August 18, 2003
Musk on SpaceX Falcon

Elon Musk gave an interesting presentation at the Mars Society Conference on Saturday morning, concerning SpaceX's efforts to develop a new launch vehicle. He also hinted at what SpaceX has in mind for the future.

Musk started out with a recap of his space-related efforts over the past two years. His initial foray into private space ventures was a plan for a privately-funded Mars lander mission. In developing this idea, he met several times with Russian space companies, intending to use their cheaper launch vehicles to help keep the mission's costs low. His assumption was that public interest in and support for space exploration could be increased with a suitably adventurous (private) mission.

However, his work on the lander project made him reevaluate that assumption, and he came to the conclusion that what is really needed is not a public-support-generating "stunt" (since the public already supports space exploration in general), but cheaper access to space to make exploration more affordable. (The Mars lander mission is currently on hold -- it may be launched by SpaceX in the future, but SpaceX will not be the company to carry out the mission itself.)

He then reviewed the current means of getting to space (manned and unmanned), describing Shuttle as plagued by serious problems and design-inherent dangers. OSP may be a good idea in some respects, but in its winged or lifting body forms is clearly not intended for use beyond LEO, which makes its worth even more questionable. He compared Soyuz to Shuttle, describing it as safer, more survivable, more reliable, and cheaper, but Russian efforts in space are limited by financial woes. Shenzhou is estimated to cost $50M per flight, and competition from the Chinese could be beneficial to the US space program -- international cooperation, on the other hand, is less desirable than friendly rivalry, because it leads to consensus stagnation rather than driving progress and innovation.

Private initiatives are underway (X-Prize contenders and others). Musk expects private companies to beat/lead OSP to market. He sees Burt Rutan's vehicle as having scalability problems (from suborbital to orbital), with the carrier airplane limiting GLOW (he also indicated Paul Allen may be the "secret financier" behind the project). Carmack's Armadillo Aerospace has a more scaleable design, but their engine needs better ISP and mass fraction, and possibly additional stage(s) to get to orbit. Jeff Bezos' "Blue" project has "non-trivial financial resources", and is a vehicle similar to DC-X. Bezos is a longtime advocate of human spaceflight, and Musk sees Blue as the primary competitor for SpaceX in the long run. Blue is likewise not an X-Prize contender, and Musk sees the rivalry between the various startups (X-Prize or not) as a friendly rivalry, with each focused on developing low-cost access to space rather than wasting resources on cutthroat competition.

He then described the SpaceX Falcon (which I'll abbreviate as SXF) vehicle and the thought behind it, and answered questions from the audience.


  • SpaceX is aiming for satellite delivery, not just X-Prize. Indeed, Musk claimed that SXF is not an X-Prize entry, and did not describe any manned missions for it (at least in the near term).
  • SpaceX's long term goals include a reusable upper stage with manned capability, and building up to a Saturn V payload class.
  • Human exploration is the company's goal, once reliability of the vehicle is sufficiently proven.
  • The standard SXF is to carry 1400lbs to LEO, the heavy lift version is to carry 4500lbs.
  • A standard SXF launch is expected to cost $6M, and $10M for heavy version.
  • The SXF will be unveiled to the public at the Smithsonian in December.
  • First launch, flying a USAF payload, is scheduled for January 22, 2004. (Not bad for a scratch-built vehicle only 18 months along.)
  • First Stage:
    • Recoverable, via parachutes to water landing.
    • LOx-RP propellants

    • Engine:
      • In-house design.
      • 72000lb SL thrust, 85000lb vac thrust
      • 265s SL Isp, 310s vac Isp
      • Ablatively cooled
      • Turbopump fed, gas generator cycle. Pump tested recently to 90% of rated flow.
      • Not throttleable, but can be modified for this.


  • Second Stage:
    • Also LOx-RP
    • Second stage engine is throttleable and restartable.
    • Little detail given here.

  • Launch sites:
    • Vandenberg arrangements are in place (launch in January will be from VAFB).
    • Working on arrangements with KSC, Marshall Islands

  • Reliability Enhancements:
    • Only two engines, only one of which is air-started.
    • Only one separation sequence (but what about payload sep?).
    • Dual-redundant GN&C.
    • No control surfaces.
    • Stage 1 TVC is accomplished by pressurized fuel fed actuators (use RP as hydraulic fluid).

  • Regulatory issues:
    • Launch/test site agreements are in place.
    • Regulatory matters are the biggest cost to the program, with more money being spent on documentation than on the development of the main engine.
    • Musk called for ground-up revision of regulations governing space launch.
    • Current body of regulation is 40-year conglomeration of ad-hoc rules, with little having been accomplished in rationalizing/streamlining them despite 15+years of talk about it. "Rules are immortal."
  • General Notes:

    • Musk expects 2-3 launches in 2004 (second payload is already lined up, and is for a foreign government, and will be launched from Marshall Islands, but he couldn't say who it was); this should grow by 1-2 per year thereafter.
    • "Falcon is designed to minimize cost per launch rather than cost per pound."
    • He believes there is a market for a $6M launch vehicle.
    • Biggest potential market in future is human launches.
    • Musk plans to personally fund SpaceX through first few flights, before seeking outside financing or selling shares. SXF development is expected to cost less than $100M, the exact figure being proprietary and thus not disclosed in the presentation.
    • SpaceX currently has 30 full-time employees, 15 part-time/consultant employees (30 = 5 techs + 22 engineers + 3 non-technical).
    • For Saturn V class vehicle, development of an F-1A-class engine is required, and it would probably be LOx-RP.
    • SpaceX may eventually upgrade to LOx/LH2 upper stage for payload enhancement.
    • Musk favors vertical landing manned vehicle over a winged one (ie: he wants a capsule or DC-X type vehicle), because such a vehicle forms the basis for landers and return vehicles for use on Moon and Mars (where wings are useless). Wings are only useful on vehicles returning from LEO, which means winged OSPs are useless for anything but ISS missions and don't support a broader human space exploration program.


Posted by T.L. James on August 18, 2003 07:25 PM

Comments

Great Summary. I enjoyed the presentation and was floored not only by his speed of development, but also that he has a paying customer for the initial launch.



Posted by: Carl Carlsson at August 19, 2003 12:15 AM


I'm a mere enthusiast, not an expert, so forgive me a somewhat elementary question:

What does "SL" stand for in the Isp rating and why is Isp higher (and thrust higher) in vacuum?



Posted by: Andrew at August 19, 2003 01:22 PM

SL = 'Sea Level'
The optimal shape for the bell nozzle depends on the outside air pressure. At a different presure, the Isp will be less than the maximum possible so the shape ends up being a comprimise with peak Isp set for a specific altitude. This is one of the things an aerospike engine doesn't suffer from since the exhaust shape automatically adjusts to the external pressure.



Posted by: Andrew P at August 19, 2003 01:38 PM

Would the Saturn V class rocket have a single engine powering it's first stage or will it have five engines like the original Saturn V?



Posted by: Dominic Pledger at August 19, 2003 05:26 PM

If it is an F-1A-class engine, and the vehicle is to have a Saturn V-class payload, it's likely. Note that the "Saturn V-class payload" and "need an F-1A engine" comments were in separate parts of the presentation, and didn't imply to me at the time that a SpaceX HLLV would resemble Saturn V.

Just speculation, though -- I don't work for SpaceX, so all I know about Falcon and future plans is what you see above.



Posted by: T.L. James at August 19, 2003 07:43 PM

Is there any chance thay anyone will build a single chamber rocket engine of about 7.5 Million pounds of sea level thrust?



Posted by: Dominic Pledger at August 20, 2003 05:32 PM

Hi All

There was a design from the early sixties with a much bigger thrust than that even... the Sea Dragon. Here's a link...

http://www.astronautix.com/lvs/searagon.htm

...its main engine would have had 80 million pounds thrust. That meant 1,000,000 lbs to LEO - three ISS at once!

Adam



Posted by: Adam at August 20, 2003 05:55 PM

Oh, THAT would be fun. Doesn't seem like it would work, but TRW and NASA (of the 1960's) seemed okay with the idea.



Posted by: T.L. James at August 20, 2003 07:25 PM

Please email Tom Mueller at tom.mueller@spacex.com and/or Elon Musk at elon.musk@spacex.com to find out wether or not SpaceX plans to develop their Saturn V class rocket with a single engine per stage. Using the minimum number of engines per stage increases reliabiliy of the launch vehicle as well as giving the best possible sound quality on liftoff for a given payload class. When you have found out, please email me at rocket04pledger@yahoo.co.uk.



Posted by: Dominic Pledger at August 21, 2003 05:09 AM

T.L. James,

Why did you say that a 7.5 Million pound sea level thrust single chamber engine wouldn't work?



Posted by: Dominic Pledger at August 21, 2003 05:00 PM

I should have been more specific -- I found the Sea Dragon concept in general a bit fanciful. I wasn't referring to the engine specifically.

The concept seems somewhat impractical to me for several reasons. First, the vehicle is floating on its side in seawater, which makes corrosion of systems and structures a serious concern, and is going to result in the thing turning into a popsicle as soon as the LOx tanks are loaded.

Second, the need to erect a fully-loaded vehicle of that size by cantilevering it from its base, without a strongback to help take some of the load, is going to make the structure inefficient. Erecting the vehicle by first sinking the base (think SRB retrieval in reverse) might be a way around this problem. Of course, if you have a million pounds of payload capacity, whats a few tens of thousands of pounds of structural inefficiency?

Third, there is the "iceberg" problem. Will it ignite with its first stage engines and tankage beneath the surface? If not, how big does the submersible base have to be to lift the entire vehicle, standing, above the water? Then there is the location of the center of gravity -- a huge mass standing above the water will require a fairly large submerged base mass (or a sophisticated active control system) to remain vertical.

Just a few things off the top of my head, related to the vehicle in general. I'm not saying that it can't be done, or that there aren't creative solutions to each of these potential problems, just that the program (at least as described in the Astronautix article) leaves open a number of serious questions.

As for the engine...it would be entertaining to watch a single huge engine like that fly, but what about engine-out capability? And how would one build such a large, single-chamber engine in the first place? And could it be done without the thing tearing itself apart from the acoustics alone? Again, I'm not saying that it can't be done, just that it would be one heck of a challenge.

Give me the development money, and I'll be happy to give it a whirl.



Posted by: T.L. James at August 21, 2003 08:37 PM

T.L. James,

If a launch vehicle has a large number of smaller engines instead of one huge one, there is a much greater risk of a catastrophic faliure of one of the engines. The Russian N-1 rocket had 30 first stage engines and all four that were launched failed! This is because when one engine explodes, it takes out the adjacent engines leading to a chain reaction which ultimately results in the destruction of the rocket. If the Keep It Simple Stupid (KISS) doctrine is followed by keeping the rocket as simple as possible, then the reliability of the rocket is as high as possible.

The acoustics could be solved by the use of sound absorbent materials in the launch pad which would reduce the intensity of the shockwaves that are reflected upwards to the vehicle. With modern materials, the launch vehicle could be made strong enough to be able to withstand the shockwaves generated by it's engine without making it too heavy.



Posted by: Dominic Pledger at August 22, 2003 04:50 AM

One thing that confuses me-why didn't he mention x-cor as a possible competitor? Their progress in engine development is every bit as impressive as spacex's, because they've done it on a shoe-string budget. Also, I would think some of their innovations would be of interest to spacex as possible cost-saving measures, specifically their fuel piston-pump(same space-saving advantages as Spacex's turbo-pump, plus it's simpler, cheaper, and more scalable)and their clear demonstration that a regeneratively cooled engine CAN be developed both fast and innexpensively. Is he really overlooking this company, and if so, why?



Posted by: Joel L. Lewis at August 22, 2003 12:32 PM

Yes, Dominic, I'm aware of that. The optimum number for a manned vehicle is typically from three to five, as one or even two engines may not give you the engine-out capability you need, while more than five is pressing your luck on reliability.



Posted by: T. L. James at August 22, 2003 07:34 PM

T. L. James,

To give you engine-out capability, the engines need to share a common propellant supply with interconnecting propellant feed lines. If the engines have independent propellant tanks like in the Delta IV Heavy which has three separate common core boosters (CCB's). The failure of the engine of one of the boosters is likely to be non-survivable. This is because the failed CCB would become dead mass and the remaining operational CCB's would have insufficient thrust to take the vehicle safetly into orbit.



Posted by: Dominic Pledger at August 23, 2003 05:19 AM

T. L. James,

A manned launch vehicle can be built with one engine per stage without compromising saftey if pintle injector engines are used. Pintle injector engines have been tested with 25 different propellant combinations with thrusts from 5 pounds to 650,000 pounds. There has never been a failure of a pintle injector engine! This is due to the inherent combustion stability of pintle injector engines.



Posted by: Dominic Pledger at August 25, 2003 05:30 PM