Right, but if the stages separate as I described, then the whole rotation maneuver would have a negative impact on getting the upper stage to orbit.
So the point of the rotation is to put some distance between the two stages, and little (or nothing) to do with pushing the upper stage to orbit.
Exactly. It just so happens that most rockets have some kind of pusher system that sends the upper stage forward and the booster back, but the direction of the push is basically irrelevant: what matters is that the stages move apart from each other.
Vehicles with side boosters like Falcon Heavy barely give the boosters a push at all. They largely just release them, and the center stack keeps accelerating while they fall behind. To the extent there’s any push, it’s to the side.
Sounds like your conflating which way they’re pointing with which way they’re going.
Let’s talk for a moment about just the upper stage.
If the centrifugal staging separation was done with the vehicle axis exactly perpendicular to the velocity vector (IOW the least helpful orientation) post-staging the upper stage would be traveling helpfully towards orbit at a couple thousand miles per hour and now also traveling useless sideways at a human walking pace. It would also be spinning slowly end over end in the plane of the the centrifugal separation maneuver. And also simply coasting since the upper stage engines are not yet firing.
So then you can wait 10-20 second for the upper stage to rotate around to being nearly aligned with the velocity vector again, fire the engines to continue into orbit and also expend a microscopic percentage of propellent firing ever so slightly sideways to cancel the walking-pace motion you no longer need and also cancel the rotation you no longer need. More likely, both those motions will be lost in the sauce of the overall nav errors of the flight. And will never be corrected out as a specific separate issue. Just keep steering towards the targeted velocity, attitude, and angle rates and all will be well.
if you don’t want to wait 20 seconds for the rotation left over from staging, just fire whatever attitude control system you have to quickly re-orient the stage along the velocity vector and start the engines.
For the lower stage the idea is almost the same. Except you’ll want to wait, or forceably align, the lower stage with however your de-orbit / return to base burn needs to be pointed. Which is definitely NOT in the direction to push the stage towards orbit.
So just a few 10s of seconds after separation, the two components have drifted far enough apart and are pointed in very different directions and then begin seriously thrusting to drive themselves towards their respective very different destinations. Which also has the effect of immediately driving them farther apart so unhappy interaction is no longer possible.
The van owner doesn’t deem all that put out:
RUD Rapid Unscheduled Disassembly
The ever-enthusiastic Chris Hadfield has a nice take on the launch (immediately correcting the presenter on his attempt to paint the launch as “putting on a brave face”):
I’ve just seen this image of the launch mount and how much the ground has been excavated by the rocket exhaust. If this image is a real one then all I can say is “ouch!”.
I still don’t see how you could call this launch "successful’. A informative failure, sure, but a success? Recall the goals of this test were to boost Starship off the pad, detach the first stage, perform a semi-controlled “landing” of the booster, take Starship up to orbital height (but not orbital velocity), and finally perform a controlled “landing” of Starship about 3/4 the way around the planet.
This was deemed to be a pretty conservative test. No orbit, no controlled landings, no effort to recover either stage intact. They accomplished exactly 1 of those goals. I agree that they learned a lot, so maybe actually my problem is that it’s fine to call this a failure, because it’s fine to fail, failing is how we learn. Science and (to a lesser extent) engineering is about learning what works and what doesn’t. Insisting that blowing up before performing a majority of the things they intended to do is a success actually just makes them seem thin-skinned.
Apollo 1 burned on the launch pad during a test and killed three people. We learned a lot from that. Does that mean it was a success too?
@Fiendish_Astronaut 2 posts up …
Assuming the pic is real as you say …
At this point the structural integrity of the OLM and its foundations becomes unknowable. Time to condemn this site & build a new one from scratch nearby. And perhaps quicker to start on a greenfield nearby than to demolish and excavate this whole thing out, even though a large excavation for a flame trench will be part and parcel of the finished new pad.
However they choose to go, this will be a major cost and schedule setback to say the least. If their eventual launch cadence goal depended on multiple launch pads, each of them just got a bunch more expensive / elaborate than expected.
Ah, yes, but those were not the “goals.” They were the scripted events if everything had gone perfectly.
“Success” and “failure” are not binary terms. And they are fuzzy terms, with no two people defining them exactly the same once we’re talking about anything more complicated than putting on your socks. Heck, if you get one on inside out unnoticed is that success or failure? Hint: it’s both.
Is everything other than 100% success a “failure”? Is everything other than 100% failure a “success”? If you disagree with both those propositions (which is IMO a very reasonable position to take), then you agree there is something in the middle which consists of some sub-successes and some sub-failures. And now where do we draw the line where there is enough sub-success to say it’s an overall “success”?
A nature of a rocket mission is it’s a chronological string of events with no redundancy or loops or anything else. From countdown start to landing or pretty explosion you keep checking off at least reasonably successful success events until suddenly you don’t and all subsequent events are not exactly failures in that you didn’t even get the chance to try them to find out if they’d have worked. They’re certainly not successes, they’re sort of the rocketry equivalent to racing’s disqualified / did not finish.
Bottom line: Trying to rigorously define a “success” / “failure” binary metric for a whole process with probably 100,000 mini milestones in it is fundamentally silly. Did they get farther than their own 50% probability expectation? Apparently, at least according to their press people both before and after. So a success? Certainly by their stated metrics. Should you or I call BS on their metrics and measurement against those metrics? If so I’d ask what you or I know better than they did/do?
Modulo the fact that there’s always room for salesmanship, puffery, ego, etc., in any organization’s claims. Managing expectations up or down so as to either over- or under-achieve vs those pre-managed expectations is a time honored tradition in human organizations.
So did they somehow just overlook this being a possibility?
With that being the case, it logically follows that it’s silly to call this launch a “success”, correct?
A “rapid unscheduled disassembly”? That’s the most eloquent way of saying something fell apart that I’ve ever seen! LOL
I’d suggest it’s equally silly to call it a “failure” for the exact same reasons you suggest.
The problem is that the public and lay press just demand it be labeled one or the other, when either label obscures about as much as it informs. The problem is not logically equivalent to “When did you stop beating your spouse?” but it has a similar no-winning-answer outcome.
So what was that? Was Starship’s launch a failure or a success?
Um … … IT EXPLODED. 
Intentionally
It’s not just the public and the lay press. SpaceX itself (actually Elon Musk) is calling the launch a success.. The employees applauded when it blew up. If we’re going to police sloppy characterizations of failure, then equal vigor should be applied to unqualified claims to success.
Personally I would say it should be judged as to whether this event helped the project stay on budget, on schedule, and within specs. AFAIK that’s mostly not public information, so it doesn’t seem like there’s much scope for the general public to judge.
No, it’s actually everyone qualified to answer. Eric Berger mentioned:
Yes, that’s a reasonable take. Was the flight an advancement or a setback? Did the information they gleaned move the development needle more than the cost of running the test? And so on.
From this perspective, it was a significant success. They retired a bunch of risk in a number of key areas. Much more to do, but it’s clear that they moved several steps forward with this launch.
That is not a crazy metric at all. And by my reading of that metric, the outcome is unequivocal “failure”.
ETA: I differ from @Dr.Strangelove just above in that I don’t think your criteria are equivalent to “Did the information they gleaned move the development needle more than the cost of running the test?” By that different metric I agree with him the test was a success. They are closer to eventual success and the money was well-spent getting as far as they got.
I will quibble that your (@HMS_Irruncible’s) metric is much more the sort of metric that traditional NASA-style government + defense contractor Space would have used. Which is largely about planning to expend far more time and effort making double-damned sure, rather than having cheaper quicker adverse surprises along the way.
If SpaceX takes a bunch longer and a bunch more money that they hoped/planned, but also beats the time/money estimates of the Traditional Space crowd by a factor of 3 or 4x, is that “success” or “failure”?