burn_at_zero

Completed Laythe cleanup. 201 science reports are sitting in the main lab. There are definitely things I missed, but I've gotten a lot of data from every biome. The two spaceplanes (66 tons of mass between them) are in elliptical Laythe orbit. I left them with enough monoprop to dock with a tanker later, if I don't fix their design issues and send new ones. Transferring out of this orbit took only 7 m/s. I had to look ahead a bit, as the MechJeb porkchop plotter doesn't play nice if the orbit isn't round. Still, I got a Tylo intercept in under a week. Didn't notice until after the burn that it was a *retrograde* tylo orbit, so I burned another ~350m/s to correct. Regardless, I'm now in an elliptical Tylo orbit. Still way under the ~1300 m/s indicated by the dV maps, though there are added propellant costs for the landers. I'll probably use this approach in the future if I ever have a complex mission and simply plan for some extra lander dV; the propellant savings for a large mothership should more than offset the extra consumption for a small lander. Next steps are to assemble a lander with boosters and use the other lander to deliver it to 30km circular (~800 m/s each way). Then I'll land, collect data, lift off, stage, circularize, dock and rendezvous with the mothership. Hopefully I can hit a biome boundary so I can collect samples from a couple of places, because I only have enough boosters for two landings. It was always a long shot to hit every Tylo biome even with a rover, so we'll have to see how that affects the final tally of science. Given what i went through on Laythe there's almost no chance of pulling off a Poles landing on Tylo. Little did I know, my boosters are welded to the docking ports on the mothership. No amount of save file tweaking will release them. Looks like my somewhat unorthodox construction techniques make ships very sad. I tried installing stockbugfix which is supposed to have a 'force undock' option in addition to other docking port fixes, but no luck. I've spent about two hours trying to manually map the ports together properly. Next chance I get to play I'll probably have to HyperEdit boosters into position and use the cheat menu to destroy the 'stuck' ones. Hopefully that's not a disqualifying act; normally I'd post the save file and ask for help but it's something like 270,000 lines of config and 67 docking ports to muck through. I've already tried the obvious things like changing state to Docked from Active, making sure each port is referred to by at most one other port and setting the active flag to true for the undock action. Once the big beast is out of the way I'll have to decide whether to base from Tylo or relocate.

I've completed science collection on Laythe. The poles were a giant pain, since I completely forgot to build in enough dV for two 90° plane changes. I have about 240 screenshots so far, but that will be a couple hours' work to crop, transcode, annotate and upload. Some general observations: First, changing the landing gear turned out to cause trouble. The medium aircraft gear can't steer, so I had to change direction on the ground with RCS or short jet bursts. Second, the parachutes were not ideal for full-fuel condition on Laythe. They were arranged 1:2:2 with one behind the cockpit, two on the engine nacelles (at CoM) and two at the rear of the fuselage. This led to about a 45° tail-first descent. I had to cut three of them to get a horizontal landing, which pushed my impact speed to about 18 m/s. The nukes give an alarming wobble but everything survives. Third, in 1.1.3 kerbals experience re-entry heating even inside a closed cargo bay. Upper-atmosphere EVA reports are awkward and dangerous but possible. Don't expect to be able to walk around inside said bay; put all your instruments within reach of the kerbal while still hanging onto the door. Fourth, it is possible to go from equatorial to 60°+ inclination with aero surfaces in the upper atmosphere. I saved about 1500 m/s of plane change dV by dropping from an elliptical orbit to 40km peri; entered and stayed rolled to the left and pitched back about 30°. Had to fly the rest on rapiers, which sucked down quite a bit of fuel (~600 LF). If I had the patience and foresight I could have chosen a higher peri and made multiple high-altitude passes; there was probably enough energy in the orbit to get all the way to a pole. Fifth, my modular lander turns out to be a great rescue tug. It's autonomous but has three seats. T/W is good and it has about 4800m/s dV. I didn't put any RCS on it, which was terribly annoying. I've had to rescue the spaceplane twice now, as I didn't want to waste the fuel to put the mothership into low orbit and the plane just doesn't have the dV to reach a high elliptical orbit. That polar inclination change is just brutal. Can't wait to finish the last of the Laythe maneuvers so I can ditch the planes. I'll leave them docked to each other in Laythe orbit, but the two of them represent something like 40% of my parts at this point. Looking forward to the FPS improvements. I wish I would have brought some other modular parts, like maybe an x16 LF tank with docking ports and a simple single-nuke propulsion module. I brought way more oxidizer than I'll need, but not nearly enough RCS fuel. (I was expecting to make four Laythe landings at 880 O each, but I only needed two landings and one of them reached orbit without using any O.)

Success. I reverted to an earlier design with more lift and a lot more vertical surface (4 rapier, 2 nuke, 1.5 shock cones per rapier), one of the craft I had gotten to orbit before. I tilted the tail horizontal surfaces downward a few degrees, which had a pervasive effect on stability derivatives and level flight angle. It also seems to have moved the aero center further back, allowing me to place canards for better control. Upgraded the landing gear to medium, threw on a ladder and it did the job. In short, I was able to reach a ~100km orbit and return, do a chute-assisted landing in hills, take off, reach mach 2.5, then land again without parachute assist. Flight path was about 40° until 12km, then gradually level out until 0° at about 20km. Altitude dips a bit while building speed, but recovers quickly. Kicked on nukes around 22km, then mode switch at about 24km and 1.0 km/s. Closed-cycle got me an apoapsis of ~80km, so I kept the nukes on and brought AP up to 95km then let mechjeb circularize. The key to landing was to start very early and set a PE of about 40km. I used mechjeb's advanced mode to hold pitch at 90° and gradually pitch down to 45° at 30km through the descent. Radiators were able to keep up with reentry heating throughout the descent. Once the air got thicker I dropped to 30°; I'd overshot KSC by several hundred km, so I had to circle a bit to find land. I tried my chutes and found the balance to be good enough. Takeoff from medium hills while nearly empty was easy. Landing was a bit harder; I had trouble with a long slope that ended in a sharp rise and had to pitch up pretty hard. At the end of it I was able to hit the ground at about 170m/s and -5m/s vertical. Braking spun me around a bit but it stayed on the ground. From here I'll probably see if I can decrease part count or mass anywhere, but that seems unlikely given the months I've already spent on the craft. I still need to put together a new mothership, and I'm seriously considering switching to ISRU. I have to recover the old mothership and parts so I can get my kerbals and my funds back as well, which won't be fun at all. Lots of work, but at least it's straightforward work.

Another successful orbit, this time with nearly 1km/s of dV in the tank: I actually landed too: Sort of, anyway. Not bad for a crash emergency landing at around 170 m/s and -10 m/s vertical with no front landing gear or nosecone.

Thank you for the craft. FAR makes a pretty significant difference in aircraft behavior. The one you posted doesn't get off the runway below 240 m/s and tends to break up below 5km and above 400 m/s due to uncontrollable yaw. I tried throttling back to 20% immediately after takeoff and was able to get to about 6km and 280 m/s before losing control. {edit: I was able to get to 19km and 1100 m/s on one attempt before spinning out. Might be putting too much thrust into it; it didn't happen until shortly after I throttled back up from 30%.} For the craft I showed above that made orbit, I started with one pair of wings and no controls. I added more surface area until I was able to take off. There are only two ailerons, which combined with torque is enough to get over 10° AoA; I had to have three tailfins or it was too difficult to keep it straight. I've tried to use two sabres and one nuke but it's not enough thrust to circularize before reentry; the four sabre + two nuke craft still struggles but can get it done. Looks like the next challenge is something that can re-enter without losing anything important.

Thank you for the craft. The first one is interesting, but a bit small. I had trouble getting off the runway but it flew quite well. The second one is about right, but it's unstable. It has some of the same problems I've been struggling with, but I think the length and balance is better. Some fairly minor adjustments might make that one a contender.

Made it to orbit: Getting back, not so much: Believe it or not, I actually got subsonic without the cockpit. If my landing gear hadn't cooked off I might have been able to land it after this. I ended up impacting at about 240 m/s and -35 m/s vertical. Takeoff speed is about 135 m/s and I have not yet been able to land it, so while it's nice to get to orbit I'm not sure this is the right track.

Thanks for the input, and for the craft. The part crewtube-airlock-125 is from Station Parts Expansion, it's a one-person airlock. I have two of them in the bay so I can store multiple science reports like soil samples and gravity measurements without having to carry a full science lab or three full-size crew parts. TPtank1mL05625 is from FuelTanksPlus, it's a 1.25m 1200-unit tank. I think the designs that use that one probably aren't worth pursuing. It seems like an SSTO with a payload of less than 2 tons should be possible. Yes, the original plan turned out to be too ambitious. Rather, too overbuilt. I just need to be more kerbal about things and not bother with redundancy. The low-gravity lander is easy; the Tylo lander is a little tricky but it's done. Only this SSTO stands between me and gigantic piles of science.

Laythe orbit would be adequate. I'd prefer it to be able to SSTO at Kerbin for testing, but if that is much more difficult than Laythe I can get it to orbit by rocket. I'll have a depot orbiting Laythe at minimum, so even if it does end up needing to fly elsewhere it can refuel along the way. Craft files: dropbox link Mods: If you have trouble with any of it let me know; I can replace the payload with an empty MK2 long bay and repost. Everything outside that part is stock, but there are airlocks inside from NF Construction (I think). Thanks in advance for looking.

Let's see if this album embed works. I grabbed top, front and side shots of most of the major forms I've tried. I didn't include shots of all the minor variations. Side A is a shot with full fuel, gear down, CoM/CoL displayed. Side B is a shot with no fuel, gear up, CoM/CoL and trans-sonic curves displayed. {edit} This isn't embedding properly. Here's a link for now, until I figure this out. {many edits later} Imgur albums apparently can't be embedded here. Photobucket is a dumpster fire, so I'm just leaving this as a link for now. Laythe planes

Is it normal to yaw out of control at high altitude? I've tried less control, more control and super-oversized tails with no effect. I've tried keeping thrust along the centerline and I've tried putting engines on wingtips, but there seemed to be no difference. I could try TCA (thrust-controlled avionics mod) to see if that gives me enough yaw authority. I'll spend some time this weekend and see what I can do. First though, I really need to post pictures of the designs I've tried so hopefully you can tell me what I'm doing wrong. I have my science payload down to a single MK2 cargo bay, which includes all science parts, two passengers and enough RTGs to run it all. The plane designs are in the 25 to 30 ton range (50-65 parts); I've tried one or two ramjets or rapiers with one or two nukes, and a heavier craft had two nukes and four ramjets, Take-off TWR has been anywhere from 0.3 to about 1.2. Rapiers seem to have better high-altitude thrust than the ramjets. My choice of MK2 cross-section means the nukes are way out from CoM and prone to tailstrike if I put one on the centerline. I don't know how much lift I have (not sure where to see that in the editor) but it is probably too much; I get level flight at low altitudes with only a degree or two AoA and landings take a lot of distance. (The ramjets help with that actually since you can reverse thrust.) Mostly I wish I had a joystick.

I quit. There's zero point in continuing this ridiculous obsession. I've put in about another 16 hours of game time with absolutely nothing to show for it. I can't get a craft that is stable above mach 3 and can't get an all-LF SSTO no matter what I do. I considered sending a subsonic cruiser aircraft and just using a massive rocket to return to orbit, so I built a couple of moderately successful designs. Then I asked myself if I'm really ready to spend several consecutive hours fending off children and family obligations while flying across Laythe to reach every biome without being able to save. No. I choose sanity. A few additional SSTO attempts followed, including my final design which has all-green stability derivatives at all relevant conditions (including the inevitable yaw and breakup around mach 3.3 and 16-18km). I've eliminated anything not essential for flight and science, no radiators or other eternal protrusions, confirmed basic alignments like CoM to CoL, experimented with a range of small adjustments of every single part on the plane. Whether or not I use FAR's assistance, whether or not I use SAS, whether or not I switch between them, same end result. SAS sends me into unrecoverable spins, while FAR tends to overcorrect and throw me into unrecoverable rolls. On two occasions I managed to break 35km, and on one of those I reached 64km before falling back to fiery doom. Oddly enough the craft is more stable with the landing gear extended than retracted; I'd forgotten to retract on my first successful mode switch, but retracting gear threw me into an unrecoverable spin and disintegration. At this point I can either uninstall FAR or KSP. Either way, I won't be completing this challenge as intended. Maybe next year, maybe without FAR. If there's a spaceplane involved it almost certainly won't be something I designed. As it is I'll be updating once the next version is released, starting a new campaign and forgetting that spaceplanes exist. Thanks everyone that offered advice. I learned a bit about supersonic aerodynamics in the process, so I suppose it was worth it. Hopefully anyone else attempting this has better luck than I did.

Thanks for the input. I will try moving some things around to get my CoT closer to CoM and get a better CoM/CoL balance. I'll post screenshots later tonight, but CoM remains ahead of CoL even at zero fuel. The canards and long delta wing section are good for that. At high altitude the plane tends to drift back and forth, corkscrew and then flip over so it feels like a loss of yaw control (or like losing yaw control first while all the control surfaces lose effectiveness). It doesn't happen on every ascent but it does happen more often on flights that needed a lot of corrections at lower altitudes. I can usually pull out of it below about 18km and try the ascent again. Smart A.S.S. helps once aerodynamics are not much of a factor and I can rely on gyros or RCS. So long as the control surfaces are effective MJ seems to have trouble with the delay between input and result. Usually it leads to positive feedback oscillations (much like stock SAS) and the plane eventually flips or spins out. I control the thing using trim inputs through most of the flight. If I had a joystick or some other analog control this would be simple, but the keyboard is a binary input and causes drastic reactions under 15km even to taps on the keys. Cutting back on control surfaces or their max deflection makes it easier to control at low altitude but impossible to control when suborbital. I can hit 400+ m/s at 2km altitude just off the runway pretty easily. I can also easily climb above useful oxygen levels and flame out if I'm not paying attention or if something else has gone wrong and I'm trying to correct. Staying at a useful altitude (22-24km with rapier, 20-21km with ramjet) to pick up speed can be challenging. My heat problems tend to come when I can't keep that altitude envelope and the plane dives below 20km at 1+ km/s. That's usually the point where I give up trying to build more speed in atmosphere and head up as fast as I can, kicking on nukes then switching mode once I run out of air. Even then it takes a couple minutes of nukes at 100% to get all the way out of the atmosphere. Sounds like I should pack more oxidizer, ditch most of the radiators and try to get better control of the craft at high altitude. Maybe moving parts around to smooth out the edges will help since there are a lot of sharp transitions in the current design. Also sounds like I should revisit the mission architecture and get rid of a lot of redundancy. I don't particularly care how long the mission takes in-game so I could just as well use a single mapping satellite that can return and refuel, maybe with a spare (~10t, ~80 parts). I could rebuild the mothership's propulsion section to be an independent tug, which would eliminate both of the heavy tugs (~400t, 140 parts). The Laythe plane would do fine for Pol and Bop landings, probably Vall as well, so no light landers (~160t, ~200 parts). I might go the other way and use a single light lander instead of a plane, though that will mean taking quite a bit more LFO. A redesign of the Tylo vehicles might allow me to combine the two, though I'd have to pull a Tylo Elcano. May as well.

The long silence is a sign of intractable problems. This means some screenshots of progress and also some ranting towards the end which is safe to skip. On the plus side my career save now has the tech tree maxed and 10 million funds banked, all without going more than a few km outside Kerbin's SOI. On the minus side I've gone another month without even assembling the stack in orbit; worse, it's beginning to look like my computer can't handle the load and I will need to redesign the mission. Time will tell. I've found a solution to the probes, but it looks ugly. I used radial-attached engines for propulsion which allows for a docking port on one end and the survey scanner on the other end. I offset it with the high-resolution ore scanner, so both instruments are canted away from the end of the probe at a crazy angle. It fits inside a 1.25m aeroshell which mounts neatly to the mothership. It's possible to undock/redock after popping the shell, so the probes can move themselves to different docking ports if needed (including hitching a ride on a tug to reach the target SOI). I also put together two ballistic Jool atmosphere probes with enough batteries for two sets of transmissions each. (There's not enough time during the descent to power the data feed with RTGs unless you get crazy; easier and cheaper to take enough battery power along for the ride.) These have to be released on the interplanetary approach unless you want to circularize pretty low over jool proper and use RCS to deorbit. Tests in sandbox were acceptable. The mothership core is indeed a 5m propellant tank, with 2.5m tank 'arms' and a forest of docking ports. Probes are attached to the mothership core on the ground and launched as a unit. It looks awkward but it flies just fine with FAR. The probes are a *very* tight fit with the heavy tugs. I don't understand why this craft flies but my planes don't. So far I have the core, habitat module, two heavy tugs and the Tylo rover docked at 150km LKO. I'm already seeing kraken warning signs like harmonic oscillations and 'floppy' engines, which is not a good sign considering I need to dock at least five more craft in the 50 to 90 part range. If my machine can't handle the load I will have to revise my whole mission profile and start over. I have a working SSTO spaceplane. It needs oxidizer. I've spent over 24 hours of game time trying to build an LF-only version and I simply can't do it. The current 'good enough' model will parachute-land on Laythe and do short horizontal takeoff; there is enough crew space to bring an engineer. I can see why some people like flying planes to orbit in KSP but I find the experience awkward, tedious and excessively dangerous. The design process with FAR I've found to be masochistic in the extreme, perhaps because I'm not a practicing aerospace engineer for my day job. Nevertheless, it flies and sometimes it survives to orbit. (For the record, I do enjoy flying planes in atmosphere with FAR up to around mach 3; hypersonic velocities and rarefied atmospheres at 20+km are nightmarish.) As the name implies, this is the fifth major structural plan and the only one that was remotely stable above 18km and ~700m/s. I've spent hours poring over how-to guides, explainers and youtube videos and even more hours making adjustments, tweaks and full-scale rebuilds then test flying. Every single part you see was essential. Every design parameter is in the green at the appropriate speed and altitude settings. The massive vertical tail structures help keep yaw under control at max altitude. The canards were critical to maintaining control over pitch. The ridiculously overpowered set of four RAPIERs is needed to have any kind of control on the edge of flameout and to have enough thrust to punch through the upper atmosphere without needing to run nukes for ten minutes. (The only other design that made orbit was all LF, but I had to run nukes for more than ten minutes to do it and the double-dip trajectory was very difficult to survive on the rare occasions that it actually reached 1+km/s and 25+km.) The wide offset of the engines further controls both yaw and roll through a large lever arm. Any fewer than six radiators and I lose either the cockpit or the engines about 90% of the time. Five chutes are the minimum to survive a landing on Kerbin undamaged with minimal fuel burned. The balance and landing gear alignment are exquisitely tuned for takeoff at any point between 70-120m/s with a little input, and you can pull back all the way while still on the ground without tailstrike. The thrust is such overkill that I usually have well over half the runway left at takeoff and I sometimes forget to retract gear until I'm in orbit. It handles rough terrain like a breeze, even the nasty hilly stuff behind KSC; taxi to a hilltop, pick up speed on the downslope and use the next upslope as a ski lift. Yet somehow it spins wildly out of control with no warning and at varying altitudes and velocities, exploding if below about 20km. If it is high enough I can usually just cut thrust, allow it to stabilize (usually upside down) and power through. It's fun to fly at low altitudes and reasonable speeds, but very frustrating when hypersonic. I gather the sharply varying profile has something to do with that but every tweak I make ends up making the behavior worse. Re-entry is usually just a long, slow, risky fireworks show at high AoA. Not too much worry for the heat (surprisingly, since it likes to try to cook the occupants on ascent) but it likes to flatspin and break up at the slightest hint of roll or yaw. Still, it will make a reasonable contingency crew return vehicle.

Thanks for the support, I appreciate it. ISRU just feels like cheating That's my usual approach but for this challenge I want to minimize overall mission time and get into the spirit of 'bring it all from home'. That's also why I don't want to reuse one mapping probe for all five moons, though I might have to cut back to just two or three. The rover/lander has more than enough dV to hop to every biome on each of Bop and Pol (five 180° suborbital hops) without refueling or driving around. I should be able to do the same on Vall, though landing near a biome boundary and doing a little driving would save fuel. I should just make it a pure lander to save on parts (60+ eliminated) and mass. I did successfully build a LF-only SSTO spaceplane. It takes a (very) long time to make orbit, but it should work on Laythe with some refinement. The crew ring looks nice from a distance but I don't like the way the parts split. 96 seats is a little excessive; I should be able to cut back quite a bit here. I wish there was a parts pack with wedge pieces to make a nice ring, but that would be hard to do without requiring specific spoke pieces. Unfortunately the probes are as simple as I can make them. They could be lighter but not with fewer parts. The survey scanner's sheer size is hard to manage and the narrow-band scanner sticks out too far for 1.25m aeroshells, which means I'm pretty much stuck with a 6t cargo bay for each one unless I get creative. The most practical thing to do would be to install KIS/KAS and take all the sensors and smaller bits inside containers. I would have to assemble the probes from parts in orbit. For performance reasons that probably means making a dockable workshop module on the mothership that separates; the probe assembly would be done with the rest of the ship off-grid. Since I have fuel tanks plus and Space-Y I should just go with a 5m or 7.5m mothership core. I'll spend some time tonight and see what I come up with. This is taking a lot longer than I thought it would, but it's nice to have real challenges for a change.