I am currently re-evaluating my decision to purchase a couple of those …
i might get the superyacht-turns-submarine instead, after all …
Even aside from the under-rated hold-downs, what I can’t fathom is why the engines didn’t have some kind of fail-safe. Like if it hasn’t received a “throttle set to 100%” message in the last 100 ms, shut down. Surely each engine has its own controller that actuates the valves, throttle, etc. That they kept running even after ripping themselves from their ground connection is just bizarre.
I wonder why they would even have under-rated hold-downs. Either this test stand was designed from the get-go to never have full-power tests, which the test engineers should have known about. Or there were different degrees of restraint that the facility could use.
But why design it like that? It would be simpler to just design for the maximum expected load, and use those restraints every time. How much time and/or money are they saving by having multiple configs and hardware instead one setup that just works? Simple is good.
I’m in software, so I can appreciate that late-binding changes in scope can totally invalidate assumptions that were made in the beginning of a project, and no one thought to re-evaluate. Not an excuse for something as basic as hold-down strength limits… but I get it.
But this isn’t the only possible failure mode for a static fire. There are a zillion smaller failures where you’d still want the engines to shut down gracefully if they lost contact with ground control. Watchdogs in various forms are bread and butter for any kind of embedded system. It’s hard to imagine many situations where being stuck at 100% is the right solution in case of failure. A clean shutdown is going to be better the vast majority of the time.
I work in software, too. I was at a company that made robots for warehouses. We specified the maximum load it could carry, and tested it up to that limit. If some warehouse only stocked pillows and stuffed animals, they were probably getting a more robust robot than they really needed, but it was a lot simpler than building different hardware for every different conceivable load.
If a rocket is a thousand feet up, and it loses radio contact with the ground, I’d want it to keep going up. Follow the pre-planned trajectory as closely as possible with the systems that are still working, rather than rather than come crashing down to earth. If an antenna breaks, or there’s interference on a particular frequency, or the ground transmitter loses power, those little issues won’t cause a catatrosphic end to the mission.
Yeah, for a ground test, maybe they should have built in some kind of fail-safe if it left the ground. But if you do that, then the thing that you’re testing isn’t exaxtly like the thing that’s going to fly.
I’m not sure I follow you. Perhaps you know more of the sequence of events than I do.
AFAIK the rocket stage itself did everything right. The engines followed the commands they had received. Including being commanded to 100% while still in the hold-down fixture. And were still receiving that command as it inadvertently climbed into the sky. A few seconds later when either a range safety human or an automated “shit’s gone horribly wrong” monitor recognized the overall situation, the engines were commanded to shut down and did so promptly. Gravity then did its oh-so-reliable thing leading to an Earth-shattering Kablooie.
What am I missing?
Here’s a thought not directly attached to the above convo …
The pad hold-down system on the Saturn V famously was just a weak link; nothing with moving parts or pyrotechnics. When the net thrust built up enough, the stack simply broke the hold-downs and headed skyward. Simple and highly foolproof.
Perhaps this pad was similar. The hold-down system is designed to break away at full net thrust during a normal launch. But somebody forgot the difference between full net thrust on a full stack with upper stages and payloads versus the full net thrust of a first stage with nothing piled on top. Oops.
As always with bureaucracy in general and Chinese bureaucracy in particular, a large dollop of face-saving will be served along with any facts that ever emerge about how this event came to happen.
At least it wasn’t North Korea, where a mistake like that would lead to face-50mm-canoning.
Is there such a difference in a purely static situation? After being freed, it would make a hell of a difference, to be sure, but while locked down I thought it was purely thrust versus retention strength.
Maybe the additional static mass (weight) of upper stages and payload would have contributed toward the “retention” side of the equation.
Nope. At least not as I see it.
Yep. Now you’ve got it.
If gravity is pulling down with e.g. 100,000kg of stack weight, and the engines are pushing up with 99,999kg of thrust, the whole stack “weighs” 1kg as far as the hold-down system can tell. It’s not doing any holding down at all; gravity is doing all the work.
Now the engines throttle up to 100,001kg of thrust. The stack now “weighs” -1kg. IOW, it’s now pulling upwards with 1kg of force. Now the hold-down system is having to bear that tension load, but it’s teeny.
This is why I kept using the term “net thrust”. Said another way, if you put a 10,010kg thrust motor under a 10,000kg stage, there’s only 10kg of force available to accelerate it upwards into the sky.
Of course with real rockets using real numbers the net thrust starts out pretty high percentagewise, not just my contrived example of 1 part in 1000.
This is because you want/need a decent rate of acceleration upwards from the git-go. The Saturn V I alluded to upthread weighed on the order of 6.5 million pounds just before ignition and generated on the order of 7.7 million pounds of thrust. So 7.7/6.5 = 1.18 to 1. IOW excess thrust is about 20% of the vehicle stack weight from the git-go.
As well, as fuel and oxidizer are consumed very rapidly, the whole stack gets lighter while thrust remains constant (net of anythrottling). So assuming constant thrust, the thrust to weight ratio only gets better over time as fuel and oxidizer leave the system out the nozzle(s).
I’ve never heard that before. What I have heard is that the multiple hold-downs had to release within a few milliseconds of each other, lest the rocket start leaning to one side or another. I didn’t think any part of that system was left to chance.
I saw a graph of the performance on the Saturn V in college; lines for weight, speed, altitude, and a few others over time. The weight was fascinating. The rocket loses millions of pounds within the first three minutes, and there were little blips in the curve as spent stages were jetisoned.
Just to add confusion, according to wiki there were two hold down systems:
Looking at the diagram here for the mechanical holddown arms, I see there’s an explosive release. That sounds like a command is sent, not just resistance overcome.
https://www.hq.nasa.gov/pao/History/SP-4204/images/m287a.gif
Aah, thank you both. I was talking about what turns out to be the second half of the hold-down system.
I was running off memory and here’s a cite which says I was wronger than I wish I was. But is informative for anyone wanting to chase down the extruded pin rabbit hole … Saturn V hold down posts: dies and pins - collectSPACE: Messages.
I once saw an amusing animated graphic that depicted the Saturn V’s takeoff exhaust in “elephants per second”- quite a few of them.
Every modern engine has a local controller. That translates incoming commands like “set throttle to 100%” to physical controls–valves, linear actuators, that kind of thing. Start-up sequencing is especially complicated for rocket engines but even just running steady-state requires constant control for throttle, mixture ratio, and such.
On a real rocket, these would be connected to the main avionics suite, which actually decides what the engines should be doing based on various other inputs. I expect that here, they didn’t have that, and the engines were connected to the ground directly.
Either way though, there’s a communication link. And that link should have been cut as soon as the rocket separated from the pad because the wires physically ripped apart. Even though the last command was throttling to 100%, the controllers on the engines should have noticed almost instantly that the link went down and then themselves shut down.
Maybe the whole thing was radio controlled? Could be, though it would be somewhat surprising since this is just a test stand. And that would just raise the question of why no one hit the big red emergency stop button. Maybe that link went down? If so, then what I said before still applies: as soon as the engines stop receiving commands, they should shut down.
A rocket doesn’t necessarily need constant radio contact with the ground (though that was traditionally the way flight termination systems worked).
However, the engines do need communications with the rest of the avionics. If that is lost, the engines should shut down.
The most important aspect of any flight termination system is to terminate thrust. And that’s important because the exclusion zone is set so that at any point in flight, you can shut down the engines and the rocket will crash somewhere relatively safe (this is the Western view, at least). Keeping the engines on is absolutely the wrong thing because now the rocket could travel outside of that exclusion zone.
Ariane 6 maiden flight in progress. It made orbit and deployed some cubesats. However, there’s been an issue:
We’ll have to see if they can get the Auxiliary Power Unit restarted somehow. Otherwise, it’s gonna reenter in some arbitrary place. And those reentry capsules are probably lost.
Did the Starliner capsule bounce down in the desert yet?
It’s still up there. It isn’t yet cleared to return except in an emergency.
Wow, okay - thanks.
NASA just put out this update:
https://blogs.nasa.gov/boeing-crew-flight-test/2024/07/10/nasa-boeing-conduct-ground-tests-ahead-of-starliner-return/
After an agency-level readiness review later this month, NASA and Boeing plan to select a new target return date for the Crew Flight Test. Following this review, NASA plans to host a televised briefing and will share more details on that when finalized.
So… probably at least a few weeks off.
Wow.
Soon they’ll draw straws with the Russians, to see who gets to kill and eat whom….
