totm january 2020 Community Caveman Jool 5 mission

[ManEatingApe]

@IncongruousGoat Thank you for the 1st contribution!
I've merged your PR containing the Tylo lander craft. The line endings config also worked perfectly.

Everyone please pull the latest version of the repo to pickup the changes.

Edited August 23, 2019 by ManEatingApe

[IncongruousGoat]

What are everyone's thoughts on graphics mods? Normally, there are a handful I'd add to an otherwise stock KSP install (Scatterer, EVE, Distant Object Enhancement, RealPlume). However, we probably want to keep our installs as similar as possible, both for compatibility purposes and to keep the pile of screenshots/footage we'll end up with at the end of this looking as consistent as possible, and I know some people don't use graphics mods for performance reasons, or just because they don't care about aesthetics.

So are there any graphics mods we'd all like to get set up, or are we going to stick to pure stock?

[ManEatingApe]

50 minutes ago, IncongruousGoat said:

So are there any graphics mods we'd all like to get set up, or are we going to stick to pure stock?

Good suggestion, as long as it doesn't affect the save file in any way, I say go for it - make things as pretty as possible!

My trusty 2011 Macbook has heard of the concept of graphics but wants nothing to do with it, especially since the GPU spontaneously de-laminated from the motherboard a few months back, so I'll be sitting this one out.

In other news, I've been testing a fuel tender design that can loft 3 tons of fuel into LKO, along with a small robot tug to dock it with the mothership. Both ends of the fuel stack have a docking port to enable daisy chaining.

Edited August 25, 2019 by ManEatingApe

[IncongruousGoat]

3 hours ago, ManEatingApe said:

In other news I've been testing a fuel tender design that can loft 3 tons of fuel into LKO, along with a small robot tug to dock it with the mothership.
Both ends of the fuel stack have a docking port to enable daisy chaining.

3 tons sounds about right for the practical limit of what can be launched with caveman tech. The maximum thus far launched is 4.25 tons (as done in this thread), but that's under ideal situations with a very low part-count payload. For our purposes, 3 ton fuel tank launches will work just fine.

On a separate note, now that we've got the mission plan laid out to (more or less) everyone's satisfaction, what's the next item on the to-do list?

[Muetdhiver]

Graphics mods are fine for me.

I am working on the mothership with on pad assembly. 2.5m center core with a poodle on the bottom, a 2.5m drop tank on top and 6 side boosters.

If everything works, that should give us a poodle with 12tons of fuel and 10 docking ports. (6 on the sides, 4 in the front, one of which is central.)

If we add 3x3 tons of fuel with the tenders, that would be 21tons, likely enough if everything else can stay under 10t. Assuming 10t of payload, that would mean about 3100 DV.

I'm also starting on the jool ejection/correction script.

Edited August 26, 2019 by Muetdhiver

[ManEatingApe]

To try and help clarity I've updated the mission plan in the OP with placeholder craft names based on the phonetic alphabet:

• FOXTROT - Fuel tender from Kerbin surface to the mothership
• CHARLIE - Crew transport from Kerbin surface to the mothership
• MIKE - Mothership from LKO to high Jool orbit then Tylo low orbit
• TANGO - Tylo lander
• BRAVO - Bop & Pol lander (same design as TANGO)
• VICTOR - Vall lander (same design as TANGO)
• LIMA - Laythe ascent vehicle
• KILO - Kerbin return vehicle

On 8/26/2019 at 8:04 AM, Muetdhiver said:

I am working on the mothership with on pad assembly. 2.5m center core with a poodle on the bottom, a 2.5m drop tank on top and 6 side boosters.

Great, I'm excited to see the results! Launch pad assembly is tricky yet extremely fun.

Let's make sure all the lander designs are uploaded to Github so that you can ensure they fit on the mothership before you launch into orbit

On 8/26/2019 at 4:38 AM, IncongruousGoat said:

On a separate note, now that we've got the mission plan laid out to (more or less) everyone's satisfaction, what's the next item on the to-do list?

Good question! We have a total of 10 contributors to this thread
This is my understanding of who is doing what along with some suggested next steps.

Participant	Done	Doing	Suggested next steps
@ManEatingApe		FOXTROT design	Use Debug menu to elevate Kerbals to 5 star experience then update save file (required for personal parachute) Finish FOXTROT design then upload craft file to Github
@IncongruousGoat	TANGO design		Tweak TANGO design to add 2nd docking port in order to allow a construction drone to grab it and dock it with the mothership. Design CHARLIE craft to transport Kerbal holding onto a ladder from surface of Kerbin to LKO.
@dvader	LIMA design		Design launch vehicle to deliver LIMA into LKO. Add extra detachable fuel tank so that LIMA can go from high Jool orbit to low Laythe orbit. Upload craft file to Github
@Pds314		KILO design	Upload construction drone design to Github Add extra detachable fuel tank so that KILO can go from high Jool orbit to low Laythe orbit. Finish KILO design then upload craft file to Github
@Muetdhiver		MIKE design	Sanity test design with landers then upload craft file to Github

To make mothership orbital assembly easier, each part (lander, fuel tank etc) should ideally have either:

• A spare docking port that a contruction drone can attach to, then use to bring the part to the mothership. This 2nd port should be co-axial with the main docking port.
• Alternatively the part is self-powered and can rendevous itself with the mothership. Ensure that empty tanks or extraneous parts can be detached to save weight on the mothership.
   

Edited August 27, 2019 by ManEatingApe

[IncongruousGoat]

@ManEatingApe Done and done. CHARLIE looks something like this:

The aerodynamics are terrible (it's a cup pointed into the airstream), but the payload mass is so low that it doesn't matter, and it keeps the Kerbal super safe. No chance of overheating or getting blown away.

A PR has been submitted with the changes.

[Muetdhiver]

I think I have mostly solved the problem of getting the orbital parameters of the transfer orbit from caveman infos. 

I now have to go from paper to code. I found a way to get the argument of periapsis without rulers. I will do a test with a real transfer as soon as times permits.

Thereare a few things I have to sort put w.r.t to the inclination correction burn, but not a showstopper.

Getting the side boosters strapped on the mothership is... an interesting exercise...

[ManEatingApe]

5 hours ago, IncongruousGoat said:

Done and done. CHARLIE looks something like this...

The aerodynamics are terrible (it's a cup pointed into the airstream), but the payload mass is so low that it doesn't matter, and it keeps the Kerbal super safe. No chance of overheating or getting blown away.

A PR has been submitted with the changes.

Fast turnaround! Craft files look great and nice tweaks on the .gitgignore file.
I merged your changes then promoted all Kerbals to 5 star experience and tagged it as "v3".
Everyone please pull the latest version.

1 hour ago, Muetdhiver said:

I found a way to get the argument of periapsis without rulers

I need to see a picture of this...

@dvader @Pds314 Could you open a PR with your craft files whenever you have time?
Craft don't have to be perfect, the goal is to give @Muetdhiver a rough idea of size of mass and size when designing MIKE.

Edited August 27, 2019 by ManEatingApe

[Pds314]

2 hours ago, ManEatingApe said:

Fast turnaround! Craft files look great and nice tweaks on the .gitgignore file.
I merged your changes then promoted all Kerbals to 5 star experience and tagged it as "v3".
Everyone please pull the latest version.

I need to see a picture of this...

@dvader @Pds314 Could you open a PR with your craft files whenever you have time?
Craft don't have to be perfect, the goal is to give @Muetdhiver a rough idea of size of mass and size when designing MIKE.

I may or may not have wifi for the next... oh.. ok. So I will probably need to Tether through my phone since I'm leaving for like 4 days in 6 hours and need to sleep... I will try to get something available. I can't promise it will be a pull request. I might just have to do a KerbalX, google docs, etc, craft drop or something and then someone with functional wifi can do the PR....

 

On an unrelated related note, I would advise everyone to do hibernate in warp set to automatic for every probe core. As a small number of solar panels will not be sufficient to keep a probe core awake forever at Jool. And it will be very possible to be stuck in the situation that you don't have the electricity to click hibernate if you warp it away. Hibernating in warp reduces power consumption by like 99%.

Edited August 27, 2019 by Pds314

[ManEatingApe]

10 hours ago, Pds314 said:

On an unrelated related note, I would advise everyone to do hibernate in warp set to automatic for every probe core. As a small number of solar panels will not be sufficient to keep a probe core awake forever at Jool. And it will be very possible to be stuck in the situation that you don't have the electricity to click hibernate if you warp it away. Hibernating in warp reduces power consumption by like 99%.

That's a very good reminder. I've updated the repo:

• New FOXTROT fuel tender craft, that can lift 3 tons of fuel to orbit
• Probe cores on CHARLIE, FOXTROT and TANGO set to auto-hibernate

Edited August 27, 2019 by ManEatingApe

[Muetdhiver]

The NavTool is moving ahead nicely after a few headaches.

The error on altitude and speeds are now under control. Under 0.01% which should be good enough.

The test was done with a random burn at a random time. The trajectory is build from one data point, two other are used as controls. Once the Keplerian elements are known, I convert them to carthesian ones, and then propagate the orbit to t0 and t1 (times of control points) and compute the altitude and speed, then compare to measures.

I have yet to verify the accuracy of the argument of Pe, but it can only be done with a full test with a correction burn and an encounter. I have a porkchop/Lambert solver for the Kerbin-Duna transfer, adaptating it should be rather simple.

 

Spoiler

For the curious, here is how it works :

First, we assume a zero inclination and set the longitude of the ascending node to 0°. That simplifies the problem to getting : eccentricity (e), semi-major axis (SMA), argument of periapsis (ArgPe) and True Anomaly (TAN) at a known time.

from Ap et Pe we have the eccentricity, from the radius (alt+kerbol radius) we get the True Anomaly at a known time.

The trick for ArgPe : what I did is : we know that kerbin has a perfectly circular orbit with no inclination or anything. it's TAN at epoch zero is known to be Pi. This means that we can compute kerbin TAN at tL (ejection burn time), and also the vessel TAN at said time. From geometry, ArgPe is simply KTAN(tL)-VTAN(tL) where KTAN is kerbin TAN and VTAN is the vessel's.

 

[ManEatingApe]

4 hours ago, Muetdhiver said:

The NavTool is moving ahead nicely after a few headaches.

The error on altitude and speeds are now under control. Under 0.01% which should be good enough.

Nice progress. I think that tiny margin of error is more than suitable! We probably won't be able to control the burn anyway near as precisely.
However as long as the intercept is reasonably close, we should be able to tweak it relatively cheaply at the edge of Jool's SOI.

If you like, open a PR with your script and I'll add it to the repo - GitHub is meant for code after all!

Out of curiosity, what do the various output terms (e.g. "rL pred", Delta for Radius, "Delta for Speed") mean?

[Muetdhiver]

The Deltas are the difference between the predicted value from propagating the orbit, and the actual ingame values. rL is radius at launch.

I have since also checked eccentricity, which is also okay, and ArgPe accuracy seems okay at ~0.1°

I have a sloppy jool ejections that I will try to salvage with the updated version using Lambert. All will be put in the repo once it's cleaned up a tad.

 

[ManEatingApe]

1 hour ago, Muetdhiver said:

The Deltas are the difference between the predicted value from propagating the orbit, and the actual ingame values. rL is radius at launch.

I have since also checked eccentricity, which is also okay, and ArgPe accuracy seems okay at ~0.1°

I have a sloppy jool ejections that I will try to salvage with the updated version using Lambert. All will be put in the repo once it's cleaned up a tad.

 

Gotcha. So is the plan to do a correction burn after each control point, t0 and t1 to bring the delta to zero?

[Muetdhiver]

10 hours ago, ManEatingApe said:

Gotcha. So is the plan to do a correction burn after each control point, t0 and t1 to bring the delta to zero?

No. Having deltas is the sign that either (or combo of) :

1) There is an error on the times used to compute

2) There is an error on the eccentricity

3) There is an error on the SMA

4) There is an error on the central body standard gravitational parameter

Previously I forgot to correct Ap/Pe for kerbol radius, so I got a ~1% error on numbers. At this point I can exclude errors on SMA & e.

I also found that the central body grav parameter was a bit off. I computed it again from kerbin speed, and then did further corrections. Speeds at t0 and t1 have now deltas of less than 1 m/s, which is remarquable.

I also got the Lamber solver working for a craft-Jool encounter, but there is an issue with times and/or ArgPe because the calculated burn is about 10 m/s off (on a 135 m/s burn). Which is enough for a miss. This is likely linked to some issues I have when the Pe is very close to the ejection burn. The error looks to be of about 3° on ArgPe.

Edited August 30, 2019 by Muetdhiver

[Muetdhiver]

After a few more headaches, I got it working. Moral of the story is that being accurate as caveman is all about being tricky

I added it in the repo along with the kerbol system definitions required by it. It's a bit of a mess and not really user friendly. Would require some refactoring and stuff as this is pretty much a pile of tricks glued together.

Other requried packages like PyKep, numpy & al would have to be installed to make use of it.

As a bonus, here is the first result.

Spoiler

 

Edited August 30, 2019 by Muetdhiver

[ManEatingApe]

58 minutes ago, Muetdhiver said:

After a few more headaches, I got it working. Moral of the story is that being accurate as caveman is all about being tricky

Good work! That result is more than accurate enough. An intercept that close can be corrected reasonably cheaply at the edge of Jool's SOI.

 

I'd like to focus on design and assembly of the mothership next. Could you share some screenshots of the proposed design?

[dvader]

Good work everyone! I was really impressed by that heavy lifter you built @ManEatingApe. You managed to lift about 20% more than I did! Also, really nice work with the calculator @Muetdhiver. I really want to have a look at the inner workings of it when I get the time.

As for LIMA, I had a couple of setbacks. I had of course designed it in a sandbox and as it turns out, neither the small delta wing nor the long air intake are available in caveman. As it turns out, the big bulky radial intake can not be used instead since it causes a lot of drag so my only option is to use the very heat sensitive small circular intake. This of course means greater care must be taken during descent so landing precision with is a bit harder.

I thought the small delta wing could easily be replaced by the Delta-Deluxe Winglet but no. The built in elevons are just for show and the entire lift vector changes direction when you try to control it. Since the lift is close to the CoM, you get almost no control at all.

So, I had to use the longer structural wing type-c. This of course increases the risk that the wing hits the poor caveman in the head during decoupling causing him to fall off the ladder. So, I have to move him a bit off center before staging. I tried using the square wings but can't remember why I gave up on that. I think I have to try again since dropping the Kerbal at 10 km is really annoying. I don't know why to worked so well with the delta wings...

Also, taking off from water is MUCH harder now. As a speed boat it works great and I can get up to about 35m/s on the surface but I have to wiggle it sideways or up/down to actually get out of the water. I haven't figured out why it is so much harder. One version I tried couldn't lift at all but would just dive no matter what I did.

All in all, the new design is more heat sensitive and thus can't be controlled as well during descent,  it is trickier to take off from water and it requires some extra fiddling not to hit the kerbal in the head during staging. But, it works.

I'm gonna keep fiddling a bit more to see if I can find a way of staging safely and I also have to redo the full test mission on Laythe just to make sure I haven't missed something important. I can get it up to about 2000m/s at 70km on Kerbin so it "should" be fine on Laythe but you never know.

Oh, and the small fairing can't be used because it is too small and the part count is just on the limit so the orbital insertion engine must be launched separately.

Which reminds me, it would be great if the Laythe descent badminton ball (or something similar) could be attached to the VIKTOR craft so we don't need another rendezvous before descending. It is 7 parts but only 65 kg. I've tested the survivability and you can drop it into the atmosphere from 440 km and survive (but not from 1000km).

I can commit the current versions if anyone wants to have a look but there will be changes.

[ManEatingApe]

35 minutes ago, dvader said:

All in all, the new design is more heat sensitive and thus can't be controlled as well during descent,  it is trickier to take off from water and it requires some extra fiddling not to hit the kerbal in the head during staging. But, it works...

...I can commit the current versions if anyone wants to have a look but there will be changes.

Very detailed and informative write-up! Spaceplanes are tricky enough normally, let alone when only low tech parts are available.
I think it's a good idea to share the current version of LIMA (even if there will be tweaks) to get early feedback and assist @Muetdhiver in the MIKE design.
You could name the craft file something like "LIMA (wip)" where WIP stands for "Work in progress" to indicate there will be further changes.

41 minutes ago, dvader said:

Which reminds me, it would be great if the Laythe descent badminton ball (or something similar) could be attached to the VIKTOR craft so we don't need another rendezvous before descending. It is 7 parts but only 65 kg. I've tested the survivability and you can drop it into the atmosphere from 440 km and survive (but not from 1000km).

I was thinking VICTOR could use purely propulsive braking and we would only use the badminton ball aero-shield for KILO.
The VICTOR design has about 4,600 m/s dV. Adding up the approximate required dV and including some buffer:

• 1,000 m/s from high Jool orbit to low Vall orbit
• 1,000 m/s to Vall surface
• 1,000 m/ to Vall orbit
• 350 m/s to Laythe intercept
• Over 1,000 m/s remaining to brake to survivable re-entry speed
• Kerbal uses personal parachute to land

There would be no rendezvous required in Laythe orbit and we could choose the landing spot more precisely.

[IncongruousGoat]

6 hours ago, dvader said:

Also, taking off from water is MUCH harder now. As a speed boat it works great and I can get up to about 35m/s on the surface but I have to wiggle it sideways or up/down to actually get out of the water. I haven't figured out why it is so much harder. One version I tried couldn't lift at all but would just dive no matter what I did.

I ran into the same problem the last time I sent a spaceplane to Laythe. What I did with that one was deliberately nose way down, dive ~10m down, and then nose up and dolphin-jump out of the ocean. Main concerns here are the Kerbal getting knocked off (of course), and having a high enough TWR to pull it off. When I last did an ascent this way I had a RAPIER strapped to the back of a tiny ~5 ton spaceplane.

[Muetdhiver]

I have been working on the Mothership a bit, and how to place the landers & stuff in a practical way.

An other thing I have been thinking about is a contingency plan if things go pear shaped, for example from a surprise encounter with a moon or something else costing quite a bit of DV.

What I propose is the following :

- make sure that the TANGO design can dock in a way such that it could be used as a backup engine with a 3T tank from a FOXTROT tender. This means that one could reconfigure the mothership to have said backup tank as the core and one of the TANGO lander as engine.

Spoiler

- Take an extra 3T of fuel as backup, with a few radial Jr ports (2 or 4) so that we can dock things to it.

Here is a the core with dummy landers :

Edited August 31, 2019 by Muetdhiver

[ManEatingApe]

3 hours ago, Muetdhiver said:

I have been working on the Mothership a bit, and how to place the landers & stuff in a practical way.

An other thing I have been thinking about is a contingency plan if things go pear shaped, for example from a surprise encounter with a moon or something else costing quite a bit of DV.

We're off to a great start! Quick question, what is the nifty tool you are using to annotate the screenshots?

For the TANGO and BRAVO landers I'm suggesting to assemble them inline once in Jool orbit, with the spare fuel like this:

The Kerbal pilot would hold on to the ladder on top during Tylo insertion.
Total dV of this arrangement is ~1,500 m/s and it takes ~1,000 m/s for a perfect high Jool to low Tylo capture, so we have some wiggle room.

Minor suggestion: Would it be possible to move the triple docking ports currently on top of the X200 to the sides e.g.
This would allow more flexibility in attaching things.

I think we'll need to plan how to get the KILO and LIMA landers into low Laythe orbit.
Aero-braking is the obvious choice as otherwise it takes ~1,300 m/s to capture into low Laythe orbit.

@dvader and @Pds314 what do you think?

[dvader]

I've sent a PR with the LIMA plane (latest version). I forgot to name it WIP but there should be no essential changes except for perhaps more fuel for the orbital insertion engine. It currently has about 750dV which is enough IF you get a good encounter. The craft can (very carefully) aerobrake from a high elliptic  Laythe orbit down to LLO. Yes, there is a "Basic Fin" at the back which burns up at 934K...

I've tried adding disposable radiators and do aerobraking directly from a high elliptic Jool orbit but you need to burn off about 600m/s which is a lot and the whole thing just flips around. I think powered capture is the way to go. It would be a fun engineering problem though to create a heat resistant shell that also keeps the craft in the right orientation.

 

Spoiler

I've been test running it a bit more and this version is much easier to get out of the water but much harder to send into space on a rocket. There's also a funny issue that I don't yet understand. It seems to make a difference if you mount the ladder parallel or sideways to the craft. I've had better luck with sideways placement but I don't understand why. I seem to get much more torque during the rocket portion of the flight with parallel placement. Both work though. I also got a new personal best, 2078 m/s at 90km around Kerbin (sideways).

Spoiler

 

I haven't followed the general construction at all but having something that can act as an emergency tug would be nice (I think that's what @Muetdhiver suggested). The Jool-5 rules allow for one refuel mission so as long as it has a ladder, engine and two docking ports many problems can be solved with a refuel mission.

Also, propulsive braking and landing with VICTOR sounds great.

[ManEatingApe]

3 hours ago, dvader said:

I've sent a PR with the LIMA plane (latest version). I forgot to name it WIP but there should be no essential changes except for perhaps more fuel for the orbital insertion engine. It currently has about 750dV which is enough IF you get a good encounter. The craft can (very carefully) aerobrake from a high elliptic  Laythe orbit down to LLO. Yes, there is a "Basic Fin" at the back which burns up at 934K...

Thank you for the contribution! You're definitely pushing the frontiers of ladder technology.
If the craft can achieve over 2,000 m/s on Kerbin then we should be fine on Laythe. I've merged your PR and tagged it as "v4". Everyone please pull the latest changes.

Some suggestions for the next iteration:

• There was an invalid part error when I loaded the craft from the save file. It seemed to be just the loadmeta file. My copy of KSP auto-updated it and now it appears OK, but it would be good if you could double-check that the craft file is still as intended.
• Let's add a 2nd docking port to the craft. This will allow the orbital construction drone to "grab" the craft then dock it with MIKE in LKO.
  Alternatively another solution is to add the capability for the craft to dock itself to MIKE (this would require 1 fewer orbital rendezvous)
• Do include your launch vehicle - we'll have to get your lander to LKO somehow!

3 hours ago, dvader said:

I've tried adding disposable radiators and do aerobraking directly from a high elliptic Jool orbit but you need to burn off about 600m/s which is a lot and the whole thing just flips around. I think powered capture is the way to go. It would be a fun engineering problem though to create a heat resistant shell that also keeps the craft in the right orientation.

Both LIMA and KILO need to capture into low Laythe orbit from high Jool orbit. This is an interesting sub-challenge as both craft have aerodynamic parts (making aero-braking tricky to control).
I'm proposing that we dock LIMA and KILO together in high Jool orbit with some extra fuel similar to the BRAVO and TANGO combo idea.
Then we would only have perform the Laythe rendezvous once for the combined craft instead of twice for each individual craft.

In this scenario, once in high Jool orbit MIKE would split into 3 parts:

• LIMA and KILO together with extra fuel attached -> low Laythe orbit
• TANGO and BRAVO together with Kerbal and extra fuel attached -> low Tylo orbit
• VICTOR -> low Vall orbit. No extra fuel needed as the craft has enough dV to insert into Vall orbit and still complete its mission.

 

What does everyone think?

Edited September 1, 2019 by ManEatingApe