By 2001 the prospects for private development of orbital launch systems looked very grim following the failures of several highly promising companies. The bankruptcies of Iridium, Globalstar and other comsat constellation projects had destroyed their targeted market and seemed to indicate that there was little to motivate new attempts to develop low cost launchers for many years ahead. Then in mid-2002 there came the surprising news that Elon Musk had started a company to develop a two stage partially reusable launcher that would begin flying within just a couple of years. The very successful internet entrepreneur had founded Space Exploration Technologies Corp., or SpaceX for short, to develop the launcher from scratch. The company has shown steady progress since then and has constructed and tested a great deal of hardware. The company plans to fly its Falcon vehicle for the first time this coming January.

Credits: SpaceX Artist view of the SpaceX Falcon launcher now in development. First launch is scheduled for early 2004.

Elon Musk brings substantial resources to SpaceX and also impressive credibility for successful business development. He co-founded the enterprise software company Zip2 in 1995 and sold it to Compaq at a substantial premium in 1999. He then co-founded PayPal, which was sold to eBay in 2002 for another highly profitable sum.

I contacted Elon via email to discuss the background and status of SpaceX. The following is that interaction assembled into a conversational interview format. [Note: I've added links to some of the items that Elon discusses. Indented sections are from the second round of questions.]

Elon Musk (Wired Magazine profile)

HS: Private rocket development by startup companies in the post-Apollo era includes projects such as Truax's Volksrocket in the late 70s, Conestoga I and AMROC in the 80s, Beal Aerospace and several other ELV and RLV companies in the 1990s. They all came up short of space and many see their history as nothing but a tale of woe and failure. To me, though, they each appear to build on what was learned before them and to provide significant advancements in the technical and strategic knowledge needed to develop a rocket business from scratch. It looks like SpaceX will be the startup company that finally makes it to orbit. When you studied prior efforts, what were some of the lessons [you] learned on what to do and, perhaps most importantly, what not to do?

The ones I'm familiar with failed on one or more of the following:

1. Lacked a critical mass of technical skill.
2. Insufficient capital to reach the finish line, particularly if an unexpected setback occurred.
3. Success was reliant on a series of technology breakthroughs that did not happen.

The above modes can obviously cross-feed one another.

Musk: Well, hard and easy are somewhat nebulous terms. I think I have high standards and would classify getting Falcon to orbit as quite difficult. Overall though, I think we have had quite a smooth development so far, which is a credit to the hard work of the SpaceX engineering team.

The design tools, such as solid modeling and finite element analysis software are substantially more powerful than ten years ago, so that's a clear advantage. Obviously, most electronics have improved a lot too, except gyroscopes and flight termination systems.

HS: Can you think of any example where "standard practice" in the design and construction
of launch vehicles is blatantly more expensive that the way you guys did it with the Falcon?

Musk: It's hard for me to comment definitively on "standard practice", as I have never worked for another aerospace company. However, based on stories I've heard, some of them sound like a Dilbert cartoon in real life. My approach is simply to seek out very talented people, ensure that the environment at SpaceX is as motivating & enjoyable as possible and establish clear & measurable objectives.

I think it is a mistake to hire huge numbers of people to get a complicated job done. Numbers will never compensate for talent in getting the right answer (two people who don't know something are no better than one), will tend to slow down progress and will make the task incredibly expensive. Also, a lot of aerospace senior managers seem to be really disassociated from and unable to do hard core engineering. I think that is a mistake and results in cloudy judgment on important technical issues -- they can't tell if something is really good or not, so they just do what everyone else does, assuming it to be the safe bet.

HS: If you had simply put the Falcon out to bid with the majors and let one of them build it, how much might a launch cost?

Musk: Probably something like $15M to $20M, although that is quite a random guess. Right now, the two cheapest commercial American LVs are Pegasus and Taurus. Their NASA list prices are $30M and $50M respectively, although rumor has it that they cost about 25% less when bought directly from Orbital.

HS: By the way, someone pointed out to me that the latest figures released for the Falcon indicate 1400lbs to LEO. That's 400 more than the number that was mentioned in previous articles. Did something turn out to provide higher performance or weigh less than expected?

Musk: Our original target was at least 1000lbs to nominal reference orbit (200km, 28.5, circular), but we had a very big sandbag in there. As the actually engine and mass numbers are coming in, we are gradually reducing the sandbag. Some publications are still reporting old numbers for Falcon. Whatever number you see for [the standard] Falcon, heavy Falcon will do slightly more than 3X.

Credits: SpaceX Test firing of the Falcon first stage engine, named Merlin.

HS: Finding, or building, a market looks like perhaps the biggest challenge for the Falcon. I saw that you made a presentation at the recent Utah Smallsat meeting. When I attended the meeting in 1998, I was very heartened to see so many young students there who seemed ecstatic that they were working on something that was actually going to go into space. Unfortunately, I'm sure most of them graduated long before their projects found rides to orbit. Despite the many excellent smallsat projects and concepts around, NASA and the Air Force have not been generous in funding them.

Musk: At this point, I'm confident that there is at least a modest market for Falcon. We have sold the first two flights (one US government customer and one foreign govt customer). My guess is we will do something like 2-3 launches next year, 3-4 in 2005 and 4-6 in 2006. However, it's worth pointing out that Falcon heavy (with the two strap-on liquid boosters) can do about two tons to LEO, which is arguably a medium satellite capability. I think roughly 25% to 35% of future launches will be Falcon heavy.

I'm told by many people that a major reason why many of the small sat programs were cancelled was the lack of a reasonably priced launcher. There needs to be a cost impedance match between satellite and vehicle. SpaceX intends to work hard to reinstate those programs now that Falcon is available.

HS: "Ticket-to-Orbit" is a concept that Kistler has proposed (see this report in pdf) for its K-1, assuming it ever gets built, in which a launch would be set to occur on a fixed date. They would sell "seats" on a multi-payload platform for that flight, just like buying a seat on plane for a given time. This differs significantly from the current "charter" system in which a rocket company arranges a launch time with the owners of the primary payload and the secondary payloads are at the mercy of that decision and whatever delays caused by the primary.

I'm sure the smallsat industry would greatly appreciate the predictability and reliability of a fixed date schedule arrangement. Do you think SpaceX might eventually try such an approach?

Musk: Certainly a possibility, although it can be very difficult coordinating multiple customers, particularly if some want to go to different places than others. At this point, our preference is to sell the whole ride. If another company wants to buy a flight and then assume the risk and complexity of reselling slots, that would be fine with us.

HS: As I understand it, you were initially interested in funding a Mars mission but the high cost of launching such a mission led you to develop your own launcher. Are you actively planning a Mars mission for the Falcon or its heavier follow-on derivatives? Don't know if the payload weight matches the launcher, but it would be cool if you launched the AMSAT-DL led P5-A Mars mission. A private launcher sending an AMSAT spacecraft to Mars would definitely signal a new era in space exploration!

Credits: SpaceX Falcon stage 1 propellant tank.

HS: In parallel with your orbital launcher development we are seeing a lot of activity with manned suborbital launchers, particularly with regard to the X PRIZE. Do you think that suborbital RLVs will provide substantial
data on how to build robust and operationally low cost orbital manned vehicles? Are you optimistic about the prospects of a suborbital space tourist business? A few years down the line, could we see a "merging" of these activities and find a manned module on top of a SpaceX booster?

Musk: I'm really glad to see all the activity in entrepreneurial space and hopefully this heralds a new era of space exploration with price and quality improvements similar to other technology arenas. I do think people will learn a lot from the suborbital RLVs, although obviously it is a big leap in both Isp and mass fraction to get to orbit. Of all the suborbital RLVs, I like John Carmack's approach the best, as it has the clearest upgrade path to orbital capability.

In the case of SpaceX, I've chosen to make a big initial bet on a semi-RLV that can deliver satellites to orbit. That I think is a comparatively dependable market and can serve as the revenue foundation for eventually going to a manned fully-RLV. It is always possible that we will do so in collaboration with one of the current suborbital ventures.

HS: Andrew Case, my "rocketry advisor" for HobbySpace, passed along these questions:

a) What is the ignition system for the first and second stage engines??

Musk: Dual GOX/RP torch ignition. Note, this means the 2nd stage can keep restarting so long as there is propellant in the tanks and enough He to do a short settling maneuver in zero g.

b) Does the first stage engine also have a pintle injector (Northrop-Grumman article in pdf) like the second stage?

Musk: Yes.

c) When the follow up vehicle is constructed will the second stage (Falcon's first stage) be recovered? This is particularly tricky since it will be moving a lot faster at burnout when it's the second stage (on the 2nd generation vehicle) than when it's the first (on Falcon), and the reentry will be a lot more demanding. If it's discarded, then wouldn't it make better economic sense to use a pressure fed engine?

Musk: We will certainly make every attempt to recover and I think it should work, although we will need to add more thermal protection to the stage.

I should clarify that the diameter will increase quite a bit on the follow up vehicle, so it will actually look like a really fat version of the Falcon 1st stage. There will also be a much bigger nozzle to take advantage of a vacuum expansion ratio, so Isp will increase a fair bit.

HS: Can you say a bit about designing the first stage to be recoverable? How much extra weight does it require in terms of extra equipment, e.g. parachute system, hardening for impact, and waterproofing compared to using a throwaway?

Musk: There are a large number of small design changes needed to make all the parts salt water and impact resistant. Tough to retrofit an existing rocket, but not too hard if designed in at the beginning. Basic steps are coating, anodizing, cladding, etc of all exposed metal parts. Also need to maintain ongoing He purges, minimize galvanic potentials and put in sacrificial anodes.

We keep detailed mass numbers close to our vest, but it is safe to say that the impact to payload capability by making the first stage reusable is not substantial. On the other hand, making the upper stage reusable would be a huge hit to payload. The economics of the process will only be fully understood once we go through it, but our projections indicate that recovery pays off. That first stage is expensive to build, so I can't imagine how we could spend more recovering it than building it. Just recovering the valves and engine would be a huge saving.

Credits: SpaceX The SpaceX engine tests use facilities on a former Navy test range in Texas that were also used by Beal Aerospace.

HS: Have you done any test drops of boosters into the water?

Musk: First one will be on day of launch. We have tested individual components for salt water tolerance though.

HS: Thanks very much, Elon, and best of luck with the Falcon.