SpaceX announces Interplanetary Transport System

First watch the video if you haven’t seen it yet.

SpaceX had a presentation at the International Aeronautics Conference today about their Mars colonization plans, and specifically about the rocket/spacecraft system they’re using for the transport.

The thing is a beast, and although only marginally larger than the Saturn V in its key dimensions, it can launch over twice as much to Low Earth Orbit… reusably (in non-reusable mode, it’s over 4 times). They achieve this, basically, by use of modern technology: methane propellant, full-flow staged combustion engines (very fuel-efficient) and composite tanks.

It uses SpaceX’s currently-under-development Raptor engine, which looks to be the among the most advanced engines ever produced. Although “only” about half the thrust of the F-1 engine on the Saturn V, the first stage of the ITS will have 42 of the engines, compared to just 5 on the S-V. Getting all of those engines to work reliably will be a challenge, and to a large extent a similar challenge doomed the USSR’s attempts at a moon landing–their rocket required 30 engines, and they never quite worked out the dynamics of the problem. But times have changed so hopefully SpaceX will have greater success.

The booster is of course only part of the system. The booster carries the Mars craft much of the way to orbit, but its fuel is expended getting the rest of the way. So there are several refueling trips with a tanker craft that autonomously docks with the transporter. They aim to have very rapid turnaround, with the tanker landing near the launch pad, and the booster landing directly on top! It would sound ridiculous if it weren’t that SpaceX has already demonstrated landing precision of a couple of meters… and on a moving barge. So maybe this isn’t as absurd as it sounds. The tanker refueling system is the key to having a manageably-sized system that is reused enough times for good amortization. Without it, the rocket would be impossible instead of merely implausible.

After several tanker trips, we now have a fully fueled Mars transport with ~6 km/s of delta V. This is enough to get to Mars and land propulsively, with most of the excess velocity shed via aerobraking at Mars. The craft has heatshields and is largely empty at this point (only cargo left, with very little fuel), so even in the thin Mars atmosphere the aerobraking should be effective.

It’s at this point that the details get a little fuzzy. One major reason for picking methane on Mars is that it can be produced there, by using water ice and the CO2 atmosphere. These chemicals and a lot of electrical input energy is all that’s needed to produce methane and liquid oxygen. Since the vehicle lands with no remaining fuel, it’s critical that it can refuel on the surface to make the return journey (both to reuse the vehicle and bring passengers bath to Earth). “In-Situ Resource Utilization” (ISRU) is the key term here and SpaceX will clearly use it, but there was little more than handwaving toward it.

Although much of the presentation was, well, visionary, SpaceX is actually making concrete progress toward their goals. The Raptor engine was recently tested, at least in some form (probably a subscale prototype). Even more impressive to my mind was their prototype composite fuel tank. That thing is absolutely enormous, and demonstrates some serious capability for manufacturing large composite tanks. It may sound simple but tanks–especially cryogenic composite ones–have killed projects in the past (the X-33 Venturestar comes immediately to mind). So the fact that they have semi-real hardware already is pretty amazing.

Anyway, it’s all exciting stuff. You can watch the full presentation here, and see the technical slides here.

I’ve left out a bunch of stuff, so please ask if you have questions.

Oh and some gorgeous renderings if you’re just looking for a desktop background. Although their primary destination is Mars, it’s the Interplanetary Transport System–and with refueling in key locations, it can reach quite a few places in the solar system.

First, a string of SuperCharger stations for Tesla cars; now a string of re-fueling stations for Interplanetary Buses.

Maybe the same thing will happen in the Space Travel industry as Tesla caused for the automotive industry: Tesla said it would create a moderate-priced EV. All of a sudden everybody wants to make EVs.

Maybe BO will develop a competing engine. Maybe some of the “use an airplane for first 50-60K feet” will graduate to serious space ability.

Speaking of Methane engines: What is the big difference in scale of the things.
If the Raptor demo was just a small-scale prototype, how much can go wrong when it is scaled up?

Gotta admit, the idea of 42 engines all sync’d together does sound problematic.
But, if the current ‘42 engines’ design is based on the performance of the current Raptor, and the current Raptor is just a proof-of-concept model, mayhaps the ultimate design will use but 20 or so engines.

A few months ago, there was a tidbit of news about SpaceX ordering truly massive quantities of carbon fiber from Toray. The total value mentioned was $2 billion to $3 billion. Given the estimated cost of $560 million to build the ITS booster, tanker, and manned spaceship, the raw carbon fiber order alone must be enough to build a lot of hardware. That, in addition to the test engines and tanks already built, shows that SpaceX is committed to building some serious hardware. This is not a mere exercise in PowerPoint engineering jam sessions…

On the other hand, I can’t see any way for SpaceX to afford to spend billions without major outside funding. Sure, they can use the profits from their commercial and NASA contracts for a significant chunk of engineering and prototype construction. At the moment that’s on the order of tens of millions per year, but that’s far short of the $10 billion development cost that Musk mentioned.

But there are some tidbits that indicate that Musk is willing to play the political games necessary to gain support from Congress and thus possibly get significant funding via NASA. Musk mentioned that the gigantic booster and spacecraft could be manufactured in Michoud, LA, which is where the SLS is being manufactured. The current congresscritters that support NASA and SLS as a jobs program might be persuaded to support SpaceX to some limited degree if it also means more jobs. Particularly if SpaceX gets substantial funding from other governments and outside sources, this means that e.g. NASA could cover a fraction of production and engineering costs but congresscritters get the benefit of far more spending in their districts…

42 engines? The N1 has 30. I sure hope I am not in Florida when this blows up.

Falcon Heavy will have 27 engines, and I have some faith that modern engineering is a bit more sophisticated than what the Soviets used for the N-1. That’s not to denigrate the Soviet engineers, but from what I’ve read they didn’t have the resources to do significant hardware testing before a launch. Each launch was a test of unproven hardware. That overall approach worked well enough for the predecessors to the Soyuz rocket, when a few dozen failures wasn’t enough to halt development. But the N-1 was just too costly.

From the vid, I would much rather have the tanker sent into orbit first then the people. This vid is made to demonstrate confidence in the reusable system, showing the people launched first, the tanker second after the successful return of the launch vehicle but practically it is bassackward.

And on that note, you want your Martian fuel-manufacturing process to be automated, so you can launch it first, re-fill the tanks, and then launch the humans. You don’t want humans waiting around on the surface of Mars for their ride home to be re-fueled.

But at this stage of the process, confident bluster is probably the right approach, so when you back down from it, you’ve still got a lot left.

Hate to be the pessimist, but I really don’t see man on Mars in our lifetime. Too costly and with the improvements in AI in the next few years, there will be nothing a man could do that a machine can’t do faster and cheaper.

Getting people to Mars and back isn’t something a machine can do instead.

I believe Musk has the willpower to push this through, but it’s a little hard sifting through the pie-in-the-sky initial plans. Can someone confirm they’re not utilizing the Buzz Aldrin cycler concept, are they?

To be fair, the same argument could be made for almost every human on the planet.

Musk’s argument for this BFR isn’t about collecting scientific information, it’s about making humanity multiplanetary. To me it looks like it’s designed to to allow people to exploit space as well as explore it.

I’m just glad someone is building something. Even a spectacular failure is a learning opportunity.

If he makes the thing all electric, he can get a $7500 rebate from the Federal government!

Interplanetary is so last millennium. Wake me up when he builds an interstellar rocket…

Raptor is actually a fairly small engine. It is physically not a whole lot bigger than the current Merlin 1D engine–most of the performance improvement is from running at a much higher combustion pressure (30 MPa vs. ~10). The vacuum version will have a large nozzle, but that’s relatively easy to scale.

So I don’t expect any major issues in their ability to physically construct the things. The design is a different question, but based on what they’ve said previously, they’ve explored a fairly wide range of size targets, and so I anticipate that they have a good handle on the scaling sensitivity and what the bounds are. They said already that the design optimized to a surprisingly small target, so it seems they explored much larger versions before converging on the current scale.

I doubt it. As lazybratsche said, they are already far along with their 27-engine design, and SpaceX has better methods than the Soviets had at the time.

Large numbers of engines has quite a few advantages, such as redundancy and economy of scale. The N-1 probably would have a great rocket once they figured out the initial design issues; they just never progressed to that point.

Besides, 42 engines is just plain good luck, especially when the first craft will be called the Heart of Gold.

I would have preferred to have seen an Infinite Improbability Drive.

This was answered in the Q&A: yes, the people will probably be launched after the propellant transfers. It actually requires several transfers (propellant is heavy…); the exact number will depend on how close they can get to their performance targets. The initial launches will probably have a relatively slow turnaround time (no landing back at the pad, etc.) so it might take months.

In the long term where flights are reliable and routine, and it takes only days to be fueled, maybe it’ll make sense to launch with people.

One aspect to their methodology which I was happy to see was a focus on sensitivity to performance shortfall. A problem that many aerospace projects have is that they can be essentially ruined by very small gaps from the targeted performance. Among other things, this is one reason why SSTO projects never seem to go anywhere; the margins are so slim that even a tiny slip can cut your payload in half, or even to zero. But SpaceX made the point that the refueling gives them margin flexibility: if there is a shortfall (the tanks are a little too heavy, or the engines aren’t quite as efficient as expected, etc.), it may require more refueling trips, but it doesn’t invalidate the overall design.

It’s not an Aldrin cycler. The same craft travels to Mars, lands, refuels, takes off, and returns. The cycler concept is good if you’re trying to optimize for propellant use; the SpaceX concept is that propellant is abundant if you can use ISRU.

This was asked in the Q&A, and Musk indicated that they believe their approach has better cost/mass, especially as transit time has significant impact on re-usability. It could be a future optimization, along with a propellant depot on the moon.

ETA: ^And what he said. :slight_smile:

The fully-fueled rocket weighs ~10,500 tons. We can guess that around 2000 tons of that are liquid methane. Methane has an energy density of 55 MJ/kg when oxidized, so that’s 110 terajoules of energy.

A “standardized kiloton” of TNT is 4.184 terajoules, so this rocket is approximately equivalent to a 26 kiloton bomb. The one that flattened Hiroshima was 15 kilotons. Not directly comparable of course, but certainly gives a sense to the energy scale involved.

Slides are here, in case anyone doesn’t want to listen to the whole presentation.

The slide labeled SHIP CAPACITY WITH FULL TANKS is pretty interesting. With a relatively light 200 t payload, they have 6 km/s of delta-V available for the MTO (Mars Transfer Orbit) injection alone. That reduces the transit time to as little as 80 days (though sometimes as high as 150 days). Huge improvement over the usual ~9 months for a Hohmann transfer.