Author Topic: NASA defends decision to restart RS-25 production, rejects alternatives  (Read 115684 times)

Offline quanthasaquality

  • Full Member
  • *
  • Posts: 146
  • Liked: 6
  • Likes Given: 2
I just keep seeing this position again, from the looks of Russia, and now SpaceX, it seems like it is important to be good at manufacturing, and pumping out large numbers of rocket engines. I guess it might also help to have a low wage workforce. The Proton, Zenit, and Angara rockets all use high performance, oxidizer rich staged combustion engines. Russia has been pumping those rocket engines out now for decades.

Presumably, kerosene oxygen oxidizer rich staged combustion engines are more difficult to manufacture than hydrogen oxygen fuel rich staged combustion engines. Russia's rd-170,180,190 series also has a higher chamber pressure than the SSME. Yet, Russia is able to manufacture the rd-180 at a reasonable price.

In the mid 90s, Marshall Space Flight Center had the idea of a low cost rocket engine, FASTRAC https://www.nasa.gov/centers/marshall/news/background/facts/fastrac.html Ideas like a low pressure, machine manufactured turbopump, ablative nozzle, and pintle injector. The rs-68 used some of those ideas, and it was cheaper, but the Atlas V ended being the cheaper rocket. We don't know how much an American rd-180 would have ended up costing. The rs-27 was the simplified rocket engine of the 60s, with heritage from the V-2, but users of it, the Atlas and Delta rockets of the 90s, were not cheap rockets. For all the hype the Saturn V gets, it was an expensive rocket. The F-1 required a lot of skilled human labor.

Maybe the rs-25 just needs some expendable, machine, big factory love.... and a Congress willing to pay for jobs for the rs-25.

Offline envy887

  • Senior Member
  • *****
  • Posts: 8144
  • Liked: 6801
  • Likes Given: 2965
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

Offline PahTo

  • Full Member
  • ****
  • Posts: 1698
  • Port Angeles
  • Liked: 271
  • Likes Given: 1194
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Offline envy887

  • Senior Member
  • *****
  • Posts: 8144
  • Liked: 6801
  • Likes Given: 2965
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Like Saturn 1/1b with it's 8 H-1s and perfect launch record? Or Saturn V with multiple engine failures but also a perfect launch record? Falcon 9 also hasn't had any propulsion issues leading to LOM, despite having a Merlin blow up on one flight.



Offline WulfTheSaxon

  • Full Member
  • *
  • Posts: 184
    • #geekpolitics on DALnet
  • Liked: 29
  • Likes Given: 1034
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Like Saturn 1/1b with it's 8 H-1s and perfect launch record? Or Saturn V with multiple engine failures but also a perfect launch record? Falcon 9 also hasn't had any propulsion issues leading to LOM, despite having a Merlin blow up on one flight.

Could you define “perfect launch record”? I can sort of see exempting it because it was a test flight, but Apollo 6 was a partial failure, and would have resulted in LOM had it been a crewed lunar flight…
« Last Edit: 03/04/2017 06:10 pm by WulfTheSaxon »

Offline Flying Beaver

Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Like Saturn 1/1b with it's 8 H-1s and perfect launch record? Or Saturn V with multiple engine failures but also a perfect launch record? Falcon 9 also hasn't had any propulsion issues leading to LOM, despite having a Merlin blow up on one flight.

Could you define “perfect launch record”? I can sort of see exempting it because it was a test flight, but Apollo 6 was a partial failure, and would have resulted in LOM had it been a crewed lunar flight…

Not loss of mission but at least an abort from low earth orbit. The crew would of been safe.
Watched B1019 land in person 21/12/2015.

Offline envy887

  • Senior Member
  • *****
  • Posts: 8144
  • Liked: 6801
  • Likes Given: 2965
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Like Saturn 1/1b with it's 8 H-1s and perfect launch record? Or Saturn V with multiple engine failures but also a perfect launch record? Falcon 9 also hasn't had any propulsion issues leading to LOM, despite having a Merlin blow up on one flight.

Could you define “perfect launch record”? I can sort of see exempting it because it was a test flight, but Apollo 6 was a partial failure, and would have resulted in LOM had it been a crewed lunar flight…

Primary payload was injected in a useful orbit. It may not have reached orbit at all if the S-II used fewer engines. And it probably would have reached TLI if the S-IVB used multiple redundant engines.

The pogo on Saturn V wasn't due to engine count, and modern simulation methods are substantially better at predicting resonant modes which makes it a lot easier to design around that particular issue.

I'm really curious what "practical experience" NASA has had with many engines that leads them to think it causes more issues than it solves.

Offline WulfTheSaxon

  • Full Member
  • *
  • Posts: 184
    • #geekpolitics on DALnet
  • Liked: 29
  • Likes Given: 1034
Maybe NASA needs to get over their irrational aversion to having many small engines on a booster, so that automated highly tooled production is worthwhile.

I don't know that it is irrational--there is some practical experience that many engines complicate things to the point the odds are there will be an "issue".

Like Saturn 1/1b with it's 8 H-1s and perfect launch record? Or Saturn V with multiple engine failures but also a perfect launch record? Falcon 9 also hasn't had any propulsion issues leading to LOM, despite having a Merlin blow up on one flight.

Could you define “perfect launch record”? I can sort of see exempting it because it was a test flight, but Apollo 6 was a partial failure, and would have resulted in LOM had it been a crewed lunar flight…

Not loss of mission but at least an abort from low earth orbit. The crew would of been safe.

How would an abort from LEO not be LOM? I didn’t say anything about LOC.

Offline sdsds

  • Senior Member
  • *****
  • Posts: 7202
  • “With peace and hope for all mankind.”
  • Seattle
  • Liked: 2050
  • Likes Given: 1962
I agree with those who feel the approach to booster design which uses a large number of engines has now been proven. It works for the booster in flight; it works for the manufacturing processes, and it works for the business model. Nine makes a particularly good number for what can be called similar redundancy!

Granted, RS-25 isn't going to be flown like that. Point taken.

That isn't sufficient to make the claim that RS-25 should be abandoned, though. It still works in flight. The manufacturing is expensive but known. The business (or at least funding) model has an extraordinary track record.
« Last Edit: 03/04/2017 09:53 pm by sdsds »
— 𝐬𝐝𝐒𝐝𝐬 —

Offline TomH

  • Senior Member
  • *****
  • Posts: 2938
  • Vancouver, WA
  • Liked: 1868
  • Likes Given: 909
How would an abort from LEO not be LOM? I didn’t say anything about LOC.

Abort from LEO would be loss of mission. He is just confusing loss of mission with loss of crew. He may be old like me; I spent 5 minutes this evening trying to put the parking garage ticket into the credit card slot.

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39218
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32738
  • Likes Given: 8196
And as I pointed out up-thread, and Steven Pietrobon showed in his paper, an SLS core with six engines and a large upper stage will use up its propellants long before it reaches 7.5 km/s. From Steven's charts, I came up with about 4.5 km/s. I'm sure Steven could provide his exact calculations.

Core stage burn out was at 4771 m/s inertial speed.
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline TomH

  • Senior Member
  • *****
  • Posts: 2938
  • Vancouver, WA
  • Liked: 1868
  • Likes Given: 909
Core stage burn out was at 4771 m/s inertial speed.

Any chance you're willing to run calcs on core with 5 RS-25s and a center J-2X, two Dark Knights, and a larger US optimized to match, staging with enough prop for the core to land downrange on the (modified deep throttleable) J-2X (or RTLS on lofted trajectory)? I know, lots of unknowns, but ballpark would be nice to know.

I know this is very unlikely to actually happen, just interested in performance numbers as a hypothetical.

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39218
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32738
  • Likes Given: 8196
Any chance you're willing to run calcs on core with 5 RS-25s and a center J-2X, two Dark Knights, and a larger US optimized to match, staging with enough prop for the core to land downrange on the (modified deep throttleable) J-2X (or RTLS on lofted trajectory)? I know, lots of unknowns, but ballpark would be nice to know.

If you're willing to pay me my consulting fee of US$100 an hour, I'll be glad to do that for you. :-)

I have simulated 5xRS-25s, Dark Knights and 2xJ-2X upper stage. That gets 144.1 t to LEO. Core stage burnout speed is 4774 m/s.
« Last Edit: 03/05/2017 05:17 am by Steven Pietrobon »
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline Lars-J

  • Senior Member
  • *****
  • Posts: 6809
  • California
  • Liked: 8485
  • Likes Given: 5384
I just keep seeing this position again, from the looks of Russia, and now SpaceX, it seems like it is important to be good at manufacturing, and pumping out large numbers of rocket engines.
...
In the mid 90s, Marshall Space Flight Center had the idea of a low cost rocket engine, FASTRAC https://www.nasa.gov/centers/marshall/news/background/facts/fastrac.html Ideas like a low pressure, machine manufactured turbopump, ablative nozzle, and pintle injector. The rs-68 used some of those ideas, and it was cheaper, but the Atlas V ended being the cheaper rocket. We don't know how much an American rd-180 would have ended up costing. The rs-27 was the simplified rocket engine of the 60s, with heritage from the V-2, but users of it, the Atlas and Delta rockets of the 90s, were not cheap rockets. For all the hype the Saturn V gets, it was an expensive rocket. The F-1 required a lot of skilled human labor.

Well the spiritual (& technology) successor of FASTRAC *is* in use now - SpaceX's Merlin. It has IMO validated the thesis of the benefits of a mass produced smaller engine.
« Last Edit: 03/05/2017 05:42 am by Lars-J »

Offline robert_d

  • Full Member
  • ***
  • Posts: 356
  • Liked: 72
  • Likes Given: 118

Core stage burn out was at 4771 m/s inertial speed.

Steve,
Did you ever model a 7 RS-25 core? If so could you point to it?
I was thinking NASA could focus on Hydrogen powered vehicles, building an optimized core for 7 engines if that could be made reusable by following the flight profile of the Falcon 9 first stage. They could work on a restartable RS-25 second stage or maybe do the J2-X if that was cheaper.

I find it criminal that just as NASA had made impressive upgrades the the RS-25, they want to dump its most impressive feature, the fact that it was designed to be reusable.

Edit: I am not asking about an SLS direct replacement/upgrade. I am asking about a single core reusable booster powered by RS-25. One that would only run for about 2:30 just like the falcon 9 1st stage so it could be recovered and re-used.   
« Last Edit: 03/05/2017 03:56 pm by robert_d »

Offline spacenut

  • Senior Member
  • *****
  • Posts: 5181
  • East Alabama
  • Liked: 2587
  • Likes Given: 2895
This has been modeled before somewhere on here.  A five core was the optimum with Dark Knights or two 5.5m core kerolox side boosters and SLS could get 150 tons to LEO or more.  This size rocket as the 70 ton version was a 3-RS-15 core and two 4 seg solids from the shuttle stack, was the lowest cost quickest version to field.  It was called Jupiter Direct.  It could have stretched tanks and 5 RS-25 core with a J2-x second stage with the two 4 seg solids and get 130 tons to LEO.  This is where the 70-130 ton Congressional mandate came from.  Instead, NASA chose to develop longer larger solids and a 4 engine core compromise to get about the current 105 ton expensive beast.  They did develop the J2-x and shut down production afterwards.  Why??, the larger solid development cost a fortune. 

Anyways a 12-15 RD-180 clean sheet booster with a two engine J2-x upper, from what I can remember got 150 tons or more to LEO.  Of course, we how have the "Russian problem" and we never developed our own manufacturing capabilities for the RD-180's. 

Seven RS-25 core would require a huge core, and probably not fit in the VAB, and be too wide for the side booster attachments.  Hydrogen takes huge tankage for boosters. 

Offline robert_d

  • Full Member
  • ***
  • Posts: 356
  • Liked: 72
  • Likes Given: 118
So thinking a bit more: My question above asked about a 7 engine core because 1) they would fit comfortably under the 8.4 meter core and 2) there needs to be residual fuel for boostback & landing. Didn't do any math.

But I see a really rough calc. shows 9 engines might work, although there would have to be a skirt of some sort (falcon v1.0 anyone?)
calc: 3 ssme x 8.5 minutes = 25.5 engine/minutes
        9 ssme x 2.5 minutes = 22.5 engine/minutes, leaving 3.0 engine minutes for recovery.
        So a SMALLER core than SLS could work (?)
Gives 9 x 400k lbs. thrust = 3,600k lbs. at liftoff. If a second stage SSME vacuum version is too big, there are alternatives in production such as the Ariane's Vulcain or the Japanese H2, if they could be made restartable.

Mount crane on new mobile launch structure. Landing pad at approximate location of where pad 39c would have been with crawler way. Lift booster back on MLP and roll back to VAB for 1-2 month refurb.
Dump solids once and for all.
Between this far cheaper vehicle, F9 Heavy, New Glenn and Vulcan, we should have plenty of launch capacity.

Offline AncientU

  • Senior Member
  • *****
  • Posts: 6257
  • Liked: 4164
  • Likes Given: 6078
Core stage burn out was at 4771 m/s inertial speed.

Any chance you're willing to run calcs on core with 5 RS-25s and a center J-2X, two Dark Knights, and a larger US optimized to match, staging with enough prop for the core to land downrange on the (modified deep throttleable) J-2X (or RTLS on lofted trajectory)? I know, lots of unknowns, but ballpark would be nice to know.

I know this is very unlikely to actually happen, just interested in performance numbers as a hypothetical.

This would be interesting. 

The core would be traveling 3x the velocity (4,771 vs. ~1,500m/s) and carrying 6x the mass (dry mass at 119.9 vs. ~20tonnes without considering the mass of residual fuel, landing legs, reaction system to flip the stage, thermal insulation/shields, and grid fins or equivalent to guide the descent) of the other stage that lands downrange.  To land, we are looking > 20x the energy that has to be dissipated.  Propellant required to be carried along for recovery would be proportionately large -- both impacting velocity at staging and mass that has to be decelerated.  Upper stage delta-v also has to be used to make up the shortfall of lower staging velocity.

According to Steve's spreadsheet, this 5 RS-25/2 J2-X version achieves around 2.5% payload mass fraction.  A large proportion of that pmf (half?) would be lost to down range recovery.  There might be none of it left (negative pmf) if RTLS was used.

Also, reusability triggers other expenses:
Five RS-25s per core would reduce inventory available to three flights from four.
The RS-25E development effort would have to be shifted from making the engine more expendable, to making it more reusable (which probably would take longer)
Dark Nights would have to be developed (now, instead of in 2030)
A deep throttleable J2-X would have to be developed and go into production
A 2 engine oversized upper stage would need to be developed
Landing legs, and other hardware listed above would need to be developed
Core stage would need to be modified to accommodate these changes
A BFB (big friggin barge) would have to be developed and deployed way down range

Oh, and practice, practice, practice... to perfect that landing.
« Last Edit: 03/05/2017 05:05 pm by AncientU »
"If we shared everything [we are working on] people would think we are insane!"
-- SpaceX friend of mlindner

Offline jgoldader

  • Full Member
  • ****
  • Posts: 760
  • Liked: 322
  • Likes Given: 171
Am I correct in recalling from threads years ago, that a 5-engine first stage would be a major redesign because the thrust structure is very different?  And would there be cooling issues with the extra SSME, or does the RS-25's hydrogen cooling more or less make that a moot point?

If money and time were not issues, adding in a fifth engine or some sort of recoverability for the engines or entire stage might be reasonable, but given the history of the current iteration of SLS, you'd need 10+ years and many billions just for the design work, plus changes to construction hardware and maybe even GSE, etc.  I don't know that the investment is sensible.

SLS is what it is.  If you're going to use it for more than 4 (5?) flights, you need new RS-25's.
Recovering astronomer

Online Chris Bergin

Post that followed had nothing to do with RS-25. If you wish to discuss other engines, please use the appropriate threads - which ain't this one if it's not RS-25.
Support NSF via L2 -- Help improve NSF -- Site Rules/Feedback/Updates
**Not a L2 member? Whitelist this forum in your adblocker to support the site and ensure full functionality.**

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1