Suicide Burns (Kerbal Space Program [external])
Martin mentioned what is known as a “suicide burn”. In the context of a “lunar lander” it is seeing how late you can fire the thrusters without causing the lunar lander to crash. It’s fine in a simulation, nobody dies, and the virtual landers are free. In real life, no one in their right mind would attempt it, please don’t.
In his article, Martin achieved a 1.66 mph impact speed (not perfect, as it exceeds 1 mph and he calls this out). His burn was 0 until 47 miles, then 164.3146124, followed by 200 for the remaining descent. It sounds logical, minimal burn, then burn like your life depended on it to avoid adding to the already cratered surface.
Can we do better?
It depends on what our goal is.
Less than 10 mph is deemed a good landing, albeit the goal (in keeping with Jim’s ratings) is less than 1. Was the goal to have the most fuel left whilst landing at less than 10 mph? I believe it was to maximise remaining fuel, he’s just not clear on the maximum impact speed allowed (or I failed to notice).
One of the best landings (see conclusion) was 0.17 mph, but it was not a suicide burn – it was allowed to burn immediately. A suicide burn is harder.
Using AI, I easily achieved 0.20 mph from a suicide burn(!), with 653.81 lb of fuel left.
The fuel rates were:
0
0
0
0
0
0
0
193.62253477887234
193.22874292238433
192.7381306843956
192.1178845936319
191.32145579849572
190.28193470044968
188.90214485823577
187.04027885633354
184.49052111379075Double-precision numbers are required.
That’s much better than 1.66 mph from a comfort-wise standpoint for the poor virtual astronauts. It does, however, use 21 lb more fuel. It is always going to be a trade-off between impact speed and fuel consumed – I didn’t need to tell you that, did I?
Please note: If you want to try any figures I quote, you need to have applied the fix that Martin proposed. e.g. use my code from GitHub or convert it yourself; then fix it.
