Inspiration Kerbal

[Jasonden]

"Inspired" by Dennis Tito's proposed Inspiration Mars, I thought that we might try something similar in KSP. The Inspiration Mars people propose a crewed mission with 2 astronauts to fly by Mars on free-return trajectories. There's a very nice (and quite readable to someone with KSP knowledge) paper on the concept. The idea here is to just use either a Space-X Dragon or a NASA Orion capsule with attached hab space. This mission wouldn't do any science, but would get humans outside of Earth's SOI for the first time and provide an amazing test cruise for the Orion vehicle. Here's an artist's conception of the mission vehicle:

and another:

The original trajectory launches in 2018 and would be a ~500 day flight with a 150km closest approach (C/A) to Mars.

That 2018 launch is looking nearly impossible, though. And the next known free-return trajectory to Mars and back isn't until 2033. But it turns out there's a 589-day free-return trajectory that includes a Venus gravity-assist first on the way to Mars, then back to Earth. Sweet!

Inspired by a thread in add-on requests, I used the KSP Trajectory Optimization Tool (caution: awesome) and discovered at least one free-return trajectory to a planet (Eve in this case) in KSP:

With this existence proof that such things are possible, I'd like to propose an interplanetary free-return challenge. This is going to be a really hard challenge, I think -- the hard part is going to be discovering the free-return trajectories, though, not in designing and executing the mission.

RULES

Fly a mission with 2 Kerbals on an interplanetary free-return trajectory, passing by another planet in the Kerbol system and returning those 2 Kerbals safely back to the surface of Kerbin.

Requirements:

1. Spacecraft must consist of a Mk1-2 capsule and 2 total Kerbals (no lander cans, command seats, or Kerbals on ladders)
   
2. Spacecraft must include some sort of additional hab space, at the challenger's discretion. I recommend maybe a Hitchhiker module, Science Lab, or one of Porkjet's inflatables.
   
3. Since the idea is that NASA or someone might do this in the near-term future, no jets or nukes allowed. Solids, liquids, ions, and asparagus are all kosher (one of the earlier plans was to use a Falcon Heavy which is asparagus)
   
4. All rocket engines, including ion engines, must be separated prior to leaving Kerbin's SOI on departure. ALL COURSE CORRECTIONS EN ROUTE MUST USE MONOPROP ONLY!
   
5. No cheating, hyperedit, or unbalanced rocket part mods (ask if you're unsure -- using any mod for your 'hab' space is always okay). Mods precleared: KW, NovaSpace, StretchySRB
   
6. All information and autopilot mods (including MechJeb) are okay. NASA is not going to fly this mission by the seat of their pants without a computer and just hope for the best.
   

As is apparently the norm with me, I have a thorough and complex scoring system. Mostly this challenge is about an entirely new type of mission in KSP and what kinds of free-returns exist out there. But in order to put some competitive fire into people, your final score will be:

That first term is a SUM over all of the non-Kerbin planets that you visit. So if you do an equivalent of the Inspiration Mars Venus-Mars trajectory, your values for both Eve and Duna will be added together.

Let me try to score out an example based on that Inspiration Mars 2018 trajectory. They have a Mars periapsis of 100km, so their score on the first term is 5000*1*(5-log_10(100))=5000*(5-2)=15000. In the paper they describe the departure in terms of C3, which is the specific kinetic energy outside Earth's SOI. Put another way, once you're out of Earth's SOI, it's just v^2. Representing the Earth departure in this way allows you to remove the Oberth effect from the equation. Anyway, with some rough assumptions I'm estimating that this corresponds to a dV of 5650 m/s (let me know if I did that math wrong -- you won't have to do that for your mission, of course, provided you have MechJeb or KER). So subtract that from 15000 and the score is 9350 so far. Then subtract off 3 times the travel time in days, which is 501 daays, leaving Inspiration Mars at 7847. Okay, I can't score their use of monoprop, so I'll assume that they had a perfect initial injection and didn't need any. The total mass of an SLS is in the vicinity of 1500t, leaving their score at 6347. That's the one to beat! I have a trajectory in mind for my career game out in 2022 to Eve that I'll be working on soon to post.

LEADERBOARD

1. von Ziegendorf (Eve/Moho/Moho/Moho/Eve/Eve/Eve/Duna/Eve) -- 146205
   
2. von Ziegendorf (Eve/Duna/Eve) -- 47884
   
3. PLAD (Moho) -- 45054
   
4. metaphor (Duna) -- 29705
   
5. PLAD (Duna) -- 24783
   
6. kdonfede (Dres) -- 19391
   
7. Jasonden (Eve) -- 6730
   
8. --
   
9. --
   
10. --
    

Edited March 18, 2014 by Jasonden

[Nahim]

thats seems realy hard. i've just landed my first Duna probe... so i think i gonna be here just as a listeneer.

[Mr.Rocket]

Hmm.... I think i'll try this this weekend. And we don't have to do a flyby of Duna, it can be any planet?

Edit: Is docking before departure okay?

Edited March 7, 2014 by Mr.Rocket

[Jasonden]

Hmm.... I think i'll try this this weekend. And we don't have to do a flyby of Duna, it can be any planet?

Right -- I looked really hard and didn't find any free-returns to Duna over the first 25 years of game time, but the search wasn't THAT thorough. So yup, any planet. I'll bet that Jool free-returns are no big deal, but I haven't tried that yet!

Edit: Is docking before departure okay?

Sure. Total launch mass in that case will be the sum of all vehicles launched.

[ihtoit]

that's gonna be like firing a gun at a bullet that's speeding toward you, firing over your shoulder blindfold on a mechanical bull set to 9.

[Jasonden]

that's gonna be like firing a gun at a bullet that's speeding toward you, firing over your shoulder blindfold on a mechanical bull set to 9.

Actually, once you know that there's a trajectory there to find (from KSP Trajectory Optimization Tool), it's not too bad. Remember that you can make course corrections as you go -- they just have to be using monoprop.

[Jasonden]

Okay, here's my entry. It's an Eve free-return in year 22 (in my career game):

Javascript is disabled. View full album

As for the score, my C/A to Eve was 1528km, total time 118 days, 1627 m/s dV after parking orbit, and 17 units of monoprop consumed.

So my total score would be 8943-1627-118*3-17-215=6730.

[metaphor]

Inspiration Mars's 2018 ejection delta-v is 4.86 km/s.

I tried to do an Inspiration-Mars styled mission some time ago, but it didn't quite work out. I think Duna in KSP just doesn't have the mass required to bend a spacecraft's trajectory back towards Kerbin fast enough. Maybe if combined with an Eve flyby.

[PLAD]

Here's my opener. First the pictures:

Javascript is disabled. View full album

I feel I might be violating the spirit of the challenge with that 228 m/s burn even though I did it using only monopropellant. In my defense I give the example of the way Cassini was flown to Saturn with it's VVEJ path. On 3 Dec 1998 it made a "Venus Targeting Manoeuvre" of 452 m/s to allow its 2nd flyby of Venus. Since it was in deep space at the time the low thrust didn't matter.

There is free return path to Duna available every time Kerbin approaches Duna. You just go for a 2-Kerbin-year orbit that will bring you right back to Kerbin in 213 days. Just time it so you pass Duna at one of the 2 points you cross its orbit. If you make the flyby at a far enough distance Duna won't affect your path enough to matter. For a better flyby you can then tweak the path to go closer to Duna and correct the timing to get you back to Kerbin. If you allow a fair-sized burn at ApoKerbol there is a large window of opportunity. It's easy with Duna because it's orbit is coplaner (close enough) to Kerbin's and has a small enough SMA. Eve is much tougher because of it's high inclination. I think your flyby of Eve was much more interesting than this flyby of Duna. However, the scoring system heavily favors Duna. I think this gets: B=2 (Duna), P=192.214 km, DV=1252 m/s, D=225 days (or 215?) U=270 units, M=182 tons. Score 24783?

I love the complicated scoring systems, they make the analysis more interesting. I bet there's a 60000 point path to Duna/Dres out there.

[donfede]

Nice show metaphor and PLAD, thanks for sharing the pics.

PLAD, I don't see any "dates or times" on your pics, though admittedly I'm not familiar with the mods used (KER/mechjeb), so maybe I missed them [Kerbal Alarm Clock user here ]. Those would be useful to others interested in repeating the mission. (similarly, it would be curious to see resources on the pics, if you happen to do another run).

I've enjoyed reading the info on Inspiration Mars mission, and am currently trying to make my way through some of the papers referenced for that mission...

Tito "Feasibility Analysis for a Manned Mars Free Return Mission in 2018" paper referenced by Jasonden above

Patel "Mars Free Return Trajectories" with detail on Free Return Trajectory types and some equations; I've yet to try applying any of these to KSP

p3 Free Return sub-classes

* cycler , visit 1, visit2

* cycler, up/down escalator

* but these have too long Time Of Flight

p3 mentions Patel sw to find FR; ref11 "Automated design of gravity-assist trajectories to Mars ... - Springer. ... Automated Design of Delta-V Gravity-assist Trajectories for Solar ... Moonish R. Patel, Purdue University, 1993"

p8 collision orbits def - passing a body twice (at the same or different locations) - good approach for FRT

p8 Henon, Howell, and others defined methods for finding collision orbs and points; using Lambert alg (pork chops), can find for many orb types.

Edited March 16, 2014 by donfede
added notes from papers

[Jasonden]

metaphor & PLAD:

Nice job guys! I love it! metaphor, I estimated some values for you like your initial launch mass, for instance. Let me know if you have more precise values.

PLAD -- yeah, I realized that my exception for monoprop maneuvers could be a giant loophole, and I decided that I was okay with it. A few hundred m/s isn't too bad, at any rate. I did get the Trajectory Optimization Tool to generate a Duna free-return that had like a 250 m/s burn at Duna periapsis at one point -- maybe I should try to find that again

[PLAD]

donfede- good idea there, I've installed Alarm Clock to give a clear indicator of the UT of events. My Duna mission had an MET of 0 at UTY1, D50, H21, M46; and it started the trans-Duna boost from LKO at UTY1:61:18:52. Is there an agreed-upon way to indicate the KSP UT of an event? Or is this OK?

Metaphor's Duna mission had a nice short flight time and close flyby- much superior for a mission carrying living things.

There might to be a 'cycler' mission possible for Duna. 2 Kerbin years is 213 days, and the Synodic period between Kerbin and Duna is 227 days. If we can stretch a 2-year conjunction class mission by 14 days by raising the apiKerbol and dropping the periKerbol enough to cross Kerbin's orbit twice... but Duna might not have the mass to pull it off...

[PLAD]

I noticed that a transfer orbit to Moho has a period of about 67 days or so. 3 times that isn't much less than 2 Kerbin years, so if one could pick up a little energy from Moho there might be a 2-Kerbin-year round trip possible. And here it is.

Javascript is disabled. View full album

I was laughing like a nut during the flyby, for an instant Mechjeb's 'altitude above surface' was less than 300 meters. At 4.4 km/s. I wasn't quick enough to catch that or the KSP indicated periapsis of 1594 but I'm happy with the 1602 meters I recorded. I needed every m/s I could get from Moho so the low pass was necessary, but no sane pilot would do this. And yes, I used F5 before my first try but we shall not speak further of why an F9 was needed. B=2 for Moho, Periapsis=1.602km, U=303 units, D=226 days, M=182 tons, LKO dV=1623 or 1746 if you count that correction out past Minmus. Score 45044?

Now on to multiple flyby missions.

[von Ziegendorf]

Hi, this challenge motivated me to fly a mission that I planned almost a year ago

Triple Flyby

1st Eve Pe= 1630km - 8939 points

Duna Pe= 42km - 33768 points

2nd Eve Pe= 3235km - 7451 points

dV= 1323m/s + 7m/s

D= 256d

U= 74

M= 102t

Total score = 47884 points

[PLAD]

von Ziegendorf- Um Gottes willen!!

3 flybys and return is spectacular by itself, but you did it with only 88 m/s (74 units)of rcs dv! That's a real free return. Could you tell me what KSP UT day the mission launched? I want to try that.

[von Ziegendorf]

Thanks With a little more care it could be less RCS

Launch was at 164d:2h:50m (14179800s) but the calculated parking orbit inclination is a bit off, you will need to burn some m/s also in the normal direction, not only prograde

[donfede]

Hi, this challenge motivated me to fly a mission that I planned almost a year ago

...

Thanks With a little more care it could be less RCS

Launch was at 164d:2h:50m (14179800s) but the calculated parking orbit inclination is a bit off, you will need to burn some m/s also in the normal direction, not only prograde

Thanks for sharing the mission details von Z! You mentioned the long planning for this mission. Can you provide more details on the planning and how you came across these orbits?

I would figure that there must be a way to update/alter the porkchop plotter to quickly identify these orbits, but am still reading and trying to grasp these concepts.

kdonfede

[donfede]

Hi, this challenge motivated me to fly a mission that I planned almost a year ago

Triple Flyby

1st Eve Pe= 1630km - 8939 points

Duna Pe= 42km - 33768 points

2nd Eve Pe= 3235km - 7451 points

dV= 1323m/s + 7m/s

D= 256d

U= 74

M= 102t

Total score = 47884 points

How did you fit the quad-coupler above the orange tank with the mainsail? Impressively small vessel (without minmaxing).

[von Ziegendorf]

I came across this: Triple flyby info, there is a drawing of ship's orbit around the Sun. I assumed (hoped ) that Kerbol system is well scaled model of the real one. Then I took all the angles from the drawing and calculated phase angles at launch (Eve +18, Duna -139). Also I roughly calculated KSP trajectories that fit those angles.

At the beginning of the game all planets (except for Jool) are almost at their Ap. Knowing orbital elements (LPe) it's easy to tell where any planet is at any date. So using my assumption I moved phase angle values directly to Kerbol system and found the nearest launch window at 172d.

During a test flight with a probe I couldn't catch Eve for the second time (only went for a double flyby). It seemed I was about a week late, that's why 164d So corrected phase angles are now Eve +2, Duna -124.

The inclination was set to allow only prograde burning and performing ejection burn and a (Kerbol orbit) plane change at the same time. So no need for deep space manouvers to make this first encounter with Eve. But today I was flying using my results that are year old. Maybe it was inclination for that 172d launch? I don't remember

About quad coupler? I was suprised it fits there when I built that stage for the first time But apparently it can be placed there

Edited March 14, 2014 by von Ziegendorf

[Jasonden]

I noticed that a transfer orbit to Moho has a period of about 67 days or so.

I was laughing like a nut during the flyby, for an instant Mechjeb's 'altitude above surface' was less than 300 meters. At 4.4 km/s. I wasn't quick enough to catch that or the KSP indicated periapsis of 1594 but I'm happy with the 1602 meters I recorded.

Holy crap, this is awesome! Your Kerbals look scared at your Moho closest-approach -- and for once I agree with them that they should be Good idea that your peri for airless bodies can necessarily be lower than for Eve or Duna!

[Jasonden]

Hi, this challenge motivated me to fly a mission that I planned almost a year ago

Triple Flyby

1st Eve Pe= 1630km - 8939 points

Duna Pe= 42km - 33768 points

2nd Eve Pe= 3235km - 7451 points

Wow! This is awesome! I am so pleased that people are finding trajectories that I hadn't even realized were possible!!!! I really enjoyed the Wired article, too. I'd say this is hard to top, but I would have said that about PLAD's Moho run if this hadn't already been posted

[von Ziegendorf]

Please don't faint, I couldn't resist The penalty for mission duration should have been non-linear function of time (propably exponential or so) to discourage such flights that are against the spirit of the challenge: Flyby Frenzy

1st Eve - 11142, 1st Moho - 40362, 2nd Moho - 26421, 3rd Moho - 22893, 2nd Eve - 8155, 3rd Eve - 1107, 4th Eve - 10441, Duna - 20397, 5th Eve - 7716

dV = 1461m/s, D = 560d, U = 563units, M = 185t

Score = 146205

Mission launched at 87789000s ~2y 286d

//edit

What I found out is that the Triple Flyby launch window should repeat every ~860 days starting from day 164

Edited March 15, 2014 by von Ziegendorf

[donfede]

...Is there an agreed-upon way to indicate the KSP UT of an event? ...

I typically note it the way that KAC does - year $X, day $Y, $HH:MM. Reading through alex moon's porkchop plotter, he uses a similar format. File porkchop.coffee ~line 52 , function kerbalDateString returns :

 52   "Year #{year}, day #{day} at #{hourMinSec(t)}"

Alternately seconds into the epoch could be used; though more precise, it's typically less human-friendly (at least KAC allows for easy switching between the time formats display).

[Jasonden]

Please don't faint, I couldn't resist

1st Eve - 11142, 1st Moho - 40362, 2nd Moho - 26421, 3rd Moho - 22893, 2nd Eve - 8155, 3rd Eve - 1107, 4th Eve - 10441, Duna - 20397, 5th Eve - 7716

dV = 1461m/s, D = 560d, U = 563units, M = 185t

Score = 146205

*wakes up from smelling salts*

Wow. I am totally blown away -- that is like flyby on crack. While shooting heroin. Skydiving. Most impressive.

I take your point about the flight duration -- I had originally planned on having harsher duration penalties, but was worried that there wouldn't be enough viable free-return trajectories at all. I was wrong! You guys are helping to show that there are a lot more of these out there than I had realized.

[PLAD]

Um.. I don't think von Ziegendorf should be called 'bottle rocketeer' anymore. Is 'Master Astrodynamacist' or maybe 'Gauss's Daddy' available?

I do suggest a scoring change though. How about in the event of multiple passes of a given planet, only the best of those passes is used? Thus you can only get flyby points once for each planet. A spectacular flight like the Flyby Frenzy would still win by a mile. And you could avoid brute force flights like what I'm looking at now, where one could fly by Moho and get into a 51.3 day orbit that keeps returning to Moho, you can keep the period at 51.3 days by flying by above and below Moho (as seen from the sun) so that you are only changing the inclination of the orbit rather than the SMA. Another 40000 points every 51 days until you run low on monopropellant, then take the next window home. It could get high points but is quite boring.

As a last word on Moho flybys, if your periapsis was to be at Lat 35S, Long 174W it could be as low as somewhere between 150 and 200 meters. I looked around for a place where the periapsis could be zero or lower, I was going to claim that the log of zero is minus infinity and thus request a score of infinity, but nowhere on Dres or Moho is low enough.