totm nov 2023 SpaceX Discussion Thread

[StrandedonEarth]

6 minutes ago, magnemoe said:

This, nobody has done a full burn with SRB for for good reasons, heavy lift orbital rockets here not military stuff. But srb add stress to craft , side boosters do same is true for falcon heavy, who was delayed by years so spacex dropped side boosters from their plan. 

The large 4-segment STS and 5-segment SLS SRBs have had full-duration static fires (can’t make them shorter), but in a horizontal, not vertical position; one at a time (not in pairs), and not mounted to a core stage, firing or not. 

(I probably read your post wrong; the wording of the first sentence is a touch ambiguous on second reading. )

[Exoscientist]

1 hour ago, zolotiyeruki said:

A few counterpoints:

1) You have written many, many words, but you still haven't answered the simple question: "what benefit would ground testing give you that modern flight testing doesn't?" 

2) Sure, SpaceX and NASA have done integrated tests in the past.  That doesn't mean they absolutely must adhere to that same practice for every engine and every vehicle.  A wise young guy once told me "tradition has to claim on sanctity."  Just because a full-up, full-thrust, full-duration static test was called for under a previous program a decade or a freaking half century ago doesn't mean that the same test is appropriate for this program at this stage with SpaceX's current development philosophy.

 Consider, that is the standard industry practice. It is not just SpaceX and Nasa doing in that way in the past. Every company in the industry does it that way. Every company could just test single engines on test stands if they wanted to. That would be much cheaper. Instead they follow the accepted practice of doing full flight analog testing prior to test flights.

 That the Soviet N-1  experienced not just engine shutdowns but engine explosions on every test flight and the  SuperHeavy/Starship also experienced the same thing, including the engine explosions, suggests this is not a better approach to getting reliable engines. 

  Robert Clark

[zolotiyeruki]

14 minutes ago, Exoscientist said:

 Consider, that is the standard industry practice. It is not just SpaceX and Nasa doing in that way in the past. Every company in the industry does it that way. Every company could just test single engines on test stands if they wanted to. That would be much cheaper. Instead they follow the accepted practice of doing full flight analog testing prior to test flights.

 That the Soviet N-1  experienced not just engine shutdowns but engine explosions on every test flight and the  SuperHeavy/Starship also experienced the same thing, including the engine explosions, suggests this is not a better approach to getting reliable engines. 

  Robert Clark

You still haven't answered the question: What benefit would static fires bring over test flights?

"Because that's how it has always been done" is not an argument.  It's a logical fallacy called "Appeal to tradition."  Tradition can result from rational thinking and sound logic, but is not proof of such.

"Suggests this is not a better approach" - Hang on.  About the only similarities between the programs are that they both involve big rockets and have suffered some failures.*  Literally nothing else is common between them.  Different countries, gov't vs private, a half-century of technological advancement, different goals, different politics.  You're trying to compare apples to oranges here.

[tater]

36 minutes ago, Exoscientist said:

 Consider, that is the standard industry practice. It is not just SpaceX and Nasa doing in that way in the past. Every company in the industry does it that way. Every company could just test single engines on test stands if they wanted to. That would be much cheaper. Instead they follow the accepted practice of doing full flight analog testing prior to test flights.

Again, so what? Accepted practice would make SS/SH an expended rocket, and grossly more expensive. Accepted practice on a huge rocket in the 2020s? We have an example. SLS.

Answer my question above: Should SS/SH find itself in orbit in the next few flights, with flight testing proceeding as expected, losing vehicles on the recovery side until they sort out S1 landing, then S2 EDL, would a new company choose the "old way," or the SpaceX way, and why might they have reason to pick one over the other? Both would now be "industry practice" where the most recent version (SpaceX) is also the current most successful rocket company on the planet.

 

36 minutes ago, Exoscientist said:

 That the Soviet N-1  experienced not just engine shutdowns but engine explosions on every test flight and the  SuperHeavy/Starship also experienced the same thing, including the engine explosions, suggests this is not a better approach to getting reliable engines. 

One, if you have to go back decades for an example in this case, it's a bad example. What was the Soviet CFD simulation work like? How about all their other detailed computer simulation? Oh, yeah, not a thing. Every single thing about the N1 comparison is unserious, IMO. Two, IFT-2 had no such problem on the booster. It was flawless until after MECO/sep. It failed a reuse/landing test—which other industry standard rocket unsuccessfully boosted back and attempted a test landing? Oh yeah, 100% of them failed this test since rockets. SLS core also blew up after sep, raining rocket parts in the south Pacific.

The only failure on IFT-2 that matters was the Ship. We have no idea what the actual failure was, so <shrug>. Engine? Plumbing? Something else?

Edited January 8 by tater

[Kartoffelkuchen]

5 minutes ago, tater said:

Again, so what? Accepted practice would make SS/SH an expended rocket, and grossly more expensive. Accepted practice on a huge rocket in the 2020s? We have an example. SLS.

Answer my question above: Should SS/SH find itself in orbit in the next few flights, with flight testing proceeding as expected, losing vehicles on the recovery side until they sort out S1 landing, then S2 EDL, would a new company choose the "old way," or the SpaceX way, and why might they have reason to pick one over the other? Both would now be "industry practice" where the most recent version (SpaceX) is also the current most successful rocket company on the planet.

 

One, if you have to go back decades for an example in this case, it's a bad example. What was the Soviet CFD simulation work like? How about all their other detailed computer simulation? Oh, yeah, not a thing. Every single thing about the N1 comparison is unserious, IMO. Two, IFT-2 had no such problem on the booster. It was flawless until after MECO/sep. It failed a reuse/landing test—which other industry standard rocket unsuccessfully boosted back and attempted a test landing? Oh yeah, 100% of them failed this test since rockets. SLS core also blew up after sep, raining rocket parts in the south Pacific.

The only failure on IFT-2 that matters was the Ship. We have no idea what the actual failure was, so <shrug>. Engine? Plumbing? Something else?

Hey guys, my blocker for *individual persons* does not have the effect it should when all of you pick up on *the persons* arguments again and again.

I am by no means a moderator, but I think that literally all of the counterarguments have at least been made twice now. Perhaps just let the person post their statements into the void, until something with a bit more substance to it comes up?

[sevenperforce]

On 1/6/2024 at 3:15 PM, RCgothic said:

Even if it was a knife-edge of mostly exactly 120s tests with a few falling short (which this data isn't), that would *still* not say anything about raptor's reliability because 3rd party observers have no idea what's being tested or what the abort criteria are.

It could be GSE faults. It could be test aborts more conservative than flight. It could be testing above 100% throttle. It could be deliberate tests to failure.

To expand on this point, let's say that @Exoscientist's speculation was 100% true -- the majority of these tests are aiming for exactly 120 seconds, and some significant number of those tests fall short of 120 seconds because the engines spontaneously fail somewhere between 110 and 115 seconds (or something like that). Raptor is clearly unreliable, right?

Nope, this still doesn't provide any meaningful evidence that Raptor is too unreliable for flights, let alone flight tests. For example, let us imagine that the Raptor manufacturing process has some chance (let's say 3%) of introducing a fatal defect in the turbopump exhaust injection manifold, and that defect is undetectable except through static testing.  Let us further suppose that engines with the defect have a 99.999% chance of failing before 120 seconds, and that engines without the defect cannot develop the defect by static fire testing. Because SpaceX is hardware-rich, they could simply elect to static test each of their engines a few times and throw out the bad ones. This would be a completely reasonable way to eliminate the defect, if that's the way they chose to do it.

Absolutely nothing can be inferred from these observations, other than the fact that SpaceX has an active, healthy, aggressive test-firing program.

[Exoscientist]

1 hour ago, zolotiyeruki said:

You still haven't answered the question: What benefit would static fires bring over test flights?

"Because that's how it has always been done" is not an argument.  It's a logical fallacy called "Appeal to tradition."  Tradition can result from rational thinking and sound logic, but is not proof of such.

"Suggests this is not a better approach" - Hang on.  About the only similarities between the programs are that they both involve big rockets and have suffered some failures.*  Literally nothing else is common between them.  Different countries, gov't vs private, a half-century of technological advancement, different goals, different politics.  You're trying to compare apples to oranges here.

 

 On the static stand you can gradually build up to full thrust and full flight duration. If you see out of nominal state for one or more engines, you can shut the test down, examine the engines not operating properly and compare to the ones that are. Then try again gradually building up to full flight conditions. In contrast if one or more engines fail in flight, even if it didn’t explode but had to be shutdown, usually you have to destroy it by FTS for fear it will stray too far out of the safety zone.

  Robert Clark

[Terwin]

2 hours ago, Exoscientist said:

 Consider, that is the standard industry practice. It is not just SpaceX and Nasa doing in that way in the past. Every company in the industry does it that way. Every company could just test single engines on test stands if they wanted to. That would be much cheaper. Instead they follow the accepted practice of doing full flight analog testing prior to test flights.

 That the Soviet N-1  experienced not just engine shutdowns but engine explosions on every test flight and the  SuperHeavy/Starship also experienced the same thing, including the engine explosions, suggests this is not a better approach to getting reliable engines. 

  Robert Clark

As best as I can tell, only a small percentage of rocket launches, especially in the last few decades, have had a full stack, full flight-like duration static fire before launch as you describe.  (none of the launches with solid boosters would count for example, and you showed that Falcon 9's don't do that before each launch, etc)

If only a small number of actual launches have performed that procedure before the launch, how could it possibly be a standard procedure?

 

14 minutes ago, Exoscientist said:

 On the static stand you can gradually build up to full thrust and full flight duration. If you see out of nominal state for one or more engines, you can shut the test down, examine the engines not operating properly and compare to the ones that are. Then try again gradually building up to full flight conditions. In contrast if one or more engines fail in flight, even if it didn’t explode but had to be shutdown, usually you have to destroy it by FTS for fear it will stray too far out of the safety zone.

  Robert Clark

That makes sense if you have unlimited funds and very little hardware, but SpaceX is very hardware rich, and is trying to do this on a commercial budged(as opposed to cost-plus).

Come to think of it, the type of static-fire you describe seems almost like a 'standard procedure' to squeeze maximum funds out of the government on a cost-plus contract with minimal costs to the provider.  Maximizing the opportunities to add delays(and costs) to the project by minimizing the number of problems you uncover with each test(by shutting it down as soon as you find a discrepancy).

As opposed to launch testing, which generally covers more scenarios per test, but has a material cost to the rocket company(hardware instead of just (billable) time and fuel).

 

I guess that is the real difference, SpaceX is following a fast, cost-effective approach, as opposed to a 'maximize the cost-plus contract' approach.

[zolotiyeruki]

42 minutes ago, Exoscientist said:

 On the static stand you can gradually build up to full thrust and full flight duration. If you see out of nominal state for one or more engines, you can shut the test down, examine the engines not operating properly and compare to the ones that are. Then try again gradually building up to full flight conditions. In contrast if one or more engines fail in flight, even if it didn’t explode but had to be shutdown, usually you have to destroy it by FTS for fear it will stray too far out of the safety zone.

I'm glad you addressed the actual question, but it still sounds like you're grasping at (and missing) wildly speculative straws to support your "SpaceX is doing it wrong" thesis.

1) You might have missed it, but SpaceX are waaaay past the point of gradually ramping up Raptors to full thrust and full duration.  See IFT-2.  Clearly, their approach is working.
2) It sounds like you're suggesting that recovering the pieces of a failed engine would aid in identifying design issues.  Do you have any evidence that SpaceX has collected any useful data from their static tests in McGregor?
3) The engines that have been flown to date have already been obsolete when they flew, so a full-duration test fire would be pointless anyway, as would collecting engine debris.
4) Gradually increasing the thrust and duration over multiple burns is silly, because A) the cumulative engine time isn't going to be representative of real-world use, and B) SpaceX is, as @Terwin said, "hardware rich."  There's no need to be preserving engines.

I'm still not seeing any comparative advantage to a full-duration static ground fire at this stage of the program.

Edited January 8 by zolotiyeruki

[Ultimate Steve]

Okay. That I can believe. Thank you for answering the question. Big static fire series to catch issues and protect hardware, in a way that, per test and per booster, in an ideal world, is somewhat more effective at catching issues than flight testing.

Then it becomes an economic problem. Cost (and feasibility) of a test stand versus cost of however many full stacks you end up exploding.

I will add that startup and shutdown are the most iffy parts of such a test, steady state operation (although starship throttles around a lot so not a perfect comparison) is less likely to cause issues in most cases. A 10 second test will catch many of the issues that would also be caught on a longer test.

You can't uncover most vibration and acceleration issues this way because the vibration and acceleration environments are different.

So now it's an even smaller number of exploded stacks on that side of the equation due to doing the half measure testing that they are doing. Given the unique difficulties SpaceX faces on their plot of land, I could see that equation closing. 

[Terwin]

18 minutes ago, Terwin said:

If only a small number of actual launches have performed that procedure before the launch, how could it possibly be a standard procedure?

Perhaps you meant that it is standard procedure for rocket development?

To that I would counter that it is only standard procedure for developing an expendable rocket. 

Falcon 9 was expendable with an eye for reuse, Starship is being designed as a fully reusable rocket, and not many of those have gotten far enough to be able to do a static fire,

 

Edited January 8 by Terwin

[tater]

Assuming SS eventually succeeds, then add that Stoke might follow the same Test by flying standard, and Nova becomes fully, rapdily reusable, would a yet newer player think, "The standard of the 2 companies who have done the best rockets in history was to test by flying, not the broken, old method of the past, we should follow the standard and fly!"

[Exoscientist]

 I have been accused of being anti-SpaceX because of my criticism of the Starship. Actually, after a calculation I'm convinced the Starship can be operational, like, tomorrow, with relatively small design changes:

Towards advancing the SpaceX Starship to operational flight: SpaceX should lower the Raptor chamber pressure and thrust level.
https://exoscientist.blogspot.com/2024/01/towards-advancing-spacex-starship-to.html

The Raptor engine has shown continued failures on all of test stands, Starship low altitude landing test flights, and the two orbital test flights. But the Raptors on the booster on the last test flight were able to complete the ascent part of the flight without failures. They failed only after the attempted to relight. 

Multiple-lines of evidence suggest that on that last test flight SpaceX throttled down the Raptors on the booster to less than 75% while those on the Starship were run at ~90%. I've suggested this is why the booster engines were able to fire reliably during the ascent and those on the upper stage were not.

If this is the case, then it suggests a method to get Raptor reliability: run them at ~75% throttle on both stages. But if keeping the same stage dry masses this would result in the payload of the reusable version being reduced to approximately in the range of 100 tons from 150 tons.

Instead, I advise first start with reducing the dry masses by optimally lightweighting the expendable versions of both stages. Surprisingly this gives a greater expendable payload than the expendable payload of the current version. Secondly, I suggest using winged, horizontal approach to reusability gives a much reduced payload loss due to reusability. Thirdly, basic orbital mechanics shows high delta-v missions such as to the Moon or Mars are done more efficiently by using more stages. Then a third stage is suggested for the Superheavy/Starship, a mini-Starship as it is called by Robert Zubrin.

This allows single launch and fully reusable missions to the Moon or Mars. No refueling flights required.

    Robert Clark

Edited January 9 by Exoscientist

[sevenperforce]

15 minutes ago, Exoscientist said:

If you see out of nominal state for one or more engines, you can shut the test down, examine the engines not operating properly and compare to the ones that are.

This is very helpful if you do not have a well-instrumented engine with modern realtime telemetry and must physically inspect parts to see what went wrong.

This is much less helpful when you have so much advanced instrumentation and realtime telemetry that you typically don't perform physical inspections of your parts even after a static test because you already know what went wrong before the vehicle can even be detanked.

17 minutes ago, Exoscientist said:

SpaceX should lower the Raptor chamber pressure and thrust level.https://exoscientist.blogspot.com/2024/01/towards-advancing-spacex-starship-to.html

[snip]

I advise first start with reducing the dry masses by optimally lightweighting the expendable versions of both stages. Surprisingly this gives a greater expendable payload than the expendable payload of the current version.

It does not.

You cannot take the mass ratio of an aspirational upper stage and imagine that it will simply scale to the same mass ratio for a lower stage. Lower stages have to be stronger than upper stages. That's basic.

20 minutes ago, Exoscientist said:

I suggest using winged, horizontal approach to reusability gives a much reduced payload loss due to reusability.

If you're removing mass left and right for an expendable version, it's not going to be able to support itself horizontally.

[tater]

1 hour ago, Exoscientist said:

Towards advancing the SpaceX Starship to operational flight

Whatever you did starting there, the font changes and it's practically unreadable in dark mode (an my forum is always on dark mode). There are a few of your posts where you do this. Do you intentionally change font/color, or is it a non-plain text quote or something?

1 hour ago, sevenperforce said:

If you're removing mass left and right for an expendable version, it's not going to be able to support itself horizontally.

I'm not changing my forum theme to read the post (can see it in your quote), but is that the booster, or the ship? In the case of the ship—doesn't work on Mars. Off the table. For the booster, you are of course exactly right.

I'd add that SpaceX has vertically landed more rockets than ULA has flown rockets (by a lot at this point). They know something about landing rockets this way, they know nothing about landing aircraft. not sure why they'd go in a direction they don't already have experience with—and it would require a huge runway... somewhere. Not sure where they have the land for it. The they have to build an airport, which I'm sure involves years of regulatory hurdles, etc. then they have to be able to move that beast from there back to the barn. A nonstarter on many levels.

[Ultimate Steve]

You know, if I had a nickel for every time we would go back and forth on this thread with someone who doesn't buy Starship and uses dark mode incompatible text, I'd have two nickels, which isn't a lot, but it is weird that it happened twice.

Side note, would it be bad faith to dig up some of those old posts and tally how many of them have been proven wrong?

[Ultimate Steve]

Back to actual SpaceX discussion, I just remembered that it actually takes less propellant to fling one of these bad boys to Mars than it does to land on the Moon. Deep space communications, making the heat shield not fall apart and Mars capable, and near zero boiloff aren't trivial problems by any means, but once one of these lands on the Moon, I expect them to fling one at Mars at the next window (or at the next next window if the next one is really close). Most of the stuff required to do Mars will have been demonstrated at that point, and far less flights are required to do a Mars landing test than a Moon landing.

[CatastrophicFailure]

9 hours ago, Ultimate Steve said:

Most of the stuff required to do Mars will have been demonstrated at that point, and far less flights are required to do a Mars landing test than a Moon landing.

Though this be madness, yet there is method in't.
 

—Billy S., esq

[Nuke]

10 hours ago, Ultimate Steve said:

You know, if I had a nickel for every time we would go back and forth on this thread with someone who doesn't buy Starship and uses dark mode incompatible text, I'd have two nickels, which isn't a lot, but it is weird that it happened twice.

Side note, would it be bad faith to dig up some of those old posts and tally how many of them have been proven wrong?

wait, there's dark mode, you mean ive been searing my retina for nothing?

[Shpaget]

47 minutes ago, Nuke said:

wait, there's dark mode, you mean ive been searing my retina for nothing?

The very bottom of the forums.

It's quite fugly and looks unfinished, but at least it's dark.

Edited January 9 by Shpaget

[Nuke]

37 minutes ago, Shpaget said:

The very bottom of the forums.

It's quite fugly and looks unfinished, but at least it's dark.

thanks.

thats a weird place to put a setting.

[sevenperforce]

22 hours ago, tater said:

I'm not changing my forum theme to read the post (can see it in your quote), but is that the booster, or the ship? In the case of the ship—doesn't work on Mars. Off the table. For the booster, you are of course exactly right.

You can always Ctrl+A to select all text on the page and read it quickly without changing your forum theme.

On a related note, thanks for pointing out that there is a dark theme. Much easier on the eyes.

Our polymath fellow is proposing wings for Superheavy, which is a nonstarter in just so many ways.

[tater]

13 minutes ago, sevenperforce said:

You can always Ctrl+A to select all text on the page and read it quickly without changing your forum theme.

I can drag over it to select as well, I was just curious about the how on some of his posts this happens. Never see it on any other posts, unsure if an explicit font/color change, or maybe a cut and paste that keeps formatting? And I'm just being difficult

15 minutes ago, sevenperforce said:

Our polymath fellow is proposing wings for Superheavy, which is a nonstarter in just so many ways.

So yeah, a huge runway... someplace. Which means designing a new vehicle, yada, yada, yada.

[Shpaget]

10 minutes ago, tater said:

maybe a cut and paste that keeps formatting

Most likely that. When you paste something, for me at least, there's a popup at the bottom of the text box that says "Pasted as rich text. Paste as plain text instead". The second half is clickable and removes formatting that rich text contains.

[tater]

Just now, Shpaget said:

Most likely that. When you paste something, for me at least, there's a popup at the bottom of the text box that says "Pasted as rich text. Paste as plain text instead". The second half is clickable and removes formatting that rich text contains.

My guess as well. I never paste my own typing (though given some errors I make that I have to come back and edit, maybe it would be better if I did so, lol).