SpaceX Falcon Mission Simulations

[Lar]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

[acsawdey]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

That makes the issue worse ... if the side boosters throttle down, and the side boosters also shut down 2 engines at the end, the core has to be throttled down even more in order to be able to burn a whole minute longer.

OK, I'm dense ... the answer is that the boosters have to do a big boost back burn, and the core just coasts.

[Slarty1080]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

Yes I heard them talk about that during the FH launch we just saw. The question is why? From a purely performance angle it would be better to leave the side boosters running flat out and throttle the core further to retain propellant in the core for later use.

If only 3 engines on the core are throttle capable then they might not be able to provide sufficient throttle range, which begs the question why not make more engines throttle capable on the core stage if it would translate into a performance increase?

There must be a good reason why they don't do this. It might be structural perhaps that level of thrust from the side boosters might be too much stress for the core?

[OneSpeed]

Here's a plot of velocity and acceleration over time for Arabsat-6A, with annotated events. The core can only throttle to 57%, so the side booster shutdown points are being determined by structural loading, rather than g forces. Full thrust of 22,819kN is only achieved instantaneously at around T+00:10.

[ATPTourFan]

For Side Shutdown 1, 2 and 3, do we have a good idea what combination of engines are disabled at these times?

[Wolfram66]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

Yes I heard them talk about that during the FH launch we just saw. The question is why? From a purely performance angle it would be better to leave the side boosters running flat out and throttle the core further to retain propellant in the core for later use.

If only 3 engines on the core are throttle capable then they might not be able to provide sufficient throttle range, which begs the question why not make more engines throttle capable on the core stage if it would translate into a performance increase?

There must be a good reason why they don't do this. It might be structural perhaps that level of thrust from the side boosters might be too much stress for the core?

Has more to do with the G-Loads on the payload. G-Limits are to not damage 1) the Payload and 2) the center core.  Jerry always mentions, on the broadcast, that the throttling is to reduce g-loads on the payload and the vehicle.... there is a fully fueled 2nd stage + Payload being pushed by an increasingly empty booster stages.

[ugordan]

For Side Shutdown 1, 2 and 3, do we have a good idea what combination of engines are disabled at these times?

Judging from ground footage, my guess is for shutdowns #1 and #2 that it's engines 6 and 7 on the boosters. No idea about the sequencing and pairing (i.e. if it's 7-7 and 6-6 or 7-6 and 6-7), though.

For thrust reduction #3, exactly 3 seconds prior to BECO, implied to also be a shutdown event due to a slight pitch disturbance to the vehicle, I have no idea which engine, but it's the only event where you could see the shutdown vapor transient in onboard cameras.

[OneSpeed]

For Side Shutdown 1, 2 and 3, do we have a good idea what combination of engines are disabled at these times?

Judging from ground footage, my guess is for shutdowns #1 and #2 that it's engines 6 and 7 on the boosters. No idea about the sequencing and pairing (i.e. if it's 7-7 and 6-6 or 7-6 and 6-7), though.

I also can't tell which engines are being shutdown from the videos, but 6 & 7 would minimise the torque about the z-axis, so they seem as likely as any. However, I can confirm that for shutdowns #1 and #2, shutting down one engine from each side booster matches the change in acceleration we saw in the raw data.

I'm surprised by how much time is spent with thrust well below that available. Perhaps SpaceX is still being somewhat conservative with this vehicle, and we'll see some performance optimisations in the future. Nevertheless, the 6465kg payload was delivered to 10.7 km/s. The sim predicts that there was enough propellant remaining in the second stage to get to 11km/s, enough to reach the Moon or Earth escape velocity.

[Slarty1080]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

Yes I heard them talk about that during the FH launch we just saw. The question is why? From a purely performance angle it would be better to leave the side boosters running flat out and throttle the core further to retain propellant in the core for later use.

If only 3 engines on the core are throttle capable then they might not be able to provide sufficient throttle range, which begs the question why not make more engines throttle capable on the core stage if it would translate into a performance increase?

There must be a good reason why they don't do this. It might be structural perhaps that level of thrust from the side boosters might be too much stress for the core?

Has more to do with the G-Loads on the payload. G-Limits are to not damage 1) the Payload and 2) the center core.  Jerry always mentions, on the broadcast, that the throttling is to reduce g-loads on the payload and the vehicle.... there is a fully fueled 2nd stage + Payload being pushed by an increasingly empty booster stages.

Yes I think your right. The 10% extra thrust on the engines for this FH compared to the first is effectively wasted.

[Proponent]

Here's a plot of velocity and acceleration over time for Arabsat-6A, with annotated events....

Would you be willing to post the data behind this plot?  Would you have altitude as well?  Through stage 2 burn?

[envy887]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

Yes I heard them talk about that during the FH launch we just saw. The question is why? From a purely performance angle it would be better to leave the side boosters running flat out and throttle the core further to retain propellant in the core for later use.

If only 3 engines on the core are throttle capable then they might not be able to provide sufficient throttle range, which begs the question why not make more engines throttle capable on the core stage if it would translate into a performance increase?

There must be a good reason why they don't do this. It might be structural perhaps that level of thrust from the side boosters might be too much stress for the core?

Has more to do with the G-Loads on the payload. G-Limits are to not damage 1) the Payload and 2) the center core.  Jerry always mentions, on the broadcast, that the throttling is to reduce g-loads on the payload and the vehicle.... there is a fully fueled 2nd stage + Payload being pushed by an increasingly empty booster stages.

Yes I think your right. The 10% extra thrust on the engines for this FH compared to the first is effectively wasted.

More thrust from the sides is very helpful during the period between center core throttle-down and side engine cutout. That is the lossiest part of the gravity turn, and the higher thrust booster reduces gravity losses while still allowing the center core to conserve fuel until after staging off the dry mass of the boosters.

[Barley]

The booster engine cut out is interesting.  If they throttled booster engines before cutting engines (as at the end of the second stage burn) they should be able to "fill in" the saw teeth in the acceleration, retaining more thrust for longer.

Perhaps we'll see operational improvements on later flights to eke out a little more margin (or payload).

Also what causes the kink in second stage acceleration at about 550s?

[Robotbeat]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

Don't forget that the side boosters throttled down at various points as well. They were not running flat out.

Yes I heard them talk about that during the FH launch we just saw. The question is why? From a purely performance angle it would be better to leave the side boosters running flat out and throttle the core further to retain propellant in the core for later use.

If only 3 engines on the core are throttle capable then they might not be able to provide sufficient throttle range, which begs the question why not make more engines throttle capable on the core stage if it would translate into a performance increase?

There must be a good reason why they don't do this. It might be structural perhaps that level of thrust from the side boosters might be too much stress for the core?

Has more to do with the G-Loads on the payload. G-Limits are to not damage 1) the Payload and 2) the center core.  Jerry always mentions, on the broadcast, that the throttling is to reduce g-loads on the payload and the vehicle.... there is a fully fueled 2nd stage + Payload being pushed by an increasingly empty booster stages.

Yes I think your right. The 10% extra thrust on the engines for this FH compared to the first is effectively wasted.

Gravity losses right at first are the worst. So even if you only use the extra thrust right at first, that's a win.

[OneSpeed]

Here's a plot of velocity and acceleration over time for Arabsat-6A, with annotated events....

Would you be willing to post the data behind this plot?  Would you have altitude as well?  Through stage 2 burn?

Sure, but take them with a grain of salt. The acceleration figures are scaled up by 100 for display purposes, and the program I've written to scrape the numbers is very much a work in progress. SpaceX keep changing the resolution and frame rate of their webcast videos. Very happy to have them at all though!

[OneSpeed]

This is a simulation of the NASA IXPE mission, currently scheduled for April 2021. The 'Imaging X-ray Polarimetry Explorer' (IXPE) is unusual for a Falcon 9 launch because of its low earth equatorial orbit, which cannot be directly entered unless the satellite is launched from the equator.

If not launched from the equator, it requires a plane change from the initial orbital inclination (28.5° from the Cape). For this mission, the plane change will require an enormous 3.66 km/s of ΔV. To optimise performance in this sim, I'm utilising the recent Starlink 0.9 first stage throttle profile, which limited the period and shape of the throttle back for MaxQ.

Due to the extremely light 315kg payload, staging occurs at a relatively high 2.45km/s. This leads to the drone ship being positioned 765 kms downrange, further than usual for F9. In order to cross the equator at the required 540km apogee, the primary second stage burn has to be somewhat lofted. This results in a 540km x -68km sub-orbital coast phase, before the combined plane change and circularisation burn over the Gulf of Guinea.

What you might already know from the recent NSF article on IXPE, is that this mission was initially scheduled to launch on the NGIS Pegasus. What you night not realise is that a part of the reason IXPE is now flying on a Falcon 9 is because of the persistence and determination of NSF member and industry professional Comga.

More than two years ago Comga asked in the NSF 'L2 SpaceX Rocket & Spacecraft Simulation Thread': "can someone with a good model calculate the payload capacity of an F9 to an equatorial (0 degree inclination) 540 km altitude orbit?". https://forum.nasaspaceflight.com/index.php?topic=37599.msg1672059#msg1672059
There have been various L2 contributors to that discussion since, but the takeaway is that Comga was able to present those figures to NASA, and the result is a $50M launch award to SpaceX. I can't think of a better advertisement for L2 membership than that!

NSF members S.Paulissen and livingjw have both suggested adding crossrange to the SpaceSim flight model, and IXPE is the perfect excuse. For the plane change burn there is now additional data: crossrange, lateral velocity, acceleration, and inclination are all displayed in real time. From this, the final mass estimate is 7,786kg. Subtracting the 4,500kg dry mass of the second stage, and 1,000kg of ullage gives an effective payload to 540 x 540km equatorial orbit of 2,286kg. Plenty of room for some additional subsystems

Edit: Whilst I did ask Comga if I could credit him in this post for helping SpaceX to obtain this launch, it looks like I've oversold it. Although we did establish that Falcon 9 could likely perform the mission, the launch award was as a result of the usual competitive process, not because of our figures. Apologies for any confusion this may have caused.

[Barley]

Why is there a step change in "velocity" and heating rate at around 225 seconds?

[OneSpeed]

Why is there a step change in "velocity" and heating rate at around 225 seconds?

This question gets asked about SpaceSim every now and then. SpaceSim is currently a 12-body simulation of the solar system. That is, it models the Sun, Mercury, Venus, Earth, Moon, Mars, Jupiter, Callisto, Europa, Ganymede, Io, and Saturn. The orbital data for these bodies is calculated from the JPL ephemerides 2016 epoch, and has the Sun moving at 59km/s through space, rotating at -2.9E-06 rad/s. The Earth is orbiting the Sun at about 29.8 km/s, and at 28.5° latitude we would be orbiting the centre of the Earth at 408.7m/s.

SpaceSim tries to take account of these combined velocities, so that at launch, SpaceSim displays zero velocity, i.e. in Earth's inertial frame of reference. However, once the spacecraft reaches Earth orbit, the Earth's rotation is no longer an influence on velocity, so SpaceSim displays velocity in the orbital frame of reference, which for an inclination of 28.5° is 408.7m/s faster. Hence the step change you see as the spacecraft crosses Earth's atmosphere height of 150kms.

There are similar step changes in displayed velocity when moving from one celestial bodies sphere of influence to another. E.g. on a trip to the Moon, there is a point at which SpaceSim considers the spacecraft's gravitational parent to have changed from the Earth to the Moon. Internally, SpaceSim is still calculating velocity with respect to the entire Solar System moving through space, but for display purposes it changes from the Earth's orbital frame to the Moon's orbital frame, and there is another step change in the displayed velocity.

For a trip to Mars, there is a long period where SpaceSim considers the spacecraft's gravitational parent to be neither Earth nor Mars, but the Sun, and there is a massive increase in the displayed velocity to around 33km/s (about 3km/s faster than the Earth relative to the Sun).

The step change in heating occurs because it no longer makes sense to attempt the calculation once above the atmosphere height.

Interestingly, SpaceX webcasts always appear to show velocity in the inertial frame of reference, but I find this misleading once in orbit. Conversely, I was watching the Apollo 11 launch the other day, and there was a callout at T+000:01:08 "Downrange 1 [nautical] mile [1.8 km], altitude 3, 4 [nautical] miles [7.4 km] now. Velocity 2,195 feet per second [669 m/s]". 669m/s is around Mach 2, which is odd because it was 15 seconds before MaxQ at T+000:01:23. Perhaps they were calling out velocities in the orbital frame of reference all the way to orbit?

[OneSpeed]

Here is an analysis of the Falcon 9 AX-2 and Crew-7 mission webcast telemetry.

Unlike the Crew-7 mission, the AX-2 webcast showed telemetry for both the booster and the second stage. The first plot compares the booster telemetry common to both missions. Although there are numerous gaps in the Crew-7 telemetry, the plots are close to identical where data exists.

The timelines predicted by SpaceX were:

-	AX-2	Crew-7
MaxQ	00:01:02	00:01:02
MECO	00:02:26	00:02:26
Separation	00:02:29	00:02:30
Boostback start	00:02:39	00:02:43
Boostback end	00:03:28	00:03:30
Entry start	00:06:25	00:06:21
Entry end	?	00:06:32
Landing start	00:07:31	00:07:27
Landing end	00:07:58	00:07:44

The actual timelines were :

-	AX-2	Crew-7
MaxQ	00:01:02	00:01:02
MECO	00:02:30	00:02:30
Separation	00:02:33	00:02:33
Boostback start	00:02:46	<= 00:02:50
Boostback end	00:03:29	?
Entry start	00:06:23	00:06:22
Entry end	?	00:06:25
Landing start	00:07:27	00:07:22
Landing end	00:07:47	00:07:39

So, both boosters landed a few seconds earlier than predicted. The Crew-7 entry burn length however, was the most notable difference, from a predicted 11 seconds down to only 3 seconds. It is not possible for me to say if this was intended, but what I can do is model the shorter burn on the complete AX-2 booster telemetry, and observe the difference in dynamic pressure and heating.

The next plot is a comparison between the live AX-2 telemetry, and the output of my attempt at a simulation of the AX-2 booster flight using SpaceSim. I have calculated the AX-2 orbital acceleration in the same way as for the telemetry, i.e. from the difference in velocity over the preceding second. Although not perfect, I suspect that it is good enough to gain a relative if not an absolute understanding of the dynamic pressure and heating on the AX-2 and Crew-7 missions.

The next 4 plots are of the values calculated by SpaceSim for AX-2 and Crew-7, a third hypothetical plot (Crew-8?) with the entry burn deleted, and a generic Starlink launch for comparison.

The dynamic pressure and heating peak values are:

-	AX-2 -	Crew-7	Crew-8?	Starlink
Dynamic pressure (kPa)	120.8	140.8	150.6	120.2
Heating (kW/mē)	70.4	86.8	93.5	109.8

So, to summarise, the maximum dynamic pressure on re-entry is similar for AX-2 and a generic Starlink launch, but was 16% higher for Crew-7, and would be 25% higher if the re-entry burn was deleted. The maximum heating was least for AX-2, but both Crew-7 and Crew-8? would still have less heating than a Starlink launch.

There is a small saving in propellant, 480kg for Crew-7, 860kg deleting re-entry entirely. Deleting re-entry might enable an additional 80kg of payload to orbit. It depends whether the Crew-7 short burn resulted in damage to the booster. If Crew-7 was undamaged by the higher than usual dynamic pressure, then perhaps the re-entry burn could be deleted for some future RTLS missions?

Here is a simulation of the three RTLS missions flying simultaneously in a space ballet.

[ChrisC]

I just came across this latest analysis thanks to you crossposting it over to the Crew-7 thread.  Thanks OneSpeed!  I've added it to my own collection of observations about this reentry burn here.