Author Topic: Shutdown of center core of FH might be the conservative choice.  (Read 9788 times)

Online LouScheffer

  • Full Member
  • ****
  • Posts: 1556
  • Liked: 1758
  • Likes Given: 200
For the Falcon Heavy, the most conservative choice may be to shut down the engines of the center core early in ascent, then use the saved fuel after the side boosters separate.

Intuitively this seems like a bad idea.  What if the engines don't restart?   Then the mission would fail.

The counter argument has 4 parts:

(a) re-lighting Falcon engines in flight is by now a known quantity.  The odds of a re-light failure are low. 

(b) The impact of an engine (or even 2 or 3) that fails to re-light is small, assuming this is detected and the engine shut down.  This follows from the rocket equation.  In the absence of gravity losses, the delta-V is determined only by the initial mass, the final mass, and the ISP.  Each of these is unchanged with one or more inoperative engines.  The acceleration is less, but the stage burns longer, and these exactly cancel.  Now with gravity included there will be some performance loss, but this is late in the first stage burn where the gravity losses are comparatively small, and the difference between 8 and 9 engines smaller yet.

(c) On the other hand, with complete shutdown you get considerably better margins (I get about 200 m/s better a wide range of assumptions).  This is much more than the loss if one or two engines don'r restart.

(d) Higher margins help overall mission success.  The Delta-4 with the RL-10 leak succeeded due to margins.  The Atlas to the station succeeded since the margins were enough to overcome engine troubles.  Likewise with the Falcon to the station - and on this flight the secondary failed since the margins were not quite enough.

You could even put some numbers on this argument.   The above examples (and I'm sure there are more) show that margin matters in a few percent of all launches.  Conversely, the SpaceX tests to date have had perhaps 50-60 air restarts, and as far as I know no engines have failed to re-light.  So the odds of two or more engines failing to re-light should be well under a percent.  (Plus, of course, the odds of some common glitch that affects multiple engines as they re-light.  That's hard to quantify but seems small.  Boost-back and re-entry burns start 3 engines at time routinely already.)

The main arguments against this seem to be (a) it's intuitively scary, and (b) that's not the way we've always done it.  These are basically the same arguments against loading astronauts before fueling.  But in both cases you need to run the numbers, and perhaps the engineering argument gives a different result than intuition.

Note that this argument does not apply to the Delta Heavy, for several reasons.  The margins would indeed be better, for the same reasons as above.  However, the core engine, as far as I know, cannot restart, much less have a history of many successful air restarts.  Also it has only one engine on the center core, and if that fails to re-light, it's catastrophic.


Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31362
  • Cape Canaveral Spaceport
  • Liked: 9638
  • Likes Given: 299

The main arguments against this seem to be (a) it's intuitively scary, and (b) that's not the way we've always done it.  These are basically the same arguments against loading astronauts before fueling.  But in both cases you need to run the numbers, and perhaps the engineering argument gives a different result than intuition.


No, not the same.  There is no equivelent benefit for early loading vs air restart

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31362
  • Cape Canaveral Spaceport
  • Liked: 9638
  • Likes Given: 299

(b) The impact of an engine (or even 2 or 3) that fails to re-light is small, assuming this is detected and the engine shut down.  This follows from the rocket equation.  In the absence of gravity losses, the delta-V is determined only by the initial mass, the final mass, and the ISP.  Each of these is unchanged with one or more inoperative engines.  The acceleration is less, but the stage burns longer, and these exactly cancel.  Now with gravity included there will be some performance loss, but this is late in the first stage burn where the gravity losses are comparatively small, and the difference between 8 and 9 engines smaller yet.


You are ignoring that only 1-3 engines are set for restart. TEB tankage has to be increased and also lines have to run to all the engines.  Also, a lot more He has to be carried to restart 9 engines.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 7767
  • N. California
  • Liked: 4059
  • Likes Given: 831
For the Falcon Heavy, the most conservative choice may be to shut down the engines of the center core early in ascent, then use the saved fuel after the side boosters separate.

Intuitively this seems like a bad idea.  What if the engines don't restart?   Then the mission would fail.

The counter argument has 4 parts:

(a) re-lighting Falcon engines in flight is by now a known quantity.  The odds of a re-light failure are low. 

(b) The impact of an engine (or even 2 or 3) that fails to re-light is small, assuming this is detected and the engine shut down.  This follows from the rocket equation.  In the absence of gravity losses, the delta-V is determined only by the initial mass, the final mass, and the ISP.  Each of these is unchanged with one or more inoperative engines.  The acceleration is less, but the stage burns longer, and these exactly cancel.  Now with gravity included there will be some performance loss, but this is late in the first stage burn where the gravity losses are comparatively small, and the difference between 8 and 9 engines smaller yet.

(c) On the other hand, with complete shutdown you get considerably better margins (I get about 200 m/s better a wide range of assumptions).  This is much more than the loss if one or two engines don'r restart.

(d) Higher margins help overall mission success.  The Delta-4 with the RL-10 leak succeeded due to margins.  The Atlas to the station succeeded since the margins were enough to overcome engine troubles.  Likewise with the Falcon to the station - and on this flight the secondary failed since the margins were not quite enough.

You could even put some numbers on this argument.   The above examples (and I'm sure there are more) show that margin matters in a few percent of all launches.  Conversely, the SpaceX tests to date have had perhaps 50-60 air restarts, and as far as I know no engines have failed to re-light.  So the odds of two or more engines failing to re-light should be well under a percent.  (Plus, of course, the odds of some common glitch that affects multiple engines as they re-light.  That's hard to quantify but seems small.  Boost-back and re-entry burns start 3 engines at time routinely already.)

The main arguments against this seem to be (a) it's intuitively scary, and (b) that's not the way we've always done it.  These are basically the same arguments against loading astronauts before fueling.  But in both cases you need to run the numbers, and perhaps the engineering argument gives a different result than intuition.

Note that this argument does not apply to the Delta Heavy, for several reasons.  The margins would indeed be better, for the same reasons as above.  However, the core engine, as far as I know, cannot restart, much less have a history of many successful air restarts.  Also it has only one engine on the center core, and if that fails to re-light, it's catastrophic.

The nice thing here, if I may borrow a phrase, is "dial-a-rocket".  Depending on how much performance you need, you shut down more and more engines.

---

Just out of curiosity here, totally random:

If you want to spin up more than one turbine, could you route a bleed from an already started engine instead of an He line for that purpose?

And I'm not even talking about "borrowing a light" for ignition :)

ABCD - Always Be Counting Down

Online abaddon

  • Full Member
  • ****
  • Posts: 1519
  • Liked: 953
  • Likes Given: 770
You are ignoring that only 1-3 engines are set for restart. TEB tankage has to be increased and also lines have to run to all the engines.  Also, a lot more He has to be carried to restart 9 engines.
Jim's point is a good one.  Would burning only three engines be sufficient?  If not, while the FH core is already different from the F9 core / FH boosters, that's a lot of weight and complexity to add.

Of course, the other factor not mentioned is center core recovery.  The higher and further the core gets the further out an ASDS would have to be to catch it, and the longer it would take to get back with it.  This all makes for a complicated trade space.

If the Block 5 numbers touted by SpaceX are accurate, I am not sure this is very beneficial.  But it's an interesting question to ask.
« Last Edit: 01/30/2017 01:59 PM by abaddon »

Offline John Alan

  • Full Member
  • ****
  • Posts: 516
  • Central IL - USA - Earth
  • Liked: 253
  • Likes Given: 1245
Just out of curiosity here, totally random:

If you want to spin up more than one turbine, could you route a bleed from an already started engine instead of an He line for that purpose?

And I'm not even talking about "borrowing a light" for ignition :)

I have also thought that... use bleed gases from a running rocket to spin up a non runner...
I also thought... put a separate RP1/LOX gas generator onboard (burner can) to generate CO/CO2 to pressurize the tanks...
but that's crazy talk I am told...  ::)  ;)
« Last Edit: 01/30/2017 02:36 PM by John Alan »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31362
  • Cape Canaveral Spaceport
  • Liked: 9638
  • Likes Given: 299

I also thought... put a separate RP1/LOX gas generator onboard (burner can) to generate CO/CO2 to pressurize the tanks...
but that's crazy talk I am told...  ::)  ;)

It would be.

CO/CO2 would condense and freeze in LOX.  And there can't be any unused O2 in the RP-1 tank

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31362
  • Cape Canaveral Spaceport
  • Liked: 9638
  • Likes Given: 299
Hot gas valves.  Who has experience with them?
« Last Edit: 01/30/2017 02:46 PM by Jim »

Offline John Alan

  • Full Member
  • ****
  • Posts: 516
  • Central IL - USA - Earth
  • Liked: 253
  • Likes Given: 1245

I also thought... put a separate RP1/LOX gas generator onboard (burner can) to generate CO/CO2 to pressurize the tanks...
but that's crazy talk I am told...  ::)  ;)

It would be.

CO/CO2 would condense and freeze in LOX.  And there can't be any unused O2 in the RP-1 tank

Agreed... but how much would condense/freeze in the short time S1 is actually firing all 9...
And would the left over gaseous O2 fraction in the RP1 tank reach flammability levels...
(note... tanks pressured on the ground with He up until engine start)
(burner has a cat on it to catch soot and oxidize/clean up the gasses generated)

Things that make me go ....HMMMmmmm...   ???  ;)
« Last Edit: 01/30/2017 02:56 PM by John Alan »

Offline envy887

  • Senior Member
  • *****
  • Posts: 2777
  • Liked: 1281
  • Likes Given: 806
About half the output of a gas generator is water, which would condense and freeze in both the LOX and RP-1 tanks, since SpaceX loads RP-1 at -8 degrees C. Also, carbon monoxide and other incomplete combustion products are fuels... which you do not want in the LOX tank.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31362
  • Cape Canaveral Spaceport
  • Liked: 9638
  • Likes Given: 299

Agreed... but how much would condense/freeze in the short time S1 is actually firing all 9...

Doesn't matter, the CO/CO2 would immediately contract as it cooled.  Along wth the steam and water, which would freeze and go to the bottom of the LOX tank.

A cat is not going to be able to handle the volume of gases.

The "Hmmmm" is Hmmm, no.
« Last Edit: 01/30/2017 03:01 PM by Jim »

Offline John Alan

  • Full Member
  • ****
  • Posts: 516
  • Central IL - USA - Earth
  • Liked: 253
  • Likes Given: 1245
About half the output of a gas generator is water, which would condense and freeze in both the LOX and RP-1 tanks, since SpaceX loads RP-1 at -8 degrees C. Also, carbon monoxide and other incomplete combustion products are fuels... which you do not want in the LOX tank.

Oh... good point... my bad... forget I ever mentioned the idea...  :-[   

And Jim... thanks for correcting me...  ;)

SHUTDOWN of FH center core... back on topic... I like it...
« Last Edit: 01/30/2017 03:14 PM by John Alan »

Offline ClayJar

  • Full Member
  • ***
  • Posts: 326
  • Baton Rouge, LA, USA
  • Liked: 675
  • Likes Given: 62
You are ignoring that only 1-3 engines are set for restart. TEB tankage has to be increased and also lines have to run to all the engines.  Also, a lot more He has to be carried to restart 9 engines.

Would the numbers work out to show a potential benefit of shutting down only the three restartable engines (the others only having throttling available)?  Those three are already plumbed and ready, after all.

Of course, the other half of the question would be whether such a method would be hypothetically possible.  Would a six-engine burn be workable, and could you restart three engines while six (or more) others are firing?  (We know that restarts work in conditions approximating 1g and free fall, but the acceleration at the point of a mid-flight relight are certainly not something *I've* modeled.)

Offline Lee Jay

  • Elite Veteran
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 6646
  • Liked: 916
  • Likes Given: 137
Is the point of this idea to mitigate the need for cross-flow?  To increase fuel saving during cross-flow?

I thought the original idea was to feed the two sets of three outboard engines from each of the side boosters while running only the center three (possibly throttled down at some point) off the center booster until the side boosters are empty.

I'm not following where this came from.

Online abaddon

  • Full Member
  • ****
  • Posts: 1519
  • Liked: 953
  • Likes Given: 770
Is the point of this idea to mitigate the need for cross-flow?
The idea is an alternate way to keep more fuel and oxidizer in the center core for longer, since SpaceX isn't implementing cross-feed.  Ideally it would be somewhat effective and simpler to implement than cross-feed.

Seems unlikely SpaceX will bother with any of this, but it is an interesting through experiment while we wait for the next launch and/or something more concrete to talk about.

Online guckyfan

  • Senior Member
  • *****
  • Posts: 6421
  • Germany
  • Liked: 1613
  • Likes Given: 1436
The most simple thing they could do is switch off the three engines they can relight without modifications. Together with throttling the other engines of the central core they could retain a lot of the fuel while the side cores burn full throttle.

Offline Lars-J

  • Senior Member
  • *****
  • Posts: 3479
  • California
  • Liked: 2727
  • Likes Given: 1729
The most simple thing they could do is switch off the three engines they can relight without modifications. Together with throttling the other engines of the central core they could retain a lot of the fuel while the side cores burn full throttle.

You guys are overthinking it... The M1D can already throttle down pretty deep. It is far simpler and safer to run a low throttle than to restart all 9 engines.

Online abaddon

  • Full Member
  • ****
  • Posts: 1519
  • Liked: 953
  • Likes Given: 770
You guys are overthinking it... The M1D can already throttle down pretty deep. It is far simpler and safer to run a low throttle than to restart all 9 engines.
ISTR the center engine is the only one that can throttle down deep, which requires special hardware (a special valve?).  That information may be old/outdated though.

Offline old_sellsword

  • Full Member
  • ****
  • Posts: 601
  • Liked: 492
  • Likes Given: 438
You guys are overthinking it... The M1D can already throttle down pretty deep. It is far simpler and safer to run a low throttle than to restart all 9 engines.
ISTR the center engine is the only one that can throttle down deep, which requires special hardware (a special valve?).  That information may be old/outdated though.

Seems like that would go against the SpaceX ethos of identical everything where possible.

Offline envy887

  • Senior Member
  • *****
  • Posts: 2777
  • Liked: 1281
  • Likes Given: 806
SpaceX only lists the SL throttle limits of the Merlin in the F9 users guide, which they have at 70%. The Mvac is listed as throttling to 38.5%.

However, we know from Musk's tweets that at least the center Merlin on the booster can also throttle down to 40% for landing, so it doesn't seem like flow separation is an issue.

Tags: