Jump to content

Starwaster

Members
  • Posts

    9,174
  • Joined

  • Last visited

Posts posted by Starwaster

  1. @AnFa Your rocket may not have the control authority to fly the path that MJ is trying to fly. In fact, the Classic Ascent is basically defining a static flight path based on the settings in the Ascent Path Editor.  A rocket with sufficient thrust AND  control authority will get to orbit over Kerbin with the default settings. You say you can get it there manually.  If that's so then adjust the shape of the curve in the editor to match your manually flown path. But understand that the classic ascent guidance doesn't do anything more than follow a statically defined curve. It's on you to either build a rocket that can fly within those parameters or to change the parameters to something your rocket can fly.

  2. On 9/20/2023 at 5:55 AM, cfloutier said:

    hey you modified MechJeb !!! so cool. feel free to send pullrequest to K2-D2 if you have suggestion. I tried to have a clear code. and I hope anyone can find how to enhance it.

    I'll take a look and see if I can contribute something useful.

  3. 56 minutes ago, cfloutier said:

    for the moment it just try to lower whole speed.
    A ratio is used to compute speed depending on altitude. It is always facing retro speed direction . 

    A better approach would have been to tring to remove horizontal speed by changin the attitude of the vessel. the drone mode have got such a possibility and I will add this optionnal feature to the landing pilot. it could avoid trouble when the horizontal speed is really high  

    I wrote something for MJ2 that I never committed that tapped into the suicide burn timer and pitch up enough to spread the necessary burn over the entire estimated time to impact/landing.  (player can input a margin to add to the timer if the lander doesn't seem to be  pitching up enough). It relied on calculations already being performed by MJ. Could something similar be useful here?

    Quote

    anyway : a new version of K2-D2 is available with a first draft of the docking pilot.

    First draft because  the UI is awfull and there are meny missing features but it can help a lot your final approch by automatically rotating the vessel in the dock direction and centering to the proper approach line

    Can't wait to try it!

  4. On 8/19/2023 at 6:25 AM, Kampfsanni said:

    The Docking Autopilot has a glitch that let him lock up right before the final push from time to time. Switching it off and directly on again  normaly solves that and will dock immediately.

    https://forum.kerbalspaceprogram.com/topic/154834-112x-anatid-robotics-mumech-mechjeb-autopilot-2143-4th-march-2023/page/168/#comment-4247646 

     

    I managed to reproduce this when I had speed limit at 0.1. Changing the limit to 1 immediately caused it to finish the docking.

    I'm not sure to what extent speed limit was a factor or not as in your previous screenshots, you always had your limit set at 1m.

    I'll take a look at the code since I was the last one to make major changes to the docking code. (some minor change by someone else but I don't think that's a factor, I managed to repro with an older branch predating that change)

    One thing that might be a factor is the undocking and redocking to a different port. If you're initiating docking to another  port from within the bounding box then the docking AP could get confused somewhere along the way because it has to first back out of the box in a safe manner, move around the station and then re-enter the box from a new angle. That's alone involves some complex maneuvering by the AP to get out, go around and then back in. Before I did some revamping of the docking code, it had a nasty tendency to try to fly through the station to get to the new docking port.

    I think it might be getting hung up on the corner of the bounding box. It would be trying to maintain the safe distance while simultaneously moving laterally to the docking axis and that could produce contradictory velocity changes.

  5. A more likely configuration would be a compressor fan with bypass doors that close as the engine transitions to supersonic / ramjet speeds. Then later transitions to scramjet and then rocket. There are research documents proposing this. Having it also be a rocket engine might not be realistic or feasible, but I wrote up a config that does turn one of the OPT parts into a hybrid scramjet / rocket engine using the data from the research document for the Mach curve data. I'll post it later today.

    Edit: Ooops, I kind of skimmed over parts of your question because it was early and I had to go make breakfast, so I'm looking at it now and I'm not really sure what you were asking here... I could still post my config but I'm not sure it's what you're looking for?

  6. @cfloutier When braking for landing, does it give separate consideration to vertical velocity from horizontal velocity or is it just trying to null out the velocity vector as a whole? 

    (A use case would be an Apollo style lunar module 'shallow' descent where maybe 90% of the thrust is cancelling horizontal velocity and the remainder used to manage descent velocity)

  7. On 7/30/2023 at 10:43 PM, SpacedInvader said:

    I have tried using hydrolox to fuel Duna missions in the past and things didn't really go so well because I couldn't really prevent boiloff, though IDK if it had anything to do with this specific issue or just failing to properly account for it. Is there an easy(ish) way to test this? Seems like at the very least, if it really is going to be an issue for me, it can be quashed with a MM patch, but I'd like to be sure its needed for my applications.

    What it looks like is this: A lower level of boiloff at the start of the mission ramping up to excessively higher boiloff over time. Time warps high enough to trigger analytical heat will make it happen a lot faster.

    Lower/higher being relative to before the changes were made (March 2022 IIRC).

    What happens is that the part has more thermal inertia than it should have and the cooling effect of boiloff is diminished. You're basically having to cool down not just the tank part, but the cryogens inside. Imagine that your LH2/LOX are being heated to around 172+ Kelvin. IRL that can't happen because they never go above their boiling temperature until after they've changed state into a gas.

    @RESOURCE_DEFINITION[*]:HAS[#vsp[*],#hsp[*]]:FINAL
    {
    	@hsp = 0
    }

    That will patch every resource to set specific heat to 0 if both specific heat and latent heat are present.  (latent heat of vaporization should only be present for cryogenic resources)

  8. 11 hours ago, SpacedInvader said:

    That all said, what's different about your branch compared to the official? Looks like its been several years since your branch has been updated based on the github info?

    It lacks certain changes that were made in the official branch. Some of it is just general: I feel that RF has become more and more bloated with features that were made to serve Realism Overhaul, which  RF predates. RF used to be a much simpler mod introducing a variety of real resources + boiloff for cryogenic resources + normalizing real world volume units.

    But also, there have been specific changes that I have problems with such as to the Unmanaged Resources feature. The changes made break that feature as it was originally designed. (and I should know since I designed it)

    Another such change is adding specific heat (hsp) back into cryogenic resources. (I removed specific heat from cryo quite a few years back). Without going into too much detail, the net impact is that cryo resources will over time absorb more thermal energy than is realistically possible. That means more heat energy that needs to be removed by radiators/cryocoolers and it's energy that shouldn't be present. The full answer is a good deal more complex than that and more complex than I feel like dealing with here right now. (I will say this though: In terms of heat, there's sensible heat and latent heat and you should never be dealing with both at the same time. Cryo should only ever be dealing with latent heat. Never sensible because it is assumed to be at its boiloff temperature at all times.)

    That second item is something you will probably never notice if you only deal with cryogenic resources for launch and not long term storage, such as ISRU or cryo depots. If you play with either of those then you may have noticed that over time you built up more heat than you could easily get rid of.

  9. @SpacedInvader I'll take a poke at it and see if I can find out what's happening, but I have to warn you, I'm not involved with Real Fuels development anymore and only use my own branch for gameplay.

    I'll have to install the current official branch and the other mods you mentioned. I can do that, but you might have better luck going into the Realism Overhaul discord channel and bugging them about it directly. I don't know if they even monitor these forums anymore. You could try making an issue on the github repository page but I don't think they really look at those anymore either. They definitely don't care about the last issue I raised.

  10. On 6/16/2023 at 10:40 PM, SpacedInvader said:

    So I know its been a couple of years since any of the mod creators for this have replied, so I'm not too hopeful for an answer, but I am curious if we're supposed to be using "useRealisticMass=false" or not for this? I've tried multiple combinations and it feels like leaving the settings on "true" leads to fairly small rockets, while setting either of the options to false leads to comically oversized rockets (i.e. just to put a single Kerbal in orbit takes an absurdly large rocket and is nearly impossible with early tech). In searching through the thread I noticed at one point there was some talk about an update to rebalance things and remove the reliance on the setting, but looking at the patch notes there was only one update after that before it development stopped and its unclear if that rebalance happened in that update. So basically, I'm left with the following question: Have the configs been reworked such that we're just supposed to use the default RF settings, or are the current configs not rebalanced and we're supposed to set realistic mass to false and really work hard to engineer workable rockets?

    Thanks

    When useRealisticMass is set to false, engines are scaled 4x and tanks are scaled 4.8x. If set to true then the part's configured mass is used as is. 

    There is an implicit assumption there that engines are scaled realistically already and if useRealisticMass = true then don't touch engine/tank masses. It also assumes that they are set to 1/4th of what they would be in stock, so the simple answer to your question would be that if you're using a stock Kerbol star system then the setting should be set to false. A realistically scaled star system (Real Solar System, Kerbin 10x, JNSQ) should have the setting at true.

    BUT. (caveat time) This mod, while having scaled down engine masses, does not necessarily have engine masses at 1/4th of their stock sizes. For instance, MassiveBooster (I think that's the shuttle SRB styled one...) has a stock mass of 4. But this mod sets it to to 2.4. If you set useRealisticMass = false then you're going to end up with a booster that is 9.6 tons. YOU DO NOT WANT THAT. That's more than double the stock mass for that part. I'm not sure of the relevant config file's history, maybe the ratio between mod and stock used to be different.  But I would set useReaiisticMass = true to avoid unnecessary headaches. 

     

  11. On 6/18/2023 at 11:44 PM, dOnut14 said:

    is it possible to make planets smaller than stock, as I am trying to  recrate This Mod for modern versions of ksp, and I am also not quite sure how to go about making a config for that kind of thing.

    Yes absolutely. You could even do it like The Little Prince if you wanted. 

  12. On 4/9/2023 at 6:40 PM, magnemoe said:

    This. Its like using an mainsail engine in an smallest rocket to orbit contest.  Its designed to push heavy stuff hard. Nuclear thermal and especially the SWERV is for pushing heavy stuff hard. 
    In late game they would be niche as they can melt ice for hydrogen, who makes sense if you are a billion km from civilization. 
    We will get orion pulse nuclear engines who are much heavier but beat them in trust and isp, but the fuel is nuclear bombs, not something you mass produce if you don't have an serious industrial base. 
    This is that you need to build an industrial base at Jool who you need to go interstellar. 
     

    Also, another thing to consider is that NERV is based on late 50s technology and SWERV represents technology that is still on the drawing board for us even now.

  13. On 3/24/2023 at 11:51 AM, sisyphean said:

    Can anyone tell my why im still getting random engine failures when i specifically turned the feature off and how i can go about stopping them from occurring in future?

    these are the settings for those preferences from my save file so i don't see why i keep getting engines just shutting down mid flight

            PreferencesReliability
            {
                highlights = True
                mtbfFailures = False
                criticalChance = 0
                safeModeChance = 0
                incentiveRedundancy = True
                engineFailures = False
                ignitionFailureChance = 0
                engineOperationFailureChance = 0
            }

    Could the failure be from something other than Kerbalism? (which is what that looks like it's from)

    For instance, Real Fuels engines can fail if they ingest vapor, which can happen due to unstable ullage. In orbit it might happen because the fuel drifted around and you need to settle it with RCS or solid fuel rockets. (ullage motors). Stage separation during launch can cause it. In which case you need to either delay ignition  while ullage motors settle the tank or use hot separation. (where you ignite the next stage motor just before separation)

  14. Not a question but:

    Please PLEASE for god's sake, don't combine part thermal mass with resource thermal mass. (if it's implemented at all)

    Resource thermodynamics  should be separate from part thermodynamics with heat transferring between each other. Vitally important for boiloff mechanics, whether it's because you choose to implement boiloff or to allow mods to properly an realistically implement boiloff mechanics.

    That's been a pet peeve of mine ever since KSP1 implemented resource  thermal mass.

  15. On 3/14/2023 at 8:28 AM, Fullmetal Analyst said:

    i dont get the point of using nuclear engines, they seem to need huge tanks to get a reasonable amount of dV, and seem to be worse than other engines in general

    here some example:

    two rockets with same payload, both have almost same dV, but the hydrogen version is a lot bulkier while actually having 100 less dV
    eo8C3Y1.png
    (hydrogen / terrier engine)
    p1TRec9.pngTlX8kNS.png

    so whats actually the point of these nuclear engines?

    they dont seem really useful compared to other fuel types

     

    when i remove the truss and SAS the stats are still ridiculous
    wOKpH36.pngImmuQtt.png

    You're taking a bare minimum design there which is not good for comparison purposes. You don't really have a payload except for the capsule. Nuclear isn't a PnP option that's going to be great for all designs and it's definitely not always better. 

    It's going to be orbital only (or upper stage, at a minimum). It'll probably be assemble-in-orbit where you don't want to be transporting tanks of dense propellants into orbit. In the real world we'd be talking 8-10m diameter. Not sure what that would be for Kerbin.

    To give a real world example of where nuclear could have been used for an upper stage, it was considered for use in Saturn V's third stage, the S-IVB. It would have been less powerful than the J2 but would have ditched the LOX tank. The tank would have been lengthened a bit but would have been lighter overall because they would also have ditched the common bulkhead between the H2/O2 tanks. The end result would have been either greater dV because of reduced third stage mass or more payload.

    So there are times and designs you'd use nuclear, but you can't just drop it in any design and expect it to be better.

  16. On 3/3/2023 at 4:26 PM, wreckreation said:

    Of course.  

    Though I feel I should point out that any misbehavior is far more likely to come from the decision to use a mod on a version of ksp the author has not verified it works on, and that the mod itself explicitly warns against.  :D

    As far as that goes, it'll work but there were (IIRC) UI bugs due to changes that were made.  Nothing that recompiling from the source wouldn't fix.

×
×
  • Create New...