SpaceX and RLV

 guest blogger John Hare

SpaceX has done an amazing job of creating and flying a new launch system from scratch. There have been a few snipers criticizing their designs, procedures, schedules, and anything else the anonymous crowd can dream up. The fact remains that a relatively new team has done a tremendous job with less cash expended than many traditional launch people thought possible.

The very excellence of SpaceX points out one of the fundamental differences between ELVs that might be reusable, and RLVs designed around continuing operations with each airframe. A really good company splashed 3 vehicles before getting one in orbit with a test payload. The bottom line is that they are building at least 5 complete launch vehicles before collecting revenue from a single delivered payload.

I am an RLV fan. If it were to cost 5 times as much to build an equivalent payload RLV as a Falcon 1, then they would be equal in total cost and revenue returned after the first completed revenue RLV flight. It will almost certainly cost more than 5 times as much money per airframe for the RLV vehicles, the question is “how much more?”. Just on operations, weekly flight RLVs could fly 50 times a year, so even if it were 50 times as much per airframe, it just might be worthwhile to build the RLV for the revenue it would return the second, third, and fourth years.

That would be if all else were equal. It is not. Development money up front is frequently far more important than operational costs in the misty future. The Shuttle is the classic example of skimping on development at the cost of unaffordable operations. The conventional wisdom is that it will cost far more to develop a minimal RLV than a conventional ELV with tried and true technology.

The conventional wisdom has a few problems. Testing is the major cost in spacecraft development. An RLV test schedule could have flown bunny hops with minimum propellant of the first stage only to characterise early flight handling. Falcon 1 flight 1 was flown all up for good reason, once it leaves the pad, it is expended whether it reaches orbit or the next island. If it is not coming back, why waste the airframe on a tenth of a test. The problems of flights two and three might have been ironed out without losing an airframe if tiny steps could have been taken.

It is my opinion that a sufficiently funded and well run RLV vehicle development effort could easily cost less than 10 Falcon 1s. It could afford to fly dozens of dummy payloads to make very sure of most uncertainties. At the end, it could be superior competition to a vehicle type that has flown a total of ten times with 70% mission completion, or even 100 times with 97% mission completion.

The following two tabs change content below.


I do construction for a living and aerospace as an occasional hobby. I am an inventor and a bit of an entrepreneur. I've been self employed since the 1980s and working in concrete since the 1970s. When I grow up, I want to work with rockets and spacecraft. I did a stupid rocket trick a few decades back and decided not to try another hot fire without adult supervision. Haven't located much of that as we are all big kids when working with our passions.

Latest posts by johnhare (see all)


About johnhare

I do construction for a living and aerospace as an occasional hobby. I am an inventor and a bit of an entrepreneur. I've been self employed since the 1980s and working in concrete since the 1970s. When I grow up, I want to work with rockets and spacecraft. I did a stupid rocket trick a few decades back and decided not to try another hot fire without adult supervision. Haven't located much of that as we are all big kids when working with our passions.
This entry was posted in Uncategorized. Bookmark the permalink.

19 Responses to SpaceX and RLV

  1. anonymous says:

    I think the main thing a RLV development effort needs to emphasize is reliablilty. It will take another 10 years or so for the spacecraft industry to adapt to the relatively low prices of spacex, so trying to convince them with even lower prices migth be futile.

    But reliability is something that ELVs simply can not provide. It is almost impossible to get a demonstrated reliability of 99% for an ELV, because even if you have 1000 successful launches, each launch uses a new vehicle. (There are literally millions of expendable rockets produced each year for weapon systems. They are much simpler than the typical RLV, yet their reliability is not that high)

    Developing a RLV has just become much easier. If spacex succeeds with the falcon 9, they will probably flood the market with refurbished merlin engines that traditional satellite operators do not trust to launch their multi-million satellite. So a RLV development company should be able to buy a bunch of merlin engines for a few million USD and then concentrate on the airframe and control software.

  2. Axel says:

    Wouldn’t it be likely to loose some prototypes during RLV development, too?

    I’m a RLV fan too. It may be wishful thinking, but… maybe Elon Musk is aiming for a RLV, but decided to go a different development path, an evolutionary one which gets paid by serving the satellite market from the beginning? How serious is SpaceX about reuse? Currently they seem to be more interested in Falcon 9 development than in making Falcon 1 partially reusable.

    Could they hope to make F9 first stage reusable? The engine cluster can be throttled down to 1/9th by using just one engine, and single engines can be throttled to about 1/3rd, if I read that correctly. How far away is F9 stage 1 from vertical landing? Does it have enough thrust vector control at that low a thrust?

    It does not have landing gear, but maybe it doesn’t need one. It could land on a crane or on a wire rope. Traditionally we assume vertical landing must be done with the bottom, where the engines are and landing gear attached to the bottom must dampen shock of touchdown and keep engines away from the floor, which makes landing gear difficult and heavy. Instead put a hook or something similar on the top, where the parachutes would be attached and fly this hook onto the crane or wire, which does the shock dampening.

    Hook landing (or maybe you have a better name?) combines many of the good parts of vertical landing and aerocapture. Low weight gear on the vehicle, low shock on landing, easy handling after landing. You can land with engines as far away from the floor as you like. Precision constraints for a smooth landing can be much lower and are therefore much easier to meet by the vehicle.

  3. Jonathan Goff Jonathan Goff says:

    I agree with you somewhat, increasing reliability is going to be a key. But as I mention in my most recent post, I don’t really think satellites are going to be one of the key markets that RLVs need to focus on. The other markets I mentioned are far more elastic with respect to price than satellites. You need reliability for those, but you also need cost reductions. Fortunately, those two often go hand in hand for an RLV. For instance, if you aren’t crashing your RLV frequently, you’re getting to amortize it over more flights, which should help keep costs down.

    As for whether the availablity of Merlins will effect RLV development, I’m not sure. I’m not sure I’d want to use a salt-water soaked Merlin for my RLV. But with the large number being produced, I could see buying a fresh Merlin, then reusing it a bunch of times in a less traumatic manner.

  4. Jonathan Goff Jonathan Goff says:

    Wouldn’t it be likely to lose some prototypes during RLV development, too?

    Sure, but it’s less likely, especially if the RLV is designed as an RLV and not just a reusable ELV (there is a difference). Once you realize that in an abort the vehicle doesn’t have to be lost every time, you start making different design decisions that lead to a much higher probability of recovery in event of an abort.

    Second, I don’t think a Falcon 9 can throttle down enough to hover. The engines only go down to about 2/3 not 1/3 IIRC.

    Lastly, I don’t like landing methods that can only work if you make it back to your landing site. Having legs means being able to bring the ship down if something goes wrong.


  5. john hare says:


    I tried to reply this morning when my computer wigged out. What Jon said, (except he said it better).

  6. Axel says:

    do you have a source for the Merlin throttle limits? Its not easy to find info on that. The 1/3 number I guessed from your(?) comment on
    “Yeah, while our pintle injector engines turned out to be throttleable down to at least 30% without any problems (or any real effort to make them deep throttleable) […]”.
    Neither can I find numbers on the F9 first stage dry weight. What is your guess?
    You say: “I don’t like landing methods that can only work if you make it back to your landing site.”
    Its hard to argue with “don’t like”. All I can say is, that SpaceX starts from traditional space technology, i.e. ELV with paying customers. Every time they can invent a way to reuse something, they win. Most of their launch happens over the ocean. If they can’t make it to a dedicated landing location (may be the launch site or a down range recovery site), the vehicle will splash/crash into the sea and is most likely be lost or at least severely damaged by impact shock and salt water. (Well, even over land legs will save the vehicle only if you still have hovering capability when you reach the ground. How likely is that in case of a serious engine failure? And you need flat enough ground with few enough people around to endanger by doing an emergency landing.)
    Um, thinking about failure modes I wonder: what happens to a hot engine when falling into water? Did anyone ever try this? Hover to the surface of the water and turn off the engine?
    P.S.: forgot to say in my last comment that hook landing will prevent the vehicle to fall over after landing. Its kind of obvious, but important because it allows to use the traditional long and narrow rocket structure for vertical landing.

  7. john hare says:


    While I don’t think hook landing would work for SpaceX, it might be useful for certain other classes of vehicle. Like a semi expendable sounding rocket. Have a circle of hooks at the top of the vehicle that extend like on carrier based aircraft. Have a net suspended very high off the ground and let it catch the rocket at terminal velocity. With enough net altitude, perhaps with ballons, the catch could be gentle. Eliminates parachutes, reserve fuel, landing gear, and throttling requirements. Needs accurate desent control though. Also a possible market for recovering expendable military munitions for reuse during training.

  8. Axel says:

    maybe a small sounding rocket can survive a hook connection at 50 to 100 m/s (or whatever terminal velocity of the rocket is). A large structure designed for minimal mass most certainly can’t. I was thinking more along < 0.5 m/s at connection time. More like the Vertijet but with a larger hook and a much larger catching wire. I feel that hook landing should be a good alternative for all vertical landing designs, except maybe for the military ones which need to be able to land without assistance at any cost.

    did you consider that most commercial air transport and almost all long-distance air transport can only land and take off at locations with special facilities? If those planes have to land outside an airport they will most likely be damaged severely. So why not land a RLV only at locations with a wire to hook on available?

  9. Jonathan Goff Jonathan Goff says:

    You can’t extrapolate from our experience with a pressure fed pintle engine to someone else’s pump-fed pintle engine like that. I’m pretty sure they aren’t designed to throttle that deeply. I can always ask, but they’re not made for the rapid throttling you need with a VTVL vehicle.

    I agree with you that ocean recovery is going to be a challenge for SpaceX. Not impossible, but it’s not a very ideal way of recovering complex aerospace hardware.

    The problem with them using your hook landing system (other than my previous voiced concern about having absolutely zero abort landing survivability) is that then they would have to launch over land. With a vehicle that can’t make a safe abort landing at any point during its burn….I just don’t see that happening. It’s an interesting idea, but I think way too limited for my taste. I think having a vehicle that can survive aborts well is going to be a key to making it reusable enough to be commercially effective.


  10. Jonathan Goff Jonathan Goff says:

    The problem is that there are a lot more airports in the world than places setup for your proposed hook landing technique. While it’s still possible to have a failure far enough from an airport that you’d lose the plane, you’re a lot more likely to have a failure far enough from a hook-landing facility that you’ll lose the rocket. It’s pretty much a guaranteed way to wreck a lot of vehicles in development and operations–and that’s assuming that the concept could be made to work in the first place.

    I’m not trying to diss your idea, just trying to point out challenges with it. There are all sorts of tricks and gimmicks for recovery that you can do to reduce weight, and make things simpler–when everything else works perfectly. A good design is one that can still do its job even when things don’t work perfectly.

    I really think that either wings and wheels or powered landing with landing gear are the two main approaches that pass the robustness test. Sure it’s extra weight, but the added flexibility, and the added odds of recovering your airframe are worth it IMO.


  11. Axel says:


    my intuitive assessment of the /added/ risk differs from yours. You may be right for vehicles designed for reuse from the beginning, especially when winged and capable of glide landing on an airport. With vertical landing I feel having an extremely reliable engine systsem is a must.

    Back to SpaceX. Falcon does not have the option to go to an airport. But it can go to space, occasionally. Going to space is dangerous.

    This weekend I was looking for some numbers on the F9 first stage landing weight. Found a video which says the tank plus mount for engines is about 4 t, 9 Engines are another 4.5 t, so it is at least 10 t plus interstage, some plumbing, wiring, avionics, residual fuel on tank walls and in tank etc… estimates 15 to 23 t dry mass. A Merlin engine gives 40 t thrust (with the old pump), so the numbers may match in the end, if SpaceX did have vertical landing in mind as an option during the design.

    To illustrate vertical hook landing, I did some sketches:

    Is fast throttling really that important? To hover with an accuracy of maybe a meter in height, you need to adjust thrust carefully, but no need to change it quickly. What must be quick is thrust vector control. If it is done by differential throttling, it is critical. But SpaceX has hydraulic actuators for that (even though I don’t know if they will use them on Falcon9).

    Mabe we just ask SpaceX…


  12. Jonathan Goff Jonathan Goff says:

    You do need the ability to throttle fairly quickly. I really don’t think that a stock Merlin would be capable of something like this. More importantly you still have to have much more precise landing accuracy for the first stage than you’re going to get with a parachute system like they’re using. For recovering an ELV stage, I still think that mid-air recovery makes a lot more sense. No need to do VTVL type hovering, the landing pad can move to rendezvous with your stage, and the technology is already demonstrated.

    Once again, I’m not saying your idea couldn’t work, I just haven’t been convinced yet that it’s really the best approach even if it could be made to work.


  13. Did I miss something?

    I thought SpaceX’s plan was to recover the first stages of both the Falcon 1 and the “9.” Has that changed?

  14. Here we go:

    Dated May, 2007:

    “””””The Airborne Systems team will design and manufacture parachutes andrelated equipment to recover both the First and Second Stages of the Falcon9 Rocket and the Dragon Capsule.”””””

  15. Jonathan Goff Jonathan Goff says:

    You are correct that they want to be able to recover the first stages of Falcon 1 and 9. What people were debating here was whether that would actually end up working in the end. I’m pretty sure they can eventually get parts back, the question is how reusable it will really end up being.


  16. Jonathan:

    I got the following reply from Roger Gilbertson (Media Coordinator) at SpaceX:

    “””Yes, first stage recovery is a high priority for SpaceX going forward. Now that we’ve demonstrated the system end-to-end by achieving orbit, enhancements such as recovery become high priority, with the requirement that we don’t add too much mass or jeopardize primary mission success.

    Yes, F9 2nd stage is also definitely in the plans, but first things first.”””

    By the way, have you seen this?

  17. Jonathan Goff Jonathan Goff says:

    Yeah, I was just saying that “high priority” and “will actually work” are not the same thing. I’m not as pessimistic as most, but there is still a lot of work they have to do, with no guarantee that their approach will actually work in the end.

    As for the RBCC SSTO paper. No, I hadn’t seen it. Once again, I’m somewhat skeptical. I think there are better ways of doing reusability than RBCC or splashdown recovery.


  18. Axel Walthelm says:

    A recent tweet reminded me of this discussion:
    “We’re going to try to catch the Super Heavy Booster with the launch tower arm, using the grid fins to take the load”

    At the time I wrote an email to SpaceX. No idea if it got any attention. It has been a long time. Re-reading the discussion a lot of it still sounds valid and some of it even has become reality. Looking forward to see if the grid fins turn out to be good hooks or if they become anchors for a dedicated hook mechanism like I imagined in

  19. johnhare john hare says:

    I had to go back and read the post and comments as I have forgotten much of the discussion and what I did remember was quite hazy. I agree with your analysis of the discussion. One thing I am grateful for is the knowledge that many people can have a similar idea without copying or stealing concepts. Inverse aerobraking is one that I thought was unique until I learned that you had done it earlier and better.

    It is interesting the number of ideas that were kicked around back then, and the percentage of them still valid or in use today. I like to think the people here had some positive influence on the industry, though I’ll likely never know for sure.

Leave a Reply

Your email address will not be published. Required fields are marked *