Jump to content

arq

Members
  • Posts

    373
  • Joined

  • Last visited

Reputation

19 Good

Profile Information

  • About me
    Sr. Spacecraft Engineer
  1. Thanks for the relocate and the link. Sorry I misplaced this. Glad it's already a known issue.
  2. It seems that the bug-report section of the forums was moved or deleted at some point after release, and this was the closest I could find. There is a bug where control surfaces can end up with inverted pitch controls. I have attached a picture of a craft that produces this issue. It seems to stem from having control surfaces mounted at an angle too near the CoM. In the image, (counting inside to out) the 1st and 3rd control surfaces have the proper pitch deflection while the 2nd and 4th are inverted. All have the proper roll deflection. Counting outward from the fuselage: The 1st (innermost, inside the engines) set are mounted horizontal to the direction of flight and deflect properly. The 2nd set are mounted at ~30deg (whatever the swept wing angle is) and have inverted pitch control. The 3rd set are shifted further back using the offset tool. Any further forward than their current location will result in inversion, but they deflect properly where they are in this image. The 4th (outermost) set are angled as well (but now using the Control Surface 1 part instead of 5) and experience inversion. In fact, any surfaces mounted at that angle along that entire troublesome line (slanting back at 30deg) will experience inversion. An un-angled control surface causes no issue, and an angled surface sufficiently far back is fine. Placing a large fuel tank on the front of the craft (to shift the CoM far forward) resolves the issue. So it appears that this issue only occurs for control surfaces mounted with a rotation, and only when applying pitch input and if they are sufficiently close to the CoM. It does not seem to affect the roll axis, though it is possible an analogous configuration could produce this issue on that axis. Removing all mods from my install did not resolve the bug.
  3. The argument (shared by Scott Manley) that "LF+O tanks are fair for LV-Ns because of reduced propellant density" breaks down as soon as you notice that LF only tanks exist. Either tank fuel types should be tweakable (or, as a weaker option, add more LF only tanks) or the aircraft fuel tanks should be removed for 'balance' reasons (and what a wreck that would be for airbreathers).
  4. Here I am! While 10deg sounds exceptionally reasonable to me (a little conservative, even), the OPness of 15G+ 'pancake' re-entries (entering at stall AoA, where you can drop from orbital speeds to maneuvering speed in 10 seconds) is a bit much. The reason that I call it OP is because it completely subverts any 'challenge' introduced by re-entry. Humans don't handle such stresses very well, airframes don't, and I can't imagine it's great for kerbals either. Usually I play FAR which sorts this out (though I'm waiting to play stock 1.0.3 aerodynamics before I make the switch). It puts aerodynamic limits on wings (they'll shear off in this sort of re-entry) and implements hypersonic lift reduction and CoL changes that make holding a large AoA very difficult even if you wanted to. And DRE adds G-tolerances for kerbals. That said, I'm not sure what I would want Squad to do about it. The hypersonic aerodynamic changes are difficult to understand as a beginner and greatly complicate design, so I don't much like that idea. Aerodynamic stresses are an option but are possibly sidestepped with moar struts, not to mention they could be annoying in other contexts. So it may just be that pancake re-entries are here to stay and for people to use if they so desire.
  5. Even better. Thanks! I hadn't seen this before. I found it hard to believe that nobody had done this yet... I guess most of my questions should be answered by digging through that source code.
  6. Setup --- One persistent point of annoyance for me is the inability to see contract markers in flight view. They are visible in the map view and even add a marker to the navball, but missions such as 'perform seismic scan at XXX' often require me to land once near the target (after using the map and navball to get very close) and then a short hop to actually make it to the proper zone (which I usually overshoot because I don't know when I'm almost there). Trying to walk to a zone on EVA is even more difficult because there is no navball to give a bearing, so it involves stumbling around making rough guesses from the map. For these reasons I would like a mod that allows targeted contract markers to be visible in the flight view. This would make landing in a zone much easier to do on the first try, and considerably easier for EVA kerbals. This could be done by spawning a part at the location so that it may be targeted, though ideally this would be possible to do partless-ly. My questions --- If someone gets the strong desire to make this mod, go ahead. Otherwise I'm going to start plugging away (but I have limited modding experience). First off, does anyone know how to retrieve contract objective coordinates? Or, even better, the coordinates of an objective with "activate navigation" enabled? I really know nothing about accessing contracts through the API, as most of the wiki documentation and tutorials don't touch on it (most pre-dates contracts). Secondly, does anyone know how to create a dummy target? Spawning and despawning a vessel (ie a cubic octagonal strut or a flag) at the desired location should be easy enough, but it would be cleaner if I could force the target marker to a specific location without a vessel to target, or setting the marker itself to be targeted using the normal system. Any answers or references regarding the above information would be greatly appreciated. I don't expect I'll have the patience to finish this without a little guidance.
  7. Do satellite and rescue missions until you unlock patched conics in the tracking station (not strictly necessary but much easier this way), then do a mission to Minmus. Even though it *looks* 3x further away than Mun, because of how orbits work it is actually just a tiny bit further away form LKO (roughly 900m/s versus 850m/s). The dramatically lower gravity makes orbiting, landing, and returning from there much easier than Mun.
  8. It is definitely possible to build a simple plane with just the first aircraft tech, and for the sake of forcing a simple design probably easier. Yes, KSP aerodynamics are a crude approximation of the real thing, but they get the gist (and the aerodynamics now are waaaaay better than they were in beta). As it turns out, making 'flyable' planes in a flight simulator is a lot easier because they can have all their parameters input directly whereas KSP has to mash something together from a bunch of blocks stapled together and then try to figure out the aerodynamic implications. However, there is a mod called FAR (check the modding section) that dramatically improves the aerodynamics model in the game. At the moment I believe it's still ironing out a few minor bugs, but it will make aircraft behave much more realistically. But FAR usually makes design and flight considerably more complicated so you may want to get the hang of stock aircraft first. By the way, aircraft stall above 30deg AoA or so. It becomes obvious from a dramatic reduction in speed and (IIRC) a shuddering noise.
  9. cybersol has more good suggestions for you. You'll still need to move the CoL back more. Larger controls surfaces will allow you to pitch up and get off the runway easier. Also, if you need to use the T400 tanks at least make sure you remove the oxidizer by right-clicking in the editor - that will dramatically reduce the weight which will make the plane easier to handle. As for landing, trying to land anything above 80-100m/s is risky if not suicide. A plane this small can probably fly at 40m/s and that would be a good speed to try to landing. This may involve gliding for awhile to burn off speed. Try landing on the runway (if possible) or the flat area immediately surrounding KSC. Even small hills require some special consideration for landing and will be more difficult - the flatter the easier.
  10. Move the CoL back so that is maybe a quarter of one of those 400 tanks behind, at least (in this picture somewhere slightly ahead of the rear landing gear). Personally I would change it down to a single fuel tank to start (jet engines are VERY efficient) and use one of the plane ones (liquid fuel only) if you can. If that still doesn't work try swapping out the Delta Deluxe Winglets for one of the canards or other surfaces where the entire thing moves.
  11. If your plane is flipping out of control, chances are that your center of drag is ahead of your center of mass. This means that if you go past a certain angle (that gets narrower and narrower at higher speed), the forces trying to flip your plane will be too much for your control surfaces to handle and your plane will turn around backwards and likely crash shortly afterward. In the spaceplane hanger, enable both the center-of-mass and center-of-lift indicators. Make sure that the CoL is behind the CoM. Or try flying some of the stock planes before you do your own. Start with small, simple planes: a cockpit, single Mk1 fuel tank, basic jet engine, air intake, landing gear, wings, and a horizontal and vertical stabilizer. Make sure to use control surfaces for the stabilizers. Once it's built, move the wings backward until the CoL is a bit behind the CoM. Once you get the hang of building and flying basic planes you can get more elaborate. Larger planes with more engines can be interesting and eventually RAPIER-equipped spaceplanes are a fun challenge. Just remember that your CoM moves as fuel is spent and if it gets behind the CoL your plane can become very unstable.
  12. Go for the LV-N. It is the best for moving heavy payloads around in space. The aerospike is mainly a strong contender for Eve launches. The LV-T30/45 are usually preferable over the aerospike in the 1.5m category, because of the higher TWR of the 30 and the gimbal of the 45. The strength of the aerospike is that it retains thrust/ISP at much higher pressures than other rockets. However, only the bottom 5km or so of Kerbin's atmosphere is really thick enough for the aerospike to be any better than the other engines, and you don't spend very long there when launching. As mentioned, the exception is Eve (or Jool, if you have some reason to be deep there), which has a much denser atmosphere that renders most engines exceedingly weak. The aerospike has the best ISP (and maybe TWR? I'd have to look closer) in the lower Eve atmosphere by quite a bit. Many people also like the aerospike for its short profile that looks good on spaceplanes. However, from a performance standpoint the T30 is usually preferable (if you don't need the gimbal), since it is both lighter and more powerful (which can offset some of the ISP loss). The LV-909 also works well in this capacity, if less thrust is needed.
  13. A 2.5m nuclear rocket, modular Mk3 wings, fuel type-tweakable fuel tanks, a 0.625m and maybe 2.5m jet engine, some in-flight-addable version of struts (can't re-secure a part in a cargo bay or strut an assembled space station) and fuel pipes, possibly a (weak) electric propeller for low-altitude atmospheric flight, and an intermediate-sized solar panel would all be good in my book.
  14. Given the presence of spaceplane parts, the LF-minus-O tanks are a terrible balance 'solution' to the LV-N. If the LV-N needs to be brought in line then it needs less ISP (maybe set it to 600s) or a dedicated fuel type to make it consistent, and if it doesn't then we need tweakable fuel types in tanks in the stock game (and those would be welcome regardless). In the meantime, tweakable fuel types are also available through http://forum.kerbalspaceprogram.com/threads/121321-Stock-Fuel-Switch. This mod can supposedly be uninstalled without breaking stuff if this feature ever becomes stock.
  15. Thank the takspayers. Besides, it's for science.
×
×
  • Create New...