Duna Outpost Mission Architecture Challenge

[Norcalplanner]

6 hours ago, Laie said:

How to break this challenge: do a super-massive launch on day one, deploy all hardware and ISRU the hell out of it.

Late night thought. Sorry, couldn't resist posting it.

This is the reason I'm working on a 150 ton payload SLV.  The way the math works out, you can hurl up to 630 tons into LKO, plus the NPM of your initial launch. With a 150 ton launcher, you get 750 tons of total payload, while still having one launch prior to each transfer window.

That said, part of the challenge is based on efficiency, so effectiveness per ton still counts. And part count and complexity go up with a larger launch vehicle.

[Kerbolitto]

I think that part count is higher with a light payload because you have to make little modules (more probes, more RCS), add some little fuel tanks here and there, in the end it's more weight-effective (in my opinion) but with larger fuel tanks and 1 or 2 more engines, a large payload would be more user friendly (and require less dockings).

 

Edited July 29, 2018 by Kerbolitto

[Mad Rocket Scientist]

I made a spreadsheet for planning missions:

https://docs.google.com/spreadsheets/d/1fJP9cZntjaAWeXIlD-A2nt1iIQphlqFImMWEs7TVLyk/edit?usp=sharing

It lets you put in the mass of your launcher and see how often it will launch, change transfer windows (comes with ones from the transfer window planner, use at your own risk), adjust the resolution while dynamically scaling notes, launch windows, and SLV launch frequency.

Next I'll try to make it so you can input dates in KSP format for transfer windows and notes, and increase the number of available notes to a much larger size.

Edited July 29, 2018 by Mad Rocket Scientist

[Norcalplanner]

2 hours ago, Kerbolitto said:

I think that part count is higher with a light payload because you have to make little modules (more probes, more RCS), add some little fuel tanks here and there, in the end it's more weight-effective (in my opinion) but with larger fuel tanks and 1 or 2 more engines, a large payload would be more user friendly (and require less dockings).

I may not have been clear - the part count goes up because that big initial initial launch includes (in my case, anyway) the miner, several smart fuel tanks, a reusable nuclear transfer stage, three multi-part comm/scan satellites, the DAV, the EAV, Duna/Ike LFO transfer tug, reusable lander crane, and then all the ground modules, rover, and Duna surface ISRU equipment.  

Edit: I almost forgot - the first stage of the lifter is so large it takes 72 parachutes for recovery.

Edited July 29, 2018 by Norcalplanner

[Alchemist]

58 minutes ago, Mad Rocket Scientist said:

I made a spreadsheet for planning missions:

https://docs.google.com/spreadsheets/d/1fJP9cZntjaAWeXIlD-A2nt1iIQphlqFImMWEs7TVLyk/edit?usp=sharing

https://drive.google.com/file/d/12vtVJyzRlhRBcMZthNo5Hlbkss9zMiFd/view?usp=sharing (.xlsx file)

It lets you put in the mass of your launcher and see how often it will launch, change transfer windows (comes with ones from the transfer window planner, use at your own risk), adjust the resolution while dynamically scaling notes, launch windows, and SLV launch frequency.

Next I'll try to make it so you can input dates in KSP format for transfer windows and notes, and increase the number of available notes to a much larger size.

Looks like you forgot that first launch is on day 40, not day 1

 

Ahh, all this talk of optimizing payload capacity for exact window timing, and I just did it the Russian way:

1. take the best applicable development
2. add another stage on top
3. build the mission around it

 

Well, here are my first two launches:

1: Durability - mobile Duna base, 4 crew (also delivered 3 satellites, 2 of them mapsats)

(Arrival date Y2D34)

 

2: Messenger - Minmus mining outpost (with fully self-sufficient life support!), the transfer stage also doubles as a tanker

That's one way to stop wasting precious payload capacity on bringing fuel to LKO

[Mad Rocket Scientist]

14 minutes ago, Alchemist said:

Looks like you forgot that first launch is on day 40, not day 1

Whoops, I'll get to work on fixing that. 

14 minutes ago, Alchemist said:

Ahh, all this talk of optimizing payload capacity for exact window timing, and I just did it the Russian way:

1. take the best applicable development
2. add another stage on top
3. build the mission around it

Well, here are my first two launches:

1: Durability - mobile Duna base, 4 crew (also delivered 3 satellites, 2 of them mapsats)

[...]

(Arrival date Y2D34)

2: Messenger - Minmus mining outpost (with fully self-sufficient life support!), the transfer stage also doubles as a tanker

[...]

That's one way to stop wasting precious payload capacity on bringing fuel to LKO

Nice looking base and Minmus fueling station.

I'm just going to try to use a nuke tug that refuels from an Ike base to avoid launching fuel into LKO.

Currently I have developed the lander and mothership for the first Duna landing, going along with commsats, a refuelling base, and possibly some space station or surface base bits:

I'm not sure how I'm going to transfer to and land on Duna, since I'm waiting until I've built everything I want to go on the first window before I make a transfer stage. I kept this just under 10 tons, leaving 30+ tons for everything else, which is pretty good. I might want to use part of the launch window 1 transfer stage to build my advanced deep space transit tug, which will be using the Kerbal Atomics "Emancipator" open cycle nuke. But it weighs like 16 tons, to I may launch the tanks and add the engine on later.

Hmm, that means it has to go back to Kerbin, which means I may not need the DAV mothership if I add hab space...

[Mad Rocket Scientist]

Alright, spreadsheet is updated:

https://docs.google.com/spreadsheets/d/1fJP9cZntjaAWeXIlD-A2nt1iIQphlqFImMWEs7TVLyk/edit?usp=sharing

[Kerbolitto]

Chapter 5 is posted :

 

 

 

I took time to explore Duna a bit, test the outpost rolling capabilities, and resolved a problem with life support !

 

 

Edited July 31, 2018 by Kerbolitto

[Ozelui]

A very ambitious challenge, count me in.

I'm afraid I didn't have the time to figure out my launch schedule, yet I'm already theorycrafting some of the vehicles.  For now, these are a pair of prototypes made for research ( i.e. learning how TACLS and other mods works ): 

Spoiler

A 30 ton self sustainable mobile base for 8 kerbals. It can produce food for up to 12? kerbals, but my plan is to use the excess production for other vehicles. Laughable top speed of 10 m/s (intended).

and a 22 ton electric deep space transit vehicle for 10 kerbals.  It won't be used unless the KSA scientists find a mod way to collect xenon gas on Duna. Of course these may change depending on the final planning.

 

And this is the mod list I'm currently using, please let me know if any of them is forbidden.  Anyway, since 1.4.5 came out this week, I'll probably upgrade because of kopernicus.  I love how Duna looks with the stock visual terrain mod, but kopernicus is a requirement and was not released for 1.4.4 (my current version).

 

[Mad Rocket Scientist]

Welcome...

To spreadsheet hell:

=IFS(A32 < 39, "", MOD(A32 - 40, $I$31) = 0, CONCAT(CONCAT("Launch number ", INT(A32/$I$31) + 1), IFS(AND(A32 >= 40, A32 <= $H$5), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1) * $I$28), AND(A32 > $H$5, A32 <= $H$6), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$5 - 40)/$I$31)) * $I$28), AND(A32 > $H$6, A32 <= $H$7), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$6 - 40)/$I$31)) * $I$28), AND(A32 > $H$7, A32 <= $H$8), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$7 - 40)/$I$31)) * $I$28), AND(A32 > $H$8, A32 <= $H$9), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$8 - 40)/$I$31)) * $I$28), A32 > $H$9, CONCAT(", Tons Launched: ", INT((A32/$I$31) - (($H$9 - 40)/$I$31)) * $I$28), TRUE, "")), TRUE, "")

FYI my previous spreadsheet doesn't work for launch prep time values of less than 40, and hence NPM values of less than ~5 tons.

Edited July 31, 2018 by Mad Rocket Scientist

[hoioh]

19 minutes ago, Mad Rocket Scientist said:

Welcome...

To spreadsheet hell:

=IFS(A32 < 39, "", MOD(A32 - 40, $I$31) = 0, CONCAT(CONCAT("Launch number ", INT(A32/$I$31) + 1), IFS(AND(A32 >= 40, A32 <= $H$5), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1) * $I$28), AND(A32 > $H$5, A32 <= $H$6), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$5 - 40)/$I$31)) * $I$28), AND(A32 > $H$6, A32 <= $H$7), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$6 - 40)/$I$31)) * $I$28), AND(A32 > $H$7, A32 <= $H$8), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$7 - 40)/$I$31)) * $I$28), AND(A32 > $H$8, A32 <= $H$9), CONCAT(", Tons Launched: ", INT((A32/$I$31) + 1 - (($H$8 - 40)/$I$31)) * $I$28), A32 > $H$9, CONCAT(", Tons Launched: ", INT((A32/$I$31) - (($H$9 - 40)/$I$31)) * $I$28), TRUE, "")), TRUE, "")

FYI my previous spreadsheet doesn't work for launch prep time values of less than 40, and hence NPM values of less than ~5 tons.

I used to manage a virtual soccer team with a group, there were some formulas that stretched my 1440p screen... Something with a quadruple self-replicating index (double match) formula that would fill a list of players of an opposing team in descending order, now that was fun to make adjustments to 

This one is pretty hard to decipher as well, especially out of context, what's it for?

Edited July 31, 2018 by hoioh

[Mad Rocket Scientist]

1 hour ago, hoioh said:

I used to manage a virtual soccer team with a group, there were some formulas that stretched my 1440p screen... Something with a quadruple self-replicating index (double match) formula that would fill a list of players of an opposing team in descending order, now that was fun to make adjustments to 

Oof, that must have been impossible to debug.

1 hour ago, hoioh said:

This one is pretty hard to decipher as well, especially out of context, what's it for?

Theoretically, it shows numbered launches of the SLV on each day that it could launch, even when the day step size is changed, along with how many tons could be launched to orbit since the last window. In practice, it's a awful collection of compounding off-by-one errors, strange bugs for such edge cases as "using odd numbers anywhere," tacked on additions somehow breaking things that already worked, and terrifyingly kludgy rounding to integers.   I think I'm going to have to break up its functions over more columns.

[Mad Rocket Scientist]

Would the Kerbal Atomics mod be OK to use? The most powerful engine (and of course the one I want to use) has an Isp of 2850s. However, it weighs almost 16 tons, burns 0.077 units of uranium a second, and its core only lasts 3 days at thrust, which isn't a huge problem, since it produces over 1.5 MN of thrust. 

[hoioh]

10 hours ago, Mad Rocket Scientist said:

I think I'm going to have to break up its functions over more columns.

That's what I do these days, mostly since I'm not working on shared sheets that much. That said, I'm working on the KEA judging sheet, which is shared, but in order to overcome the issue I'm simply working with a calculations tab within the sheet, which is where the magic happens in order to display coherent info on the actually used page. Maybe that will work for you as well?

[Death Engineering]

On 7/31/2018 at 12:18 PM, Ozelui said:

A very ambitious challenge, count me in.

[...]

And this is the mod list I'm currently using, please let me know if any of them is forbidden. 

Awesome! Nothing verboten there that I can see. If you haven't read the whole thread, there have been several klarifications along the way that may help.  

On 7/29/2018 at 7:30 PM, Mad Rocket Scientist said:

Alright, spreadsheet is updated:

[...]

Would the Kerbal Atomics mod be OK to use?

Far out. I'll link the spreadsheet into the OP. Thanks!  

I've not used Kerbal Atomics, but the Nertea-made mods I've used are balanced. I think I have one in my build, so sure why not.

 

[Mad Rocket Scientist]

18 minutes ago, Death Engineering said:

[...]

Far out. I'll link the spreadsheet into the OP. Thanks!  

I've not used Kerbal Atomics, but the Nertea-made mods I've used are balanced. I think I have one in my build, so sure why not.

You're welcome. I'll try to improve it in the future.

OK, thanks.

[Kerbolitto]

Hello This is a preview of SLV 03 docked to some modules waiting on KerbinSpaceStation since SLV02. The ship is "quite" large at 180t. and almost 400 parts .. I've never done something like this before.

 

 

It has something like 2k3 dV with Lf tanks half full and a TWR of 0.27 (8 nukes) so it's not too bad this time !

 

 

[meyst]

This is good to see and I admit temptation to join you all...

 

Quote

funk

"Finally I have a reason to try out an Aldrin-Cycler to Duna". I'm by far not sure how to do it, especially how to find the right trajectory. So far I was playing around with some flybys, but it breaks after some circles. I'm not even sure yet, if it is possible at all. But I'll try.

I have not heard anything about those for a while!

I used an outbound Aldrin-Cycler (S1L1) cycler myself. Some problems I encountered ironing out the bugs of that outbound orbit;

Expecting to always see a Ballistic Trajectory on the Kerbin gravity assist (usually but not necessarily always the case due to the angle needed on the gravity turn. [This is even more true of an Earth<->Mars Aldrin-Cycler])

Not recognizing/correcting course/orbit deviations early enough. Errors escalate every orbit.

Orbital duration, drifting/not in sync with the Synodic orbital period of Kerbin and Duna, 19 645 697.3 seconds. I settled on an outbound Kerbol (solar) orbit Ap of 31 505 200 km and Pe of 13 400 000 km but anything that matches the Synodic period should work.

 

 

I strongly suggest you test that inbound Duna>Taxi>Aldrin-Cycler>Kerbin transfer before committing to it. That one is a surprisingly high Delta V intercept for the taxi and not the easiest thing to fly either.

 

 

 

 

 

 

 

 

 

 

[Kerbolitto]

4 hours ago, meyst said:

I strongly suggest you test that inbound Duna>Taxi>Aldrin-Cycler>Kerbin transfer before committing to it. That one is a surprisingly high Delta V intercept for the taxi and not the easiest thing to fly either.

Intercepting a ballistic encounter might be very hard ... I tried it by catching an asteroid passing between Mun and Kerbin, so quite close, it requires 3 corrections from the lander + 3/4 adjustments with RCS at very low thrust... 910m/s to get away from Kerbin, 475m/s to catch the target. Hardest thing / heavier cost would be plane change in my opinion (the asteroid I chose was on a polar encounter with Kerbin so it was easy to launch at 0°) but catching a ballistic between Mun and Minmus for example might be very very  precise!

[funk]

7 hours ago, meyst said:

I used an outbound Aldrin-Cycler (S1L1) cycler myself. Some problems I encountered ironing out the bugs of that outbound orbit; 

Expecting to always see a Ballistic Trajectory on the Kerbin gravity assist (usually but not necessarily always the case due to the angle needed on the gravity turn. [This is even more true of an Earth<->Mars Aldrin-Cycler])

Not recognizing/correcting course/orbit deviations early enough. Errors escalate every orbit.

Orbital duration, drifting/not in sync with the Synodic orbital period of Kerbin and Duna, 19 645 697.3 seconds. I settled on an outbound Kerbol (solar) orbit Ap of 31 505 200 km and Pe of 13 400 000 km but anything that matches the Synodic period should work.

 

 

I strongly suggest you test that inbound Duna>Taxi>Aldrin-Cycler>Kerbin transfer before committing to it. That one is a surprisingly high Delta V intercept for the taxi and not the easiest thing to fly either.

 

Thx for your recommendations. So far I'm done with both cyclers and the taxis. Thx to this I found the trajectory easily. Two things need to be adressed.

1. Don't enter Dunas SOI, it's screwing up everything

2. My test gravity assists at Kerbin are somewhat of inaccurate due to wonky SAS and I guess low TWR

Further the dv requirements for the taxis are simply ridiculous. For a Kerbin -> Cycler transfer it's ok with 1500m/s + 100m/s to enter Dunas SOI. But Duna -> Cycler transfers are about 2600m/s. Because I want alternating taxis I need nuclear propulsion, so no aerobraking. Makes 4300m/s dv for the taxis. It's viable and the test results where ok, but the margin of error is small (~200m/s) on one leg. Further there will be refuel missions for the cyclers every two periods with a tanker also alternating between the cyclers.

I found several ways in some papers to reduce dv, but it didn't fit the challenge well. E.g.doing two Kerbin grav assists conflicts with the timeframe, adding more cyclers could fix that, but conflicts with mission efficiency.

Another way would be repeating small burns along the trajectory, but I don't have the time to code a solver for this in KSP and it's just tedious to perform so many burns.

Overall I think my designs and mission plans are 80-90% done and hopefully I can start in a few days... rip spare time.

 

Edited August 3, 2018 by funk

[Kerbolitto]

Funny thing about this challenge, every step is harder than the last ..

 

Thought the 1st launch was hard to plan with fuel restrictions, non-hohmann transfer and fast aerobraking .. Now I'm scratching my mind with SLV03 to plan the quickest injection (non hohmann if possible), decoupling half of nuke pods to recapture on Kerbin and leaving enough fuel to retroburn the whole stuff around Duna..

 

I think I'm starting to see a common thread here .. There will be 0 simple aim and inject in this mission

 

edit : Should I put the whole SLV03 to Minmus and refuel tanks to the max (3k9dV - 0.22TWR) or launch from Kerbin just a little time before the transfer window (1k1dV injection / 1kdV capture - 0.27TWR) ? Round trip to Minmus with ISRU and fuel tanks take approx. 25 days, launching from Minmus would reduce that time to 10 11 days before being able to inject with full tanks but if Minmus is on the bad side I'm screwed (got 66 days before optimal transfer).

What do you guys think ?

Edited August 3, 2018 by Kerbolitto

[funk]

12 hours ago, Kerbolitto said:

There will be 0 simple aim and inject in this mission

Recommending this as a second title of the challenge...

12 hours ago, Kerbolitto said:

Should I put the whole SLV03 to Minmus and refuel tanks to the max (3k9dV - 0.22TWR) or launch from Kerbin just a little time before the transfer window (1k1dV injection / 1kdV capture - 0.27TWR) ? Round trip to Minmus with ISRU and fuel tanks take approx. 25 days, launching from Minmus would reduce that time to 10 11 days before being able to inject with full tanks but if Minmus is on the bad side I'm screwed (got 66 days before optimal transfer).

Minmus has as a period of what - 50days or so? You should be fine. But is the additional fuel needed?! The pilot ususally knows best.

[Norcalplanner]

I'm going to have to back out of this for now - this challenge occupies the same mental space that I need to devote to work right now. 

[Alchemist]

Well, I'll just leave this here

 

[hoioh]

2 hours ago, Norcalplanner said:

I'm going to have to back out of this for now - this challenge occupies the same mental space that I need to devote to work right now. 

I know that feeling, got a different kind of job these days