I wouldn't be surprised if the first one (and several beyond) costs more to refurbish than to build new. But once they get the procedure down and figure out what parts actually need attention, it should come way down.

That would actually be incredibly surprising, shocking, in fact. There's really only so much work that can be done inspecting and refurbishing the rocket stages, they only have so many systems. They are precisely the opposite sort of beasts as the Space Shuttles which were some of the most complicated systems ever created by mankind. The Falcon 9 boosters are pretty much just engines, propellant tanks, avionics, and some miscellaneous bits. There's just not a ton to inspect and not a ton that can go wrong. More importantly, the vast majority of the cost of the vehicle is in the engines and tanks. The tanks are unlikely to degrade from launch to launch and are practically binary in their serviceability (they either work or they don't). Engines are incredibly expensive to manufacture and realistically the amount of inspection work that would need to be done to match that cost is just mind-boggling, and there's no good reason to imagine they'd need to do more than bore scoping. Especially since at this point if they find a potentially problematic engine, they just unbolt it and bolt on a brand new one. This increases costs a bit, of course, but it still saves a tremendous amount of costs. If the incremental flight cost of a first stage goes from the full manufacturing cost down to the cost of an engine or two plus recovery, inspection, and refurbishment, that's at a minimum tens of millions of dollars saved.

"Engines are incredibly expensive to manufacture"

Well, the Shuttle engines were, on the order of $50M or something like that. But the Merlin-1D, according to a former SpaceX employee on Reddit, is in the ballpark of $1M per unit. They've really streamlined the design for mass production, so even if thrown away after a single use, the Merlins are actually much cheaper than the Shuttle engines even if you take RS-25 reuse into consideration (9 launches per one RS-25 unit in average - so about $5-6M per engine-launch in just amortized manufacturing costs?). They're also much simpler and sturdier, much like the F-1 engines - despite not being designed for it, it turned out that those could have been used for about twenty-thirty times without major problems and without major component replacements. The RS-25s pretty much required general overhaul after every flight.

And the Falcon 9 first stage has nine engines, so that's about 9-14 million dollars worth of engines for the stage. That's in the ballpark of a quarter of the total hardware cost of the booster, and that almost certainly doesn't include the testing overhead for each engine.

Anyway, my point is, even if all that SpaceX had to work with for reuse was the main rocket body and a pile of the majority of the engines in working order, that would easily add up to tens of millions of dollars worth of equipment, and thus tens of millions of dollars of cost savings on reflight. There's practically no way that reflight could be more expensive than a new rocket, there's just not enough complexity in the vehicle for that to realistically be possible.

True. I've long held the view that low costs of launches basically require two things: the equipment should be cheap to manufacture and cheap to service. Shuttle violated the former big time and in all likelihood also the latter. F9 certainly satisfies the former already and there's no reason to think that it won't eventually satisfy the latter, too.

Note that the SSME puts out about 3x the thrust and has a substantially better specific impulse, so you should probably be comparing one RS-25 to 3-5 Merlins.

On the other hand, that SSME cost figure is probably in 1970s-era dollars....

I figure a lot of the extra time and money spent is confirm expectations on how well the individual components held up to the stress. Throw in a few flights and you might start to see more fatigue issues creeping in and the test is, are they occurring where expected or not. that temperature range they operate in has to have some strong effects too

They would be well served by doing destructive tests on a wide range of components even if they pass inspection. Further, while a simple design this also means they are relatively cheap to manufacture.

It's possible that they've already reflown returned engines without telling anybody. There are some indications that this has happened, and it would be a fairly safe thing to do on some flights (since the Falcon 9 can deal with an engine out situation, even at T+0). It's likely they'll do just that with engines and other components in the future as well, such as landing legs and grid fins.

Not only that, but they can then add diagnostic components for subsequent launches onto the rocket, which you don't generally need if you're going to toss it after one launch.

As nobody has done it before I believe they are going to spend as much or probably more in refurbishing the old stage than building from scratch.

As an engineer you get a lot of know how just doing something. The first time I used a CAD program professionally it took me like 10 times more effort and time than what it takes me now. Now is second nature.

Nobody knows what it takes to relaunch a rocket. The exciting news is that they will discover it. In the end it will be way cheaper.

Space shuttle engines have flown all the way to orbit multiple times. However the technology was a bit different. Also see x-17 and dc-x which did multiple rocket powered flights, or all the way back to Me-163 Komet.

> Nobody knows what it takes to relaunch a rocket

Well, apart from Blue Origin, I guess. They've done four sub-orbital missions now, with their production hardware. I know there are differences, but it's a similar problem-space, so they are an existence proof that it's possible.

Initially, I think you're right.

Eventually, I think they're aiming for airliner-style reuse. Engineer any frequently-breaking parts to be better, then check every X flights once they've got the kinks worked out.

I am guessing it will be more like an engine rebuild on an engine for a 777 - tear down and xray of key parts. The turbopumps and compressors in a rocket engine are very sensitive, high precision parts.

> The turbopumps and compressors in a rocket engine are very sensitive, high precision parts.

SpaceX deliberately engineered lower performance, higher reliability engines. Think minivan instead of Ferrari. The SSMEs were genuine marvels of engineering, but that came at significant cost to reusability.

Also I think 2 of them can flame out and Falcon can still achieve the primary mission. Minivan with an extra tire.

That depends very much on what the primary mission is. For LEO missions like the CRS flights, then sure. For GTO launches, they probably don't have the margin needed.

For a GTO mission the Falcon 9 can have a single engine out and still complete the primary mission.

Is this still true at max GTO payload weight?

Remember that with reusability in play, the guidance system can decide "screw the first stage landing, the payload is more important" and simply switch to an expendable launch profile if it turns out that you have more engine problems on the first stage than you designed for. Then you use the reserve fuel to get the $300M satellite safely to orbit instead of getting the $30M stage safely to the ground. The customer is happy and you've only lost a unit for which you have a dozen of direct substitutes in storage anyway (SpaceX already doesn't have enough storage space, apparently!).

It might depend where in the flight the engine fails, assuming we're talking about the first stage. If I had to guess, a GTO mission (placing a payload in a 36,000 km x 350 km elliptical orbit) could still be successful with an engine failure in the first stage after 65-75% of the fuel had been burned, and the vehicle was generally lighter, allowing the burn time to be adjusted longer to reach the same velocity at the point of first stage cutoff and separation to ignition of the second stage. But this guess comes from dozens of hours spent playing Kerbal Space Program....

Or a minivan with runflat tires. They'll still get you where you're going, but a bit slower and not quite as efficiently.

Isn't that the kind of detailed tear down that forced the shuttle in to being more expensive than traditional systems?

Indeed. Hopefully SpaceX can do less, since their hardware is more robust. The Space Shuttle engines were driven at the edge of failure because they needed every last bit of performance. SpaceX's designs tend to go for robustness at the cost of efficiency, so they ought to need less work.

Possibly of interest, the giant engines used in the Delta IV Heavy have 80% less part count than an SSME, but they were never intended to be human-rated: https://en.wikipedia.org/wiki/RS-68

Not that it has brought launch costs down much, considering how the ULA loves their juicy government contracts for billion dollar NSA/NRO satellite launches.

IIRC the Falcon 9 engine design is the opposite of an RS-68, which is the "one huge fucking engine" design philosophy, which is why a falcon9 has a cluster of many smaller engines and is tolerant of an individual engine failure during launch.

> IIRC the Falcon 9 engine design is the opposite of an RS-68, which is the "one huge fucking engine" design philosophy, which is why a falcon9 has a cluster of many smaller engines and is tolerant of an individual engine failure during launch.

Indeed! A Falcon 9 mission has lost an engine mid-flight, and the flight system adjusted the burn duration on the fly to compensate:

http://motherboard.vice.com/blog/spacex-s-non-fatal-failure

Also, every shuttle flight is human-rated. I would guess that SpaceX wouldn't use "refurbished" engines for manned flights, while they could cut down on the thoroughness of checks for lower value payloads.

That seems likely at first, but it'll be really exciting when that switches around and reused engines are the preferred choice for manned flights since they've been proven.

After all, you never put passengers on the very first flight of a new airliner. You test it out first, and then once it's demonstrated you start using it for real. We'll know that reusable rockets have really made it when the first flight of a new one is done as a test flight without risking any lives, and only afterwards is it put into normal use.

Plus for the orbiter refurb they had to check the entire heat shield, flight control systems, crew habitats, etc. etc. Easier when it's just the rocket + fuel tank.

Indeed. The heat shield in particular was extremely delicate and tended to get beaten up a lot coming back, so that was a huge amount of work to get ready for the next flight.

The entire spaceplane concept drastically complicated things. Capsule-based return systems are vastly simpler to operate, which is despite everything the Soyuz has been flying continuously this whole time.

Downside is you can't capture enemy satellites and return them to ground for study/reverse engineering.

This is how NASA got many billions of funding from the USAF for part of the space shuttle program - look through the list of flights in the 1980s where the payload was an unspecified national security satellite. The capability was promised and it was funded for it - but almost never used. The actual ability to grab a big thing and return it was hardly ever used, such as the LDEF mission. It doesn't help that a huge percentage of Soviet satellites that they might want to have grabbed would have been unreachable from a space shuttle launch site (IMINT satellites in 90 degree inclination polar orbits, satellites in molniya orbits, radar naval observation satellites in MEO orbits) all beyond the orbit/delta-V capabilities of a shuttle mission. And of course geostationary orbit where no human has ever gone.

The first few, probably yes.

But I'm pretty convinced they're trying to avoid this. The engines have been test-fired many, many, times (30 starts is a figure that has been mentioned a few times. Remember that the first recovered stage (Orbcom) was returned to the launch pad and went through a static fire just a few weeks after landing. It seems clear to me that they've got their eye on reasonably-quick-turnaround reuse.

I bet they're less concerned with # of firings and more with fatigue/wear from atmosphere reentry. Very hard to test on ground.

They've now got 45 intact engines back, from a variety of re-entry regimes. We know they've done an all-up static fire. Several of the returned stages have had engines removed and I imagine some of these have been tested.

If there's a major problem, they already know about it.

> however, that will still be cheaper than completely building a 1st stage from scratch.

For a lot of companies this sort of thing is not at all cheaper. They're set up for production, not for disassembly, inspection and refurbishment. It'll be interesting to see how SpaceX tackles this as their launch service solidifies.

Indeed. With airliners, there are whole separate companies doing things like modifications.

The research on used components would also immensely help in the design of new rockets for spacex.

They have already done a full-length static fire of a recovered first stage. That means a full simulation of a flight with the engines running and the stage bolted down to a test stand. Obviously you can't simulate the dynamic and aerodynamic loading on the airframe, but the engines and tanks have shown that they can hold up and that was the biggest difficulty with the space shuttle. I think it's safe to assume at this point that recovered stages can be reflown without refurbishment beyond what is required to prepare a newly-manufactured stage for flight. I would put money on SpaceX putting dummy payloads or heavily-discounted commercial payloads on recovered stages and just reflying them over and over until something shows signs of stress or breaks in flight.

If they have to go through all those checks, then that is a failure for several aspects of the program. The goal of reusability has been parts that don't need to be checked. The engines have been run over and over on test stands. They shouldn't have any faults after a single launch. If they do, then the test program wasn't accurate. If hardware needs to be redesigned to meet that goal, then it also needs to be re-tested and much of the program is back to day one. I'm sure they will do checks, but I am equally sure they dread finding something.

I;m reminded of the story of Shuttle's tires. They were designed to last four or five flights, but famously were replaced every time. SpaceX wants to end such things. Parts are to do what they are designed to do. A part designed to last multiple flights without the need to be x-rayed every time, isn't going to be x-rayed every time.

It's not a failure. It would be sheer negligence to not have a full check for the _first time they've ever done this._

If they have cracks in things that are designed to go through multiple flights without cracks, and those cracks haven't appeared in testing, that's a failure. Any such finds in checks today means the necessity of checks again next time, and the hundreds of times after that. SpaceX needs to find nothing so that they can then decide at some later date to stop doing the checks. Not cracking open the fuel pumps after every flight is a necessary part of the planned cost reductions. If inspections suggest that the design needs to be checked every time, as with Shuttle, then new design is needed and the testing program hasn't met its goals.

> If they have cracks in things that are designed to go through multiple flights without cracks, and those cracks haven't appeared in testing, that's a failure.

No, it's real world engineering. You can't simulate everything when you're doing something totally new. You need to make tests on real hardware to confirm that your model is good.

Exact. It is not really any different than going from in-house to beta testing in software to get those real world bugs that always come up -- it is a known step of the design process before even starting.

Especially aerodynamics. Modeling fluids alone is very hard and intensive, and that's before you model surface interaction, possible ablation, temperature, shock, and all the other forces present in a dynamic situation.

Not doing any x-ray testing would be completely negligent.

"those cracks haven't appeared in testing" - this is the testing. You don't just simulate the stresses, you have got a bunch of devices that have experienced atmospheric reentry in different conditions, and now you could do the tests on them.

True, but if this additional testing reveals issues not revealed earlier, they'll just iterate again until they won't have to X-ray every plate after every flight, because that is clearly not the MO they have in mind. In the worst case scenario, they might need one extra vehicle generation (BFR?) to actually get there, but I'm quite certain they will eventually get there. Otherwise there's just no point to that company.

Even if you design a component to last for the life of the entire system, that doesn't mean there's no point checking it through that life. There will always be the pieces at the bottom end of the bell curve. You must catch them.

Look at the aviation industry for a good example. A pilot might check certain parts of the aircraft before starting it up each time. They'll likely check more things more closely for their first flight of the day. Then it'll get ever closer inspections every X days by maintenance personnel. These checks catch faults.

The idea of the program certainly is, that the rocket can be flown several times without requiring extensive checking/maintenance. However, as there are no experience with reusing the Falcon 9, it is the proper scientific/engineer approach to extremely carefully inspect the whole rocket to document its conditions to analyze possible failure reasons. The engines have been endurance tested on the ground, but not after reentry. So one has to learn about this. Ideally, all those inspections show that the rocket is just fine, but one can know only doing the proper science. With those experiences gained, it is possible determine the service intervals for standard use.

That's true, but it's not at all clear that scarygliders was talking only about these first few re-used rockets as it was a completely unqualified statement. I certainly read that post as talking about refurbishment generally, in which case sandworm101's reply seems to me to be perfectly reasonable and measured.