Author Topic: Starship performance improvements  (Read 14416 times)

Offline steveleach

  • Senior Member
  • *****
  • Posts: 2140
  • Liked: 2735
  • Likes Given: 952
Re: Starship performance improvements
« Reply #20 on: 11/24/2023 04:55 pm »
My understanding is that gravity losses decrease as the vehicle gains velocity, so would the ship at some point cross a threshold where the increase in ISP from just using the vac-raps outweighs the increase in gravity losses from the reduced thrust?

In theory shutting the SL raptors down seconds early would improve performance slightly but that's probably not feasible in practice because the 3 fixed engines would not provide roll control or engine out capability.
Thanks deltaV. A few more questions, because I'm genuinely trying to learn; I'm not challenging your assertions.

Why just seconds (out of the ~6 minute burn) early?

How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

How resiliant to late vac-rap engine out would a Starship mission be anyway? You'd presumably have to compensate with more use of the sea-level engines, and have more gravity losses, so would you have the margins to handle it? If you do have the margins, could you just relight one of the sea-level engines?

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37428
  • Cape Canaveral Spaceport
  • Liked: 21411
  • Likes Given: 428
Re: Starship performance improvements
« Reply #21 on: 11/24/2023 05:15 pm »

Thanks deltaV. A few more questions, because I'm genuinely trying to learn; I'm not challenging your assertions.

Why just seconds (out of the ~6 minute burn) early?

How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.
« Last Edit: 11/24/2023 05:16 pm by Jim »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37428
  • Cape Canaveral Spaceport
  • Liked: 21411
  • Likes Given: 428
Re: Starship performance improvements
« Reply #22 on: 11/24/2023 05:17 pm »
you just relight one of the sea-level engines?

They are never shutdown while the vacuum ones are firing.

Offline steveleach

  • Senior Member
  • *****
  • Posts: 2140
  • Liked: 2735
  • Likes Given: 952
Re: Starship performance improvements
« Reply #23 on: 11/24/2023 10:08 pm »
you just relight one of the sea-level engines?

They are never shutdown while the vacuum ones are firing.
They are if you shut them down, which was what I was asking about.

Offline steveleach

  • Senior Member
  • *****
  • Posts: 2140
  • Liked: 2735
  • Likes Given: 952
Re: Starship performance improvements
« Reply #24 on: 11/24/2023 10:12 pm »

Thanks deltaV. A few more questions, because I'm genuinely trying to learn; I'm not challenging your assertions.

Why just seconds (out of the ~6 minute burn) early?

How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.
Jim, I appreciate that these may seem like stupid questions to you, but your answers simply aren't helping. If the question is worth your time, I'd appreciate it if you give more detail; if it's not, just ignore it.

Offline deltaV

  • Senior Member
  • *****
  • Posts: 2101
  • Change in velocity
  • Liked: 568
  • Likes Given: 1864
Re: Starship performance improvements
« Reply #25 on: 11/25/2023 02:03 am »
Thanks deltaV. A few more questions, because I'm genuinely trying to learn; I'm not challenging your assertions.

Why just seconds (out of the ~6 minute burn) early?

How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

How resiliant to late vac-rap engine out would a Starship mission be anyway? You'd presumably have to compensate with more use of the sea-level engines, and have more gravity losses, so would you have the margins to handle it? If you do have the margins, could you just relight one of the sea-level engines?

I'm afraid I don't know the answers to your questions. Quantifying when it would make sense to shut down the SL engines requires trajectory optimization software and it's hard for an amateur like me to get access to that software due to ITAR laws. I've considered writing that software myself but it would be a fair amount of work for something that would be hard to share with anyone else due to ITAR.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37428
  • Cape Canaveral Spaceport
  • Liked: 21411
  • Likes Given: 428
Re: Starship performance improvements
« Reply #26 on: 11/25/2023 02:15 am »

Thanks deltaV. A few more questions, because I'm genuinely trying to learn; I'm not challenging your assertions.

Why just seconds (out of the ~6 minute burn) early?

How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.

Differn
Jim, I appreciate that these may seem like stupid questions to you, but your answers simply aren't helping. If the question is worth your time, I'd appreciate it if you give more detail; if it's not, just ignore it.

There will be some sort of roll induced by those variances that differential thrusting can't stop.  Also, differential thrusting is not quick reacting.

You lose control because

Offline Brigantine

  • Full Member
  • ***
  • Posts: 303
  • NZ
  • Liked: 146
  • Likes Given: 445
Re: Starship performance improvements
« Reply #27 on: 11/25/2023 02:32 am »
How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.
Jim, I appreciate that these may seem like stupid questions to you, but your answers simply aren't helping. If the question is worth your time, I'd appreciate it if you give more detail; if it's not, just ignore it.

To interpret Jim a bit... to compensate for engine misalignment and thrust differentials, to maintain the right attitude (pitch/yaw/roll) it takes more than an occasional puff of RCS, you need a trim - a constant baseline correction to offset a constant bias.

Suppose 3 RCs do this with a trim of just 0.6⁰ i.e. 0.01 radians at 50% thrust. That's correcting forces totaling 38 kN applied constantly for the rest of the 2nd stage burn (assuming similar leverage). If using cold gas, that's a ~1.8% hit to your ISP (7s).

Or if you assume you have very fine throttle control of RVacs (pitch & yaw sorted), and your roll RCS has 4x the leverage of the RCs, that's still 10 kN. Assume hot gas thrusters and that's a ~0.2% hit to your ISP (1s).

What force can the current Ship RCS manage?

Also compare with:
Running two sea levels at 40% thrust and 6 Vacuum at full thrust, we get an Isp of 6*378 + .8 * 363 or an average of 376.2

However it's silly to plan a mission on what is likely to happen, you have to plan on single failures.

So if it's a VR that fails it's 5*378 + .8*363 => average of 375.9

The difference between that an a differentially throttled Vacuum Raptor system is .5%, or with a deltaV of 4km/sec 22 m/sec.

There's enough other stuff going on to not worry about 22m/sec.
« Last Edit: 11/25/2023 02:41 am by Brigantine »

Offline Nomadd

  • Senior Member
  • *****
  • Posts: 8831
  • Seoul
  • Liked: 60391
  • Likes Given: 1295
Re: Starship performance improvements
« Reply #28 on: 11/25/2023 02:47 am »
 I keep looking for updates on hot gas thrusters, but haven't seen anything since the early lunar lander award days. Those could relate to several things from lunar landing to killer rcs to putting excess tank gas to good use.
Those who danced were thought to be quite insane by those who couldn't hear the music.

Offline steveleach

  • Senior Member
  • *****
  • Posts: 2140
  • Liked: 2735
  • Likes Given: 952
Re: Starship performance improvements
« Reply #29 on: 11/25/2023 09:01 am »
How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.
Jim, I appreciate that these may seem like stupid questions to you, but your answers simply aren't helping. If the question is worth your time, I'd appreciate it if you give more detail; if it's not, just ignore it.

To interpret Jim a bit... to compensate for engine misalignment and thrust differentials, to maintain the right attitude (pitch/yaw/roll) it takes more than an occasional puff of RCS, you need a trim - a constant baseline correction to offset a constant bias.

Suppose 3 RCs do this with a trim of just 0.6⁰ i.e. 0.01 radians at 50% thrust. That's correcting forces totaling 38 kN applied constantly for the rest of the 2nd stage burn (assuming similar leverage). If using cold gas, that's a ~1.8% hit to your ISP (7s).

Or if you assume you have very fine throttle control of RVacs (pitch & yaw sorted), and your roll RCS has 4x the leverage of the RCs, that's still 10 kN. Assume hot gas thrusters and that's a ~0.2% hit to your ISP (1s).

What force can the current Ship RCS manage?

Also compare with:
Running two sea levels at 40% thrust and 6 Vacuum at full thrust, we get an Isp of 6*378 + .8 * 363 or an average of 376.2

However it's silly to plan a mission on what is likely to happen, you have to plan on single failures.

So if it's a VR that fails it's 5*378 + .8*363 => average of 375.9

The difference between that an a differentially throttled Vacuum Raptor system is .5%, or with a deltaV of 4km/sec 22 m/sec.

There's enough other stuff going on to not worry about 22m/sec.
Yep, I get that RCS won't cut it for pitch and yaw, but I was under the impression that differential thrust does provide that. It's specifically roll-control that I'm not really understanding. How do single-engine upper stages handle roll control? As I understand it, you can't control roll with a single gimballing engine.

And yes, InterestedEngineer's numbers showed that its not worth it, though I'm not sure where his ISP numbers came from. Regardless, it looks like keeping 3 sea-level engines running at 40% (assuming 6 vac-raps at 100%) only drops overall ISP by 2% or so.

Offline catdlr

  • Member
  • Senior Member
  • *****
  • Posts: 10705
  • Enthusiast since the Redstones
  • Marina del Rey, California, USA
  • Liked: 8044
  • Likes Given: 7393
Re: Starship performance improvements
« Reply #30 on: 11/25/2023 10:13 am »
Ok, I'll go out on a limb;

My prediction:
1) Starship will be lengthened by one ring segment (part of Version 2).
2) Three additional R-Vacs are added to Starship (part of Version 2).
3) Extra Starship engines means no need to gimble inner three sea-level raptor engines at Hot staging but they will be running (like they did for IFT-2) along with the Vac-Raptors.
4) Tower extension (allow the Chopstick's added length necessary to lift the lengthened Version-2 Starship up another 10 meters):
  a) The upper tower parts delivered this week from Robert's road will be utilized to increase the overall tower height   
       along with added tower rails.  All will be assembled at the Sanchez site.   
  b) When completed, the new tower top section will fitted on top of the existing tower but the existing pully will stay
       as is until Version-1 starship test is completed (i.e., no interruptions to testing)
  c) When Version-1 testing is completed (timing allows), the cabling will be restrung from the old tower top through
       the pullies of the new tower top
  d)  Old pullies and cantilevered arms supporting them from the original tower top will be set aside for a second tower.
5) SQD will not need to move any higher but will be beefed up to provide added protection from rocket blasts (can be completed before Version-2).

Erata: The chopstix chain control cables may need extra links and cableing extended (possible, I'm not sure).

Correction: Changed length of Version-2 starship to one additional ring segment (instead of the incorrect 10 meters - what was I thinking?)
« Last Edit: 11/25/2023 06:37 pm by catdlr »
Tony De La Rosa, ...I'm no Feline Dealer!! I move mountains.  but I'm better known for "I think it's highly sexual." Japanese to English Translation.

Offline Brigantine

  • Full Member
  • ***
  • Posts: 303
  • NZ
  • Liked: 146
  • Likes Given: 445
Re: Starship performance improvements
« Reply #31 on: 11/25/2023 10:16 am »
How do single-engine upper stages handle roll control? As I understand it, you can't control roll with a single gimballing engine.

Well, yes RCS. But AIUI being single engine with the thrust straight through CoM, they don't need to run continuously to trim the roll. Occasional puffs suffice. (no RVac misalignment effects to compensate for)

(I don't think anyone actually uses reaction wheels, KSP style?)

Offline Nomadd

  • Senior Member
  • *****
  • Posts: 8831
  • Seoul
  • Liked: 60391
  • Likes Given: 1295
Re: Starship performance improvements
« Reply #32 on: 11/25/2023 01:42 pm »
Ok, I'll go out on a limb;

My prediction:
1) Starship will be lengthened by one ring segment (10 meters) (part of Version 2).
4) Tower ex8tension (allow the Chopstick's added length necessary to lift the lengthened Version-2 Starship up another 10 meters):
I don't think the word "meter" means what you think it means.
Those who danced were thought to be quite insane by those who couldn't hear the music.

Online chopsticks

  • Full Member
  • ****
  • Posts: 1057
  • Québec, Canada
  • Liked: 1075
  • Likes Given: 164
Re: Starship performance improvements
« Reply #33 on: 11/25/2023 02:08 pm »
How much roll control is required once out of the bulk of the atmosphere? Too much for RCS to handle?

Because engines are not aligned perfectly, they don't have the exact same thrust and engines shutting down induces torques is why gimbaled engines are needed for the whole burn duration.
Jim, I appreciate that these may seem like stupid questions to you, but your answers simply aren't helping. If the question is worth your time, I'd appreciate it if you give more detail; if it's not, just ignore it.

To interpret Jim a bit... to compensate for engine misalignment and thrust differentials, to maintain the right attitude (pitch/yaw/roll) it takes more than an occasional puff of RCS, you need a trim - a constant baseline correction to offset a constant bias.

Suppose 3 RCs do this with a trim of just 0.6⁰ i.e. 0.01 radians at 50% thrust. That's correcting forces totaling 38 kN applied constantly for the rest of the 2nd stage burn (assuming similar leverage). If using cold gas, that's a ~1.8% hit to your ISP (7s).

Or if you assume you have very fine throttle control of RVacs (pitch & yaw sorted), and your roll RCS has 4x the leverage of the RCs, that's still 10 kN. Assume hot gas thrusters and that's a ~0.2% hit to your ISP (1s).

What force can the current Ship RCS manage?

Also compare with:
Running two sea levels at 40% thrust and 6 Vacuum at full thrust, we get an Isp of 6*378 + .8 * 363 or an average of 376.2

However it's silly to plan a mission on what is likely to happen, you have to plan on single failures.

So if it's a VR that fails it's 5*378 + .8*363 => average of 375.9

The difference between that an a differentially throttled Vacuum Raptor system is .5%, or with a deltaV of 4km/sec 22 m/sec.

There's enough other stuff going on to not worry about 22m/sec.
Yep, I get that RCS won't cut it for pitch and yaw, but I was under the impression that differential thrust does provide that. It's specifically roll-control that I'm not really understanding. How do single-engine upper stages handle roll control? As I understand it, you can't control roll with a single gimballing engine.

And yes, InterestedEngineer's numbers showed that its not worth it, though I'm not sure where his ISP numbers came from. Regardless, it looks like keeping 3 sea-level engines running at 40% (assuming 6 vac-raps at 100%) only drops overall ISP by 2% or so.
Yes, as Brigantine said, roll control on single engine upper stages is controlled by RCS. But it works fine in this case because there's only one source of thrust and its alignment is always straight through the center of mass. With multiple fixed engines not thrusting through the center of mass (Raptor vacuum engines) but rather around the periphery, there will always be slight misalignments and a roll motion will be induced. This doesn't happen with single engine upper stage because the thrust always pointing through the center of mass. And relying of differential thrust as your only yaw and pitch control isn't optimal either since turbopump engines don't react as fast as a gimballing engine does.

Offline alang

  • Full Member
  • ****
  • Posts: 403
  • Liked: 213
  • Likes Given: 7
Re: Starship performance improvements
« Reply #34 on: 11/25/2023 02:26 pm »
I don't know about the OP, but I wasn't asking whether differential thrust is best in all situations, or claiming that it didn't have downsides.

I was asking about a very specific situation: the ship is on a relatively nominal trajectory and already at a high enough velocity that gravity losses are less of a factor.

Even if in theory differential thrust could steer Starship, it would be something they would only want to do in a crisis.
Very confidently stated, but do you know whether SpaceX agree with you?

Elon Musk has been asked this before, and his answer was that differential thrust was out because it can't handle an engine out scenario.

Sorry, no source. Internet searchability has collapsed like the Roman Empire. It might be one of the Everyday Astronaut interviews.

The empire strikes back:
https://twitter.com/elonmusk/status/1184732755294162944?t=zaFX5s_symy084-aH0K2hg&s=01

Online ugordan

  • Senior Member
  • *****
  • Posts: 8507
    • My mainly Cassini image gallery
  • Liked: 3519
  • Likes Given: 757
Re: Starship performance improvements
« Reply #35 on: 11/25/2023 02:33 pm »
(I don't think anyone actually uses reaction wheels, KSP style?)

On an upper stage of an LV? I'd be hella surprised if anyone ever did.

Offline steveleach

  • Senior Member
  • *****
  • Posts: 2140
  • Liked: 2735
  • Likes Given: 952
Re: Starship performance improvements
« Reply #36 on: 11/25/2023 03:38 pm »
The empire strikes back:
Thanks for digging that out.

Sounds like they considered it and concluded it was a bad trade.

Thanks also to everyone who helped me understand all this.

Offline catdlr

  • Member
  • Senior Member
  • *****
  • Posts: 10705
  • Enthusiast since the Redstones
  • Marina del Rey, California, USA
  • Liked: 8044
  • Likes Given: 7393
Re: Starship performance improvements
« Reply #37 on: 11/25/2023 06:35 pm »
Ok, I'll go out on a limb;

My prediction:
1) Starship will be lengthened by one ring segment (10 meters) (part of Version 2).
4) Tower ex8tension (allow the Chopstick's added length necessary to lift the lengthened Version-2 Starship up another 10 meters):
I don't think the word "meter" means what you think it means.

Yes, Sir, you are correct.  That's what happens when responding at 3:30 in the morning.  I corrected the post.  (to one ring segment).  Have a great day.  Thanks for the catch.
Tony De La Rosa, ...I'm no Feline Dealer!! I move mountains.  but I'm better known for "I think it's highly sexual." Japanese to English Translation.

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 5303
  • Florida
  • Liked: 5003
  • Likes Given: 1430
Re: Starship performance improvements
« Reply #38 on: 11/25/2023 10:12 pm »
Ok, I'll go out on a limb;

My prediction:
1) Starship will be lengthened by one ring segment (10 meters) (part of Version 2).
4) Tower ex8tension (allow the Chopstick's added length necessary to lift the lengthened Version-2 Starship up another 10 meters):
I don't think the word "meter" means what you think it means.

Yes, Sir, you are correct.  That's what happens when responding at 3:30 in the morning.  I corrected the post.  (to one ring segment).  Have a great day.  Thanks for the catch.
Only one segment stretch to the prop tank section only solves the increased prop load to get to 1600t. To solve the needed extra volume to stack up to 140t of Starlinks (~100 total full size V2s [50 X 2 stacks])you would need to add also 2 more ring sections to the payload section as well. And now we are at the need for a 10m tower extension that has been alluded to.

A NOTE here is that just 2 Starships so equipped Starlink launches average per month is 2,400 full size V2 Starlinks a year. Also even if such Starship launches costs as much per launch as $75M. The cost of launch of each of these Starlinks would be only$750K each. A capability savings over even the V1.5 costs on F9 of $750K but having 8 times the throughput. Thus an equivalent cost for throughput V1.5s vs the V2 Full of $94K.

Offline catdlr

  • Member
  • Senior Member
  • *****
  • Posts: 10705
  • Enthusiast since the Redstones
  • Marina del Rey, California, USA
  • Liked: 8044
  • Likes Given: 7393
Re: Starship performance improvements
« Reply #39 on: 11/26/2023 04:36 am »
Ok, I'll go out on a limb;

My prediction:
1) Starship will be lengthened by one ring segment (10 meters) (part of Version 2).
4) Tower ex8tension (allow the Chopstick's added length necessary to lift the lengthened Version-2 Starship up another 10 meters):
I don't think the word "meter" means what you think it means.

So I made a mistake in my previous statement. I didn't mean to say 1 ring segment, but rather a 10-meter increase in length (approximately 30 feet). I apologize for the confusion. It will require multiple rings to achieve this extension. However, increasing the length will affect the tower's capacity. My prediction is that the tower will need to be modified to accommodate this additional length. Please note that the 10-meter measurement is an approximation and not exact.  IMHO.  So, let's just wait and see, it's just my prediction, I will be wrong somewhere.

https://twitter.com/elonmusk/status/1644868577265172481
« Last Edit: 11/26/2023 10:51 am by catdlr »
Tony De La Rosa, ...I'm no Feline Dealer!! I move mountains.  but I'm better known for "I think it's highly sexual." Japanese to English Translation.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0