Jump to content

Muetdhiver

Members
  • Posts

    219
  • Joined

  • Last visited

Everything posted by Muetdhiver

  1. Hi Kurgut, I will add a detailed install/use guide in the githhub readme next time I update the scripts.
  2. Excellent ! I am in the process of assembling a JNSQ Duna Expedition vessel, and I had a few ideas that came along : 1) SSPX has a bunch of cargo containers, would it be possible to alter the B9 config to allow storing sludge/consumables into them ? 2) SSPX has a bunch of "utility" station modules. In the idea of having modules to turn bad air to air, I think it would be nice if those parts had this capability. This is just ideas I got, feel free to discard them, as I know that things that appears "simple" can be quite hard to mod in Anyway, you mod is good fun to ramp up the logistics planning.
  3. I have been enjoying the mod quite a bit. However, there is one thing unclear to me : Is there a way to turn bad air into air using anything else than a greenhouse ? I was expecting the chemical plant to offer this function, since it is placed in the recyling tech node, but it seems that it has no function. This would be very usefull to be honest.
  4. Wayfinder https://github.com/Muetdhiver-lab/Wayfinder Wayfinder is a python based Multiple Gravity Assist search tool for KSP, including support for the JNSQ planet pack. It allows for efficient search of gravity assist sequences using state of the art tools using pykep/pygmo packages from ESA. At the moment it's a set of python scripts with no GUI. Requires python 3.6 or 3.7 How does it work : Wayfinder uses a job batch system with the results of said jobs saved in an xslx format for storage and readability. Jobs can be added with the desired parameters (fly-by sequence, insertion type, search space bining, optimizsation level and so on). Once added, jobs can be run in batches, and results will be saved. Once results are saved, the results can be searched and accessed with a few utilites as : - finding the best result in a given set (sequence + launch dates) - display several sequences for comparison in terms of DV cost and Time of flight. - compare different sequences in a DV cost vs launch date line plot - display a job result as a flight plan in a text format References and thanks: This work would not have been possible without - pykep : https://esa.github.io/pykep/ - pygmo : https://esa.github.io/pygmo2/ - transfer planer for JNSQ : https://github.com/LouisB3/ksp-lwp-jnsq - the original transfer planer : https://github.com/alexmoon/ksp Special thanks to ESA to make those package available to the public, it's awesome. An other big thank you to the Jool5 Caveman Challenge , as it kickstarted the whole project.
  5. Cargo bays are pretty much the caveman challenge special sauce magic ingredient Thermal shield, aerobrake and may also serve to protect delicate gear, all for a measly 100kg ? Best part ever.
  6. I do not think there is a pad weight limit. During tests I went to ~50t on the pad. The biggest lawn assembly I know off was during the jool5 caveman insanity to lift the core stage forthe jool vessel.
  7. What the people above said : the challenge is a tad ill defined. Either it's 3333 on the pad, and then it's a matter of putting an ion engine to get anywhere, or it's 3333 in vaccuum, which means that jets and roton style contraption become the way to go. I will add that in theory if one can get to Eve, on can get anywhere with slingshots and enough time.
  8. It works okay. on the plus side you save at least a part with the integrated decoupler, potentially 2 with the heat shield. However, for me it has two big downsides : poor aerodynamics, and no reaction wheel. I tend to use service bays as heat shields, works amazing, protects all the precious science and can act as a poor-man aerobrake by opening the doors.
  9. Diamond and NCD are brutal. With regard to funds, the money maker is mun temperature/crew survey missions, re-usability* and exploration missions, and never, ever let a science instrument burn *It is possible to recover reliant/swivel based rockets in diamond/ncd from LK orbit. Mun returns are trickier but doable. Recovering terrier based upper stages is also possible from Mun, but it's really tricky and getting to 0EC is death.
  10. Indeed, the spark is better than the terrier (in terms of DV) till you go above FLT800+capsule. If TWR is no issue, then it's a darn good caveman engine (say if you want to do Gilly, or Ike).
  11. Note that if you have a service bay, drogue chutes are not needed. Just open the doors and you'll bleed speed real quick. If you aim at a water body l found that you may be able to recover the terrier stage. This type of recovery is possible from a mun return, but the reentry window is very small. For caveman docking : youyou need to launch a bit before the target flies over the LAX, and aim an orbit that shares Pe but has Ap a tad higher than the target if it is behind you. (Or a tad lower if before you, swapping Pe and Ap above) Then, trim relative inclination. After that look at the distance between the craft in the map and look at how much it shrinks with each orbit. Then adjust Pe or Ap to get a close encounter, select target then kill relative velocity (which should be low). The enjoy the docking
  12. I do not think that LKO stations make a whole lot of sense in a caveman run. That being said, I have been using a refuel station in low polar mun orbit in a JNSQ caveman advanced run to great effect. Simply put : re-using the lander several time before returning to kerbin saves a lot of fuel down the line. I just have a mk1 based lander with a Jr-port and dock a fuel tank on the nose. Do a landing, RDV and then swap for a new one. The heat shield is replaced by a service bay with a science box to stack the science. Note that "station" is kind of a lie. It was more a train of tanks docked in a string. @fourfa your minmus optimized lander looks oddly familiar '^^ I see that you too have opted for the front mounted drop tank on the upper stage
  13. I have done on pad assembly many times for caveman plays. 1) using orbital assembly is the best way in 90% of the case. Faster, easier and cheaper. 2) mass fraction of such pad assembled rockets is poor due the atrocious woblyness requiring *ahem* suboptimal gravity turns. 3) if you do need a payload above 4t then go for it. But it is a nutty thing btw, there are ship files for a pad assembled rocket in the github repo of the community jool5 caveman insanity. It might prove usefull to you. It was used to lift the core engine/docking hub of the jool ship. Despite all their troubles and ugly, I love those silly things. And yes, if you play with MH the bobcat is amazing. Not in the sense that allows to lift more mass to LKO, but in the sense that it allows to use more parts for the payload since the reliant+3-4 sparks+3-4 nosecones becomes a bobcat plus the 1.875 to 1.25 m tank, a net gain of 5-7 parts...
  14. There is no github to follow for now. I use pickle for serialization/storage and until I change it to something safe there will not be. However a replacement is in the works with JSON. I'm testing it on a Kerbin-Duna-Duna-Edna mission at the moment. Non-ballistic grav assist sequences are really tricky to nail down, even when you know exactly where and when to look. So hopefully, soonTM
  15. I have been working on a pretty nice tool to find flybys in KSP/JSNQ and I'm finally getting to the in-game application stage with my JNSQ caveman evolved mission to Edna. Here is all the best gravity assists found from Kerbin to Edna with 2 to 3 grav assists on the way, for a launch window between day 750 and 950. ToF is color coded. It's not release ready material, but it's getting there. JNSQ is perfect for this stuff with the high DV costs.
  16. The LK is frankly amazing work. It's looking gorgeous.
  17. Yes you can. It is indeed possible to do a moho pump down via multiple flyby's. Keep in mind that in order to chain them, you need to aim for partial resonant orbits. I usually stop when I get down to 11:10 as it becomes very long for a low gain. 3-4 flyby's pump down is doable and worth it, so long as you stay very precise. Moho is my favorite solar pinball target. You can lower the DV needs quite a lot. Here is a ~3.85 km/s flight plan for the Kerbin-Eve-Eve-Moho route : With a bit of pump-down you might be able to fit it in the 3.5 km/s enveloppe.
  18. The new vostok/soyuz parts are fantastic, as is the solar panel toggle. My only gripe is with the fuel tanks dry mass which is way too high. This is an issue with the toroidal tanks and 5/6/8 core tanks both. I would suggest setting it in line with the regular tanks dry masses.
  19. I'm using the the pygmo library from ESA to tackle problems beyond the scope of lambert. For example finding moho transfers with multiple assists, which is immensly expensive w.o. high quality search algorithms. I will meditate on it for a few days. I might come up with some ideas.
  20. This is impressive work. Any chance to couple such a thing to a python based flyby sequence finder or triple impulse transfer solver ?
  21. I would recommend Moho. Definitly the hardest one to get to. It's a great opportunity to mak use of gravity assist to soften the DV costs and difficulty of capture.
  22. All I can say is that this is a top notch example of caveman tech. Kudos
  23. TBH I find that building vetically is still the easier way. But it's still quite awefull, and separating Jr-docked side boosters is glitchy as hell. On pad assembly is also very inefficient in terms of mass fraction delivered in LKO fom to the terrible ascent profile you have to use to avoid losing the rocket to wienerli instabilities altogether. IMO the "easiest" on pad assembly would be one 1.25 core held by the clamps on which two 1.25 mobile side boosters are "docked". This should allow for a ~7-8T 1.25m payload. 2.5m core is horrible but necessary for the core. TBH the 1.875 parts would alleviate some issues, maybe.
  24. The lander with-a-can is a lot more doable than the relay idea. However : the lander can will need to be maned, which means that we need two return seats in capsules. This adds yet more weight. In the best case scenario, if the total payload mass increases by 3.2x, it means that the transfer stage will need to be massive with all the wobblyness and wienerli behaviour that it implies along with the nightmare of multiple on-pad-assembly. It can be "dealt with" with doing a very low TWR ejection (~0.1) but that will make the ejection imprecise. An alternative would be to take the K-E-K-K-J route, which would cut the transfer stage mass a lot. For a Jool high elliptical orbit, the DV is ~2150 DV, or 6.14*ISP with a poodle, meaning a ~1.9 mass ratio. Let us round to 2 for margins. With a tank+engine dry mass fraction of 1/5th of the fuel load (optimistic-ish) we get 1.2T of fuel per ton of payload and 2.4T of vessel in LKO for each 1T of vessel in Jool HEO. Using a K-E-K-K-J route would lower the DV to ~1300 DV or ~1.5 mass ratio. This gives 0.55T of fuel per 1T of cago in HEO or 1.65T of vessel in LKO for each 1T of vessel in Jool HEO. Downside being having to pull off a triple gav assist without nodes and a kerbal on a ladder The gain is that the vessel to assemble in LKO is only twice the size of the fist mission, rather than 3.2x.
×
×
  • Create New...