totm nov 2023 SpaceX Discussion Thread

[kerbiloid]

Quote

Why only light 2 engines for landing? Any engine failure means loss of vehicle, so you have two single points of failure. Why not light all 3, do the flip, then pick the best two and turn off the other?

Only six can be stable.

Six engines, six legs, six attachment points...

Wait.... Oh, ...!

[tater]

1 hour ago, GuessingEveryDay said:

@tater. You need to get some sleep.

It was only around midnight.

[RCgothic]

RIP SN5. I had hoped you'd be used for the mock up.

 

[RealKerbal3x]

SN5 was an extremely important prototype, the first real Starship tank section to fly. F.

[Delta dart]

So does this mean the end for Starhopper and SN2?

[sevenperforce]

24 minutes ago, Spaceception said:

Well, isn't the engine flip supposed to be replaced with a thruster flip at some point? And then one engine would work once it's already upright?

Originally it was supposed to be a thruster flip but they determined that the pitch authority on the engines is so much better than on the thrusters that it's better to just use the engines for the whole thing. 

[RealKerbal3x]

Just now, Delta dart said:

So does this mean the end for Starhopper and SN2?

Starhopper is still used as a weather station and camera tower, so it probably won't be scrapped. SN2 may have been modified to be a water tank, so that might stick around too.

SN5 and 6 were just laying around, taking up space.

[kerbiloid]

3 minutes ago, RealKerbal3x said:

Starhopper is still used as a weather station and camera tower, so it probably won't be scrapped. SN2 may have been modified to be a water tank, so that might stick around too.

SN5 and 6 were just laying around, taking up space.

An importance of a secondary profession.

[Delta dart]

Also it was suggested that it might be being cut to mount the HLS mockup pieces although this is unlikely. 

[RealKerbal3x]

3 minutes ago, Delta dart said:

Also it was suggested that it might be being cut to mount the HLS mockup pieces although this is unlikely. 

Some people suggested that when they started to cut up SN6. What remains of SN6 is now in the scrapyard.

[mikegarrison]

Are people watching the same test flights as I am? It's very clear that there is no time for anything to go wrong. The idea that the engines are redundant is valid if you know in advance that one has failed. Like, say, it failed in orbit. OK, then use a different engine for landing.

But if it fails during landing? There is just no time to spare.

[paul_c]

7 minutes ago, mikegarrison said:

Are people watching the same test flights as I am? It's very clear that there is no time for anything to go wrong. The idea that the engines are redundant is valid if you know in advance that one has failed. Like, say, it failed in orbit. OK, then use a different engine for landing.

But if it fails during landing? There is just no time to spare.

AIUI for this type of "suicide"/landing burn, the later it is done, the more efficient it is. So you'd aim for as late as possible. If they reevaluated the risk of non-relight of an engine, knew the likely time to sense it and command another engine to throttle up (or have to start - but then I think they'll move to 3 engines start) they could trade efficiency for 'contingency time period'.

In the same way we might see a helicopter landing, they don't routinely do suicide slam-landings even though it would save a bit of fuel.

[JoeSchmuckatelli]

11 minutes ago, mikegarrison said:

Are people watching the same test flights as I am? It's very clear that there is no time for anything to go wrong. The idea that the engines are redundant is valid if you know in advance that one has failed. Like, say, it failed in orbit. OK, then use a different engine for landing.

But if it fails during landing? There is just no time to spare.

Which is why we are speculating as to the decision to only light two vs three. 

 

I guess the question is whether it's easier to turn off a raptor you don't need, or to turn one on. 

 

Have to presume they thought about it and had a reason to go with only two... Unless it was for the insurance money 

[RCgothic]

Ultimately, the engines have to relight.

They can't depend on 3/2 redundancy because whatever it is giving problems is probably a common mode failure. If one won't light, chances are more than one won't light. Or possibly none will light.

That's not something you can really band-aid with extra engine ignitions and sooner starts.

They just have to get it right. Whatever the problem is, they'll find it and fix it. Just needs more flight tests.

[tater]

17 minutes ago, mikegarrison said:

Are people watching the same test flights as I am? It's very clear that there is no time for anything to go wrong. The idea that the engines are redundant is valid if you know in advance that one has failed. Like, say, it failed in orbit. OK, then use a different engine for landing.

But if it fails during landing? There is just no time to spare.

Yeah, that's why I suggested that they light all the engines, and turn them off as required. There is no possible time to see an error in 1, then light a replacement. That flip is just too late for that. They need to flip perhaps earlier (more props required), and with all engines lit, then turn them off with the vehicle looking at engine performance to make sure the best engine is the one left running (and if all are good, then the one best positioned).

Also:

 

LOL:

 

[paul_c]

4 minutes ago, RCgothic said:

Ultimately, the engines have to relight.

They can't depend on 3/2 redundancy because whatever it is giving problems is probably a common mode failure. If one won't light, chances are more than one won't light. Or possibly none will light.

That's not something you can really band-aid with extra engine ignitions and sooner starts.

They just have to get it right. Whatever the problem is, they'll find it and fix it. Just needs more flight tests.

That was my thought initially, but in history and today, there is a principle that you "don't put all your eggs in one basket" and have an amount of redundancy. For example, twin engine airplanes can still perform on a single engine, etc.

Apollo 6 had 2/5 engines fail in its 2nd stage during ascent. It was able to complete the mission, albeit with a different resulting orbit and an alteration to some of the mission objectives. Space Shuttle STS-51-F also had an in-flight engine shutdown, did an abort-to-orbit and it was still a mostly successful mission.

[tater]

To be sure there are other places in the flight profile where they have redundancy, just not on landing on Earth unless they light all of them.

The propulsive phase of Mars EDL is presumably longer (have to check that old animation).

@mikegarrison what are the certs like for airliner engine out performance? They need to be able to have an engine out once committed to take off, and be able to go around, right?

For SpaceX to even think about "airliner level safety" for Starship you'd think they'd need to be able to land with any single engine, or 50% of the engines at least (not to mention liftoff failures). (not that I take P2P very seriously, but THEY talk about it, so it's legit to nitpick them)

[RealKerbal3x]

SN7.2 venting.

[tater]

[mikegarrison]

17 minutes ago, tater said:

@mikegarrison what are the certs like for airliner engine out performance? They need to be able to have an engine out once committed to take off, and be able to go around, right?

It depends on the type of airplane. For instance, small general aviation airplanes have no requirement for this (which should be obvious because some have only one engine).

Part 25 airplanes (large transport) have be able to maintain a positive rate of climb even with one engine out. This is usually the sizing criterion for the engines and also the vertical stabilizer, which has to be able to handle the asymmetric thrust.

Obviously all engines out is a problem. (cf. US Airways flight 1549)

There are also requirements for what is known as ETOPS, which is how the airplane performs if an engine fails while a twin-engine airplane is out over the ocean and has no place to land.

Edited February 4, 2021 by mikegarrison

[tater]

7 minutes ago, mikegarrison said:

Part 25 airplanes (large transport) have be able to maintain a positive rate of climb even with one engine out. This is usually the sizing criterion for the engines and also the vertical stabilizer, which has to be able to handle the asymmetric thrust.

Yeah, I was talking about airliners, not GA. Thanks, I thought there was some requirement.

Airliners can as you pointed out, even have survivable events with all engines out—not a thing ever with Starship. So engines in fact need to be substantially more reliable than the jet engines used in airliners (though bird strikes/etc are likely less of a concern—hmm, wonder if any Merlins have ever ingested a bird coming down?

Edited February 4, 2021 by tater

[RealKerbal3x]

 

[mikegarrison]

25 minutes ago, tater said:

Yeah, I was talking about airliners, not GA. Thanks, I thought there was some requirement.

Airliners can as you pointed out, even have survivable events with all engines out—not a thing ever with Starship. So engines in fact need to be substantially more reliable than the jet engines used in airliners (though bird strikes/etc are likely less of a concern—hmm, wonder if any Merlins have ever ingested a bird coming down?

This is part of why I consider the Starship point-to-point commercial passenger idea laughable. The level of safety and reliability required for commercial passenger transport is so far higher than that demanded for something like flight to the ISS that either Musk has no clue or else he's just trolling for interest.

[tater]

2 minutes ago, mikegarrison said:

This is part of why I consider the Starship point-to-point commercial passenger idea laughable. The level of safety and reliability required for commercial passenger transport is so far higher than that demanded for something like flight to the ISS that either Musk has no clue or else he's just trolling for interest.

Yeah, while a fascinating concept (who wouldn't want a 30 min flight to Australia?), it seems pretty disconnected from reality. It's a fun thought experiment, but I can't even imagine that level of reliability given how insanely safe air travel is—and the vast majority of air crashes are pilot error, so there is room to improve air safety (which is hard to imagine).

[Delta dart]

Is NSF streaming the SN7.2 test?