will happily give them credit for that as well.

Funny, because it reads more like you've been trying to downplay their achievements as inconsequential. SpaceX does research, and no matter how close it formerly was to Fastrak the Merlin currently isn't one.



Now, more interestingly, does anyone know what the wording for the mission-to-Mars earmark was? NPR made me wonder if NASA is supposed to do a cycler (I really want an automated Moon base to build habitat hulls for 4+ year Mars orbiters, but there's no way that could get off the ground as fast as conventional sourcing).

[Mjolnir]

hi hi

I got the numbers from Former ULA CEO Michael Gass, and the ULA's 2014 launch contract. 164 million is the cost for a low end Delta rocket, not a low end Atlas rocket.

No, it is the cost for an Atlas V 401, according to ULA's current president and CEO, figures he gave in testimony to Congress, and ULA's web site. And that's not even the cost of a separate launch, it's as part of the 36-launch block buy, and when the block buy was just 27 launches the it was $184 million for a 401...buying a single launch would cost much more.

[Nemo]

I still dont understand why we want to bother going to Mars in person anyway. Let the rovers go down in the gravy well. Use the people to man stations on the moon and start up some manner of resourcing and manufacturing capacity. Even that can be largely auto or tele-operated.

[Mjolnir]

Not to mention the fact that the Merlin gets its excellent thrust to weight ratio at the expense of specific impulse.

As long as it can achieve orbital insertion of the payload, this is perfectly reasonable. You'll never get good Isp from a chemical rocket because pure chemical rockets just can't achieve the enery / mass ratios needed, so you're better off focusing on ion engines & such as soon as you get your orbital characteristics to a state that allows them to achieve the job. I won't say that higher Isp chemical rockets are inherently bad, but if that cuts into your reusability quotient (just as an example), then you need to ditch the extra Isp, because the rockets currently cost much more than the fuel, so if you cut rocket costs by half (which a single reuse of a lower stage probably won't quite achieve) then you might get around 40% cost reductions. Fuel, meanwhile, commonly costs ~10%. Thus, reusability is more important than Isp while your reuses are few (or none), because each reuse makes more of an impact.

If they were obsessing over such minor differences in Isp, they wouldn't be using a kerosene-burning gas-generator engine in the first place.

Also, increasing thrust to weight on a first stage engine reduces gravity losses. Given the limited delta-v required of the first stage, trading Isp for increased T/W can easily be a net gain. Any increase in thrust or reduction in engine mass on the first stage also translates directly into more mass available for the second stage and payload. As for the second stage, the Merlin 1D Vacuum gets a specific impulse of 348 s, considerably higher than the vacuum Isp of the Fastrac.

Now, more interestingly, does anyone know what the wording for the mission-to-Mars earmark was? NPR made me wonder if NASA is supposed to do a cycler (I really want an automated Moon base to build habitat hulls for 4+ year Mars orbiters, but there's no way that could get off the ground as fast as conventional sourcing).

What mission-to-Mars earmark? There's no manned Mars mission being funded, just the ExoMars lander and the 2020 rover.