totm nov 2023 SpaceX Discussion Thread

[Nuke]

were gonna need a bigger engine

[tater]

~2 min

[Codraroll]

16 hours ago, Exoscientist said:

 IF it is scrapped would that include the engines? That would not speak well toward Raptor reusability. 
 

  Bob Clark

You've got an absolutely fascinating ability to arrive at the same conclusion regardless of the direction and magnitude of the evidence presented to you.

Also, the forums have a signature function, if you absolutely need to sign your name under every post.

Edited 12 hours ago by Codraroll

[Exoscientist]

44 minutes ago, Codraroll said:

You've got an absolutely fascinating ability to arrive at the same conclusion regardless of the direction and magnitude of the evidence presented to you.

Also, the forums have a signature function, if you absolutely need to sign your name under every post.

 The proof is in the pudding, as they say. Two key things to look out for is how well the single reused engine performs on this flight, and how many engines from this flight will be reused on the following one. 
 

  Robert Clark

[SunlitZelkova]

11 minutes ago, Exoscientist said:

 The proof is in the pudding, as they say. Two key things to look out for is how well the single reused engine performs on this flight, and how many engines from this flight will be reused on the following one. 
 

  Robert Clark

They are saying your conception of what constitutes evidence is wrong.

Right now the vehicle itself is in development. Although reuse is obviously still an aspect, the main focus is getting it to perform as well as possible and lift things to orbit.

Reuse is taking the back seat right now. Just as the first aspect of Falcon 9’s development was to get it flying in the first place, and bonafide reusability came much later, it may be a relatively long time before the evidence you are looking for starts to show.

What is more important to search for at this time is: Is this rocket reliable? Are the clustered engines doing fine? etc. etc., I’m sure folks here can think of better grades for how the rocket is doing.

After SpaceX checks all those boxes: then we get into reusability.

This is not an all in one test program like SLS, where it is expected to do everything it can on the first flight. It is staged, perhaps vaguely similar to military aircraft development in the 1930s. You fly the clean airframe first usually and then work out all the equipment needed to fire weapons or communicate.

[tater]

10 minutes ago, SunlitZelkova said:

Right now the vehicle itself is in development. Although reuse is obviously still an aspect, the main focus is getting it to perform as well as possible and lift things to orbit.

This.

They have zero reason to reuse an entire vehicle until they have a candidate reusable vehicle. For this test they are reflying an engine on the booster. At a certain point it's worth considering that maybe—just maybe—the dolts working on Staship might actually have some clue what they are doing.

[Deddly]

3 hours ago, Nuke said:

were gonna need a bigger engine

This doesn't bode well for Raptor, right, @Exoscientist? 

[Nuke]

1 hour ago, Deddly said:

This doesn't bode well for Raptor, right, @Exoscientist? 

right now its the starship of theseus. we still dont know what the end product looks like, we just know it will be big and reusable.

[AckSed]

Speaking of development, overexcitable YT thumbnail said that they might be switching to new, orange version of the tiles, but I haven't seen any hint of this.

Any credibility?

[Exoscientist]

21 hours ago, Ultimate Steve said:

As for the current topic, while it would have been really cool to see an immediate reflight, the expectation was thorough inspection and then either display or dissection. For a historical point of reference, the first and third Falcons to successfully land were not reflown. Of everything to successfully land before Block 4, none flew more than twice and several did not refly. Those that did never turned around faster than 160 days. So the current performance is still Falcon comparable in that regard.

While I would also say that the warped nozzles during re-entry also preclude reflight, I can't, as the engine being reflown on flight 7 was on the outer ring (though I'm not sure if it was one of the warped ones).

But while I'm in the mood, let's do another Raptor reliability tally for the fun of it:

  Hide contents

Flight 1

I am straight up going to skip most of this one, partially because it is the oldest and is more or less irrelevant to modern Raptor stats, and secondly, there is a LOT of ambiguity as to how many engines failed, why they failed, and how many (if any) in flight engine restarts were attempted (there is an okay amount of evidence for at least one). There is a shot from NSF convincingly (but blurrily) showing 8 engines out. The icon on the SpaceX stream is inaccurate, as to be expected with a fire in the engine bay severing the connections between the engines and the flight computer. Official SpaceX camera views only ever show a maximum of six engines out, and any engine failures after the tumble would not have been caught. The fire could have also caused the engine failures rather than the engines themselves, and last I checked we don't know for sure if the three engines out at liftoff were pre-emptively disabled or attempted ignition and failed.

It is folly attempting to draw any conclusive numbers from flight 1 with current information unless SpaceX comes out and gives us a timeline showing us every notable engine event and why it happened. The only thing we can for sure say is that at least six engines went out at some point, likely more.

Flight 2

49 attempted engine burns (I will only count burns scheduled to occur before loss of vehicle)

30 engine burns successfully completed

Of the 19 failures:

• 1 boostback burn ignition failure
  • Possibly as a result of the clogged liquid oxygen filter
• 12 boostback burn failures
  • These have been attributed to a clogged liquid oxygen filter starving the engines of liquid oxygen. A few gradually failed, and then one went boom, and then the rest of the vehicle rapidly went boom.
  • I would not call these failures Raptor's fault
• Six second stage engine burns did not reach completion
  • This was caused by a liquid oxygen vent setting stuff on fire, causing communications wires to be severed and the vehicle to self destruct, so 100% not Raptor's fault

All of the failed burns except for 1 have either hard-confirmed or soft-confirmed non Raptor causes, with the only outlier being the one that did not restart for the boostback burn. I believe this is also due to LOX filter clogging as it is in the same area as the engines that started failing almost immediately afterwards.

I don't know what I would call this reliability. 48/49 and 48/48 are reasonable, or if you don't want to count Raptor burns that were cut off early for non Raptor reasons (as Raptor could have failed later in the burn and including partial burns would inflate the statistics), 20/20 is also a reasonable number.

Flight 3

63 engine burns planned (that the vehicle made it to timeline wise)

56 engine burns attempted

36 engine burns successfully completed

Of the 27 engine burn failures:

• 6 engines failed due to a clogged LOX filter during the boostback burn (not Raptor's fault)
• 7 engines "benignly shut down early," completion of the boostback burn was aborted likely due to the earlier failures. This is a simultaneous shutoff of engines that were later commanded to ignite, and is likely not a Raptor issue. This is speculation, but I would guess that the vehicle deemed itself not healthy enough to safely continue with the full boostback and did a kind of "offshore divert" similar to what we saw on flight 6.
• 6 engine ignitions for the landing burn not attempted as those were the 6 that got clogged during the boostback and were deemed to be unhealthy
• 5 Raptor ignition failures during the landing burn. This might also be LOX filter clogging, but is likely something unexpected with Raptor in that high dynamic pressure and g force startup environment. SpaceX says "And utilizing data gathered from Super Heavy’s first ever landing burn attempt, additional hardware and software changes are being implemented to increase startup reliability of the Raptor engines in landing conditions." I am content to call these Raptor failures, though I will note that assuming these failures were related to these conditions specifically, they would not endanger any hypothetical payloads.
• 2 Raptor burns were not completed as the booster slammed into the Ocean or broke up just above the surface before landing burn cutoff (not Raptor failures, though the other 5 Raptor failures caused the situation that caused the burns to be cut short)
• In space engine restart was not attempted/was skipped due to the vehicle losing attitude control (not a Raptor issue)

I would call that 51/56 but other numbers are reasonable.

Flight 4

65 engine burns planned (see note)

63 engine burns successfully completed (see note)

Of the 2 failures:

• One engine failed shortly after liftoff
  • Cause unknown
• One engine did not ignite
  • Cause unknown

Note: The engine indicator was not functioning for Starship's landing burn, but SpaceX confirmed that three Raptors ignited in a post flight update, so we can be reasonably certain that three successful Raptor burns took place.

63/65, could possibly change if failure details are made public.

Flight 5

65 attempted burns

65 successful burns

No failures

Same note from flight 4 applies, engine indicator was not functioning for the landing burn.

65/65

Flight 6

66 attempted burns

66 or 53 successfully completed burns

• I believe that the offshore divert was called out after the completion of the boostback burn, but internally it could have been before, which would mean 13 Raptor burns were cut short for non Raptor reasons.

Thus, 66/66.

 

 

 

If I count from flight 2 onwards, counting partial Raptor burns that were cut short for non Raptor reasons as successful Raptor burns, I would count it as 293/300, but there are a *lot* of more or less valid ways you could count it. That puts the reliability at 97.67%, which is, well, not the greatest. RS-25 is indeed about 99.76%, but the RS-25 is, well, the RS-25. Anything against the RS-25 is going to look unreliable.

And it is worth noting that up to 6 of those 7 failures are correlated to the intense landing burn environment (and I don't think we have detailed information on any of those 7). I point that out not to say that "It's actually 99.66% reliable!" but to point out that the RS-25 has to start up once per flight on the launch pad at 1 G, in one orientation (well, 2 slightly different ones considering the shuttle and the SLS), at one atmosphere, at a time where ignition failures (and there were several) could be aborted.

Meanwhile, Raptor has to start up in the following environments:

• On the launch pad (1g 1atm)
• Maybe arbitrarily throughout ascent (variable g, variable atm) (An emergency inflight engine restart may have happened during flight 1 but this is not certain)
• During hot staging (<1g, <1atm, at a large gimbal angle, and nearly up against a solid wall)
• During boostback (>1g for the later engines, in a near vacuum, with a somewhat sideways acceleration vector, with fuel inlet pressures that might vary a decent amount)
• During landing (Significantly more than 1g, a stupid high dynamic pressure, with a decent amount of atmospheric heating)
• In space (0g, 0atm)
• During ship landing (1g sideways, high gimbal angle, subsequent unique acceleration vector and gimbal profile)

Raptor's job is significantly harder than the RS-25's, and it has to operate in some environments that can really only be tested in flight, so the landing burn failures aren't completely unexpected. There's a bunch of comparisons you can do here, Raptor burns in a typical RS-25 profile, counting only Raptor burns that the payload cares about, but those aren't really valid comparisons IMO so I won't make them.

There's also the "Rolling average" argument which would help offset early developmental failures, but is not especially valid as we don't yet have enough later flights for it to be anything other than "If you ignore the ones where it failed, its reliability is pretty good!"

The only somewhat valid metric coming to my mind is that if we get through the next three flights and liftoff of the fourth flight without any further engine failures (Which I don't think is particularly likely at this point in time), Raptor will beat the RS-25's streak for consecutive successes (which I believe to be 352 flight firings, counting from after the STS-51F failure through Artemis 1).

TLDR: Let it cook.

 

 Thanks for that. That’s a pretty detailed appraisal of the test flights in regards to the Raptor. I have mentioned this before, but I really do not like this approach SpaceX is taking by testing Raptors all together for full mission burns only in flight tests. The typically way this is done in the industry is you construct a separate test stand for the full stage burn. And you test that over and over until you are confident all engines can burn reliably for the full lengths of the full mission burns. 

A separate full up test stand is an extra expense, but these now 7 test flights without an operational paying flight yet aren’t cheap either. An estimate is ca. $100 million construction cost for Superheavy/Starship, so ca. $700 million just in construction costs. When you add on operational costs this could total over $1 billion. A full up test stand would not have cost that much.

  Bob Clark

[.50calBMG]

To my knowledge (which I would consider above average), no other rocket has actually been built up and tested like that aside from Falcon 9. Falcon Heavy didn't, but that's a bit of a stretch all things considered. Before you say SLS did, no, they didn't have the boosters on it. Same with shuttle, Saturn V, Titan, Saturn I, any of the Soyuz family, etc. Literally the only rocket that I know of that was given a full duration burn while on the ground in flight configuration was Falcon 9, and I'm sure you can find that video on YouTube. 

As for Raptor reliability, flight one used the Raptor 1, which in fairness kinda sucked, but it was more just to prove that the idea worked, and is still the first engine of its kind to be flown even with its teething troubles. Nobody says Raptor 1 was the final evolution, it was literally the starting point, which is why it's been replaced. Follow how Merlin evolved, and Raptor 1 was the Merlin 1A. Right now, we are phasing out Raptor 2 (Merlin 1C) and moving to Raptor 3 (Merlin 1D). There have been three versions of Merlin 1D, and now Merlin 1D++ is probably the most reliable rocket engine ever flown. As it currently stands, there have been over a hundred Raptor 2s lit and relit in flight, and only a small percentage of those have failed to relight, which is in all probability not a fault of the engine, more likely fuel slosh, clogged filters, GSE faults, yadda yadda. 

Raptor isn't perfect, but it's certainly not bad considering how new it still is, and it has plenty of room to grow if you base it off Merlin's history. 

Edited 6 hours ago by .50calBMG

[.50calBMG]

As for building a test stand for testing that many engines at once, I'b bet buying the land, paying the contractors to clear the land, then building a test stand strong enough to hold up against something with THREE TIMES THE THRUST OF A SATURN V (in the early prototype aka smallest version, no less) for minutes on end, multiple times, will be much more expensive than actually just launching the rocket. The concrete or steel alone just to build that stand probably costs as much if not more than just building the rocket. Then you have to pay for all the licenses to make that much noise for that long, just to get data on how that cluster of engines behaves for maybe the first 5 seconds of a roughly 7 minute flight, and no data on how it behaves with the change in atmospheric pressure, temperature, speed, G forces... Why wouldn't you just fly at that point? That test makes no sense. The only way a static test like that makes sense is doing what SpaceX already does, the short static tests before launch that last 5-10 seconds. After that time, the rocket isn't sitting still anymore, so why bother running the engines longer in an environment that is A: already understood and B: not relevant to any actual flight?

Edited 6 hours ago by .50calBMG

[Nuke]

4 hours ago, AckSed said:

Speaking of development, overexcitable YT thumbnail said that they might be switching to new, orange version of the tiles, but I haven't seen any hint of this.

Any credibility?

thats clickbait, i think they are just putting a few of the metallic tiles on to see how they do. like 5 or 6 of them. i guess they want to know how they do before they make thousands of them.

Edited 5 hours ago by Nuke

[tater]

 

[Exoscientist]

3 hours ago, .50calBMG said:

To my knowledge (which I would consider above average), no other rocket has actually been built up and tested like that aside from Falcon 9. Falcon Heavy didn't, but that's a bit of a stretch all things considered. Before you say SLS did, no, they didn't have the boosters on it. Same with shuttle, Saturn V, Titan, Saturn I, any of the Soyuz family, etc. Literally the only rocket that I know of that was given a full duration burn while on the ground in flight configuration was Falcon 9, and I'm sure you can find that video on YouTube. 

  …

 Traditionally in the industry, individual stages are tested for full up, full mission length, full power burn on the test stand.

  Bob Clark

[tater]