Before I go into detail on any of the two stage to orbit (TSTO for the uninitiated) approaches that I mentioned in my post last week, I’d like to briefly discuss what I think is the key issue that drives the design and development tradeoffs for reusable TSTO launch vehicles. That issue is: how do you get the first stage back after a mission, and ready to fly again?
This article will focus on the key tradeoff that stems from this question: whether to try and recover the first stage downrange, or whether to try and perform some sort of return to launch site maneuver. The answer to this question is probably the number one driver of what approach one takes for developing a TSTO vehicle.
RTLS vs. Downrange Recovery
As I pointed out in my brief discussion about SSTO vs. TSTO approaches in Part I of this series, attaining orbit is mostly about building up a lot of horizontal velocity, and only a little bit about gaining vertical altitude. For performance optimized TSTO ELVs, the first stage often imparts a significant portion of the overall delta-V (especially for ELVs delivering satellites to GTO or GSO). This means that it ends up coming in hot, fast, and a long way downrange from the original launch site. Now, there are several different approaches to deal with this problem (or avoid it altogether).
One option is to just let the stage come down where it wants to, and recover it downrange. Downrange recovery can take several forms including recovering a stage out of the ocean after a splashdown, landing the stage at a downrange site and ferrying it back (either by rocket flight, a carrier plane, or by truck, train, or barge), or it could involve mid-air recovery of part or all of the first stage. While downrange recovery may is the general approach that probably imposes the smallest performance penalty, each of the actual approaches to down-range recovery have some pluses and minuses.
Let’s take splashdown recovery first. Falcon-1 is an example of the splashdown recovery. The stage separates where a typical ELV would want to have a staging event, and then (hopefully) it’s fished out of the ocean and refurbished for reuse. Some of the benefits of splashdown recovery:
- Splashdown recovery is probably one of the easiest and best understood methods for recovering a traditional ELV-like first stage.
- There’s a large experience base to use as a foundation for carrying out such a design.
- Even if your flight rate is low enough that it isn’t saving you much money, you’re still able to learn a lot from being able to perform post-flight inspection on the propulsion hardware. Thus, even if you aren’t flying enough to save a lot of money via recovery, it will help your reliability.
- Ocean splashdowns don’t require anywhere near as heavy of recovery equipment as land parachute landings.
But they also have several drawbacks:
- Trying to make a complicated rocket engine sea-water compatible, especially a turbopump-fed rocket engine, is not a trivial task. Material selection, and getting the stage out of the salt water (and cleaned out) as quick as possible are all required.
- There’s a lot of time and labor involved in hauling the stage back, cleaning it out, making sure nothing got damaged on reentry or splashdown, testing everything to make sure it’s still in working order, etc. This fundamentally limits how frequently you can refly a given stage. It also translates into a lot of extra personnel and labor-hours required above and beyond what you would normally need to just build, test, and fly an expendable vehicle.
- The wear and tear from ocean recovery, splash down, etc. are likely going to limit the number of reflights you can get on a stage or engine before major overhaul or outright replacement.
- Your potential launch sites are limited, since you need a large body of water on which you can drop big heavy hardware. Most likely (for US entities) that means flying out of one of the existing ranges like Wallops, Vandenberg, or Canaveral. These locations, while excellent for flying missiles, and while also improving their commercial friendliness over time, are still a long way from the environment you want to be operating a reusable launch vehicle out of.
- While it’s possible to design a launch vehicle splashdown recovery first stage in such a way that a first stage failure doesn’t necessarily imply the loss of your cargo, it is much harder to design such systems for graceful abort modes. Unless the upper stage is also designed for splashdown recovery (with the payload designed for it as well), a stage failure probably will result in loss of payload. This loses you one of the big potential advantages of reusability–graceful and intact aborts.
The idea behind mid-air recovery is that instead of allowing the stage to crash down into the water, you instead snatch it (or a high-value part of it) out of the air using a helicopter or other sort of aircraft aircraft. This is similar to how Genesis was supposed to be recovered, and was the method used for recovering a lot of the film capsules from early spysats. There are actual serious players looking at this idea, but I don’t know if it’s supposed to be public knowledge yet, so that will have to be a post for another day. There was also a paper floating around by a company that does mid-air recovery work, including work for the SpaceHab ARCTUS project. If I can dig it up again, I’ll probably post about that as well.
Anyhow, here are some of the benefits of mid-air recovery:
- No salt water contamination in the rocket hardware! This greatly cuts down on the amount of work that needs to be done to turn a stage around. No need for decontamination. No need for stripping down hardware. Probably eliminates the need to “requalify” the propulsion system before reflight.
- Gentle, low-shock recovery is much less likely to damage stage or propulsion hardware, also making it more likely that the hardware can just be reused after some inspection.
- There are companies that specialize in this sort of thing, and you can just rent their services instead of trying to do this in-house. They aren’t cheap, but they’re a lot cheaper than building a new stage every time.
- Your propulsion system is going to be in about as close to the same condition as it was when the engine shut down as you’ll get for any recovery technique–this makes it a lot easier to get good reliable data on wear-and-tear on the engine, so you can improve the quality over time.
But here are some of the challenges:
- Complex recovery technique. Sure, you can practice it a whole lot for not too much money, but there is some increased risk of failing with the rendezvous or recovery operations, which could occasionally cost you a stage.
- Weight limits. Even with the latest techniques, which can recover payloads up to 80% of the maximum cargo capacity of the helicopters, you’re still limited to around 22klb or less. Depending on the size of your stage, this may mean that you can only recover part of the stage (like say the engines). That’ll still likely save at least some money, but it’s not as big of a win as getting the whole stage back intact.
- There may also be issues with trying to recover a big, but fluffy stage. Depending on the weight distribution, there could be some real oscillation issues (like what happened when they tried to move the Roton ATV under helicopter).
- Range issues. Depending on how far downrange your stage comes back, you might need to also rent not just a helicopter, but some sort of barge to operate the helicopter off of. This will increase the amount of time it takes to turn a stage compared to if you could just fly it back.
- Like with splashdown recovery, this method of recovery still doesn’t give you graceful and intact recovery methods in the case of a first-stage failure. With dump valves and two helicopters, and a mid-air recoverable upper stage, you might be able to recover the payload over part of the trajectory, but you’ll still have zones where a failure means sure loss of the payload (or a launch escape abort if you’re flying people). It isn’t a showstopper, but it does reduce the upside somewhat.
- Due to challenges #1 and #5, you probably still need to launch out over the ocean, which means that once again you’re still going to face the issue of launching out of an existing missile range. Basically, since there’s a chance you could biff the in-air recovery, you have to do this over an unpopulated area. And since your vehicle doesn’t likely have graceful failure modes, it’s more like an existing ELV than a more traditional RLV, and will probably be treated as such by the FAA and the ranges. Not a showstopper either, and it might just be possible to pull this off with an over-land launch if you can find a sufficiently deserted area, but definitely a challenge.
Mid-air recovery is probably too weight constrained for something like a complete (but dry) Falcon IX first stage, but might be an interesting option for recovering the Falcon IX upper stage or the Falcon I first stage. It’d also probably be just the right size for recovering the first stage if they hadn’t canceled the Falcon V. Other than the weight limit, there’s some real benefits of this approach over the traditional splashdown technique.
Downrange On-Land Recovery
This type of recovery can take several forms. It could be a powered VTVL landing at a downrange pad. It could be a powered or glide landing for a HTHL. It could be a parachute and airbags landing (like Kistler, just downrange). But basically you have the thing land, on the land, downrange, and then fly the thing back, or ship it back.
Here are some of the benefits:
- Much more efficient, performance-wise, than any of the RTLS approaches. You can still stage at the most optimal staging velocity, therefore making your upper stage design a lot easier. You also get a lot more payload per given takeoff (and dry) mass.
- At least some of the RTLS approaches can also sometimes use this as a performance enhancing option–in case you need to launch a bigger payload than you can handle with a normal RTLS trajectory.
- Unlike mid-air recovery, this recovery approach can scale up to fairly large sizes.
- In emergency cases for RTLS approaches, you may want to be able to land your vehicle at alternative downrange sites anyway.
- Unlike the other two downrange recovery options, this option is a lot more compatible with intact and graceful aborts.
And here are some of the challenges:
- A given launch site will typically have its launch azimuths (directions in which you can launch) restricted a lot more for downrange land recovery than it will for an RTLS vehicle. This is because you need to have a suitable place downrange where you can actually land. This makes downrange recovery vehicles less flexible than RTLS capable vehicles.
- You need facilities at both ends, especially if you intend to fly the stage back after landing.
This may entail having almost as many launch support people at the downrange site as at the initial site, which greatly increases the fixed costs of such a system. Probably not quite double (since you don’t have payload processing facilities there), but it’s a non-trivial expense.
- If you do a rocket powered return, you’ve now effectively halved both your flight rate (as you have to do two launches, two landings, two ground preps, etc. per a single paying flight), and halved the number of revenue generating flights you can get out of a given airframe. Both of these directly affect the bottom line.
- If the return flight is a rocket-powered suborbital flight (as per AST’s definitions), I think that each of your downrange sites will need to be an FAA licensed launch site, and you will need launch licenses for all of the return flights. Now, once you have one launch license to base things off of, getting additional ones should be easier, but its still extra paperwork. Also, your Ec and MPL calculations are going to be different for the return flight, because your IIP will move at different rates over different areas under your groundtrack for the two trajectories (not to mention mission-critical operations will occur with your IIP over a different location). All of this stuff has to be taken into account.
- If you have a jet powered return (either using a carrier aircraft, or if the stage has built-in jet engines), you now need to deal with the aircraft side of FAA, which may entail getting the vehicle type-certified. I’m not certain, but having a vehicle that operates under both regimes is likely going to make things a lot harder, not easier. Being unusual is not a virtue when dealing with regulators. If you’re using an existing carrier craft, that’ll make things easier however, as it is purely a subsonic aircraft, and thus a lot closer to what FAA is used to dealing with.
- If you try to return the stage via trucking or train, now the stage has to be “roadable”. Which means making it skinny enough to fit on existing transports. While this may be feasible for some smaller, dense-propellant RLV stages (after all I think that Falcon IX is roadable), it is a constraint on the size of stage. And the aspect ratio roadability forces you into is not as ideal for VTVL stages. VTVL stages want to be shorter and squater than typical rocket stages.
- If you return the stage via trucking or train, you now need heavy moving equipment at any downrange sites, experienced heavy equipment personnel there, and it’s going to cost you a lot of extra time. All of these things add cost, and slow down your turn time.
Conclusions: The Case for RTLS
Now, I probably ought to clarify something. I don’t think any of these downrange recovery ideas are stupid. If done right, they can save a lot of money compared to a purely expendable system, while also increasing reliability by allowing for post-flight inspections and the like. Especially with the downrange land-landing techniques, you can get all of the benefits of traditional RLVs.
In other words, while there are some challenges with downrange recovery, there are often some real benefits. There are some cases where using these downrange recovery approaches really is the best choice. SpaceX and the others looking at these approaches aren’t being foolish by pursuing them. I just think that the inherent limitations of this sort of reuse (especially the first two options–splashdown and mid-air recovery) will probably prevent it from being a revolutionary as opposed to a modest, evolutionary improvement over a purely ELV approach. Now, in the near to medium term, even when RLVs first start flying, they’ll likely be relatively quite small compared to the EELVs (for reasons I’ll go into in a later post). Which means that approaches that allow existing ELVs to become somewhat more reusable, and improve their economics somewhat are actually useful. I think that while small RLVs will bump ELVs out of the people, light cargo, and propellant markets very quickly after they enter the field, there’ll still be payloads that are too big for RLVs that are small enough to be economically viable in the near-to-medium term. The ULA’s, SpaceX’s, and Sea Launch’s of the world will still have a useful role to play for some time yet. Particularly for launching bigger payloads like Bigelow stations, transfer stages, etc. So, having ways to improve them is good, even if they aren’t necessarily going to change the world all by themselves.
As for the last options–land recovery downrange, it actually does have the potential to be revolutionary. But the approach still has some serious economic and regulatory drawbacks that are sufficient to make one start looking at RTLS approaches, even though they may be less optimal from a purely performance-based standpoint. There are four primary (and one somewhat oddball) RTLS techniques/trajectories: pop-up, glideback, boostback, flyback, and once around return. Of these five approaches, the most well known (until recently) and thus most thoroughly studied is the flyback approach. However, the first three are the ones (pop-up, glideback, and boostback) that I think are the most promising and relevant to near-term orbital RLV endeavors, and thus will get the bulk of my focus for the remainder of this series (Parts III-V). But I’ll probably spill a few electrons discussing the last two as well. They are interesting after all.