totm nov 2023 SpaceX Discussion Thread

[magnemoe]

2 minutes ago, kerbiloid said:

The long and narrow aerodynamic detail from the cabin glass to the fin ("гаргрот") - is it a "fairing", too? Or what's her English name?

  Hide contents

 

Think we are in the same boat as non native English users, newer seen fairing used outside of rockets. 
Granted I'm no expert on aircraft design but not heard it on planes, might be that parts tend to have more specific names 
 

[tater]

24 minutes ago, Meecrob said:

...which is a fairing.

Yeah, technically it is. The trouble I suppose is that "fairing" WRT rockets is generally the bit that protects the payload.

 

[sevenperforce]

On 9/9/2020 at 7:30 AM, magnemoe said:

On 9/9/2020 at 7:09 AM, totalitor said:

Is it: the higher chamber pressure, the more nozzle extension you need.

Correct me if I am wrong.

Logic indicates you are correct with higher pressure you can have an larger nozzle before pressure become to low at the end. 
However you might want to go with smaller nozzles than optimum because of space constrains. 

Not exactly.

The higher your chamber pressure is, the greater expansion ratio (and therefore the greater nozzle exit area for a given chamber/throat size) you need. This is true. 

The expansion ratio is the ratio of the nozzle exit to the area of the throat. The throat is the point of highest dynamic pressure; the exit is the point of lowest dynamic pressure, and the larger your pressure drop from throat to nozzle exit, the more efficiently you're converting thermal energy into kinetic energy. RaptorSL, for example, has a nozzle expansion ratio of 40:1, meaning that the area of the exit is forty times the area of the throat. This is dramatically higher than the MerlinSL, with its expansion ratio of just 16. This makes intuitive sense; the Raptor's chamber pressure is 3.4x greater than the Merlin's and so you need a bigger expansion ratio in order to translate as much of that pressure differential as possible into a higher exhaust velocity. The throat of the Merlin is actually 2.8 centimeters larger in diameter than the throat of the Raptor!

However, when we are talking about nozzle extensions, we are talking about the difference between the expansion ratio at sea level and the expansion ratio in vacuum. And so then it actually goes in the opposite direction.

Remember that the efficiency with which you convert thermal energy into kinetic energy is a function of the pressure drop between any two points. An infinitely long, perfectly efficient nozzle would convert 100% of the thermal energy into kinetic energy. However, at sea level you must truncate the nozzle to avoid overexpansion and resulting flow separation. The MerlinSL cannot expand from 1410 psi to 0 psi, but from 1410 psi to a minimum of 14.7 psi. Likewise, the RaptorSL cannot expand from 4800 psi to 0 psi, but from 4800 psi to a minimum of 14.7 psi.

But if you think in terms of efficiency, what does that tell you? Well, the MerlinSL is quite efficient -- it converts 98.96% of its chamber pressure into exhaust velocity at sea level; only 1.04% of its pressure is wasted (14.7/1410 = 1.04%). RaptorSL, on the other hand, is even more efficient -- it converts 99.69% of its chamber pressure into exhaust velocity at sea level, with only 0.31% of its pressure being wasted (14.7/4800 = 0.31%). This is why higher chamber pressure is so important. It's not just a matter of thrust per unit area (although that's important), but of specific impulse. The higher your chamber pressure, the more work you can extract from your propellant in the pressure drop from chamber to sea level pressure.

What does all this mean? Well, it means that the MerlinSL has a significant amount of "work" left to extract, and so the MVac needs a large expansion nozzle (the MVac's expansion ratio is 10.3x greater than the MerlinSL's). The RaptorSL, on the other hand, doesn't have as much leftover "work" in its exhaust at nozzle exit, and so the RVac doesn't need as large of an expansion nozzle relative to its SL nozzle. The RVac's expansion ratio is only 3.2x greater than the RaptorSL's! 10.3/3.2 = 3.22, which is remarkably close to how much more chamber pressure the Raptor has in comparison to the Merlin.

tl;dr -- the higher your chamber pressure, the more efficiently you can extract work from your propellant while expanding to sea level, and so the less remaining work there is to extract while expanding to vacuum, requiring a proportionally smaller vacuum expansion bell.

 

Edited September 11, 2020 by sevenperforce

[sevenperforce]

On 9/10/2020 at 2:04 AM, tater said:

 

What Tim may or may not understand is that the RS-25 was designed in order to have maximal efficiency in vacuum without risking flow separation at sea level. If the RS-25 had been built as a pure sea level engine, it would have had an expansion ratio of roughly 25:1, making its nozzle only 1.4 meters across instead of 2.4 meters across. The expansion bell was far larger than it needed to be at sea level and so it had to be tapered in order to prevent flow separation. The RS-25's vacuum specific impulse is almost 10% greater than the RS-68 because it is a staged-combustion engine (which is intrinsically more efficient than a GG engine) and has a chamber pressure more than twice as high as the RS-68, yet its sea level specific impulse was only 0.3% greater. A dedicated sea-level version of the RS-25 would have a higher sea level specific impulse than the RS-25 actually has, but a markedly lower vacuum specific impulse. 

So that is why. Expansion ratio is not everything. Because the RS-25 was a sustainer engine which needed to provide the bulk of the Shuttle's dV in vacuum, Rocketdyne built an altitude-compensating nozzle which prioritized vacuum specific impulse over sea level specific impulse. The Superheavy booster does not need to provide very much vacuum dV and so it is a much better idea to optimize for sea level specific impulse (and it's much easier because of the ridiculous chamber pressure).

Elon obviously knows all this, but he is giving Tim the simple answer -- booster too small, nozzles too big, can't fit more.

 

[tater]

2 minutes ago, sevenperforce said:

Elon obviously knows all this, but he is giving Tim the simple answer -- booster too small, nozzles too big, can't fit more.

Yeah, agreed. But Elon answers Dodd's questions all the time, so...

 

[sevenperforce]

Just now, tater said:

Yeah, agreed. But Elon answers Dodd's questions all the time, so...

I don't mean to crap on Tim. Really, I don't. But asking "why don't more sea level engines use tapered vacuum nozzles" suggests he does not have nearly as deep an understanding of rocket science as he seems to hold out. It is kind of a waste of time to ask Elon dumb questions. If I had 1/10th the access to Elon that Tim has, we could REALLY learn some good stuff.

[tater]

2 minutes ago, sevenperforce said:

I don't mean to crap on Tim. Really, I don't. But asking "why don't more sea level engines use tapered vacuum nozzles" suggests he does not have nearly as deep an understanding of rocket science as he seems to hold out. It is kind of a waste of time to ask Elon dumb questions. If I had 1/10th the access to Elon that Tim has, we could REALLY learn some good stuff.

Yeah. It's a strange question, "Why no vac engines on a booster that barely enters space?"

SH with not enough sustainers and SRBs? LOL.

[sevenperforce]

10 minutes ago, tater said:

Yeah. It's a strange question, "Why no vac engines on a booster that barely enters space?"

SH with not enough sustainers and SRBs? LOL.

I always laugh whenever I go back and watch Elon trying to explain why aerospikes on Starship are a bad idea and how he has to basically explain it like Tim is 5:

"Um, so uh, you've gotta get your combustion efficiency, uh...so you know, there's really two parts to...like, when you have a rocket engine, what are you trying to do? You want to shoot things out, uh, as fast as possible, in a straight line. So you've got your combustion efficiency. What percentage of max theoretical combustion efficiency are you, um, and then what sort of nozzle efficiency? Are you straightening the flow? You have to shoot it out in a straight line so then you go in the other direction."

And then later...

"I would love it if someone could show us how an aerospike is a smart move, because then we would just do an aerospike! You just have to show that your combustion efficiency is not affected, and that you're straightening the flow sufficiently. And this is the other thing: if you've got a two-stage rocket, where your boost stage is primarily in atmosphere and your upper stage is primarily in vacuum, then you can specialize for a vacuum nozzle and for a sea level nozzle, and then why do you need the aerospike?"

Any altitude-compensating nozzle is by definition less efficient at sea level than a specialized sea level nozzle and less efficient in vacuum then a specialized vacuum nozzle.

[tater]

4 minutes ago, sevenperforce said:

Any altitude-compensating nozzle is by definition less efficient at sea level than a specialized sea level nozzle and less efficient in vacuum then a specialized vacuum nozzle.

The only utility is in an SSTO, really.

WRT SS, I suppose SS is an SSTO on Mars (where it doesn't matter enough to merit them). The only other place is landing, and again, it's a few seconds of burn to kill terminal velocity.

I'd be more interested in air augmentation on ascent

 

[Ricktoberfest]

1 hour ago, sevenperforce said:

I don't mean to crap on Tim. Really, I don't. But asking "why don't more sea level engines use tapered vacuum nozzles" suggests he does not have nearly as deep an understanding of rocket science as he seems to hold out. It is kind of a waste of time to ask Elon dumb questions. If I had 1/10th the access to Elon that Tim has, we could REALLY learn some good stuff.

Maybe Elon answers his questions cause he doesn’t go on public forums making an ass of himself by saying how dumb other people are. Just sayin

 

[mikegarrison]

36 minutes ago, sevenperforce said:

Any altitude-compensating nozzle is by definition less efficient at sea level than a specialized sea level nozzle and less efficient in vacuum then a specialized vacuum nozzle.

No, this is not true. What's true is that a nozzle designed as a compromise is by definition a compromise. But let's say you had a nozzle that could morph to any shape. Then it could be made ideal for every ambient pressure.

It's not the "altitude-compensating" part that is compromising. It is the way that altitude compensation is achieved that can be compromising.

[kerbiloid]

Spoiler

"It's a free country, and it's a private rocket, and it may have any shape of nozzles it wants, unless they are legally prohibited in this state."

 

[sevenperforce]

22 minutes ago, tater said:

WRT SS, I suppose SS is an SSTO on Mars (where it doesn't matter enough to merit them).

A "sea level" version of Raptor for Mars would be just 0.7% smaller in diameter than the RVac. 

22 minutes ago, tater said:

I'd be more interested in air augmentation on ascent

I could see it being used to great effect by solid-fueled multistage smallsat launchers. They develop high atmospheric speeds due to their high thrust and they stage rapidly, so you can shed the extra weight of the augmentation system. Something like the SS-520 could probably be cut nearly in half with air augmentation.

If you want to go reusable, it's tougher. John Bucknell is in love with air augmentation for ascent but honestly EDL is always the killer. I can't think of any way to make reusability and air augmentation play together due to EDL issues, unless you were to go with a HOTOL design.

6 minutes ago, Ricktoberfest said:

Maybe Elon answers his questions cause he doesn’t go on public forums making an ass of himself by saying how dumb other people are. Just sayin

Smart people can still ask dumb questions. Just saiyan.

[mikegarrison]

1 hour ago, sevenperforce said:

I don't mean to crap on Tim. Really, I don't. But asking "why don't more sea level engines use tapered vacuum nozzles" suggests he does not have nearly as deep an understanding of rocket science as he seems to hold out. It is kind of a waste of time to ask Elon dumb questions. If I had 1/10th the access to Elon that Tim has, we could REALLY learn some good stuff.

It's a game. This is a "softball" interview. Tim asks leading questions so that Elon can look good answering them.

This, by the way, is *why* Elon works with Tim. If Tim asked him tough questions that perhaps made him look bad ("hardball" interview), Elon would be a lot less willing to talk with Tim.

[sevenperforce]

4 minutes ago, mikegarrison said:

No, this is not true. What's true is that a nozzle designed as a compromise is by definition a compromise. But let's say you had a nozzle that could morph to any shape. Then it could be made ideal for every ambient pressure.

It's not the "altitude-compensating" part that is compromising. It is the way that altitude compensation is achieved that can be compromising.

Fair point. I should have said "realistically" not "by definition".

Just now, mikegarrison said:

It's a game. This is a "softball" interview. Tim asks leading questions so that Elon can look good answering them.

This, by the way, is *why* Elon works with Tim. If Tim asked him tough questions that perhaps made him look bad ("hardball" interview), Elon would be a lot less willing to talk with Tim.

I don't want to ask Elon tough questions; I want to ask Elon interesting questions that I can't learn the answer to by googling for twelve seconds.

Sigh. Not everyone can nerd out on this stuff as much as we do.

[mikegarrison]

2 minutes ago, sevenperforce said:

I don't want to ask Elon tough questions; I want to ask Elon interesting questions that I can't learn the answer to by googling for twelve seconds.

In that case, then you don't want to talk to Elon anyway. You want to talk to the folks who work for him.

[tater]

16 minutes ago, sevenperforce said:

If you want to go reusable, it's tougher. John Bucknell is in love with air augmentation for ascent but honestly EDL is always the killer. I can't think of any way to make reusability and air augmentation play together due to EDL issues, unless you were to go with a HOTOL design.

I prefer the "reuse in space" nuclear variant.

[sevenperforce]

10 minutes ago, mikegarrison said:

In that case, then you don't want to talk to Elon anyway. You want to talk to the folks who work for him.

Eh, I don't know about that. Elon's position as chief engineer is not just window dressing. He might not be THE injector expert or THE turbopump expert or THE combustion efficiency expert or THE metallurgy expert, but he probably knows as much diverse information about his rockets as anyone at SpaceX.

Just a few examples of the many questions about Starship I'd ask if I could:

Spoiler

• Has SpaceX started designing the gas-gas RCS thrusters for Starship and Superheavy? What kind of Isp are they expected to develop? Will they run directly off tank ullage pressure or do you think Starship will use an accumulator?
• The heat shield on Starship seems like a tough problem to solve, even though the tiles are much lighter and thinner than the ones on the Shuttle orbiter. What are the different ways SpaceX has attempted to affix the tiles to the skin? Is there anything that seems particularly promising? How many tiles could be lost on re-entry without risking burn-through on the skin?
• P2P flight with Starship seems like it would be amazing but it also seems risky. Do you think there's enough demand for hypersonic transcontinental flight? Could Starship use extra sea level engines to do single-stage P2P? 
• For powering Martian ISRU, do you expect to use nuclear, solar, or both? How long do you expect it will take to refuel Starship on Mars? Will you do proof-of-concept on Earth?
• What has SpaceX done to mitigate or eliminate failure modes for Starship's entry, descent, and landing? What remaining failure modes bother you or keep you up at night? Are there any contingency landing modes?
• Currently, do you expect Superheavy's landing legs to be fixed, telescoping, or folding? How fixed is that design?

These are all questions that I am confident Elon could easily answer off the top of his head. They're also all questions that we don't have the immediate answer to.

Of course, they're also all niche questions that only extreme nerds would care about. Maybe that's why Tim has a channel and I don't.

 

10 minutes ago, tater said:

I prefer the "reuse in space" nuclear variant.

In which case air augmentation is wasted mass

[tater]

11 minutes ago, sevenperforce said:

In which case air augmentation is wasted mass

Is it? That was his original design, and chemical version was really the variant.

No EDL, air aug was to get large payload as I recall his talk.

[mikegarrison]

6 minutes ago, sevenperforce said:

Eh, I don't know about that. Elon's position as chief engineer is not just window dressing. He might not be THE injector expert or THE turbopump expert or THE combustion efficiency expert or THE metallurgy expert, but he probably knows as much diverse information about his rockets as anyone at SpaceX.

IIRC, you're not an engineer, right? I've been one for 30 years now. Trust me on this.

The higher the executive you are pitching to, the simpler the powerpoint charts have to be. It's not because the execs are not smart. It's because knowing the details isn't their job. Their job is to be able to understand just enough of the details, for just long enough to make the right decision, and then to move on to something else. Depending on the issue that can mean anything from "doing it this way is cheaper and just as good" to a four-hour deep-dive into really technical issues. But four hours of an exec's time better be worth it -- you don't go in with a four-hour pitch when the bottom line is simply "doing it this way is cheaper and just as good".

[sevenperforce]

3 minutes ago, mikegarrison said:

Their job is to be able to understand just enough of the details, for just long enough to make the right decision, and then to move on to something else. Depending on the issue that can mean anything from "doing it this way is cheaper and just as good" to a four-hour deep-dive into really technical issues. But four hours of an exec's time better be worth it -- you don't go in with a four-hour pitch when the bottom line is simply "doing it this way is cheaper and just as good".

I agree...which is why I wouldn't ask Elon stupidly technical questions like, "From a materials integrity standpoint, what are the primarily failure modes if Raptor's turbopumps were pushed too far in an emergency abort?" or "What are the pros and cons of using a thorium salt reactor to operate the Martian ISRU plant?"

But I am confident he could speak to Starship's broad failure modes, demand for hypersonic passenger transport, and RCS thruster design.

[tater]

Yeah, broad questions he can plausibly answer are better than in the weeds questions.

That said, asking "Why not put a hybrid sea level/vacuum engine on a rocket stage that doesn't need it?" is not a good use of access

 

[magnemoe]

3 hours ago, sevenperforce said:

What Tim may or may not understand is that the RS-25 was designed in order to have maximal efficiency in vacuum without risking flow separation at sea level. If the RS-25 had been built as a pure sea level engine, it would have had an expansion ratio of roughly 25:1, making its nozzle only 1.4 meters across instead of 2.4 meters across. The expansion bell was far larger than it needed to be at sea level and so it had to be tapered in order to prevent flow separation. The RS-25's vacuum specific impulse is almost 10% greater than the RS-68 because it is a staged-combustion engine (which is intrinsically more efficient than a GG engine) and has a chamber pressure more than twice as high as the RS-68, yet its sea level specific impulse was only 0.3% greater. A dedicated sea-level version of the RS-25 would have a higher sea level specific impulse than the RS-25 actually has, but a markedly lower vacuum specific impulse. 

So that is why. Expansion ratio is not everything. Because the RS-25 was a sustainer engine which needed to provide the bulk of the Shuttle's dV in vacuum, Rocketdyne built an altitude-compensating nozzle which prioritized vacuum specific impulse over sea level specific impulse. The Superheavy booster does not need to provide very much vacuum dV and so it is a much better idea to optimize for sea level specific impulse (and it's much easier because of the ridiculous chamber pressure).

Elon obviously knows all this, but he is giving Tim the simple answer -- booster too small, nozzles too big, can't fit more.

This also SH need lots trust so it need lots of engines, also as its RTLS its smart to have high trust to you can accelerate SS up to suborbital speed fast and drop it before getting far downrange and the working area is limited to 9-10 meter in diameter. 

Now for manned starship having an sort of altitude-compensating nozzle for the vacuum engines would make sense for abort modes.
Sort of as only requirement is that the engine does not explode or damage other critical parts. 
 

13 minutes ago, tater said:

Yeah, broad questions he can plausibly answer are better than in the weeds questions.

That said, asking "Why not put a hybrid sea level/vacuum engine on a rocket stage that doesn't need it?" is not a good use of access

Yes for SH as said above, but on the other hand you don't stay at sea level for long, at 10 km atmosphere is 0.1 bar. For most rockets with SRB or the Delta heavy core having an altitude-compensating nozzle should make some sense as it takes you halfway to orbit, but yes my only experience is KSP  
 

Edited September 11, 2020 by magnemoe
SS not SH

[mikegarrison]

21 minutes ago, magnemoe said:

Yes for SH as said above, but on the other hand you don't stay at sea level for long, at 10 km atmosphere is 0.1 bar. For most rockets with SRB or the Delta heavy core having an altitude-compensating nozzle should make some sense as it takes you halfway to orbit, but yes my only experience is KSP

Keep in mind that efficiency isn't everything. Raw thrust is needed most desperately right at the start, so if you can't generate sufficient thrust with your upper-altitude sustainer engine, you end up losing all your gained upper-altitude efficiency and more due to "gravity losses" near the pad.

[sevenperforce]

38 minutes ago, magnemoe said:

Now for manned starship having an sort of altitude-compensating nozzle for the vacuum engines would make sense for abort modes.
Sort of as only requirement is that the engine does not explode or damage other critical parts. 

The altitude-compensating nozzle with the highest vacuum efficiency is a dual-bell nozzle, and that is what the RVac has. It is more efficient in vacuum than the adjusted-exit-angle nozzle of the RS-25, but less efficient at sea level due to parasitic plume recirculation within the extension.

That's ok, though, because the RVac would only fire at sea level in the event of an abort, which would be a time for prioritizing thrust, not efficiency. 

21 minutes ago, mikegarrison said:

Raw thrust is needed most desperately right at the start, so if you can't generate sufficient thrust with your upper-altitude sustainer engine, you end up losing all your gained upper-altitude efficiency and more due to "gravity losses" near the pad.

Precisely. An altitude-compensating sustainer engine will be more efficient than a sea-level-optimized over the course of the entire ascent, but it has lower thrust than a smaller engine at sea level AND takes up more space.