Constant Altitude Mixed-Isp Landing Calculator

[Lirtosiast]

Ever since the days of @tavert and @Kosmo-not we've known how to execute an efficient low-TWR landing: throttle to 100% and keep the ship at constant altitude by pitching down. But what if you're designing, say, a reusable mission to Tylo that uses both NERV and RAPIER engines? When landing on Tylo, how far do you pitch up, when do you switch on the RAPIERs, and at what thrust?

@FleshJeb and I wrote a spreadsheet calculator for this purpose. It assumes a constant-altitude ascent/landing (pretty close to the optimum), and works for ascent or landing, and any types of engines. Simply enter parameters in the green boxes and your mass fraction, total dV expended, and ascent/landing profile appear in the blue boxes. Accuracy is within about 1-2% due to the CAL assumption and Euler's method.

Calculator is here.

-----

To implement:

* Add table of engines and celestial bodies

* Use a less inaccurate algorithm than Euler's method

* non-CALs (hopefully the math is not difficult)

[FleshJeb]

Pfft, you did 99% of the work. I just provided a little feedback and discussion. I still don't understand exactly how the Lambda column works--I'll have to pick that apart for myself.

Of note, I spoke to Sarbian on Discord a couple weeks ago, and he said if I (or anyone else) can get a CAL script working in KRPC, he's willing to refactor it for MechJeb.

That's been a goal of mine for a couple of years now, because I know how badly MJ can fail a low deorbit burn on the Mun if your TWR is less than 2.5 or so.

I've spent a number of hours digging into MJ's landing code, and some slight notion of how to integrate it within the larger framework.

To implement:

• Empirical verification of method.
  • Single engine type
  • Multiple engine types
• Python script for KRPC  using the spreadsheet's predictive method to time and execute the burn. I've played with KRPC a fair amount, and have a decent understanding of the framework.
• Calculator in Python is complete: https://www.dropbox.com/s/2r4ruljkcux0qqw/CAL_Euler.py?dl=0
• Integrate Normal/Anti-Normal burns into the prediction. I think it should be fairly easy to add to this method, compared to what I was going to do.
• Find out if it's more efficient to do the plane change to the landing site at the AN/DN or during the CAL. (lower average velocity, and Pythagorean efficiency, so it might be much more efficient).

Stretch goals:

• A generalized landing script that will transition from CAL to suicide burn when the TWR is high enough. (Or combo-suicide/CAL?)
• Integrate into MechJeb.

EDIT: @Lirtosiast This might fit better in the Tutorials subforum.

Edited March 25, 2018 by FleshJeb

[FleshJeb]

On 3/23/2018 at 10:39 PM, Lirtosiast said:

To implement:

* Add table of engines and celestial bodies

DONE. I also finalized the Python calculator. It's linked in the post above.

https://docs.google.com/spreadsheets/d/13miu6GKftD1K5SRyD_SNMIwZl7N4ofTfBduUsK70Vqo/edit?usp=sharing

Ripped the vacuum engine data from 1.3.1 using KRPC

Copied the relevant celestial body info from the KSP wiki.

I didn't want to mess with the main sheet, but you can add fuel quantity by type to the spreadsheet:
    fuel_1 = (f_1_tot / v_e_1) * dt             # LF  for NERV
    fuel_2 = (f_2_tot / v_e_2) * dt * f_2_ratio # LFO for RAPIER

Edited March 25, 2018 by FleshJeb