SpaceX Falcon 9 : Hispasat 30W-6 (1F) : March 6, 2018 - DISCUSSION

[BeamRider]

No doubt about the primacy of catching up to committed customer demand!  And yes, sometimes no ship is big enough, although I still remember being stunned to read that the first F9 ASDS landing to succeed was in a 50kt wind! Light aircraft are typically grounded in such conditions, at least by older pilots. 

Anyway, I wonder how the Spacex business model may evolve from recovery being a (stated) secondary objective, to being part of the expectation for BOTH them and the customer.  Will that happen?  How long will it take? I could see commercial arrangements that say "Spacex would like to recover this booster, so we will discount your launch by x% if you give us the right to delay up to x days.  If we try and fail within that timeframe, you still get the discount." OTOH a customer might go for "Spacex will guarantee launch within x days of the target date for a full rate with no discount, on a best efforts basis. If we can recover the booster without risking that date or your payload, we will have the right to do so. Customer still pays full freight." Question I would have is, if Spacex doesn't want to recover the booster, does the customer expect a discount, or perhaps NOT expect one?

Obviously there are all kinds of gotchas in there, but I am sure there are legal precedents for how such arrangements can be made fair to both parties.  It will be interesting to see how such business practices will put pressure on "unrecoverable" launchers.

[acsawdey]

No doubt about the primacy of catching up to committed customer demand!  And yes, sometimes no ship is big enough, although I still remember being stunned to read that the first F9 ASDS landing to succeed was in a 50kt wind! Light aircraft are typically grounded in such conditions, at least by older pilots. 

F9 S1 is hardly a light aircraft, being approximately the same dimensions as the fuselage of the longest 737. That said, 737 crosswind limits are in the 35kt range ...

[cscott]

The other piece is RTLS.  If you're returning to the landing site, launch weather is landing weather (more or less).  The goal AIUI is to migrate those expendable and downrange ASDS F9 flights to 3-core RTLS FH flights as much as possible.  That is, the goal is to use the ASDS as little as possible once they have alternatives.

I think that also explains why they have not invested more in an ASDS armada. They view it as a stepping stone, a bridge to FH (and BFR), not part of the ultimate goal.

[zubenelgenubi]

Of course, there is precedent...Foul weather at TAL sites played a role in STS launch scrubs.

[Basto]

Two objects related to today's #Falcon9 launch tracked in a sub-GTO orbit
2018-023A: 184 x 22,261 km, 26.97°
2018-023C: 186 x 22,215 km, 26.92°

https://twitter.com/Spaceflight101/status/971074423108358144

From this I find about 320 m/s to raise apogee to GEO, then 1800 m/s to circularize.   Total about 2120 m/s to go.

So performance was typical for a block 4, and customer accepted less than GEO apogee.

So seeing as they were not going to be able to recover the booster why would they not try to remove the recovery hardware and burn S1 to depletion prior to MECO to get the payload a better orbit?

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat and spaceX would not have had to expend the Ti fins.

What am I missing here?

[AC in NC]

So seeing as they were not going to be able to recover the booster why would they not try to remove the recovery hardware and burn S1 to depletion prior to MECO to get the payload a better orbit?

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat and spaceX would not have had to expend the Ti fins.

What am I missing here?

The simplest option.  All things considered, it was preferable to proceed than introduce additional workflow.  At the risk of sounding flippant, it seems self-evident that the customer or SpaceX didn't feel it was worth it.  It's impossible to know all the considerations that went into that decision.  We could speculate be that seems somewhat pointless.

[abaddon]

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat

Would it?  From what I have read the bird has plenty of fuel, and using chemical engines it should get to station rapidly.  The obvious answer is that SpaceX delivered exactly what was promised and it wouldn't have helped the payload get on station quicker.

[speedevil]

So seeing as they were not going to be able to recover the booster why would they not try to remove the recovery hardware and burn S1 to depletion prior to MECO to get the payload a better orbit?

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat and spaceX would not have had to expend the Ti fins.

If they're going to take a few days to remove the fins, they can as easily put the ship out and recover the fairing once these winds have died down.

[Basto]

Probably the reason is that it would require a lot of new flight software effort, every bit of which would have to be re-certified, etc..  It would likely take more than just a few days.  That's why the first stage even pretended like it was landing, etc.

Makes sense.

This really shows the recovery performance penalty.  Last year, a fully-expendable Falcon 9 put 6.761 metric ton Intelsat 35e into a 296 x 42,742 km x 25.6 deg supersynchronous transfer orbit.  Here, a Falcon 9 configured for first stage recovery could only lift 6.1 tonnes to 184 x 22,261 km x 26.9 deg.


 - Ed Kyle

You are always quick to bring up the "recovery penalty" but seem to always forget about the "expendable penalty" ($$).

[LouScheffer]

Bugs me that the SpaceX press kit and coverage called this "geosynchronous transfer orbit".  It wasn't.  It was subsynchronous.

I think SpaceX is correct here.   The definition of "geosynchronous transfer orbit" is one that brings you closer to GEO.   Such orbits can be supersynchronous or subsynchronous, but are still GTOs.   To make your position consistent, you would also have to be bugged that supersynchronous orbits are called GTOs.   But I've never heard you object to this practice.

If you want to argue that a subsynchronous orbit is "worse" and should not qualify, you have to define "worse" from the customer point of view, not the orbital mechanics point of view.   If it takes more delta-V, but costs less, a customer might well consider this orbit "better".  Presumably this is the case here since the customer had other options.

In short, the customer has chosen this transfer orbit to fulfill their objective of getting to GEO.  That makes it a geosynchronous transfer orbit.

[drnscr]

Two objects related to today's #Falcon9 launch tracked in a sub-GTO orbit
2018-023A: 184 x 22,261 km, 26.97°
2018-023C: 186 x 22,215 km, 26.92°

https://twitter.com/Spaceflight101/status/971074423108358144

From this I find about 320 m/s to raise apogee to GEO, then 1800 m/s to circularize.   Total about 2120 m/s to go.

So performance was typical for a block 4, and customer accepted less than GEO apogee.

So seeing as they were not going to be able to recover the booster why would they not try to remove the recovery hardware and burn S1 to depletion prior to MECO to get the payload a better orbit?

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat and spaceX would not have had to expend the Ti fins.

What am I missing here?

Probably the reason is that it would require a lot of new flight software effort, every bit of which would have to be re-certified, etc..  It would likely take more than just a few days.  That's why the first stage even pretended like it was landing, etc.

Bugs me that the SpaceX press kit and coverage called this "geosynchronous transfer orbit".  It wasn't.  It was subsynchronous.

This really shows the recovery performance penalty.  Last year, a fully-expendable Falcon 9 put 6.761 metric ton Intelsat 35e into a 296 x 42,742 km x 25.6 deg supersynchronous transfer orbit.  Here, a Falcon 9 configured for first stage recovery could only lift 6.1 tonnes to 184 x 22,261 km x 26.9 deg.


 - Ed Kyle

Sir, I don’t wish to sound disrespectful, I really don’t.  However, most of your posts in SpaceX threads seem rather negative toward the company.  Now, maybe i’m reading your comments wrong, and if I am, I apologize.  But, what is the point of always sounding negative when you post in a SpaceX forum?

[envy887]

This really shows the recovery performance penalty.  Last year, a fully-expendable Falcon 9 put 6.761 metric ton Intelsat 35e into a 296 x 42,742 km x 25.6 deg supersynchronous transfer orbit.  Here, a Falcon 9 configured for first stage recovery could only lift 6.1 tonnes to 184 x 22,261 km x 26.9 deg.


 - Ed Kyle

I get a ~380 m/s difference between the two orbits, which shows that the supersync payload penalty for reuse is about 26%, by my math.

Which is actually a little better than the 33.8% listed on the SpaceX website (5.5 t for $62M vs 8.3 t max expendable).

[RedLineTrain]

Two objects related to today's #Falcon9 launch tracked in a sub-GTO orbit
2018-023A: 184 x 22,261 km, 26.97°
2018-023C: 186 x 22,215 km, 26.92°

https://twitter.com/Spaceflight101/status/971074423108358144

From this I find about 320 m/s to raise apogee to GEO, then 1800 m/s to circularize.   Total about 2120 m/s to go.

So performance was typical for a block 4, and customer accepted less than GEO apogee.

So seeing as they were not going to be able to recover the booster why would they not try to remove the recovery hardware and burn S1 to depletion prior to MECO to get the payload a better orbit?

Yes it could have delayed the launch by a few days but it would have been offset by a better time to operational status for the sat and spaceX would not have had to expend the Ti fins.

What am I missing here?

Probably the reason is that it would require a lot of new flight software effort, every bit of which would have to be re-certified, etc..  It would likely take more than just a few days.  That's why the first stage even pretended like it was landing, etc.

Bugs me that the SpaceX press kit and coverage called this "geosynchronous transfer orbit".  It wasn't.  It was subsynchronous.

This really shows the recovery performance penalty.  Last year, a fully-expendable Falcon 9 put 6.761 metric ton Intelsat 35e into a 296 x 42,742 km x 25.6 deg supersynchronous transfer orbit.  Here, a Falcon 9 configured for first stage recovery could only lift 6.1 tonnes to 184 x 22,261 km x 26.9 deg.

 - Ed Kyle

My notes indicate that Intelast 35e was a minimal residual shut down, whereas the Hispasat webcast stated that the orbital insertion was "very accurate" -- i.e., perhaps not a minimal residual shutdown.

[joek]

But, what is the point of always sounding negative when you post in a SpaceX forum?

This is a space flight forum with some SpaceX threads (well, many SpaceX threads).

[Lar]

- Ed Kyle

Sir, I don’t wish to sound disrespectful, I really don’t.  However, most of your posts in SpaceX threads seem rather negative toward the company.  Now, maybe i’m reading your comments wrong, and if I am, I apologize.  But, what is the point of always sounding negative when you post in a SpaceX forum?

IMHO: Ed is a skeptic, particularly of SpaceX, but of other providers as well.  He is also a stickler for getting every number correct or admitting that it is not known for sure. A healthy skepticism and a desire for accuracy are good things. We should not denigrate skepticism.  And we shouldn't call people on their opinions or why they hold them, as you did here.

That said it is reasonable to call people on their math, or their assumptions or their conclusions, to be skeptical of the skepticism, as  it were. Question the data. Question the conclusions. But please don't question people's motives.  Thanks.

(Kinda speaking as a mod there in that last part)

[LouScheffer]

I get a ~380 m/s difference between the two orbits [expendable vs recovery]

This makes excellent sense, and is a better way to look at it than percentage (which can vary a lot by mission, since the rocket equation is very non-linear).

To have enough fuel for recovery, SpaceX needs to save about 9 seconds of fuel (this is 81 engine-seconds, of which they use about 20x3 = 60 for re-entry, and about 21 for landing with 3 engines (30 if they use single engine)).  At the end of the first stage burn, the rocket is accelerating at 4-5 Gs.   4.5 Gs x 9 seconds is about 395 m/s, very close to your value.

[John Alan]

I see my speculation back on 2/22 was pretty close...
https://forum.nasaspaceflight.com/index.php?topic=43435.msg1791886#msg1791886

Ed,
Maybe the assumed normal of GEO-1800m/s days from the Cape are over... and we just need to adjust to it...

As I implied back in the linked posting above...
The SpaceX price sheet is likely driving customers to make darn sure they stay under the ASDS recovery price cap...
They can do this by either getting under the weight limit stated... OR (as I have said many times before)
Agreeing to take a less energetic boost, and put more delta-v into their own payload boosting systems...

GS had indicated discussions along those lines with customers had happened, and it seems we just saw the first of maybe many, that fly out at less then the typical GEO-1800m/s on purpose... 

[drnscr]

- Ed Kyle

Sir, I don’t wish to sound disrespectful, I really don’t.  However, most of your posts in SpaceX threads seem rather negative toward the company.  Now, maybe i’m reading your comments wrong, and if I am, I apologize.  But, what is the point of always sounding negative when you post in a SpaceX forum?

IMHO: Ed is a skeptic, particularly of SpaceX, but of other providers as well.  He is also a stickler for getting every number correct or admitting that it is not known for sure. A healthy skepticism and a desire for accuracy are good things. We should not denigrate skepticism.  And we shouldn't call people on their opinions or why they hold them, as you did here.

That said it is reasonable to call people on their math, or their assumptions or their conclusions, to be skeptical of the skepticism, as  it were. Question the data. Question the conclusions. But please don't question people's motives.  Thanks.

(Kinda speaking as a mod there in that last part)

Sir, I meant no harm to anyone at all and stated that at least a couple of times.  I also apologized in the body of my comments.

[envy887]

I get a ~380 m/s difference between the two orbits [expendable vs recovery]

This makes excellent sense, and is a better way to look at it than percentage (which can vary a lot by mission, since the rocket equation is very non-linear).

To have enough fuel for recovery, SpaceX needs to save about 9 seconds of fuel (this is 81 engine-seconds, of which they use about 20x3 = 60 for re-entry, and about 21 for landing with 3 engines (30 if they use single engine)).  At the end of the first stage burn, the rocket is accelerating at 4-5 Gs.   4.5 Gs x 9 seconds is about 395 m/s, very close to your value.

I normalized the percentages as the payload difference to the same orbit using the rocket equation and an estimated 4500 kg upper stage burnout mass.

I don't know that looking at staging velocity difference will tell the same story, since the difference in payload mass is trivial at staging (less than 0.5%) but much more significant at the end of GTO insertion (5.9%).

[ChrisGebhardt]

Two objects related to today's #Falcon9 launch tracked in a sub-GTO orbit
2018-023A: 184 x 22,261 km, 26.97°
2018-023C: 186 x 22,215 km, 26.92°

https://twitter.com/Spaceflight101/status/971074423108358144

But what’s 2018-023B?  Nothing else launched last night.  This could indicate a potential unknown ride share happened last night.