SpaceX is on a roll lately. Their Falcon 9 Reusable testbed continues to make improvements. In case it’s not apparent from the video, that’s a big rocket. Each leg is something like 25 feet long. They get a bit smoky in the video but that’s just the ablative coating burning off.
Sure, the DC-X project was pulling off similar feats in the mid 90’s. But that was a pure prototype that never would have made it to orbit. Ultimately, the project was to build an SSTO, even though it was unclear if it was possible. The SpaceX test, on the other hand, uses a rocket virtually identical to the Falcon 9 first stage, which is already flying to space successfully.
And just last week they managed a mostly-soft landing of their first stage in the ocean. Unfortunately they were not able to recover it due to storm conditions, but the telemetry looked quite good. If they continue their successes, they’ll be flying back and landing at the launch site within a year.
SpaceX has also had success in the legal arena. ULA–a duopoly turned monopoly–thought they won a multi-year contract with the Air Force for dozens of launches at very high prices. SpaceX thought it bad form to sign such a long and expensive deal mere months before they expect to be certified for Air Force launches themselves. Furthermore, ULA sources virtually all their engines from Russia, and it also seems in bad form to be sending Russia billions in light of their activities in Ukraine. A Federal judge agreed and issued an injunction against ULA buying more engines (it probably won’t last, but it’s good to see this stuff out in the public light). John McCain and others are also making public comments about the ULA/AF contract.
At any rate, it’s exciting times for SpaceX. I managed to visit their factory last week (got a quick tour from their software lead) and see some of their cool toys, like a 3D printer that prints in titanium. Feel free to ask questions if you’re interested.
Rockets like that always look like they should be totally unstable, with the reaction force coming from the back end of a long narrow object. I have no idea why they don’t just start cartwheeling violently at the slightest provocation
Impressive. It looks as if it came down almost precisely where it took off from.
And, yes, I was brung up on pictures and movies of tail-landing rockets, too. But, just for the record, although his moon rocket in the movie Destination Moon lands on its tail, in the novel that inspired it, Rocket Ship Galileo, the titular Galileo seemed to mostly land on its “belly”
There’s a difference between a Lunar Module and a Saturn V rocket.
During the first attempt at launching the Mercury-Redstone 1 mission, the booster flew 4 inches and then landed upright. But that wasn’t exactly the plan.
A couple of things about the video: the aerial footage, where the rocket rises past the camera and then descends - I assume that’s from a drone? I think video from drones is pretty freaking awesome, and I bet we are going to see that usage of UAVs absolutely explode in the near future. Heck, some of the coverage of this week’s Arkansas tornado damage on the network news was drone footage. I imagine we are going to see more and more of these aerial points-of-view really soon.
Second, I got a chuckle from the cows running away from the noise of the launch. Just seems so incongruous with the space-age flying rocket ships … run, cows! Run!
The whole thing looks so fake, like the rocket is a tiny model on a string. It’s pretty cool. Makes Thunderbird 1 look like a horizontal-landing piece of crap.
It’s cool but they’ll need to curtail the amount of payload they can get into orbit since they’ll need the fuel to land the rocket to re-use it. Anyone have the trade off numbers?
Healthy competition is what everyone says the space industry needs, so this looks like a healthy thing. Russia can make its silly little comments about the US needing a trampoline to get to space. Laugh it up, Russia. Also, the inbred US military industrial complex needs some shaking up, it seems.
I’m not sure that is the plan for space use, just for this test. It would seem to make a lot more sense to use a parachute for most of the way down then rocket the last segment, as in 1000m.
The idea is the extra fuel will weigh and cost less than the alternate landing methods. I don’t know if that’s proved out yet. But the weight cost of larger tanks has become negligible with modern materials. With the DC-X they claimed the additional fuel would weigh less than wings and the additional fuel to get the wings in orbit to start with.
I’m wonder if it will all work out in the end unless much more efficient rocket engines are developed. For instance, I wouldn’t want to land tail first under power coming in from space if there was no back up system, which would add weight and cost. Having a small escape capsule with parachute might help, but are you just going to let the rest of the rocket drop to earth somewhere over the expected landing area?
Well, they are unstable and they would cartwheel. It’s not too far off from balancing a broom on your finger: it only takes small nudges to keep it upright. Of course it’s vastly more difficult in this case, with fuel sloshing around, the weight constantly changing, the limited response time of the control systems, and so on, but the basic principle is simple.
Yep, some kind of multi-copter. I imagine you’re right unless the FAA manages to block them.
SpaceX claims a 15% payload hit for a soft-landing at sea (includes landing legs and some residual fuel) and 30% for return-to-launch (includes more fuel). Not insignificant, but full first-stage reuse will save them quite a bit more than that so it’s a very good deal overall.
This hover test isn’t quite an accurate representation of their actual reuse–they have engines going the whole time and the movement is all pretty slow and deliberate. In practice, they will (or rather have, since this has been tested already) do a few small burns to get the landing trajectory right, turn off the engines, and then do a massive burn at the very last minute to perform the landing.
Imagine standing at the launch site: a machine the size of a building is falling towards you at terminal velocity. Engines are off and the only signs of activity are a few puffs from the control thrusters. 10 seconds before impact, the engines light and the whole thing brakes to a stop at a couple gees, reaching zero velocity less than a meter above the ground.
Maybe. At 13150 kg to LEO at $56.5 mill give $4300/kg. Drop the payload to 70% (I’m assuming it’s faster to refurbish a dry rocket stage than one that’s been brined) and you’re putting 9200 kg into LEO. If they keep their $4300/kg price they’ll be asking for $39.6 million a launch.
Say they make 10% over their launch cost, then for every regular firing they’ll make 5 million but only 3.6 for the reusable one. Likely there’s a number of mitigating factors I’m not even considering. Let’s hope they make a go of it.
It seems that there is some lack of clarity on what SpaceX is attempting to do with this system. The purpose of the Falcon 9R vehicle is to test the technologies and procedures to return the first stage of the Falcon 9v1.1 (and eventually the triple-core Falcon Heavy) vehicles back to a recovery site intact so that the engines, tankage, and avionics can be refurbished and reused. This is not intended to return the upper stage (which will remain expendable), payloads/uncrewed Dragon, or the crewed Dragon capsule to a soft landing. Despite animations you may have seen showing the Dragon capsule performing a powered landing on a pad, the intended return and recovery method for the crewed Dragon is the conventional use of parachutes as is done on Soyuz and was done for the Apollo capsule. (The uncrewed Dragon capsules currently in use to provide payload to the ISS are expended and destroyed upon reentry.)
This is indeed an impressive achievement, and although it has been done previously by others (the aforementioned Delta Clipper-Experimental, Blue Origin Goddard) it is obviously not an easy thing to do. It isn’t just like keeping a broom balanced on your fingertip, because the vehicle itself has internal dynamics which have to be correctly modeled in the autopilot such that the sloshing of propellant, bending of the vehicle structure, et cetera are all accounted for in the attempt to keep the thrust vector pointing through the vehicle center of gravity while compensating for rotation, nutation, et cetera. A simple ‘stick’ model in Matlab or Python with a sinusoidal forcing function and a finite impulse filter will give a sense for how difficult it is to reduce response below unity even for a simple known input, and with a time varying stochastic input (e.g. propellant slosh, wind shear, change in structural damping due to propellant unload, et cetera) it is a fiendishly difficult problem.
However, what is actually the more impressive achievement than just taking off, flying up to 1 km altitude, and returning near the site is the return and recovery of the Stage 1 from suborbital space, where it is reentering at hypersonic speed with a fair amount of sloshing propellant, reorienting the vehicle and removing precessive and nutational effects, and maintaining control through the transonic regime. I figured they would require several attempts before coming into a successful recovery (and given how SpaceX statements on their first attempt were couched in conditional language, it seems they did, too) and yet on their first return attempt (Flight F9-6, which was also the maiden flight of the Falcon 9v1.1 vehicle) they passed through that hurdle and only lost the vehicle due to lower altitude aerodynamic instability. More than any other test, that really demonstrates the viability of soft landing their vehicle and paves the way to further enhancements which improve recovery.
However, whether that will materially reduce launch costs by itself is still highly questionable. Yes, they get to recover the hardware, but while hardware costs are not insignificant, much of the cost isn’t in fabrication per se but in testing, which would need to be redone to some extent after the engines, avionics, thrust vector system, propellant feed system, and other associated hardware is refurbished. Some items, like flight termination systems, aft insulation, fuel and fueldraulic filters, cryogenic seals, et cetera will have to be replaced as part of refurbishment. They can’t just land a stage, refuel, and be able to fly again in a few days; this will be a month or more of refurbishment, re-acceptance testing, and stage/vehicle integration. The real cost savings is and will remain in reducing ground processing and integration effort, largely by automating integration testing, simplifying systems to require less labor, and using other methods (such as the horizontal integration that the Russians have always done and that SpaceX is wisely aping) to reduce all of the labor and oversight that goes into making a bunch of hardware into an aerodynamically unstable vehicle prepared to fly into space on the breath of a temperamental fire-breathing beast with the brain of an idiot savant.
As for the hit on payload capacity, the numbers quoted by Dr. Strangelove are good ballpark estimates but the specific answer depends on how big a particular payload and all of the associated deployment hardware is, what orbit it is going into, and where you want to return the first stage(s). It should be understood that most of the impulse from the propellant throughout the initial boost is used to propel more fuel rather than the vehicle itself; that is, every second you are flying you are having to carry all of the propellant you will use in the following second, et cetera; hence why the vehicle is mostly propellant at launch and why carrying enough fuel to return the first stage(s) has a lot of impact on the overall payload availability. However, simulations for a Falcon 9v1.1-sized vehicle show that the minimum amount of propellant needed to recover and land the vehicle is actually surprisingly low. Compared to the overall propellant load the amount needed to return the vehicle is quite small, mostly because the vehicle is largely an empty can at that point. The challenge is making the landing without either running out of propellant (bad day) or carrying too much extra propellant, which means you gave up payload capability. Finding that right balance will be an exercise of simulation validated by flight data, which SpaceX is doubtlessly dialing in with every flight and recovery attempt.
Regardless, neither SpaceX nor anyone else will be launching a single stage to orbit (SSTO) vehicle to orbit and returning it under power as envisioned by Golden Age science fiction authors any time soon, although a credible design for such a vehicle does exist, albeit the vehicle carried air-breathing jets and JP-4 fuel for landing. A true reusable SSTO is still, at this point, a dream of engineers and aerospace enthusiasts, but demonstration of powered landing capability, even of just the first stage, is certainly a step closer to achieving that goal.
No, they aren’t destroyed. In fact, they are used to return cargo and trash from the station and recovered. I saw a somewhat charred one hanging from the SpaceX factory ceiling.
It’s true that they aren’t yet reused, in part due to NASA requirements, but that will probably change over time.
I’m curious about something. Supposedly, the Falcon 9 Heavy side boosters will have a mass fraction in the ballpark of 30:1. If we use the sea level Isp of 283 (i.e., no benefit from an extended nozzle in vacuum) for the Merlin 1D, we get a delta-V of 9400 m/s. That’s enough to orbit.
There’s no payload, clearly (ok, they can put a dozen cubesats onboard), and it would be a total stunt. There would be no point… except to have the very first SSTO. Aside from the obvious engineering work (software, etc.), I wonder if there’s a showstopper I’m missing.
Supposedly, the Atlas II wasn’t too far from an SSTO, due to its very light balloon tanks (Falcon 9 uses “semi-balloon” tanks–they require pressurization for launch but are strong enough to survive ground handling without pressurization). With slightly better engines, it could have done it.
)