Optimal Munar Orbit Insertion Challenge

[Tarmenius]

Having narrowed down the most efficient Munar descent from a specific orbit in my previous challenge http://kerbalspaceprogram.com/forum/index.php?topic=15407.0, the natural evolution of the topic drew conversation toward what effect the Munar Insertion has on overall efficiency. Thus, a new challenge is born.

Here, we will focus primarily on the various methods of transitioning from Munar Intercept to Descent. Is it more efficient to go directly into the descent phase after entering the Mun\'s Sphere of Influence (SOI)? Or perhaps it will save more fuel to set up a bi-elliptic transfer into a low-Mun orbit before beginning your descent. There are plenty of options to explore, but only one will prove to be Optimal. I\'m sure some of you have done the math, and that\'s wonderful for theory but let\'s see how it works in practice. As always, be encouraged to discuss that math (even though I won\'t understand most of it ), and any theories you may have for saving those precious fuel units. Even though this is framed as a challenge, the true goal is a greater understanding about how to get the most from our rockets.

With that, we come to the substance of our challenge. The goal will be to insert into a 3km by 3km orbit around the Mun. Success will be measured by how little fuel is used during this phase. To create the scenario, I have taken the original stock lander from the Optimal Descent (to theMun) challenge and placed it on a free-return trajectory through the Mun\'s SOI with a Pe of just under 1 million meters. This was chosen to allow a subsequent return to Kerbin with a Pe of roughly 30,000m (a safe re-entry altitude) and thus simulate a more realistic flight plan. You begin the challenge at the not-quite-half-way point, right after I had performed the Mid-Course Correction1 burn with fuel leftover from the Trans-Munar Injection stage. As with the previous challenge, the lander must be activated with [spacebar] as my proof to you that the lander starts with 100% fuel.

For those of you who did not participate in Optimal Descent, and are unfamiliar with how to use the attached persistance files, a brief explaination: Simply move your existing persistance file (found in the 'KSP/saves/default' folder) to another location for easy recovery, download mine and place it where yours was.

Upon starting the challenge, you are free to use whatever method you desire to reach the designated orbit. You may also land if you choose, to demonstrate the cumulative effect of your Insertion, but that is entirely optional. However, there will likely be Medals for certain extraordinary achievements. Positions on the Leaderboards will be determined by screenshot of fuel remaining (in kgs... Gubs, while amusing, just didn\'t feel right after the initial laugh was over). I chose the MCC1 location as the starting point to allow the widest variety of maneuvers while still focusing on a single aspect of the mission, so don\'t be afraid to think outside the box.

And for those of you wanting an extra level of difficulty, I have also included a lander with as little fuel as possible using stock parts. It is the second persistent.sfs file attached. I tried to create an identical free-retun trajectory for this lander and while it isn\'t perfect, it\'s pretty close (~34,000m higher from a target of 1Mil m). Be warned, it has a very tight fuel margin. And good luck getting back to Kerbin with that one! Mwahahaha! Seriously though, I barely made it to the surface. I would have made it in one piece too, but I ran out of fuel a few meters up and was going about 5-6 m/s when I landed but was bounced tens of meters back up. I just managed to right myself and save the capsule, so at least the crew made it alive, even if the rocket isn\'t intact and the tank is empty.

[NOTE]: I have made changes to the challenge\'s goal. This should now allow us to focus entirely on the Insertion phase. Leaderboard positions will be updated when new results are submitted.

Once again, Glory and Honor are there for the taking! Who will ultimately prove themself worthy of the top spot? Will it be you?

Standard Lander

1. PakledHostage: 202.9kg

2. Kosmo-not: 202.8kg

2. closette: 202.8kg

3. Tarmenius: 202.6kg

4. togfox: 201.1kg

Low-Fuel Lander

1. closette: 99.8kg

2. Tarmenius: 86.2kg

Landed

*In order of submission and unranked as it\'s optional*

Kosmo-not: 11.6kg (Low-Fuel Lander) / 109.4 (Standard)

closette: 111.9 (Standard)

REMINDER: The first attachment is the Standard Lander. Second is the Low-Fuel Lander

[PakledHostage]

Nice follow-on challenge to your last one.

I will have a go at this when I next get a chance to play the game. It might be a few days, though. I seem to manage to find more time for theorising about rocket science than I do for playing KSP… Probably because my mind drifts to it when I’m bored of doing what I should be doing... (A guy can only take so much of “it is broken, can you make it go?â€)

[closette]

Nicely set up, and it must have taken a fair amount of work. I almost missed it though - could you provide a link to this challenge from the munar descent one (perhaps put a link both in your first post and in the replies at the end)?

Did you make the 'low fuel lander' by starting with the original craft and dumping fuel while making the craft oscillate back and forth or spin around so there would be no net effect, or did you have to start over? (Just curious - I\'ll be sticking with the full fuel version!).

We should see some creative solutions to this one...

[Tarmenius]

Thanks guys. PakledHostage, I know how that goes. A few years back, I was working for a small manufacturing company doing very repetitive tasks (operating press-brakes, assembling parts, etc), and while some of my co-workers saw the repitition as a drawback, it allowed me to think about and come up with some cool stuff. Without letting it distract me, of course. Anyways, no hurry on getting to this; I don\'t plan on going anywhere

closette, the Low-Fuel Lander is basically the same craft, substituting the FL-T250 half-tank for the Mk1 Fuselage. This design won\'t work when jet fuel is separated from rocket fuel, though. So when that happens, hopefully we\'ll have better options for creating this type of scenario. And while the fuel budget is tight, it should be doable for you. You only used 108 kgs of fuel in the last challenge, and the Low-Fuel Lander has a capacity of 150. I botched my own attempt, but I\'ll get it. And so can you... if you want to.

[Zephram Kerman]

... (A guy can only take so much of “it is broken, can you make it go?â€)

Roy: Hullo, I.T., have you tried turning it off and turning it back on again?

Reginod: It\'s broken.

Roy: Is it plugged in?

Reginod: Can you make it go?

Roy: Are you sure it\'s really plugged in all the way?

Reginod: ... He\'s smart.

Excellent challenge. I appreciate these purposeful explorations of the questions on everyone\'s minds. By 'crowdsourcing' these questions, we can establish the best answer with confidence, and have a place to refer people when they ask. I\'ll pitch in a few attempts as soon as I can.

[Tarmenius]

Thanks Zephram, that\'s exactly the purpose I had in mind when I made the last one. Since it grew into this one, I\'m thinking I may compile the results from these two plus Optimal Ascent (With PakledHostage\'s assent ) into one comprehensive picture of the most efficient Mun mission.

[EDIT]: After a couple more attempts with the Low-Fuel Lander, I still can\'t manage to land with any fuel left. I\'m very close though. The most recent attempt got me to the surface with only the engine destroyed after again running out of fuel meters above the ground. I\'ve tried two methods: drop Pe to 2500m upon entering the Mun\'s SOI to descend from there, and establish a 10km orbit before dropping Pe a little below the surface, giving me a long path at very low altitude so I could find a good landing site. I think next time I\'ll establish a 10km orbit, then drop Pe to 2500m (saving a little fuel) and land as usual. Hopefully the little fuel I save will make the difference in allowing me to land safely.

[BlazingAngel665]

Definetly on to this one, but I still have to reclaim my title in the previous challenge. I will take the kerbals home again. . . somehow (you need another medal for those of us who manage to avoid killing the intrepid crew and/or stranding them in space with little to no fuel.

[Tarmenius]

By all means, feel free to reclaim it. Feel free to try it without MechJeb, too! And if anyone can manage a return trip to Kerbin, I will definitely think of the most epic medal I can and award it to them. Hell, for such an incredible feat I may have to mess around with MS Paint and create an actual image of the medal, too.

[EDIT]: Although it wasn\'t a better result than my previous attempt, while landing the Standard version I did manage a nice view of the Mun Arch.

[closette]

You set a tough standard to beat, Tarmenius. After a few crashes on the night side where I couldn\'t see the ground (thanks a lot!) and also one fly past the Arch, here is my easily beaten attempt: 90.7kg.

The KSP 'sunrise' is rendered so well it hurt my eyes a bit to look at it. Well that\'s my excuse for the hard landing.

Screenshots attached. And yes I borrowed a USB mouse just so I could right-click and get fuel state!

I am including a map view screenshot as well so others can see what I am doing. My simple approach is:

- coast into the Mun\'s SOI

- burn towards Kerbin/KSC which is roughly at right angles to velocity, to reduce periapsis to about 10 km

- coast a short while, then retroburn to reduce periapsis to around 5km. I do this also to make the periapsis closer to the daytime side of the Mun, not very successfully

- gravity turn from periapsis and try not to hit the ~1400m ground too hard.

Next time I think I\'ll circularize at periapsis instead, then orbit around to the daytime side to pick my spot, but for attempts so far I\'ve also been trying to land in the shortest possible time, which means in the dark.

[Tarmenius]

(thanks a lot!)

You\'re welcome.

Out of curiosity, once inside the Mun\'s SOI, have you found it more efficient to burn toward Kerbin instead of strictly retrograde? Also, if your orbit is 10km by 5km where your Pe falls just on the night side of the terminator, I imagine there wouldn\'t be much difference in velocity or altitude if you were to wait until the day side to begin your descent. I\'ll have to try it to be sure, though.

Congrats on getting around the no-right-clicking-on-Macs issue!

[Kosmo-not]

I decided to use the low fuel lander for my first go. Landed with 11.6kg remaining. I retro-burned at the start of the scenario to get my periapsis down to the surface. After entering SOI, I burned radially to bring periapsis to 2km. Approaching about 3.5km, I started my landing burn and noticed I was heading right for a mountain. I lost some fuel to avoiding that. It\'s so hard to see on the dark side of the Mun.

*edit*

I went ahead and did it with the other lander that has more fuel. I got down with 109.4kg remaining (second screenshot).

[closette]

Regarding lowering the hyperbola\'s periapsis, I was following the advice given me at the end of the descent challenge (which I could verify by calculation if I wasn\'t feeling lazy) to burn at right angles to velocity to 'turn into' the Mun instead of just retrograde to reduce orbital energy.

At large distances there does not seem to be a lot of difference to the initial delta-v required. The 'turn' method does add to the orbital energy, and therefore periapsis velocity, which seems less desirable, so if someone has done the calculations please share them. Otherwise I\'ll break out the pen and paper myself, and or try both experimentally starting with the persistent.sfs file for this challenge.

[PakledHostage]

At large distances there does not seem to be a lot of difference to the initial delta-v required. The 'turn' method does add to the orbital energy, and therefore periapsis velocity, which seems less desirable, so if someone has done the calculations please share them. Otherwise I\'ll break out the pen and paper myself, and or try both experimentally starting with the persistent.sfs file for this challenge.

This sounds interesting... Are you saying that burning retro-grade to lower Pe to 5 km and then landing required roughly the same Delta-V as landing after lowering the Pe to 5 km using the 'turn' method? I haven\'t had a chance to make an attempt at this challenge myself yet, but I\'ll have to make this comparison myself.

If it is done correctly, the turn method shouldn\'t add any orbital energy in the same way that turning the handle bars on your bike doesn\'t add energy. Sure you\'ll reach Pe at higher speed, but that is because you have converted potential energy into kinetic energy. The overall orbital energy is the same. Kosmo-not and I discussed this back in April.

If the two methods result in roughly the same fuel burn requirements, then I would assume that the only benefit to the 'turn' method would be that it allows us to take greater advantage of the Oberth effect at Pe. I guess the question then would be whether or not the improved efficiency of the higher velocity burn is enough to offset the delta-V 'wasted' on turning (rather than reducing orbital energy). That, in turn, would depend (I\'m sure) on the geometry of the initial Munar transfer orbit...

[Tarmenius]

After a couple tests with the Standard Lander to measure the differences between burning toward Kerbin and burning Retrograde, here\'s what I found (both burns were initiated just inside the Mun\'s SOI):

While burning toward Kerbin, I set my Pe to 2994m (aiming for 3000). This used 21.9kg and resulted in an orbital velocity at Pe of 852.3 m/s.

While burning Retrograde, I set my Pe to 3038 (again aiming for 3000). This used 40.5kg and resulted in an orbital velocity at Pe of 784.6 m/s.

Interestingly the Pe\'s were about equal distances from the terminator, with the Kerbinward on the night side and the Prograde on the day side. Sadly, my poor descent control didn\'t allow further comparison of fuel use.

[PakledHostage]

Interesting. What happens if you make the 'turn' immediately after starting the scenario? I haven\'t tried it myself, but intuitively you shouldn\'t have to wait until you\'ve crossed the Mun\'s SOI to start the turn. Also, what happens if we just insert into a 3 km x 3 km circular munar orbit rather than landing? For both capture techniques, you need to pass through a 3 km x 3 km circular orbit in order to land. The fuel burn from there to the surface would be the same for both methods.

[Tarmenius]

Good suggestions. I particularly like the idea of a 3km x 3km orbit; it takes the landing variable out of the equation. I\'ll go run a couple more tests right now.

[closette]

While burning toward Kerbin, I set my Pe to 2994m (aiming for 3000). This used 21.9kg and resulted in an orbital velocity at Pe of 852.3 m/s.

While burning Retrograde, I set my Pe to 3038 (again aiming for 3000). This used 40.5kg and resulted in an orbital velocity at Pe of 784.6 m/s.

Thank you for trying this. Since IIRC 1 kg of fuel causes a delta-V of about 6.5 m/s for this fully fueled craft, in terms of delta-V and fuel to achieve a 3km circular orbit, I think the 'turn method, i.e. burning off-axis ~towards Kerbin wins despite the higher periapsis velocity.

I just chose 'towards Kerbin' as an easy way of pointing almost at 90-degrees to velocity in the orbital plane, by the way. The results of this challenge will be very applicable to new users trying their first mission, so much like the approximate 'TMI burn at Munrise' I thought this could be a good rule-of-thumb. We\'ll see how good compared to the optimal results on the leaderboard.

The actual optimum impulse direction to lower periapsis is probably some hard-to-calculate angle depending on how far out you are. (This also comes up when trying to return to Kerbin with a low perikerb after leaving the Mun\'s SOI).

@PakledHostage, I know what you mean, but applying an impulse at exactly 90 degrees to velocity must increase orbital energy somewhat, since the resulting velocity is the hypotenuse of a right triangle, therefore higher than the original, but if the off-axis delta-V is small in comparison, it won\'t make much difference. Thanks for the reference for your and Kosmo-not\'s very relevant earlier discussion.

[Kosmo-not]

@PakledHostage, I know what you mean, but applying an impulse at exactly 90 degrees to velocity must increase orbital energy somewhat, since the resulting velocity is the hypotenuse of a right triangle, therefore higher than the original, but if the off-axis delta-V is small in comparison, it won\'t make much difference. Thanks for the reference for your and Kosmo-not\'s very relevant earlier discussion.

If you keep the craft pointing 90° from the velocity vector, it will have the same effect as adding a circular arc to your initial velocity vector thus not increasing velocity.

[closette]

Urm, not sure how I did this or whether it should qualify: 116.1385 kg remaining, but fuel tank detached on landing!

I basically tried to reduce periapsis from outside the Mun\'s SOI first, adjusted after entering (since the patched conic seems to overpredict periapsis for this case), coasted to periapsis of about 6km altitude, tried to establish a circular orbit, but ended up doing a flat-ish gravity turn to the sunny side of the Mun. I lost control just before landing, but the quicksave file lists 116.1385 kg for the tank belonging to 'Munar Landing Challenge Debris'. Apparently right-click doesn\'t work for separated tanks.

I\'ll try again later tonight - real life is pulling me away from this challenge!

I don\'t mind being disqualified from the Leaderboard, but at least this shows that you can end up with more than 110 kg remaining if you are careful - more careful than me!

[Tarmenius]

you meant 'Retrograde'

Yes, I did. Thank you for pointing that out; it\'s been corrected.

So, thinking on the suggestion PakledHostage made, it has prompted me to want a change in the parameters of this challenge, as much as I\'d hate to do it after it\'s been started. I just realized that the way it\'s set up currently does not isolate the Insertion phase. Instead of judging by fuel remaining at landing, I think it would work better to establish a standard orbit altitude and determine who can reach it on the least amount of fuel. Since closette and Kosmo-not have already submitted entries under the current rules, I\'d like their input on whether to make the change at all. And I\'d like input from everyone on what altitude the standard orbit should be. Personally, I like the 3km x 3km.

Now, on to some more test results. First, I expanded on my original set from a couple posts back to include establishing a 3km x 3km orbit. Again, I performed the initial burns from just inside the Mun\'s SOI (both at roughly 2,087,000m). Here are the results:

• 
  [li]Burning Kerbinward, I set my Pe to 2950m using 21.9kg (again). Pe changed to 3310 by the time I got there (?), with an orbital velocity of 852 m/s (again). Setting my orbit to 2900m x 3050m used 40.7kg, with a total fuel consumed of 62.6kg (187.4 remaining)[/li]
  [li]Burning Retrograde, I set my Pe to 3030m using 41.6kg. Pe changed to 3050, with an orbital velocity of 783.3 m/s. Setting my orbit to 2820m x 2730m used 30.1kg, with a total fuel consumed of 71.7kg (178.3 remaining)[/li]
  

Next I performed the same maneuvers (kerbinward/retrograde) from the scenario\'s starting point. Here are those results:

• 
  [li]Burning Kerbinward, I set my Pe to 2700m (Projected by Patched Conics) using 5.0kg. Pe changed to impact by the time I entered the Mun\'s SOI so I adjusted it back (burning directly away from Kerbin) to 3800, using 0.2kg. By Pe, I had an orbital velocity of 852 m/s (again). Game froze at that point, but given the identical velocity I assumed the same fuel use from above (40.7kg). Total fuel consumed: 45.9kg (204.1 remaining)[/li]
  [li]Burning Retrograde, I set my Pe to 2500m (Projected by Patched Conics) using 4.6kg. Pe changed to 1200m once inside the Mun\'s SOI so I adjusted it back to 3010m (burning Prograde), using 0.3kg. By Pe, I had an orbital velocity of 859 m/s. Setting my orbit to 2750m x 2940m used 42.5kg, with a total fuel consumed of 47.4kg (202.6 remaining)[/li]
  

Hopefully that\'s clear enough. If not, I\'ll ask some people I know if they have a spreadsheet program I can use. In summary, burning toward Kerbin proved to be more efficient and even more so from the scenario\'s starting point. I\'m sure Kosmo-not already knew that, though

[Kosmo-not]

I think it\'s a great idea to have a 3x3km ending orbit since the landing portion has already been covered in the other challenge.

[Tarmenius]

Thanks. I changed to original post to show the new goal, but kept your positions on the Leaderboards until we get new results. Now, I\'m off to get some screenshots that reflect the new goal as well.

[PakledHostage]

Well, coincidentally, I just completed this challenge using a 3.3 km x 3.6 km pre-landing orbit. I don\'t know if that is sufficiently within tolerances to meet the new objectives but I\'ll submit it anyway.

Regardless of whether or not it counts, I think my results are interesting. Immediately after starting the scenario, I used my spacecraft and the navball to measure the angle between the Mun and Kerbin. I then made my burn to lower my Munar periapsis by burning while oriented 1/2 way between the two (roughly 25 degrees elevation in the navball\'s brown hemisphere, on a heading of 270 degrees). It cost 6 kg fuel (39.1 m/s delta-V) to lower my periapsis to 4 km.

I then waited until reaching periapsis to close down my orbit. I pre-calculated the orbital speed required for the 4 km orbit and burned at full throttle until reaching that speed. My imprecise flying resulted in a 3.3 km by 3.6 km orbit about the Mun. Once inserted into that orbit, I had 200.5kg fuel remaining (net delta-V expended to this point: 330.9 m/s. This includes the 39.1 m/s expended to lower the Pe).

From that orbit, I made my DOI burn then coasted along until I came across some suitably high terrain. I had never seen the Mun arch before, so I was quite surprised to see it coming over the horizon. I quickly pressed F5 to save my position then proceeded to violate my own prohibition on 'do-overs' to make my landing with minimal fuel burn, somewhere near the arch. I landed with 114.2 kg fuel remaining after several attempts (net delta-V expended 958.3 m/s). The arch is in the background of the screenshot.

I will leave it up to the judges to decide whether to count this entry or not.

[closette]

OK to change it by me, but I WAS in the lead for a moment with my hard landing (but intact fuel tank)...such brief but sweet glory which I shall never experience again (Sobs quietly).

3x3km seems a bit 'tight' for a final orbit and 5x5km would be better for time-warping, but I\'ll go with the flow, as long as there is some tolerance such as +/- 0.5 km on each, althout 2.5km is probably too low.

This is already turning (pun intended) into a fascinating challenge, with applicability to many missions.

By the way I came across this online calculator for going from orbital state to orbital parameters and back:

http://orbitsimulator.com/formulas/OrbitalElements.html. Use M=9.76e20 kg for the Mun\'s mass, 1kg for the spacecraft mass, and set the 'Z' components all to zero for in-plane maneuvers.

[Tarmenius]

PakledHostage, I\'d definitely say that counts. Have you tried landing the Low-Fuel Lander yet? My next goal is to take what I learned from my tests and try that one again.

closette, don\'t count yourself out just yet. Now that landings no longer have to factor in the Leaderboards, I\'d say it\'s anyone\'s game.

Having made several attempts at a 3x3km orbit, often times botching it terribly, I realize how slim the margin of error is. Incidentally, the best way I found to pull it off is to start circularizing a good 30 seconds before reaching Pe, pitching above the horizon slightly if needed. You\'ll end up having to separate the process into a few short burns to keep Pe from running away from you near the end, though. That\'s how I finally narrowed down the 2900 x 3050 orbit.

Either way, I think a tolerance of +/- 1km is reasonable, for as difficult as such a low orbit is to achieve. And if someone is really having a hard time making it, I\'m not against making exceptions either. I really want this to be more of a challenge and less of a competition. And given the nature of this forum\'s members I\'d wager almost everyone would be okay with that.