A thread dedicated to theorycrafted relating to a specific Jool spoilers challenge

[Strawberry]

Hey spoilers for a new Jool thing ahead, dont read ahead if you dont want that

This thread is for theorycrafting relating on to how to get from the surface of Jool to Jools orbit, the end goal of this thread will be hopefully to undertake a mission that manages to undertake such a mission under this criteria:

1. No infinite fuel/propellant, everything has to work under stock ksp

2. No exploits, if a bug fix would remove being able to do this then it doesnt fit

3. This one is a light rule, part clipping may be needed but its discouraged. If it can be done without then that would be ideal.

This page will be broken down into two sections, observations and theory. The first section observations, will be dedicated towards observations regarding Jool, this will hopefully help narrow down the exact parameters of the mission. The second section will be Theory, which is theory on how to actually get to Jools orbit.

Observations:

Jools atmosphere like in ksp1 starts at 200k, if you're going too fast (testing is needed to get the specific value), your spacecraft will be destroyed due to the immense drag.

At 20k altitude on Jool the Vector engine has 75 isp, at around 12k altitude a vector engine cuts to zero isp and no longer works. The aerospike engine interestingly enough dips then peaks in performance around 50k altitude, at 20k altitude the aerospike has an isp of around 240, at 10k altitude an isp of around 90, and the engine cuts off at around 7k altitude. Other engines presumably perform worse here (also in case youre wondering, RCS and solid fuel rockets also cut off).

From observing specific jump heights, the gravity of Jools surface seems to be roughly similar to kerbins. The tracking center reports it to be .8 g and this seems right.

Jools atmosphere is 6.7 atm, Eves is 6.2 atm, Eve may be a good testing ground for some stages of launches.

The trip planner states that it takes 14k delta v to get from Jools surface to low Jool orbit, this is a very big number.

Theory:

If no rocket engines work, this challenge becomes significantly harder. This would mean that propeller propulsion is needed, with the weakened reaction wheels this may be very hard to make in ksp2.

The bulk of the delta v should probably come from electric propulsion, for specifics on this read the spoiler below

Spoiler

Electrical power would be very important for this mission. Minimizing the energy to mass ratio is critical here. Solar power wouldn't be smart at all here due to Jools far distance. The four engines that initially attracted me for this mission are the ones below:

KR4-P3 (The nuclear reactor): This engine would be the worst to here, it has a energy to mass ratio of ~7.1 U/T. It is also very large. Basically every other engine is better then this.

PB-NUK (small RTG): This one is promising, with a energy to mass ratio of 10. They're also very small. 

FCA-06 (Large fuel cell): This one has a energy to mass ratio of 75, impressive stuff. However this does not account for the needed weight of fuel. The FTL-T400 (the medium small sized methalox tank) can operate an ion engine for over a hour). When you factor in the 2.250 tons needed to power this thing, you get an energy to mass ratio of 7.2. 
Batteries: All batteries have the same energy to mass ratio, with an energy stored to mass ratio of 20000 U for every ton. Math is needed to see if this is greater or less then large fuel cells.

A middle stage will definitely be needed to get to orbit. Nuclear engines have too low of efficiency and thrust at high pressures so methalox engines will very likely be the go to. Vector engines may be the best here thanks to there high thrust, that or Aerospikes. Testing is needed.

The current plan is as follows:

A lower stage, if rocket engines work then this would probably be easiest. The trajectory should be as vertical as possible to get out of the atmosphere as fast as possible. With the high amounts of electricity generation the upper stage demands, the electricity will likely already be there for it to work fine if electric propulsion is needed.

A middle stage to get out of atmosphere and beginning to kick the spacecraft eastwards, best engines would probably be a vector or aerospike engine (testing needed). If the delta v is high enough these could begin to push the orbit eastward.

(Possibly needed): Lower Upper stage, these would probably be methalox as well due to the high density, likely orbital. These would be there to push the orbit eastwards that way the electrical engines can take effect

Upper stage: Electric propulsion, likely using small lots of small rtgs for power. Very high delta v. 

Edited February 26, 2023 by Strawberry

[Strawberry]

What Im learning through experimentation is that all sizes are equally effective in terms of mass and power for a lot of things. A large sized reaction wheel is 3x as heavy, 3x as power hungry, and 3x as powerful as a small reaction wheel. This is useful to know when trying to reduce volume. What Im also learning is that rtgs are stupidly good, but it is very hard to get propellors working

Edited February 25, 2023 by Strawberry

[Strawberry]

Spent a good bit working on trying to get propellors working, this is the result I came up with but even this spinning insanely fast didnt work. I think the aero may be borked to the point of practically disallowing propellors

[Strawberry]

At 20k altitude on Jool the Vector engine has 75 isp, at around 12k altitude a vector engine cuts to zero isp and no longer works. The aerospike engine interestingly enough dips then peaks in performance around 50k altitude, at 20k altitude the aerospike has an isp of around 240, at 10k altitude an isp of around 90, and the engine cuts off at around 7k altitude. 

Honestly despite the better atmosphere gain of aerospike, I think Vector is the way to go. These tests confirm that some type of non rocket lower stage will be needed, and with how thick Jools atmosphere is I think youll be able to push yourself up enough to get you to get the Vector engine to have decent isp and vectors produce so much more thrust. Hopefully Jools atmosphere is thick enough to where you can actually use propellors unlike Kerbin. 

[Strawberry]

Okay this right here is 17k delta v, I was worried about needing to have multiple stages of ion engines for the upper stage but I dont think that will be needed. A third of the weight is the nuclear engines, so using rtgs should help loads.

Edited February 26, 2023 by Strawberry

[Strawberry]

I just realized if push comes to shove that decouplers could be used to get the spacecraft upwards, though for obvious reasons I very much dont want to do this.

[Wheehaw Kerman]

I took Val skydiving on Jool yesterday.  She’s currently standing on the surface - didn’t even need a parachute to land safely.  Going to be following this thread with interest.  Pulling off a rescue from a gas giant would be something.

[Strawberry]

Oh yeah I forgot to mention, if anyone can get propellors working please share here and your design and I will like your comment, Im pretty confident about the rest of my design, I just cant get the propellers to work.

[K^2]

Can we do ye olde style helicopter built with reaction wheels? I know they're weaker than KSP's, but you also shouldn't need a whole lot of torque here with the right sort of a build. Procedural wings sure come in handy for building efficient blades.

[Strawberry]

27 minutes ago, K^2 said:

Can we do ye olde style helicopter built with reaction wheels? I know they're weaker than KSP's, but you also shouldn't need a whole lot of torque here with the right sort of a build. Procedural wings sure come in handy for building efficient blades.

That's what Ive been trying, I may just be doing bad technique but I cant get them to work. It feels like I can get close to taking off but not quite there. By the way tip for this, a small reaction wheel is half as powerful as a medium one, but takes up a fourth of the space. 

[K^2]

Hm, yeah, after playing around with it, I think a helicopter is a no-go on Jool. The atmo is just too soupy, so even when I hacked the rotation speed, it'd get slowed way back down again. It doesn't take a lot to get a glider going, though. If one figures out how to do bearings, a prop plane might still be viable. Another potential option is an ornithopter of some kind using bendy physics and maybe landing legs to provide flex? Going to keep experimenting, but I'm back to the drawing board on this for now.

A nice thing about KSP2's save system is that it's a lot easier to hack a starting location of a quick save.  Here's a location block for starting about 180m over the equator going North at 10m/s. Note that the orbital elements are completely bogus (Carried over from Kerbin), but they get ignored due to the LocationType set to SurfaceCoordinates. If you want to start at a higher elevation, increase the number for localPostion.x. Note that it's from planet's origin, so 6000200 is 200m above the "sea level", or about 180m above surface at the relevant location.

      "location": {
        "LocationType": "SurfaceCoordinates",
        "serializedOrbit": {
          "referenceBodyGuid": "Jool",
          "inclination": 0.6278264218015327,
          "eccentricity": 0.99479775850994834,
          "semiMajorAxis": 300841.16646634549,
          "longitudeOfAscendingNode": 130.33767680426303,
          "argumentOfPeriapsis": 90.173730836902038,
          "meanAnomalyAtEpoch": 3.1415913216744689,
          "epoch": 23094.907076562049
        },
        "surfaceLocation": null,
        "rigidbodyState": {
          "referenceTransformGuid": "Jool",
          "referenceFrameType": "Body",
          "localVelocity": {
            "x": 0.0,
            "y": 0.0,
            "z": 10.0
          },
          "localAngularVelocity": {
            "x": 0.0,
            "y": 0.0,
            "z": 0.0
          },
          "localPosition": {
            "x": 6000200.0,
            "y": 0.0,
            "z": 0.0
          },
          "localRotation": {
            "x": 0.0,
            "y": 0.0,
            "z": 0.70710678118,
            "w": -0.70710678118
          },
          "PhysicsMode": "RigidBody"
        },
        "originatingSimObject": {
          "Guid": "00000000-0000-0000-0000-000000000000",
          "DebugName": null
        }
      },

 

[Strawberry]

1 hour ago, K^2 said:

Hm, yeah, after playing around with it, I think a helicopter is a no-go on Jool. The atmo is just too soupy, so even when I hacked the rotation speed, it'd get slowed way back down again. It doesn't take a lot to get a glider going, though. If one figures out how to do bearings, a prop plane might still be viable. Another potential option is an ornithopter of some kind using bendy physics and maybe landing legs to provide flex? Going to keep experimenting, but I'm back to the drawing board on this for now.

Honestly I couldnt get one working on kerbin either, its not hard to stack reaction wheels but its just hard to have the wings produce enough lift for takeoff. The aero system generally feels better but it feels off for helicopters and I dont know why. I feel like we should work on getting a kerbin helicopter working before we design one for Jool (If youve gotten one on kerbin working please share the design).

Edited March 1, 2023 by Strawberry

[K^2]

6 hours ago, Strawberry said:

The aero system generally feels better but it feels off for helicopters and I dont know why.

Oh! I think I know why. I'm going to have to go build a new experimental craft!

Edit: Ok, I don't know exactly why, but there's definitely a bug in how the wings are processed. It seems to be using craft CoM movement to either determine the AoA, or the velocity, or both. In either case, that would imply that rotary motion is effectively ignored. I'm playing with stages and seeing if I can come up with some sort of an arrangement that lets me use the torque to generate lift, but I'm starting to lose hope. I'm sure this will get fixed eventually, but that isn't helpful short-term.

This suggest that ornithopter isn't viable either. We need an actual linear drive of some sort to maintain forward motion of the craft in order to get lift out of wings. It'd have to be a Kraken drive of some sort, I fear.  (Something, something, chosen one, something, something, balance to the force.)

Edit 2: Got something very basic sort of working. It takes off. Briefly. Since the aerodynamics of rotorcraft might be of interest beyond Jool exploration, I've split that off into its own topic: Rotorcraft Workaround

Edit 3: The small NTR works on Jool, and I have been able to build a nuclear jet airplane, which managed to climb to a little over 50m of altitude. That is currently my record for ascent on Jool. I'm sure that can be improved on, but climbing even to 7k where any chemical rockets become viable seems just as out of reach as before. Experimentation continues.

Edited March 1, 2023 by K^2

[Strawberry]

19 hours ago, K^2 said:

Edit: Ok, I don't know exactly why, but there's definitely a bug in how the wings are processed. It seems to be using craft CoM movement to either determine the AoA, or the velocity, or both. In either case, that would imply that rotary motion is effectively ignored. I'm playing with stages and seeing if I can come up with some sort of an arrangement that lets me use the torque to generate lift, but I'm starting to lose hope. I'm sure this will get fixed eventually, but that isn't helpful short-term.

This suggest that ornithopter isn't viable either. We need an actual linear drive of some sort to maintain forward motion of the craft in order to get lift out of wings. It'd have to be a Kraken drive of some sort, I fear.  (Something, something, chosen one, something, something, balance to the force.)

Edit 2: Got something very basic sort of working. It takes off. Briefly. Since the aerodynamics of rotorcraft might be of interest beyond Jool exploration, I've split that off into its own topic: Rotorcraft Workaround

Edit 3: The small NTR works on Jool, and I have been able to build a nuclear jet airplane, which managed to climb to a little over 50m of altitude. That is currently my record for ascent on Jool. I'm sure that can be improved on, but climbing even to 7k where any chemical rockets become viable seems just as out of reach as before. Experimentation continues.

Hmm okay this is bad news, kracken drive would definitely work as people have found those, but its no more challenging using one on jool then it is using it elsewhere so not really interested in that. I need to test if SWERV or dawn work on Jool, as those may be promising, but yeah this is definitely bad news. The jank rotorcraft route is interesting but seems like pain to design.

[magnemoe]

20 hours ago, K^2 said:

Oh! I think I know why. I'm going to have to go build a new experimental craft!

Edit: Ok, I don't know exactly why, but there's definitely a bug in how the wings are processed. It seems to be using craft CoM movement to either determine the AoA, or the velocity, or both. In either case, that would imply that rotary motion is effectively ignored. I'm playing with stages and seeing if I can come up with some sort of an arrangement that lets me use the torque to generate lift, but I'm starting to lose hope. I'm sure this will get fixed eventually, but that isn't helpful short-term.

This suggest that ornithopter isn't viable either. We need an actual linear drive of some sort to maintain forward motion of the craft in order to get lift out of wings. It'd have to be a Kraken drive of some sort, I fear.  (Something, something, chosen one, something, something, balance to the force.)

Edit 2: Got something very basic sort of working. It takes off. Briefly. Since the aerodynamics of rotorcraft might be of interest beyond Jool exploration, I've split that off into its own topic: Rotorcraft Workaround

Edit 3: The small NTR works on Jool, and I have been able to build a nuclear jet airplane, which managed to climb to a little over 50m of altitude. That is currently my record for ascent on Jool. I'm sure that can be improved on, but climbing even to 7k where any chemical rockets become viable seems just as out of reach as before. Experimentation continues.

We had that spinning helicopter thing in the short clip before launch who used rotating wings for lift. 

[Strawberry]

25 minutes ago, magnemoe said:

We had that spinning helicopter thing in the short clip before launch who used rotating wings for lift. 

Tried to roughly recreate that and I couldnt get it to work, Im guessing the aero model in that build is different then the one we got now

[K^2]

3 hours ago, Strawberry said:

Hmm okay this is bad news, kracken drive would definitely work as people have found those, but its no more challenging using one on jool then it is using it elsewhere so not really interested in that.

Out of curiosity, I did try building the jet/landing gear Kraken drive with NERVA, and it didn't work. That one felt less cheaty, since it at least uses fuel, but either I did something wrong, or it doesn't work with NERVA, or maybe it doesn't work at high pressure.

I'm not fully giving up on rotors. I should be able to build a larger rocket with proper hubs that can support debris-based rotor blades. It does, however, mean a much, much larger lander. So that will be an interesting challenge with all the Krakening.

2 hours ago, magnemoe said:

We had that spinning helicopter thing in the short clip before launch who used rotating wings for lift. 

Yeah, something clearly changed. Maybe someone was doing "optimizations," and "simplified," some code somewhere. Sometimes, it might look like you can take a computation out of a for-each loop to make things snappier, but for aerodynamics, you really can't.

[Strawberry]

1 minute ago, K^2 said:

Out of curiosity, I did try building the jet/landing gear Kraken drive with NERVA, and it didn't work. That one felt less cheaty, since it at least uses fuel, but either I did something wrong, or it doesn't work with NERVA, or maybe it doesn't work at high pressure.

I'm not fully giving up on rotors. I should be able to build a larger rocket with proper hubs that can support debris-based rotor blades. It does, however, mean a much, much larger lander. So that will be an interesting challenge with all the Krakening.

The design I had for my Jool take off involved an XL fariring so larger isnt necessarily a bad thing, sadly that design save got krackened which sucks. Making a space plane would definitely simplify a lot of things so if space plane is the best route to go then that could be great.

[magnemoe]

2 hours ago, Strawberry said:

Tried to roughly recreate that and I couldnt get it to work, Im guessing the aero model in that build is different then the one we got now

Yes saw your post above now this strategy did not work.

[K^2]

2 hours ago, Strawberry said:

The design I had for my Jool take off involved an XL fariring so larger isnt necessarily a bad thing, sadly that design save got krackened which sucks.

This is why I'm save-hacking for now. I'm sure I'll try a real mission once I have a working ascent vehicle, but fighting Kraken on top of long transfers only to find out that you need an adjustment to design is not working for me.  I'm happy enough just "teleporting" a craft from KSC runway to the surface of Jool and making sure I can take off from there.

Edit: Need. More. Torque. (But also completely proves that the aerodynamics is bugged, as I can spin up to much higher speeds before detaching the blades.)

About 15m altitude reached, which is not as good as 50m I got on Nerva, but then this is pure RTG power.

Edited March 2, 2023 by K^2

[Strawberry]

Dawn engine turns off on around 88k, however NERV has a minimum cap of 7.5 kn of thrust and 90 isp at any altitude, and SWERV has a minimum cap of 160 isp and 77 kn of thrust. Making a SWERV spaceplane seems like the best route