Jump to content

lemon1324

Members
  • Posts

    41
  • Joined

  • Last visited

Everything posted by lemon1324

  1. Make the Adjustable Landing Gear mod from 1.0.5 stock!
  2. I'm not sure this is true, actually. Rockets pump in their own fuel and oxidizer, so the absolute exhaust pressure is constant regardless of altitude. Jets operate at a constant compression ratio, so as the atmosphere gets thinner the absolute pressure should go down, while the pressure ratio is the same at all altitudes. I'm pretty sure air breathing jets should have similar visual expansion as a result.
  3. Holding prograde and burning with an ion engine would be a spiral, the only problem is currently you can't time warp and hold local prograde at the same time - your orientation during warp is fixed w.r.t. the inertial frame, not the orbit-relative frame.
  4. Your ship will have some velocity vector with respect to the parent body before the burn, and some velocity vector after (which is on your final orbit). For impulsive burns (i.e. burn time much shorter than orbital period), the most efficient burn should be to take the velocity vector difference and burn only in that direction. The difference in KSP1 and KSP2 maneuver nodes (ignoring bugs), is that KSP1 nodes assume the burn happens instantaneously for computing the changed orbit, but this isn't actually true, leading to the "start your burn ahead of the node" we've gotten used to. Even that isn't totally right, since your mass changes, so more than half of the delta-v comes in the second half of the burn. KSP2 nodes actually compute the time-varying thrust along your burn vector, assuming the burn starts at the node. This means (again, assuming no bugs) executing the node as planned will result in the final previewed orbit, regardless of the thrust-to-mass of your ship. This won't be true for a spiral-out type of burn with an ion engine, but at least currently that type of maneuver isn't supported under warp because orientation cannot change under warp.
  5. End-mounting the new procedural wings on the previous wing lets you split the trailing edge into multiple control surfaces for things like inboard elevators and outboard ailerons. An option to "inherit root section from parent tip section" would make this a lot easier to adjust dynamically, however.
  6. Rather than the huge error messages when parts aren't operational (e.g. the solar panel in shadow), a more functional parts manager could be used to also monitor parts issues, by adding a warning or error icon on the part names.
  7. Am I the only one where the runway actively repels me when I'm trying to land? I'm flying a small airplane with small gear and whether I set auto suspension or play around with the parameters I can't seem to get any landings without some sort of crazy bouncing happening. Friction is set very low on the nosewheel, and this happens before I even try braking. It almost feels like the auto-suspension landing gear springs are being made stiffer after the plane compresses them but before the rebound, which would add nonphysical energy to the system.
  8. This was the first thing I did as well, built a plane and flew west to find K2, which, of course, wasn't there. On the plus side it does mean you can fly a proper glideslope approaching the runways instead of having to dive over the mountains.
  9. Procedural wings! Time to try some in-atmosphere circumnavigations.
  10. What are those engines hanging off the wing? They look awesome, and non-stock.
  11. Just saying, I think I hold that record with MET of 16hr 55m 52s on my triple cirumnav with basics.
  12. If you're having problems with the controllers not being fully stable, there's two things you can do: The controller causing you the most problems is the one that has the largest error/oscillation about its own set point. You can figure this out by opening the controller parameters for each controller, and then watch how the actual value changes with respect to the set-point for that controller. Tune controllers from the bottom up; for longitudinal motion, that means tuning the AoA controller, then the VSpd controller; the Altitude controller shouldn't need tuning, as it's just proportional. I've found that (and this will be generally true for cascaded controllers) the lowest controllers need the most tuning from craft to craft. I've never once needed to touch the altitude controller apart from setting VSpd limits.
  13. As the other crazy-use-of-turbofans guy, (belated) congratulations on those! Nice to see airplanes that make it around not on fire. Good times too. I'll have to get back into it and do something new with either the 0.625 or the Goliaths... I wonder which engine you should actually use to make the most circumnavigations on jets. Or maybe most Kerbals flown on a circumnavigation? Hmm...
  14. :resources should be a read-only tool to get info from parts. I'm pretty sure (haven't done it, haven't had too much time for KSP lately) you want to take a look at PartModule interfacing, which basically lets you do everything that shows up in the right mouse button menus. It should look something like "part:getmodule("moduleName")" to get the fuel tank module and then either :doaction or :setfield depending on which the fuel enable is. There's also getter functions so you can query the modules on a part and the fields/actions on the part. (check the PartModule documentation on the kOS github documentation)

  15. I'm pretty sure Mach 2 cruise with a more optimal descent is about as fast as it gets. Of course someone could always prove me wrong, but Mach 2 is already faster than the most efficient cruise speed, and I wasn't able to get anything with enough fuel to go much faster.
  16. Muhahahahaha! So, a couple things about KSP Stock: "Basic Jets" are configured not as turbofans, but as real-life turbojets. KSP turbojets are real-life turboramjets (think J58 on the SR-71). Note how in this plot I found, max thrust to weight happens near Mach 1.8 or so for basics. The drag model is maybe a little overzealous about reducing drag after punching through the transonic regime. I present... Nike, flown by Val on a circumnavigation in 1hr 51m 24s on basics. It was nontrivial to get through the transonic regime, but once supersonic, this was supercruising at Mach 2! And actually, the hardest part about this was finding a configuration that could even get past the transonic regime on just basics. I couldn't do it with the MkII system of parts, and even Nike requires a careful flight profile to do it. Same disclaimers about mods as my last entry; recorded data, plots, and files can be found on my Dropbox. Does this qualify as a velocity entry? It's about a factor of two slower than the slowest on that board, but also about a factor of two faster than the fastest circumnavigation on basics so far. I'll note that this was very suboptimal--I had 340 fuel left at the end, and the descent was leisurely rather than aggressive, so there's definitely lots of room to beat this time, pretty easily. I'm less interested in maintaining this one than having the longest range on basics (and of course, Nike is the Greek goddess of speed.... and victory )
  17. I'm certainly calling it a GE90, both because it's my favorite turbofan and because it looks exactly like one: The pylon is distinctive in how close to wing level the top of the engine is, and no other large turbofan actually looks quite the same. (besides which, the GE90 is the largest single turbofan by more than a foot fan diameter over the next, and also by maximum thrust--so if you pick a specific engine, well, this is it.) Though it does look silly on the airliner wing we currently have in stock, because the wing isn't large enough. The Mk3 fuselages are about big enough though, so hopefully we get a matching wing! (or apply tweakscale liberally)
  18. Ah, right. This is pretty much exactly what I needed for my expedition entry. Full post is on the previous page, but essentially I used Pilot Assistant for semiautomated piloting; to nail down my ascent trajectory I entered my desired parameters into Pilot Assistant, and then manually changed the target ascent rate at various altitude landmarks, then tried variations to optimize fuel use. For data logging, kOS (again, the script I used available on my last entry post). You can use kOS to log data to a file, and more importantly for me, I could define my own measurements to log to a file (e.g. fuel efficiency). Here's one such data log showing my final ascent profile. As far as I've seen, if you want truly automated autopilot that works for airplanes, you'll have to write your own in kOS. I'm working on a kOS script to do it, but it's not going to be done anytime soon.
  19. The "Ascent path editor" is for rocket ascent trajectories. You might be able to hack it to work, but MechJeb isn't great at airplanes anyway, so you'd probably get poor results. On my triple circumnavigation I used Pilot Assistant, which lets you tune the controller to match your aircraft, and with tweaking of the various output clamps you can get it to, for instance, ascend to 12000m at a climb rate of 0.1m/s, if that's the profile you want.
  20. Well, It's done at last! Three full circumnavigations, all done using only basic jets. Flown in the Icarus IV with Jeb commanding, and Bill in the co-pilot's seat. Total flight time 16hr 55m 52s! A few notes on the flight: There was one quicksave, because I attempted to change the audio volume in-flight. This reset the screen to windowed, meaning I couldn't see the "accept" button to leave the settings dialog. I could, luckily, quicksave/force restart KSP/load to continue flying. Because most of this was flown overnight with autopilot and I took screenshots with AutoHotKey, the map-view screenshots (in this appendix album) aren't super-helpful. If you flip through them fast, you can see the planet spin around, but unfortunately it's always from the night side. To compensate, my kOS display displays the ship coordinates on-screen. Paging through the provided screenshots shows the MET and longitude stepping in half-hour and 30deg increments respectively. This should be enough data to verify that I did actually fly thrice around. On that note, kOS was used to give a live estimate of fuel efficiency eta = (surface speed) / (fuel burn rate) in meters traveled per fuel unit. kOS was also used to generate a data log at 15-sec intervals to plot the flight afterwards. No kOS shenanigans occurred. Speaking of shenanigans, all screenshots, scripts used, output data, and a craft file are available in this ZIP file on my Dropbox. Proof no kOS stuff was done other than listing data, and there's also the raw datalogs of the flight, containing lat-lng at 15 sec intervals if you need more proof. Unfortunately I accidentally turned kOS off about 200m up and on final for landing. Oops. It's not technically a complete log, but close enough. Pilot Assistant is the only mod used that affects flight controls, and honestly this is theoretically possible but practically impossible without it. MechJeb was used only to put a marker on the navball for landing. The actual landing was flown with pilot assistant, using keys to update setpoint. Telemachus was installed so I could check progress on my phone, but doesn't alter aerodynamics. A few notes on the aircraft design: For range, I want my wing loading as low as possible. Since the big wings can hold fuel, max out on those first, and have two sets of them. Both wings are lifting, so no additional drag because of a stabilizer generating negative lift. All else being equal, with a turbofan (real-life too) to maximize range you want to go as fast as possible. Real commercial airplanes are limited by the drag-divergence Mach number. KSP doesn't model this in stock (FAR does!), but there's machLimit = 0.75 in the engine .cfg, and some velocity curve which I didn't actually plot. Experimentally, I found efficiency is maximized at about Mach 0.65. For a given speed, one engine is better (again, experimentally in KSP). Since I can't push enough fuel through the air at Mach 0.65 because drag, use two engines. Most of the time, I'm at around 90% throttle. Flying due east, I'm never going to turn. Use airbrakes as yaw spoilers to avoid having a vertical tail. I'm not totally sure this reduces drag, since I can't seem to get the debug menu to let me right-click on aerodynamic surfaces to get their aero forces, only on non-aero parts. On the other hand, it looks cool, and makes sense on a real airplane so it stays. The design choices did mean tuning PA was surprisingly tricky. At once point I managed to get the pitch controller to excite the bending mode of the fuselage, resulting in pretty crazy flutter on the control surfaces. Also, I didn't realize until pretty far into testing that the yaw became unstable at very low fuel and high altitude. Once tuned though, the airplane flies rock-solid. Tiny drag matters since I'm flying nearly three Kerbin days. Originally I flew this with medium landing gear and didn't make it; I had to land on the next continent to the west. Then I did a test flight and found that the small gear have less than half the drag! Switched gear, and made it with 266 fuel left. (of 5640). These end up being too short, so the rear ones are on the engines. Likewise, the Telemachus antenna was making a surprisingly large amount of drag, so I set just that part's drag model to none in the cfg. With the spirit of the rules, I'd argue, since it's doing nothing to actually make the airplane flyable, just means I don't have to be at my desk to know how it's going. Also, the ladder was making significant drag even when stowed, so I ditched it. Thus, no beauty shot with the pilots post-flight. The flight profile was actually optimized. I flew around for about an hour each at full and empty fuel, finding optimal altitude and speed, then calculated climb rate for the cruise-climb. Ideally the climb would be nonlinear, but since I set-and-forget for the entire cruise, it's linear between the two optimal points. It turns out good enough though. Side effect of using all these wings: when empty, the stall speed is about 30 m/s! It might be even lower still; that's just the speed I rotated when taking off for the empty-fuel test flight. Advanced canard winglets: originally to look cool, then adjusted their angle to trim the airplane for the cruise climb at neutral elevator (not how you'd do it on a real airplane, but I couldn't adjust wing incidence precisely enough).
  21. It should be possible to do a triple circumnavigation with only basic jets! I have discovered a truly marvellous proof of this, which this margin is too narrow to contain. (Seriously though, it's going to be 9 hours time-accelerated, so it'll be a while for me to get time to actually fly it. Done some test flights, I think it should work.)
  22. Weird bug report: Most of the time, Pilot Assistant works wonderfully. Occasionally though, when the vessel loads on the runway, Pilot assistant thinks it's pointing HDG 360, bank +/-90, and often angle of attack is nonsensical as well. The stock navball is always pointing in the correct direction, and whenever this has happened so far it's on re-flying an airplane that previously worked perfectly with Pilot Assistant and the identical mod configuration. The last time I was able to reproduce it several times by setting full fuel load (it worked) or 10% fuel load (it behaved weirdly) using all stock parts, kOS and MechJeb installed but not on the aircraft. Of course, when I attempted to reproduce it for screenshot purposes, it worked perfectly. Have you ever seen this before? I'll try to actually find a reproducible test case for this, and try to check if it's a mod conflict, though it's intermittent enough that I don't know how easy that will be to rule out. Installed Mods: kOS, MechJeb, Adjustable landing gear (problem happened on airplanes both with and without).
  23. Was wondering why no one mentioned my earlier entry, then realized it's not actually stock because navigation lights. Sorry >.< Updated entry: Aircraft: Daedalus VI Crew: Val (Captain), Jeb (Copilot) MET: 5H 42M 13S Less than one day on just basic jets! Fewer pictures, since this is exactly the same flight, but actually all stock this time. Is this one good enough for a badge?
  24. Are you on the ground when you do this? I noticed the same thing happening and looked at the code on Git. The autopilot doesn't control heading modes when the craft is landed, I've been able to set headings in flight though. Of course, that's just my reading of the code, I haven't written it.
  25. This is wonderful work. My aircraft look so much more reasonable now. Would it be easy to allow for front/back adjustment of the leg angle? A few of my aircraft have the gear is on a sloped part of a wing/nose, and the gear not being vertical on the ground is a little weird looking.
×
×
  • Create New...