Colonel_Panic

Well I actually found the cause of the problem. In my case the plane wasn't bending enough to be a problem (it's extremely rigid and reinforced. The docking port is only for holding the payload), and the wheels were perfectly in alignment (using snap-to geometry and symmetry settings), HOWEVER, it turns out that if the front wheels are on "backwards" (aka, they open down and forward instead of down and back), the front wheels will buckle and sway, causing the plane to tilt and veer to the side and crash. At first I thought switching to the trusses from pylons had fixed it, but after rearranging the plane again the problem came back. Then I flipped the front wheels around and viola! problem solved. As for the flight instability, it turns out that kerbal engineers haven't figured out how to make fly-by-wire vectored-thrust systems work nicely together... the front engines were gimballing the wrong way and turning the nose the opposite direction from where I wanted it to go. Locking the gimbals fixed it, and I achieved SSTO in the first flight after fixing them.

^This is the way I do vertical stack fuel routing, and a few other tricks (like routing around parts that are in the way between two tanks, or tanks and docking ports.) That said, the game does treat the decoupler as a valid 'to' part for fuel transfer. If you connect a fuel line to a tank, and then the other end to a decoupler, it will provide that fuel to whatever the decoupler is attached to. Going up a stack rather than down though may be a bit more complicated because the decoupler is technically connected to the engine and not the tank above.

Wow, you were right about the engine gymballing. I'm kicking myself now that this was the problem all along. also yeah I know about the cog/col thing, some of my later versions had col far far aft, still no difference...

From those pictures I can tell you that you have a HUGE moment of inertia, because the most weight is at both ends. This means that it requires a tremendous amount of force to rotate the ship. Your only recourse is going to be patience. Use a large RCS tank, and use it sparingly. I usually reserve my RCS only for very time sensitive maneuvers, and when I do use it, I pulse it briefly to initiate a slow turn, then pulse it again at the end to stabilize the ship again. If you use it for your whole turn, you'll save only a tiny amount of time over a pulse, but it will waste your fuel quickly. EDIT: I would also remove the RCS thrusters that are near the middle of the ship. They're not doing any good there and are only wasting fuel. Put them only at the outer corners on the ends to have them be most effective.

Hit Alt+F12 and check display control locks to see if there's anything actually locking you out of controlling the ship. Start your turns to a node earlier, lock on it before you timewarp if you do timewarp, use more RCS positioned further from your moment of inertia, and don't forget to turn it on (default: R). If all else fails, pulse your gimballed engines and use their turning ability to get you started rotating toward the node right before your burn, then peg the engines as you get close to stabilize on the right spot.

I've completely rebuilt this plane into different though similar variations and configurations seven times, but it still does the same thing... shortly after takeoff, it starts pulling to the left or to the right harder than I can compensate for, or starts rolling and yawing back and forth violently until it enters a flat spin. Every. Time. This is the longest flight I have yet accomplished with this plane: Any idea how I can fix this, or is the physics engine just incapable of handling a double fuselage design?

might try connecting to the decoupler, not a truss.

That moment when you spend 5 hours making the perfect spaceplane, and find there's nothing you can do to stop it disintegrating on the runway.

It's not staged, it's SSTO. The payload is mounted directly at the COM so that I can deliver it and return without changing the balance of the plane. It fits multiple payloads, but I'm just using the LV-Ns to test it since it's the longest payload I have to carry.

Sorry, slipped my mind. I did manage to 'fix' the gear problem by replacing the pylons they were mounted to with the much beefier trusses, and mounting double gear to a plate. However I soon found that there's an underlying issue with the plane aerodynamics that makes it unflyable anyway. For some reason it will always yaw to the side harder than you can correct... and the funny thing is, it isn't always the same side. It will sometimes randomly switch which side it's pulling to mid flight. I can't figure out why. I checked and all the tanks are emptying at the correct rate, weight distribution isn't off, etc... it's just really boned. Any help would be great. :x

So I spent like 6 hours designing a spaceplane, finally ready for maiden flight... but as soon as I throttle up, she turns hard right, runs off the runway and smacks a wing into the ground and disintegrates. I retried the takeoff about a dozen times, always the same thing, regardless of control input or SAS. So I took her back to the hangar, doubled up on the landing gear and took her back to the runway. Now she turns hard to the left. So it's definitely the landing gear that's causing this. it's some kind of weird wobble feedback loop that always destroys the plane before it gets more than 50 feet down the runway. Thing is, everything is in symmetry and well reinforced, I can't for the life of me figure out why I can't get it off the ground. Anyone else ever have this problem, or have some tips and tricks to solving it? thanks. P.S., it's a twin-fuselage plane, so only the front end is truly properly linked, despite all my bracing, and the back may be oscillating in and out enough to skew the gear. I'm not sure.

I actually have a 140 ton lander in Eve orbit right now waiting to find a good place to put her down... whether she'll make it back up again though is anybody's guess. The ascent vehicle is only about 90 tons and carries 8300 atmospheric delta-v. It may depend a lot on how efficiently I staged the ascent, how well I do my gravity turn, and how well I pick and peg my landing spot, but I have only 100dv left to deorbit it before I dig too far into my landing deceleration tanks, so I need to be right on the money from a fairly long range. Like trying to drop a pickle into a barrel from 100,000m.

I had that first one happen recently... and completed the docking maneuver without them... 3 times! It was a hell of a trip. I basically had launched an SSTO fighter into orbit, and then realized my mistake when I went to refuel and didn't have RCS. I came full stop in front of the tanker, pegged it on the nose, and gave it a push... managed to make contact and dock and refuel... then I un-docked, dropped into a lower orbit, and promptly forgot about it when I went to work on a new carrier. Then after putting the carrier up, I realized that my crew was incomplete without Jeb! Looking everywhere for him, I found him on the KSS, noticed there was an SSTO fighter nearby I could use to pick him up, so I grabbed it, went to dock with the KSS to pick up Jeb... and then realized it was the very same fighter with no RCS thrusters! One very agonizingly slow and dodgey approach later, I was docked, EVA'd the pilot into the station and EVA'd jeb into the fighter. At this point I was determined to finish this, or die trying... flew Jeb over to the new carrier, but there was a problem... the only available docking port was perpendicular to the normal, and slowly rotating around the ship. Without RCS thrusters, I couldn't realign and follow it around. I made contact with the docking port, but the mass of the fighter was too great for the little clamp jr to pull around into alignment, and as fast as it did pull it, the carrier rotated further away. After something like 10 minutes of wobbling precariously against the docking port and almost smacking my wings into the other docked fighters several times, it finally docked. That third one is why I ALWAYS include at least 2 fixed solar panels on every ship now. And then there was that time I planned out an entire round-trip to Eve, and then realized on approach to Kerbin at the end of the mission that I had forgotten to add a parachute to the return vehicle.

Very nice design and write-up! I'm very much amused by the fact that your rocket configuration with the upside-down mounted lander and return vehicle is very similar to how I designed my MK I Eve lander. I eventually abandoned that design due to some technical constraints and since re-optimizing my stages required re-building it from scratch, I redesigned it to launch facing the other way the second time, as in addition to better stability, it allowed me more room for ad-hoc modifications that let me repurpose it easily for changed mission parameters.

Obing was told he was in for a trip. 42 days later the real fun began. (doctored a bit in PS, sadly didn't think to hide the HUD for the screenshots, but at least it preserved that priceless look on his face)

I've mentioned this before, but that is damned impressive. I especially like the VTOL capability on Duna. Is that the only way you can land it there, or can you also glide in and take off again normally via rocket propulsion? These designs of yours are inspiring me to work on some spaceplane challenges of my own.

I'm having some issues using LV-Ns with stack decouplers. Namely that when the engine activates, the side panels blow off explosively and destroy any other part of the ship near the engine. Is there any decent workaround for this? I really need to use a stack decoupler for my LV-N core, and unfortunately there's really no way for me to accomplish this without it being surrounded on all sides by more engines and fuel tanks.

Very nice, thanks!

You can get around this by following my tips above. ASAS wobbling your ship with RCS on (also mechjeb does it too) happens because one of two things: 1) your ship is way too wobbly, and you're controlling it from near one end, so the navball flails when the ship wobbles. 2) your RCS thruster placement is uneven around the ship's center of gravity, and mechjeb/ASAS is oversteering. This issue is compounded by ship wobble. In either case, if you correct the second problem, the effect will be dramatically reduced. By having linear RCS thrusters placed in perfect balance around the ship, your translation maneuvers won't cause the ship to turn or wobble, which in turn won't cause the ASAS to kick into overdrive to try and correct it. The biggest problem in balancing RCS thrusters is that as your fuel depletes, the center of gravity will shift away from the fuel tank and toward the overall center of gravity of the rest of the ship, which means permanently placed RCS thrusters can't be in the right place all of the time if you're using them at different levels of fuel. except they can: if you design your ship so that the center of gravity of the fuel tank, and the overall center of gravity of the rest of the ship is in the same place, no matter how your fuel changes, you'll always have your RCS thrusters spot-on and wobble-free.

Lol, that's officially the cheapest idea I've ever heard for this... I like it. That said, I'd rather do this right I do leave my return vehicle with engines and fuel probed in orbit though, so all I have to do is reach orbit with the capsule and a docking port, then re-attach to the return vehicle and burn home. Saves the weight of carrying a nuclear engine and a small tank's worth of fuel down to the surface and back up again.

A lot of really great info and discussion in this thread! thanks! So it sounds like a bigger core stage and smaller outer stages is the best thing for an ascent profile, maybe solid boosters for the start. I haven't had a problem with losing engines or tanks due to poor staging unless I'm doing something wrong. I guess some people like to pack their asparagus stages all together (and, as the namesake implies, this is probably common practice) but I always try to keep to a 6-sided radial pattern, very rarely needing more than one 'shell' of stages. When it comes to the big tanks and mainsails, separation motors are useful for keeping them away from your core engines. Where I start running into issues with lifter efficiency is when I'm dealing with really big payloads... for example the single-launch round-trip Eve lander. I can get about 6k of delta-v 'for free' with really simple staging, but after that you rapidly start approaching diminishing returns. After around 9k delta v, I can only gain maybe 200 delta v at most per additional asparagus stage. Then stuff starts to become a mess of 'tagged-on' boosters to try and get that extra oomph, and I'm wondering if there's a better approach to net me higher delta-v for ascent and escape, and save me the trouble. Generally speaking I never stage off tanks without engines on ascent, but I do in orbit, preferring to keep one engine or engine cluster for all maneuvers in space, and just ditching fuel tanks as I go, though I've considered setting up staging that would allow me to ditch engines as well with the final few stages of tanks, since I wouldn't need the extra thrust anymore of, say, 3-4 LV-Ns once my fuel weight is mostly gone, and especially if I'm parking it in orbit for use as a return transfer vehicle, since when I come back from the surface the remaining payload will be much lighter. Right now I have an Eve lander and return vehicle, which should theoretically do the job (haven't actually completed that mission yet) at under 1000 tons on the pad. But I figure I can, with sufficiently optimized staging in the later parts of the mission, do much more with much less, because every inefficiency is compounded sevenfold by the time you've pushed that capsule into orbit and transferred the second time. I'll let you know what I figure out. EDIT: my ultimate goal here is to make a lifter for a single manned capsule that can achieve 8000-12000 delta-v for an atmospheric ascent from Eve with under 50-80 tons of weight respectively. If I can hit the 50 ton mark, the transfer vehicle required to get it there and subsequently the lifter to get it off Kerbin will be dramatically reduced.

Holy hell that's goddamn impressive. .craft file? I'd really love to get a chance to pick at that ship. I didn't see a link anywhere, but this sounds interesting. more details please? I really like the idea of using air breathing engines to deploy large tankers. LV-909 is superb for light duty, but I honestly rarely find myself using it outside of things like OKVs and ICBMs. Because its thrust is so low, using it to get into orbit or circularize is a real trick, and if I'm not going to be using a stage at all until I get into orbit, I'd just as soon make it an LVN. That said, I've never had luck with long-range SSTO unless it involved orbital refueling. In my experience getting anything with that much fuel up high enough requires more than 1 or 2 jet engines, or it lacks the TWR and lift to make it up off the runway. More jet engines means more problems with fuel consumption, air consumption, and flame-outs, not to mention more weight to lug around once you get into space. You can give it more wing, but that also increases dead weight in space. It's a very difficult balance to strike. Does anyone have any tips for building really big spaceplanes in an efficient manner?

Video is up: also finished fine tuning the x-wing. How do I attach .craft files to the thread? or can I not do that? Thanks for all the feedback folks! Postimage. It's a convenient one for me that doesn't force-resize images, has really big upper limits, and can upload and provide links for multiple images at once... without creating an account. Look at the MechJeb vehicle information in the lower right on that last screenshot. He says 273.34 tons, and I'm inclined to believe him. There is a lot of structure and extraneous parts for the sake of making it rigid and aesthetic. Plus that includes the weight of the probes, rovers, fighters etc docked with it. I think it's around 100 tons dry in orbit without the fighters. The fighters are around 10 tons each (dry) so that would make the total dry weight around 140. The total wet weight is, as stated above, 273.34 tons according to MechJeb. It does lag pretty bad, down to about 2 FPS once I had it fully loaded. It's like hitting a wall when you reach the limitations of the physics engine. I don't know why it lags this bad for me, as I've had other vehicles that lagged less on launch while having (I'm pretty sure) a higher total part count. It may be a result of just my particular structure and rigidity of the ship. You can see my video now above. Interestingly, my framerate does not go down while recording... though, video work was what this machine was designed with in mind. I just overstrut in general. I like more rigidity and tend to over-engineer. I do try and limit the use of struts in the final stage, but on launch I subscribe to the MOAR STRUTS philosophy. That's an interesting video, though, incidentally, almost exactly how I originally had this ship configured when I built the MK I. However it had a number of faults that I eliminated in later versions. With the engines rotated out in that configuration, the decoupling of the engine shrouds would actually destroy 3 of my engines as the parts impacted each other. I also tried the docking clamps like that originally, but they did not have the strength to hold up the weight of the rocket and resist shear. It detonated several times on the launchpad before I removed them. Finally, I had those adapters on the ends as shown, but it turns out they make a really really weak link with the jumbo fuel tanks, and would shear off and wobble like crazy no matter how many struts I added, so I removed those too and attahed the tanks directly to the tri-couplers. I am, however, still having an issue with 2 of my engines (the left and right LVNs in the center cluster) that seem to wobble freely on the tri-coupler and overheat as if they aren't directly attached like the others. I'm really perplexed as to what's causing it. Actually I do need the fuel lines. I tried it without them first (on the side booster clusters... the middle cluster needs the lines because I feed from different tanks) and the engines wouldn't fire because they were fuel deprived. That might have been before I removed the adapters as previously mentioned, though (I had them on both sides) I want to limit my parts list to as bone stock as possible. I make an exception for MJ because he's invaluable (though, less reliable lately in staging analysis in the VAB, so, if I can find a mod to replace him....) RCS for those satellites would require adding an RCS tank, and of present there's no RCS tank suited to the task of use in small probes like that.

Hi, so for most missions so far I've been using a radially decoupled 4-stage 7-stack lifter either 1 or 2 tanks tall depending on mass needs (which I just today learned that you folks colloquially refer to as 'asparagus' staging), and often feeding fuel from the tanks of whatever ships I'm lifting into the second or third stage booster pair reasoning I can lift more with less this way and just refuel it later via orbital rendezvous. What I'm getting into now though, is trying to build some extremely low-mass, high-ÃŽâ€v lifters for lander return vehicles, and I had a few questions with that regard, since ÃŽâ€v isn't the whole picture. There's also things like gravity and air drag to deal with, and fine tuning the TWR and ÃŽâ€v of each stage can make a difference in the total ÃŽâ€v needed to get into orbit. Take 'asparagus' staging for example. Is there a 'sweet spot' I should aim for in TWR per stage? do I want TWR to start high and get lower, or start lower and get higher, or stay the same throughout the ascent? Will I get more total ÃŽâ€v by adding more fuel weight to the later stages thereby reducing the TWR/ÃŽâ€v of earlier ones, or the other way around? Is there any proven formula for maximum efficiency?