SpaceX Falcon 9 : Intelsat 35e : July 5, 2017 : DISCUSSION

[AC in NC]

I recall Ed Kyle giving an excellent contextualization of performance that delivered Inmarsat to GTO with an impressive deltaV deficit.

Here's how I was finally able to get my mind around this launch result. 

It would take an Atlas 541 to match this performance, or a Proton M Briz M Phase 3 or 4.  Ariane 5 ECA could do it, of course, as could Delta 4 Heavy.  CZ-5, theoretically though it has yet to demonstrate the capability.  H-2B, ditto.  That's it, I think, among active launchers.

Delta 4 Mediums can't do this at all.  Neither can H-2A, or CZ-3B, or CZ-7, or GSLV Mk 3. 

When Falcon 9 first began flying, I thought of it as slightly better than Delta 2 class.  The machine has evolved, and the most recent two flights have exhibited a new level of performance - to the extent that I'm convinced we are seeing at least a Block 4 second stage.

 - Ed Kyle

Is there enough information available to compare the Inmarsat mass and deficit to this Intelsat mass and deficit?  Would be interesting to see a comparison and if any additional performance has been squeezed from the the system.

[ugordan]

Is there enough information available to compare the Inmarsat mass and deficit to this Intelsat mass and deficit?  Would be interesting to see a comparison and if any additional performance has been squeezed from the the system.

Not before TLEs for Intelsat 35e are published.

[LouScheffer]

Is there enough information available to compare the Inmarsat mass and deficit to this Intelsat mass and deficit?  Would be interesting to see a comparison and if any additional performance has been squeezed from the the system.

Yes, I did this calculation here .  If the performance of the rocket was exactly the same, just a bigger payload mass, then we'd expect a 48,000 km apogee.  It got 44,000 km,  well within the errors of the calculation and the same for all practical purposes.  Certainly no big performance boost.

[LouScheffer]

If it's really 26^o as they were aiming for, I get about 1720 m/s to GEO for a 44,000  km apogee.

[jrgagne99]

Any word yet on fairing recovery?

[edkyle99]

Challenger's ill-fated STS-51L flight was the first to launch from Pad 39B in January 1986.  So all 9 shuttle launches in 1985 would have been from 39A.

We tried, really hard, to launch more Shuttles that year (1985), but there were scrubs and rollbacks and delays, etc.  SpaceX is doing this year what NASA wanted Shuttle to do back then, except for the crewed launches. 

At 10 launches so far this year, Falcon 9 becomes the first U.S. launch vehicle to fly 10 times in a calendar year successfully since Delta 2 did it in 1999, 18 years ago.  With three more launches, Falcon 9 will have us digging into the 1970s launch lists for comparison.  One or two more after that, and we'll be looking at 1960s numbers.

Of course none of those frequent flyers in those days could lift 6.761 tonnes to GTO.

 - Ed Kyle

How would these numbers look if you went by payload mass rather than number of flights? Counting the Orbiter as part of the LV for this purpose. Are SpaceX beating the payload to orbit achieved by the shuttle and ELVs?

For 1985, when STS flew 9 times, the numbers are close, but STS appears to have slightly exceeded Falcon 9's payload numbers for its 10 flights in 2017 to date. 

In 1985, STS put 14 satellites into GTO or GEO using IUS, PAM-D, PAM-D2, or other PKMs.  My estimate of the total satellite or equivalent satellite mass placed into GTO in 1985 is about 27.5 tonnes (just for the satellites, not the PKMs).  Falcon 9 has boosted about 27.4 tonnes to GTO in five missions so far this year.

In 1985, STS lifted about 43.9 tonnes to LEO, including three Spacelabs, Spartan-101, and a handful of other payloads.  So far this year, Falcon 9 has lifted about 36.63 tonnes in five LEO missions, assuming that the NROL-76 payload was relatively light.  Of note, however, is that two of the Falcon 9 missions have gone to near-polar orbit from VAFB, something Shuttle never achieved.  The other Falcon 9 LEO flights have gone to 51.6 deg orbits, something STS didn't do in 1985.

Of course STS also put 58 astronauts into orbit in 1985, something Falcon 9 has yet to achieve, and we haven't accounted for the mass involved in supporting the human flight aspect here.

 - Ed Kyle

[kevin-rf]

Space-Track isn't listing the TLE's yet.

[StuffOfInterest]

For 1985, when STS flew 9 times, the numbers are close, but STS appears to have slightly exceeded Falcon 9's payload numbers for its 10 flights in 2017 to date. 

In 1985, STS put 14 satellites into GTO or GEO using IUS, PAM-D, PAM-D2, or other PKMs.  My estimate of the total satellite or equivalent satellite mass placed into GTO in 1985 is about 27.5 tonnes (just for the satellites, not the PKMs).  Falcon 9 has boosted about 27.4 tonnes to GTO in five missions so far this year.

In 1985, STS lifted about 43.9 tonnes to LEO, including three Spacelabs, Spartan-101, and a handful of other payloads.  So far this year, Falcon 9 has lifted about 36.63 tonnes in five LEO missions, assuming that the NROL-76 payload was relatively light.  Of note, however, is that two of the Falcon 9 missions have gone to near-polar orbit from VAFB, something Shuttle never achieved.  The other Falcon 9 LEO flights have gone to 51.6 deg orbits, something STS didn't do in 1985.

Might be a bit of an eye popper to take the average cost per flight of the Shuttle and Falcon 9, multiply by the number of flights, and then divide by the tonnage lifted.  Oh, and of course, adjust for inflation.  Shuttle was great, but the costs never did come down.

[JamesH65]

Was ground software that was the issue, not the rocket.

So my friend missed his chance to view a launch due to a software glitch? Bummer.
I doubt we'll ever get details on this, but it's a bit worrisome to have software problems crop up at this point, where they're demonstrating system maturity. As a hardware guy, I hate software problems.

As a software guy, it's almost always the hardware's fault.

[D_Dom]

Unless it is on fire it is a software problem.

[stcks]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

[DatUser14]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

And how does that compare to Atlas V performance?

[gospacex]

Was ground software that was the issue, not the rocket.

So my friend missed his chance to view a launch due to a software glitch? Bummer.
I doubt we'll ever get details on this, but it's a bit worrisome to have software problems crop up at this point, where they're demonstrating system maturity. As a hardware guy, I hate software problems.

As a software guy, it's almost always the hardware's fault.

Looking at some CPU erratas...

Broadwell: "Back-to-back page walks due to instruction fetches may cause a system hang"? Doh... How am I supposed to avoid causing page walks?

Kaby Lake: "Processor hay hang under complex scenarios". Er, what?

[joncz]

Looking at some CPU erratas...

Broadwell: "Back-to-back page walks due to instruction fetches may cause a system hang"? Doh... How am I supposed to avoid causing page walks?

Kaby Lake: "Processor hay hang under complex scenarios". Er, what?

<mumbles something about microcode and how, even when it's hardware, it's software's fault>

;-)

-a software guy

[edkyle99]

Unless it is on fire it is a software problem.

Software can start fires (allegedly). 
http://time.com/4450433/iran-investigates-if-series-of-oil-industry-accidents-were-caused-by-cyber-attack/

 - Ed Kyle

[yokem55]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

And how does that compare to Atlas V performance?

It's in the ballpark of what an Atlas 531 could do.

[envy887]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

And how does that compare to Atlas V performance?

It's in the ballpark of what an Atlas 531 could do.

For $142M basic list price.

[Mike_1179]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

And how does that compare to Atlas V performance?

It's in the ballpark of what an Atlas 531 could do.

For $142M basic list price.

With gas generator engines and RP-1 on both stages. Still kind of boggles the mind.

[ugordan]

It's a little unfair to compare what another vehicle (like Atlas) can do when its listed performance is given for a guidance commanded shutdown vs a minimum residual shutdown that happened here. The former will always reserve some delta-V margin for performance dispersions like steering losses, engine underperformance, etc.

If F9 had to deliver a payload to a *specific* GTO-1720 orbit, the maximum payload mass would certainly be less than 6.7 tonnes precisely because they'd be sandbagging the performance to make sure they hit the target with some propellant margin to spare.

Either way, what F9 did here is fairly impressive as is SpaceX's turnaround record with LC-39A.

[John Alan]

2017-041A    42818    FALCON 9 R/B    775.27min    25.84°    42861km    293km
2017-041B    42819    INTELSAT 35      772.84min    25.85°    42742km    296km

About GTO-1719. Excellent performance.

And how does that compare to Atlas V performance?

It's in the ballpark of what an Atlas 531 could do.

For $142M basic list price.

With gas generator engines and RP-1 on both stages. Still kind of boggles the mind.

Old school tech...redone right...for under half the price...