Caveman: The Making of Another Nanocrystalline Diamond

[Muetdhiver]

The trick with interplanetary is all about missing

Seriously, getting an encounter for anything but Duna is extremely hard and luck dependant. So, the whole trick is to miss, see how late/too early you are for an encounter, and correct the trajectory so that you catch up the planet after X orbit.

Say you are 30 days early, and the planet has a 600 day orbital period. Assume our orbital period is 300 days and AP is exactly at Jool orbit*. If you adjust your Pe (assuming jool) so that your period is 315 days, jool will catch up 15 days at each pass. After two orbits, you meet the target. If at each revolution one registers the target body passing at AP (assuming jool) and the ship passing at AP/PE, one can further adjust the orbital period to catch up.

*This can also be corrected, just read the altitude of the planet when it passes your AP. At your next passage of PE, correct AP to this altitude. Recompute the period and correct for the time differential.

Remember : T = 2*pi*sqrt(a^3/mu) where a^3 is the semi-major axis (AP+PE)/2 and mu is the standard grav parameter. You can get mu from kerbin orbit (which is circular, so easy to compute T with it).

I hope that helps, because if done right you get encounters with really good PE. Way better than direct transfer (IMO direct transfer is only viable for Duna in caveman, because it's inclination and exentricity are low).

For Jool here is how I did : I computed the period of an orbit for Jool and lauched 2*T before the "regular" launch window. I ajusted AP / inclination so that the orbit seemed to touch jool's. I noted the date of my arrival at AP. Then noted Jools passing at my AP and it's altitude and the time differential. When getting at my PE, I corrected for the AP to match Jool's altitude and noted the date. Once at AP once more, I noted the time, computed the time differential for the next orbit and corrected PE to get the time differential with jool for next orbit to zero. Once jool passes your AP you can do a second check. (but can't do much about it at this point). If things have been done right, you get a clean encounter at the next AP.

The orbital period need to be as accurate as possible. Minutes do matter, since a few minutes off means a much worse PE, hours off means massive PE and day off mean no encounter

Edited August 25, 2018 by Muetdhiver

[The Minmus Derp]

Cool!

[dvader]

@Muetdhiver That is the method I've been using for rendezvous around Kerbin up until now. I didn't want to do the same thing with Jool since the only reasonable intercept I could find required 25.2 years of waiting.  Even though my testing says it should work, I am a bit concerned Bob might have an early meeting with the Kraken if he floats in space for such a long time. I haven't tested the orbit calculations in KSP (and I don't think I will since it is 25 years of waiting at 100,000x), but my spreadsheet says the best option would be:

I launch from kerbin when Jool is about 22 degrees (=0.06 orbits) "behind" the ships major axis. I reach Ap the first time after 2.8 years, adjust my Pe hopefully just a tiny bit then go around four times more for a total of 25.2 years which is 4.5 ship orbits and 2.06 Jool orbits.

It is probably  the most accurate way of doing it (especially with corrections at Pe as well as you  suggested) but I am concerned about Bobs well being. Losing the "target lock" on Bob while orbiting would be a quick end to the mission. Testing says that won't happen but... nobody expects the Kraken (or the spanish inquisition).

Still, the more I think about it, I realise I probably have to do it that way. My other method of estimating the time to intercept by comparing screen distance ratio to velocity ratios works but requires repeated small adjustments which means I have to pick up Bob an unknown number of times which I might regret once there. Just getting him from Minmus to Mun was more annoying than expected because of all the small burns.

 

 

[Muetdhiver]

IIRC I was able to do the whole thing with 2.5 ship orbits (i.e. still 10+ years), Jool catching me nicely. It did cost a bit of fuel, but it was worth it.

Of course having bob flying arround does not help I suppose XD

But really, I don't think there is a better way, short of implementing a lambert solver to adjust the encounter after the sloppy caveman ejection burn. Which is not very caveman and way beyond pen&paper calculations.

Edited August 27, 2018 by Muetdhiver

[dvader]

@Muetdhiver That is so weird... I'm sure I calculated the 2.5 orbit solution the first thing I did but rejected it for some reason. But, now that I reopen my spreadsheet and just enter the numbers with the exact same formulas I used for the 4 ship orbit solution and, there it is: A perfectly valid solution leaving when Jool is 53 degrees behind and hitting Jool after just two ship orbits = 1.16 Jool orbits. That is, with a Jool orbit of 12.24 years and a ship orbit of 5.6 years: (1+53/360)*12.24 = 2.5 * 5.6 =14.04 years. I must have made some kind of mistake, mentally discarded the 2 orbit solution and then fixed the mistake before calculating the 4 orbit solution.

I was thinking of using Keplers Equation to calculate the time to Ap by measuring all relevant angles either directly on screen or from a screenshot but it seemed like a lot of work. Both to get the formulas right and to use it since the angles must be measured accurately. Now I don't have to

Thanks for making me check my calculations again! Both Bob and Val owe you 10 years of their lives.

 

[dvader]

The ship has been built and the time for mission launch is coming closer. I've visually examined the inclination of the Jool orbit and there seems to be a decent launch window in just 200 days or so. I'm gonna wait until Jool is at least 55 degrees behind the major axis and then launch.

Here's the build process (and some money missions and preparing to pick up Bob):

https://imgur.com/a/Zt78RDM

Highlights:

Edwin the Engineer: We named it the Ravens Nest because you're a cool flying pilot and it will be your home for the next few weeks or months. 

Val: Ravens Nest. Fitting name. It sure looks like it was built from twigs held together by saliva. Are you sure it won't shake apart once you start the engines?

Edwin: No, the self induced oscillations were due to the SAS. Just leave it off and you'll be fine.

Val: WHAT?

[dvader]

And we are on our way! Some things went well and others not so much.

In terms of real time, having Bob around is quite expensive. Normally, when a ship is in orbit you just go full throttle and you are on your way. This time, it took 1h 50m of real time from a fully assembled ship to having Bob on the ladder. Then it took 3 ship orbits to eject because I didn't pick him up in time for the ejection burn and had to try again.

The good news is that all the space trash added a lot more dV than expected. I got to an Ap of 8.5Mm and still had trash left to burn. It must have been at least 800 dV.

The not so good news is that my central trash engine drained some of the fuel from the late stage central tanks. I discovered it pretty quickly and didn't lose too much but it could have ended the mission if I had not noticed. I had turned off crossfeed on one of the docking ports to prevent this sort of thing but it seems crossfeed might be one directional or you have to turn it off on both ports. I don't know which. I think it cost me 100-200 dV so it is not mission threatening, just annoying.

The other not so good news is that my ejection angle was too small and I ended up with Jool being less than 55 degrees behind me. During testing I always got a larger angle than I wanted but maybe I was just more carefull with this burn. It means I have to burn retrograde at Ap to go faster and catch up with Jool. Orbital energy calculations suggest I need 300 dV in the "wrong" direction so if I get a really bad Jool intercept, I will need another 300 dV extra to achieve orbit.

After some thinking, I decided to take it slow and go for 2 Jool orbits and 4 ship orbits meaning I only have to do a 80 dV retrograde burn. The ship arrived at Ap one orbit later exactly when it should (4 minutes difference over 5.5 years is pretty impressive) but... Bob is 1.2 Mm away. It would still be a pretty good Jool hit but I would like it to be better. Maybe I can reduce his drift by stabilizing the ship even more before I let go. Or maybe small compensating burns during orbit... I don't know but I have one more orbit to go before the final orbit  corrections so I still have time to practice.

Another annoying thing is that x100000 warp speed isn't 100,000 faster. The ship has an orbital period of about 50M seconds. At x100000 that should be 8 minutes so 4 minutes for half an orbit. I set a timer at 4 minutes and left but when I returned, I had almost gone one full orbit and not just one half. So, note to self: Don't leave the computer unattended during x100000 warp.

Full story:

https://imgur.com/a/2Tze39I

Not so many highlights. I really spend most of the time intercepting with Bob, calculating stuff in a spreadsheet, doing small corrections and waiting. Here's Bob at 71 Gm achieving a new altitude record for EVA cavemen. I have a video of parts of it which I might post.

 

[JAFO]

Awesome!

1 hour ago, dvader said:

I have a video of parts of it which I might post.

Please do!

[Mr. Peabody]

This promises to be quite interesting. Great job so far!

[dvader]

8 hours ago, JAFO said:

Awesome!

Please do!

Done! I couldn't figure out how to concatenate or speed up using those fancy GUI tools but good old ffmpeg could. I am not a youtuber and have no ambitions to become one so please excuse any amateurishness.

It is a concatenation of the first two sessions at 8x the speed, about 2h 50min real time. It ends a bit prematurely when I press the keypad minus button to zoom out since it is also tied to "stop recording"...

I didn't bother recording the next session since it was mostly updating spreadsheets and doing other stuff while waiting for the ship to go around its orbit.

I also have a plan for checking Bobs drift before going for the next long time warp. His Ap is almost the same as the ships but the Pe is about 600 km lower. The difference in orbital period is about 563 seconds which at a speed of 2200 m/s is 1.2 Mm so that explains the full drift. So, I can probably check the efficiency of any drift reducing method by just checking Bobs Ap/Pe after he lets go of the ladder.

But first, I have to catch him again.

 

 

 

[dvader]

Making videos was easier than expected so here's Part 2. It could be the most boring video on the internet with 12 minutes of spreadsheet calculations, minute changes, more calculations and then waiting... and then calculations, waiting and so on. Even at 8x it is slow. I'll add some screenshots later for those who can't be bothered to watch.

The good news is I may have found a way to decrease Bobs drift. By first holding on to the ladder, letting go, going to warp but quickly going back to normal before he drifts away from the ladder, gripping the ladder again and then letting go one last time it seems the drift is reduced a lot. I believe the ship is just much more stable after a short warp than it is with just SAS and that might be the difference. Or, it's just random luck. Anyway, after my last orbit Bob didn't drift more than 300m which is quite impressive over more than 5 years.

I'm down to the last orbit now and my calculations suggest that if I do nothing, I should be 12000 seconds or 27 Mm away from a perfect hit so the gravitational pull should/could make it dead center. I might just live with that since the general error in my measurements could be much greater. But, that's a decision for another day.

 

[JAFO]

8 hours ago, dvader said:

I can probably check the efficiency of any drift reducing method by just checking Bobs Ap/Pe after he lets go of the ladder.

Now that sounds like a caveman technique for checking course corrections, if ever I heard one!

[dvader]

Just now, JAFO said:

Now that sounds like a caveman technique for checking course corrections, if ever I heard one!

You can see it in action at 8m12s. It is really exciting stuff : https://youtu.be/rIynwubQWAE?t=8m12s

I let go, warp, grab and let go again. Then I go to the tracking station to check Ap/Pe since you can't tell the ship orbit from Bobs orbit in the map view.

[dvader]

A lot has happened since last. Bob's will has been carried out and there were no major disasters.

The story continues here with: The awakening!

https://imgur.com/a/q7Y1TPG

Teaser:

Here's also a video with all the details. It's a total of three real hours of game time in two (or three) sessions. The action starts at about 1:40 and then again at about 13:50. It starts with the final orbit around kerbol, then Jool intercept, Bop intercept, visiting all 5 biomes, awakening the ancient one and finally going back up for a rendezvous with my fuel tanks (which was a lot harder than expected). Now I'm in space around Bop with science worth millions... or at least hundreds...

 

Next, I have to go back to Kerbin of course. The surface missions cost more than expected but my main fuel tanks still hold much more than expected. My calculations say I might have 5800 dV left but that is so much that I don't really trust the numbers.

 

[dvader]

Done! There was some last minute drama but both Bob and Val are safely back on Kerbin. (but is Kerbin safe?)

Full story is here:

https://imgur.com/a/zs4v8yb

with the detailed, unedited video below.

A short summary of the excitement:

1. At about 9:15 video time, all is well and Bob gets hold of the ladder one last time, I check and start the engines and prepare for the final insertion burn.
2. I speed up to 4x, go forward a bit, then go back to take a few final screenshots at 9:30. All seem fine and Bob is by the ladder. Or is he... His arms are getting suspiciously long.
3. I go back and forth a bit to check the fuel in the side tanks and at about 10:00 I burn off the remaining 300 dV for the main engines and drop them. The trajectory "forks" a bit but I don't notice.
4. At about 10:05 (40 real time seconds later) I realise there is a weird fork in my trajectory... What is going on, where is Bob and what's his trajectory? If he goes below 70k he will turn into smoke...
5. At 10:11, I realise Bob must have let go of the ladder before the burn so he is on a safe trajectory but he is now 16 km away and is travelling away from the ship at 365 m/s. Also, the target lock is gone but can be re-acquired (I have no idea why. Maybe I clicked on Kerbin).
6. I try to catch up as quickly as I can but with a single Spark and full tanks it's not going very fast... I designed for dV, not emergency maneuvers.
7. At 10:17, something goes wrong. At first I can't figure out what is going on. The ship spins when the retrograde marker shifts to prograde but the engines are dead and I can't restart them. There should  be thousands of dV left but nothing happens. After a while I realise some idiot disabled crossfeed on the main tank! The smaller, later stage tanks are spent but the big tank is still full. I enable crossfeed again and continue burning.
8. At 10:55, I catch up with Bob again but I'm already at 300km, well past periapsis. I start burning frantically, while keeping an eye on the Pe. I don't want the ship or Bob to get below 70km just in case some other disaster strikes (like I run out of fuel and fail to launch a rescue mission in time). It costs a lot, almost 2000 dV or so but in the end I am in an orbit around Kerbin.
9. Now, I can finally relax a bit. I have about 800 dV left and use it to lower my Ap to well below the Mun before launching a return craft. Things fortunately go well from here and both Val and Bob end up alive and on the ground. The mission yields 730 science, much more than the 450 I need, and both Val and Bob gets promoted!

It was probably the worst possible moment for Bob to drop the ladder and I am glad I had such a ridiculous amount of fuel left. I needed every bit of it to save the situation (well, 2/3rds of it).

That concludes my NCD Caveman attempt. About 640 science came from Bop, 427 from Minmus, 101 from Jool, 100 from Mun, 88 from Kerbin and 15 from Kerbol. After researching all techs I have about 300 science left.

 

 

[JAFO]

7 minutes ago, dvader said:

Done! There was some last minute drama but both Bob and Val are safely back on Kerbin.

Woohoo!  Congratulations!

[IncongruousGoat]

@dvader Congratulations! Not only did you manage to complete NCD, but you did it with 300 science to spare. No mean feat indeed. I was wondering how long it would take for someone to find a better way to do this challenge, and it seems the answer is "not long at all".

[dvader]

Just now, IncongruousGoat said:

@dvader Congratulations! Not only did you manage to complete NCD, but you did it with 300 science to spare. No mean feat indeed. I was wondering how long it would take for someone to find a better way to do this challenge, and it seems the answer is "not long at all".

Thanks! But, I guess 3 months real time for a game challenge would be considered a long time by most ordinary people (maybe 1.5-2 not counting vacations).

At first I thought it was just more fun to do something new and see what I could do with Bob but he really proved invaluable. If I'd do it again I would probably (ab)use him again. The playthrough could be streamlined a bit by skipping the Greenoliths that didn't pay off and go for interplanetary as soon as possible. I don't think the beginning can be modified much though. I basically did the same thing you did with only minor differences (like landing on rooftops instead of early space EVA) but there are so few options that I can't imagine huge improvements. But then again, I didn't think anyone would ever beat it at all so I wouldn't trust my judgement