As usual, I have little other than interest to add. From my understanding, the quality control on rockets is phenomenally tight - even moreso with NASA. It boggles my mind, that something so incredibly huge is manufactured to such demanding parameters. (When golfing I currently use as a ball marker a titanium washer that my kid picked up off the manufacturing floor because it was determined not to be up to snuff.)
Everything I hear about the quality control involved in initial construction impresses me that it wouldn’t take much impact/damage at all to render reuse impractical.
The problem of reliability is a tricky one, because there are so many systems in a launch vehicle that are critical to reliability and not capable of being made fully redundant. Even systems that are not themselves critical, such as environments instrumentation or cameras can cause a propagating failure if they fail in some ways, e.g. come loose and cause damage. The demonstrated reliability of the current ULA EELV vehicles is the result of extensive integration testing and verification, but this obviously comes at significant cost and limits to throughput (although ULA has been also been making a large profit on every launch). Automating much of that testing would reduce both time and cost, but would have considerable NRE that would have to be distributed among a large number of launches, and ULA has argued that there just isn’t enough manifest to cover that development.
Estimating system-level reliability from a standard component rollup approach is what is generally done, but practically speaking it is a largely useless exercise because if you use conservative estimates then your reliability ends up being ridiculously poor, and if you use “realistic” estimates there typically is not enough data to firmly support the values from a purely frequentist approach. (There are alternative approaches to estimating reliability but none that are widely accepted in the industry.) The goal would be to have the same threshold of assured reliability from flight to flight, especially since a customer will likely have no control over the history of their particular launch vehicle, or else will have to bear the cost of flying a virgin launch vehicle as EELV has already specified for their launches.
I don’t see any reason why there would be reduced propulsive performance from flight to flight, and in fact SpaceX would need to maintain a performance baseline to be able to manifest a payload on any given vehicle. Although there are natural variations in engines from build to build, significantly degraded performance would be an indication of something failing.
This is very much true of current production vehicles and especially American and European launchers. Contaminated fuel or oxidizer, a connector not properly seated, stress corrosion cracking in some random fastener, falling debris, et cetera, are all common modes of failure and may be exacerbated by reuse. By how much is difficult to say, because often the things that actually fail are not well reflected in reliability estimates. In fact, one of the lessons from the Shuttle program was almost none of the things that caused aborts or failures were correctly reflected in the prior reliability estimate, even though the top-level overall failure/success ratio fell well within the bracket of estimates (1:50 to 1:100). In other words, it isn’t the problems that you know about that get you; it’s the problems that you either don’t know about or have decided are not worthy of concern, that end up causing most of the problems. We’re just not very good at estimating failure from details, which is why simple and robust design ends up trumping complex and redundant in practical application. Unfortunately, rockets can only be made so simple, and you can only achieve a limited degree of robustness without gross performance penalty.
It seems to me, a used first stage would have a high wind resistance (drag?) so its terminal velocity would probably be fairly low. (sort of like a falling empty drink can). This would suggest the amount of propellant needed to slow it to a stop would be quite small.
The figure I’ve heard is that a parachute is like jumping from about 10 feet up - you want to be careful how you land (hence the retro burst for capsules with parachutes, hence the high rate of injuries for paratroop trainees in WWII) but not exactly the speed you want to drop a fragile sealed thin metal casing, let alone the complex machinery of a rocket engine. (Also, most modern parachutes for humans, they are the flying parawing design where you can flare like an aircraft to land at almost zero velocity, but it requires careful piloting.)
Picking my kid - quality control engineer at ULA - up at the airport tomorrow. This thread will provide me plenty of conversation fodder.
One thing that often confuses me is how terribly restrictive so many of the regulations are. When I ask him why they are or are not bidding on a particular project, he’ll say they are not allowed to, or McDD is, or something. A particular platform can be approved for cargo, but not human launch.
He’s often tight-lipped about specifics, but he does consistently say that NASA is the most demanding client. At times, I get the impression he thinks ridiculously so. And I believe one of his complaints is that SpaceX is not held to all of the same requirements as the other established manufacturers. Of course, there is likely a benefit in fostering competition which might require decreased burdens for newcomers. Also, my son admittedly has a bias.
I don’t know enough about it other than to say it confuses me. But there is a huge regulatory framework, in addition to the technical/cost/production challenges.
BTW - our favorite Elon Musk quote: “Rockets are tricky!”
So, there are a number of different types of regulations that come into play, including the ULA partnership agreement between Boeing and Lockheed, the EELV contract bid structure, which may restrict or limit the types of missions or payloads which a bidder can propose on, the manifest priority given to national security payloads over commercial or NASA (quasi-commercial) payloads, NASA’s human rated requirements, range safety and operations requirements, and for DoD acquisitions, the Federal Acquisition Regulation (FAR). The latter is actually the most onerous because it dogmatically spells out what you can and cannot do in an acquisition program in great detail, but as far as I can tell no two contracting officers (CORs) agree on the interpretation, which can be interesting when there is a handoff from one COR or organization to another. (NASA is in the interesting situation where they can often procure outside the FAR for a commercial procurements but will often invoke FAR regulations when it suits them, which is one of the many infuriating things NASA does.) NASA is demanding on a number of different levels, both technical and managerial, and my limited experience with them has been uneven; sometimes programs are really well managed and sometimes there is just a thundering herd of oversight and technical “assistance” that we’d be better off without. I won’t go into detail other than to say that some of the NASA centers are clearly better managed and focused than others.
I would not presume to be sufficiently knowledgeable enough about the SpaceX EELV certification and bid to know whether they received favorable treatment or not, but it is far to point out that the Boeing and Lockheed received significant subsidy while SpaceX received more limited funds strictly to support the EELV certification process for their cert flights. There is, of course, politics involved (and SpaceX is not reluctant to invoke both political ties and public opinion in support of their bid) but it is also the case that there is a legitimate desire for an independent second party to provide EELV services. The fear on the part of ULA is that this will further dilute a manifest that already can “barely” support Boeing and Lockheed (albeit at substantial profit) and that the resulting competition may reduce reliability in an effort to deliver at minimum cost. I think there is a balance somewhere in between what ULA charges and what SpaceX advertises (as their base level manifest cost without any support services or integration & test data necessary for government missions) but time will tell.
Of course, Boeing and Lockheed are totally free (within ITAR restrictions) to manifest commercial payloads on their Delta IV and Atlas V vehicles but their costs are so high that it is rare they manifest commercial payloads. Other companies have attempted to perform in the commercial launch business but the truth is that in order to compete commercially a manifest of 10-12 launches per type per year is about the minimum necessary to break even at commercial rates (at least for US providers) and nobody has yet developed that kind of launch frequency and tempo. SpaceX is clearly trying to go for that flight frequency but their vigorous pursuit of EELV certification indicates that they know that the assured near term profits still require the higher margin government assured flights, which cost a lot more than the advertised manifesting cost.
A purely commercial launch vehicle really awaits a customer base that will flight at high enough frequency to support that rate and is tolerant to the failures that will inevitably occur with a less rigorous commercial development and mission assurance paradigm. The evolution of the commercial smallsat industry may provide that base–customers who can afford to pay a few tens of millions for a dedicated flight or a few millions for a co-hosted or rideshare flight, and build frequently enough that the loss of a single flight is not a critical blow–but there is a real chicken and egg problem in showing enough value that investors will put in enough capital and stick through the failures to allow the designs to mature to a state of acceptable reliability.
For those who’ve never seen it, here’s an absolutely amazing on-vehicle video of the entire flight of the Shuttle SRBs from launch to splashdown. Few things to notice:
[ul]
[li]During launch you can see the aerodynamic vapor effects when the vehicle breaks the sound barrier at @ 700 MPH.[/li][li]Once the boosters separate from the Shuttle, and are essentially coasting, they still keep accelerating for quite a long time.[/li][li]You can catch glimpses of the Shuttle flying away in the distance while the boosters are coasting in the upper atmosphere.[/li][li]You can also see the other booster many times during their descent together.[/li][li]The booster’s motor nozzle is jettisoned just before splashdown to stop it from being rammed up into the stack (you can both see & hear it being jettisoned and splashing down a few second before the booster itself).[/li][li]The tremendous sound the parachuted boosters still make upon splashdown at, as stated above, 50+ MPH*![/li][li]You can see the recovery ships approaching on the horizon, as well as see & hear the other booster splashdown nearby.[/li][/ul]
The description says that all the sounds are real, and during the booster’s coast and sub-orbital reentry the sounds are pretty freaky!*
Additionally, with the parachute method they have to recover it and haul it back.
In the case of parachute landings, what is the specific damage that occurs? If it has to do with the speed of transit through the lower atmosphere and its ensuing friction, how does the powered landing of the booster avoid this?
Has anyone published a flight log showing such things as the time after launch of booster separation, vehicle speed and altitude at that time, and so on? When was the booster reignited, and at what altitude?
Pretty sure it’s the land/rocket interfacing. Even if it lands slowly enough not to create impact damage, the weight distribution across the structure (which could be in any direction) might make it bend or weaken it. That means everything has to be thoroughly tested. Having to thoroughly test everything is a large part of why the Shuttle Program never delivered the promised savings of time and money.
To those who know more: What are the costlier elements in SRBs? Can’t be the propellant. Are there a lot of complex systems in an SRB or is it pretty much just a big propellant tank/combustion chamber with little more?
Unpowered landings like the Space Shuttle used are dangerous, and the wings, control surfaces, landing gear and heat shielding add a lot of weight. Uncontrolled landing using parachutes are even more dangerous. So the idea is to cut out the extra weight used in unpowered landings and put that into fuel for landing. I can’t imagine a manned craft not having a parachute escape pod just in case the vertical landing goes awry, so they’re not going to be cutting all the possible weight out. Besides that, it’s cool. That’s how rockets are supposed to land.
Early on, SpaceX was considering parachutes for recovery of their first stages from the ocean. A few of the early Falcon 9 rockets actually had parachutes installed to test this method. What they found was that the first stage didn’t even survive to the point of deploying the parachute - hitting the atmosphere was enough to destroy the stage. To survive that, they needed to restart the engine to slow the rocket enough to survive reentry. And once they’d developed the ability to reliably restart engines in-flight - a fairly challenging thing to do with pump-fed, non-hypergolic engines - vertical landing wasn’t that great of a step further.
This picture has some approximate information, including a comparison to the Blue Origin flight.
There are four burns total: the initial boost, a boostback to put it on a rough trajectory back to the launch site, a reentry burn to slow the booster before it hits the thick atmosphere, and the final landing burn. Peak altitude is ~200 km and downrange distance 95 km. Booster separation is at 1.7 km/s; I’m not sure about the other speeds, except that it was over mach 1 not too far from the landing site (since you can hear sonic booms).
I don’t see how uncontrolled parachutes landings are more dangerous than the unpowered (gliding) landings that the Shuttle used. There were an order of magnitude more things that could go wrong with the Shuttle: Damage to airfoils, control surfaces, hydraulic systems, landing gear etc. compared to simply parachutes (and retros like the Russians use to land on land). Not to mention all those controlled landing systems have to all be protected from extreme heat during reentry, as opposed to simply protecting a smaller crew cabin.
As for “That’s how rockets are supposed to land”, well yeah, if you believed '30s goofy pulp scifi magazine drawings, or cheesy 50s scifi movies (which all just used stock footage of V-2s launching, shown in reverse*!*)
You can’t quantify the risk without going into immense detail about the actual systems, but the nature of the risk is different.
Passive systems are subject to more uncontrolled risk. At least over land, parachutes might drop you onto a tree or the side of a hill or a cliff or whatever. Hopefully you have chosen your landing zone to minimize this risk, but in the end you’re still subject to some degree of dumb bad luck.
Powered landings are more complicated, but at least most of the risk is controllable. If something goes wrong, it’s probably the fault of the design or manufacture. Or you launched in conditions outside the flight envelope.
Not into a salt water sea, except the salt water itself… and the chance of a storm causing problems. There’s no way for the parachute landing system to retarget to the other ocean… its orbit sets which orbit it will hit.
Maybe they could use the Great Lakes for landing ? that would make sense.
Storms cause less problems.
But anyway, F=mA… that means that hitting something hard isn’t just a problem because of the strength of the object hit , but also because the hard thing induces a high de-acceleration… which is a high force.
You might think forest trees are blunt… but when a branch or trunk breaks, then it develops a sharp feature, which may punch into the rockets side.
But anyway, the size of heat shielding that something like the shuttle requires for a parachute re-entry means that its far cheaper to bring it in as a atmosphere skimmer…
That’s a better way to put it. Both approaches carry a lot of risk, and so does tail down re-entry. The risks are different, and we don’t have real experience with vertical landing from space yet.
