SpaceX Falcon Mission Simulations

[manoweb]

Do you think it makes sense to add a timestamp column? It could contain the seconds since T0. Also, it would be trivial to do that in the spreadsheet just by adding 0.033s (or, 0.0333667000334 as it is 1/29.97 of a second in theory) on each line, so not sure it's worth it

[OneSpeed]

This is a simulation of the NROL-76 launch created with NSF user zlynn1990. The broadcast gave us the best view yet into the first stage recovery process with amazing footage and telemetry data. It allowed us to tweak a lot of previously unknown aerodynamic properties for the first stage during the recovery process. This simulation follows the webcast telemetry data very closely throughout the launch and landing process.

This video also shows a very speculative simulation of how the fairing was landed in the ocean. We have very little information here, so everything happening is a complete guess. A supersonic ballute is deployed around mach 6 to help stabilize the fairing while re-entering the atmosphere. Nitrogen thrusters are used to maintain and correct attitude. Once the fairing is below 100 m/s a parafoil is deployed. This helps guide the fairing to the final landing zone (a bouncy castle like setup). Once the fairing is nearly above the ocean it flares up to help reduce velocity and essentially stalls into the landing structure.

[IainMcClatchie]

Fabulous stuff.  Thank you!

Do I understand correctly that this was NOT a "block 5" launch?

[manoweb]

People are not even sure it was a Block 4, whatever that means...

[deruch]

It took me multiple rewatches to figure out what was going on at ~5:47s when the video seemed to suggest that the fairing half "throttled down" during its reentry.  I was stuck trying to understand what sort of propulsion system you thought it had on it until I finally caught that the "throttle down" was connected to Stage 2.  Oops.

[manoweb]

OCR data, chart and spreadsheet, for SpaceX's Inmarsat-5

[envy887]

OCR data, chart and spreadsheet, for SpaceX's Inmarsat-5

Excellent!

Can you overlay SES-10 on the same graph, or attach the spreadsheet?

[manoweb]

I have not done SES-10. This launch was special in the fact it was expandable so they broadcast the totality of the "second stage telemetry" from launch to SECO... Let me have a look at SES-10 video

[gospacex]

OCR data, chart and spreadsheet, for SpaceX's Inmarsat-5

Great! Can you graph acceleration?

[manoweb]

gospacex, I think it should be easy once you have time and velocity data. Right now I'm doing OCR of SES-10. Maybe somebody good with spreadsheets (I am "so-so", I used to be good with Lotus 1-2-3, but I haven't practiced much since then) can share the effort and add acceleration (jerk?) data?

envy887 I did not realize the SES-10 (and older!) webcasts have full stage2 telemetry. The only exception was NROL76, the one I started with, that had full stage1 telemetry, that was my source of confusion.

[manoweb]

This is SES10 data and graphic:

[manoweb]

This is the comparison between SES-10 and Inmarsat-5

[envy887]

This is the comparison between SES-10 and Inmarsat-5

Fantastic, thank you!

You can clearly see the throttle-down for trans-sonic is at around 320 m/s for each. But Inmarsat hits transsonic earlier and higher, and accelerates through it faster than SES-10. The Inmarsat booster is going higher and accelerating faster all the way to staging despite pushing a larger payload.

SES-10 is known to be a v1.2 Block 1 booster, so Inmarsat definitely was upgraded. But was it upgraded from NROL-76? The first stage telemetry could show the difference. It would be awesome if you could overlay that data.

Interestingly, the Mvac doesn't appear to have been upgraded. The lower acceleration of the second stage is explained by the higher payload mass.

[manoweb]

I have created a small github project; this is an application that "helps" reviewing the data from the OCR phase. It's at its initial commit, it serves a purpose, but it's far from being "usable" by the masses

However if anybody is interested in using it, please let me know.

https://github.com/alesanmanoweb/SpaceXocr

[TrueBlueWitt]

This is the comparison between SES-10 and Inmarsat-5

Fantastic, thank you!

You can clearly see the throttle-down for trans-sonic is at around 320 m/s for each. But Inmarsat hits transsonic earlier and higher, and accelerates through it faster than SES-10. The Inmarsat booster is going higher and accelerating faster all the way to staging despite pushing a larger payload.

SES-10 is known to be a v1.2 Block 1 booster, so Inmarsat definitely was upgraded. But was it upgraded from NROL-76? The first stage telemetry could show the difference. It would be awesome if you could overlay that data.

Interestingly, the Mvac doesn't appear to have been upgraded. The lower acceleration of the second stage is explained by the higher payload mass.

Let's not all forget that SES 10 S1 had a lot more mass going uphill with Landing legs and Grid Fins, plus whatever other re-usability feature on First stage weren't there for Inmarsat..  Do those more than offset the heavier mass of the Satellite in terms of First Stage performance?

[OneSpeed]

Let's not all forget that SES 10 S1 had a lot more mass going uphill with Landing legs and Grid Fins, plus whatever other re-usability feature on First stage weren't there for Inmarsat..  Do those more than offset the heavier mass of the Satellite in terms of First Stage performance?

The landing legs and grid fins are certainly a factor, and SES-10 may not be the best point of comparison. It launched at only 87.5% of rated thrust, which makes sense when you consider it was the first ever re-use of a first stage, and SpaceX would have been very keen to give it the best chance of success. Perhaps Echostar 23 would be a better reference? Anyway, Inmarsat-5 appears to have launched at 90% of rated thrust, and although it went to 94% after Max Q, I'm not sure we've seen a block 4 yet. Interestingly, the second stage appears to have reverted from 107% to 98% of rated thrust for the first minute or so, and none of the stages have used any high AoA manoeuvres this time. Apologies for the step in velocity at 150kms.

[envy887]

Let's not all forget that SES 10 S1 had a lot more mass going uphill with Landing legs and Grid Fins, plus whatever other re-usability feature on First stage weren't there for Inmarsat..  Do those more than offset the heavier mass of the Satellite in terms of First Stage performance?

The landing legs and grid fins are certainly a factor, and SES-10 may not be the best point of comparison. It launched at only 87.5% of rated thrust, which makes sense when you consider it was the first ever re-use of a first stage, and SpaceX would have been very keen to give it the best chance of success. Perhaps Echostar 23 would be a better reference? Anyway, Inmarsat-5 appears to have launched at 90% of rated thrust, and although it went to 94% after Max Q, I'm not sure we've seen a block 4 yet. Interestingly, the second stage appears to have reverted from 107% to 98% of rated thrust for the first minute or so, and none of the stages have used any high AoA manoeuvres this time. Apologies for the step in velocity at 150kms.

I pulled 10-second intervals off the webcasts for both I5F4 and Echostar 23. The difference in burn time appears to be close to the difference in acceleration, suggesting I5F4 had about 2% additional prop load but ran at the same thrust except around transonic.

[manoweb]

The landing legs and grid fins are certainly a factor, and SES-10 may not be the best point of comparison. It launched at only 87.5% of rated thrust, which makes sense when you consider it was the first ever re-use of a first stage, and

Out of curiosity, how did you make this graphic?
1) Did you use your own data or you used the one I provided (I am trying to see if my work is actually useful to people )
2) How did you graph the acceleration (this is basic spreadsheet usage, I know, but I'm still curious)
3) How did you calculate the thrust, this one seems tricky

[manoweb]

I pulled 10-second intervals off the webcasts for both I5F4 and Echostar 23. The difference in burn time appears to be close to the difference in acceleration, suggesting I5F4 had about 2% additional prop load but ran at the same thrust except around transonic.

Hello Envy887, pulling values manually from the video seems a pretty boring task... If you need just point me to which mission you like and I can extract the telemetry data easily, all ~16000 points of it. In the previous posts I always provided either a CSV file or LibreOffice spreadsheet, is the format not good enough? I'd rather spend time on the telemetry extraction tool, and see you guys do the spreadsheet analysis

[envy887]

I pulled 10-second intervals off the webcasts for both I5F4 and Echostar 23. The difference in burn time appears to be close to the difference in acceleration, suggesting I5F4 had about 2% additional prop load but ran at the same thrust except around transonic.

Hello Envy887, pulling values manually from the video seems a pretty boring task... If you need just point me to which mission you like and I can extract the telemetry data easily, all ~16000 points of it. In the previous posts I always provided either a CSV file or LibreOffice spreadsheet, is the format not good enough? I'd rather spend time on the telemetry extraction tool, and see you guys do the spreadsheet analysis

I haven't seen data for Echostar-23, a CSV of that data would be great