SpaceX’s first stage landings on the barges have all been autonomous, I’m guessing the reaction time needed is far beyond what any human pilot could do. Could this have been achieved in the 80’s or 90’s using the level of computing tech we had then? Or is there some other advances in materials science that they are using we didn’t then have?
Lets forget if those first stage landings and renewability are going to turn out to be a good idea economically or not, I’m just interested for this thread in knowing if it could have been achieved if they’d had the budget / will to try and do so.
The Delta Clipper was a proof of concept for a similar tail-first landing in the early 1990s. The powers that be certainly seemed to think it was doable at the time.
As for sea landings, there were, apparently, systems capable of landing a F-18 on a carrier without human intervention in 1985.
It doesn’t seem to be a question of having enough computing power to do the job; managing fuel vs throttle vs gimbal vs trajectory seems (to my untrained eye) similar to what happens during ascent.
So I’m pretty sure it would have been possible to do it, at least from a technical standpoint. It would probably have been more costly, though.
The main problem would have been navigation and terminal guidance, not materials, propulsion or computation. High-precision GPS was not available to commercial users until May 2000. WAAS augmentation was not available until 2003: Wide Area Augmentation System - Wikipedia
Today the landing barge may also use differential GPS (DGPS) which would enable the stage to obtain a positional fix within a few centimeters. DGPS was not really widely available until the 2000s. Differential GPS - Wikipedia
There were solutions before this such as Microwave Scanning Beam Landing System (MSBLS), but it was fairly short range, which would have required inertial guidance or other radio nav methods to get close enough to enter the MSBLS “basket”: Microwave landing system - Wikipedia
In the 1980s, cruise missiles used digital scene matching correlation (DSMAC) to locate the precise target within a few meters. Such techniques could also have been used for terminal approach to the landing barge, assuming a day landing.
Given the limited cross-range and propellant capability of the SpaceX landing stage, I’m not sure how feasible it would have been using pre-GPS methods. It could have been done but would have been much harder. Precision GPS was available to military users around 1990, so if it were classified as a military system they could have used that. MSBLS was available and the techniques for DSMAC were known or could have been developed. It was theoretically possible just more expensive, more difficult and would have been less reliable.
Ultimately, the stage will be HIGH and likely descending in a pretty small area where they’ll have parked the landing barge, so all the stage has to do is actually find the barge and course-correct to land on it.
Seems like they wouldn’t necessarily even need GPS, but rather just some sort of thermal imaging system, and an easily detectable target. Maybe even with some kind of ability for the descending stage to beam a laser pulse at it and verify that it’s the target, and not some hapless container ship.
While commercial GPS certainly facilities the targeted landing capability of the F9 Stage 1, it isn’t really necessary. Even in the 'Eighties, commercial inertial navigation systems were available that were suitable for medium fidelity trajectory solutions, and this combined with terminal beam guidance (e.g. having a directional radio beam and guide marker pattern analogous to the Instrument Landing System used for commercial airliners) should provide sufficient accuracy for terminal guidance. Nothing about the guidance & control or propulsion systems onboard the vehicle require more than 'Eighties computing and materials technology, and the art of powered landing on legs is not without challenge but doesn’t require any novel innovations that were not available in that era.
The real success for SpaceX seems to have just been practice and refinement by trial and error; in other words, being willing to try and fail, and try again after learning from the failure. The problem with many development programs today is an almost complete intolerance for any kind of failure, even those that occur specifically because they are pushing the edge of the envelope of experience and capability. By dint of basically not caring and even anticipating failure (i.e. deliberately testing beyond prior experience) SpaceX learned quickly what worked and what needed development. Or, as I tell people when they start bitching about failures in dev test or planning out a schedule that includes contingency for unplanned failures, if you aren’t failing and blowing shit up on a fairly regular basis, you aren’t learning much.
Except the stage is falling into its own plume, so any thermal or visual imaging is going to be substantially obscured.
I was involved with military robotics in the 80’s, and it seems to me that it would be possible, and much easier after the MEMS sensor revolution of the nineties. It still seems to me like a bit of a waste considering parachutes are lighter than the amount of fuel you need for such a landing…
Parachutes are lighter for terminal landing, but they can only be used in atmosphere, which wouldn’t be useful for controlled return of the vehicle from an exoatmospheric trajectory. (The vehicle will reenter, of course, but likely tumbling and not at a controlled speed and orientation.) Large parachutes of this kind can’t just be packed like personnel chutes; they have to be packed and loaded in stages with reefs (lines that limit opening of the canopy, and therefore the loads experienced, until they are intentionally released), and are loaded using a large hydraulic press. Because of the amount of aerodynamic drag required, a very large total canopy surface is needed, which generally requires multiple (three or more) separate canopies, and each of those will require a drogue chute to extract and control inflation. All of this needs to be packaged into a trunk somewhere at the top of the structure and then attached to a pivot point with sufficient strength to carry the load of the entire vehicle. And being flexible elements there is a lot of variability and potential for failure, so parachute deceleration systems are designed for high redundancy with the expectation that some damage may well occur during deployment, inflation, and stabilization. There is also a limit to how slowly a parachute system can effect a descent and the resultant damage resistance or tolerance the vehicle needs to survive landing. A somewhat more simple alternative to parachutes are ballutes (inflatable balloons which increase the drag surface) but they provide even less control of terminal speed and altitude.
For what SpaceX intends to do–recovery and reuse of complete stages–the powered landing mode makes sense. Whether it will be technically feasible with the current architecture, or fiscally viable regardless of design changes, remains an open question, but that they’ve repeatedly achieved controlled landing at all is a pretty significant accomplishment and has potential applications beyond stage reuse.
SpaceX actually did originally plan to use parachutes to recover the boosters. What they found during the first few Falcon 9 launches is that the first stage simply didn’t survive reentry - it broke apart on coming back down and hitting the atmosphere. They needed to develop engines that were reliably restartable in flight, and use them to slow the rocket during reentry. And once they had that, it wasn’t much of a step further to fire the engines again for braking just before hitting the ocean to bring the rocket to a stop just above the waves. There were a few early launches that made ‘landings’ on the ocean this way. Unfortunately, they then discovered that even if you bring the stage gently down to the ocean, just falling over into the water was enough to break it apart. So, they added legs to be able to land on a barge (or back on land for flights with enough spare fuel to make it back) and grid fins to have enough steering fine control to actually land on a barge. Parachutes just weren’t a viable solution.
That’s really interesting. Rather analogous to how the precursors most complex organs in biological evolution developed (by natural selection) for a different purpose than what they ended up becoming.