Zeiss Ikon

Incremental improvement. You're on the right track. If you haven't done so, you might look at the one or two stock airplanes supplied in the game. They're tested and known to fly well...

Beyond the wind at altitude, there's the weight of 4 km of whatever the tether is. Most weather balloons can't lift four kilometers of any sort of "rope". The tether material has to be strong enough to support four kilometers of its own weight. The longer it gets, the thicker it needs to be just to hold its own weight. It's possible some kind of fishing line (one of the synthetics) or kite line (also modern synthetic -- Spidewire or Dyneema, for instance) could manage this without a tapered line. Based on my experience with kites, there also comes a point where the line is horizontal where it leaves the tether point and forms a catenary curve up to the lifting device (balloon or kite) -- and you can't go out further than that without some means of increasing either the available lift or the tension on the line. With small Eddy style kites and common nylon kite line (and in manageable wind speeds), this happens before you have as much as seven hundred meters of line out, an the kite will be only two or three hundred meters above ground. There are good reasons weather balloons aren't normally tethered. First is that one of the measurements they're routinely used for is measuring wind speed at altitude as they rise at a known rate. Second is, as noted, wind. Wind will actually push a tethered balloon down, because of the downward component of the tether tension (you can observe this with a party balloon on a string and a fan). For lower altitudes (up to a kilometer or so) kites are a better choice for tethered observations. As noted, once you're up a few hundred meters, wind is very much a given, direction and velocity are the only questions. Trains of kites have been flown roundly as high as the OP wants to go, but this is complex, depends strongly on a very steady ground wind so each kite in the train can be attached and launched in the same direction the line points, and further, depends on winds aloft agreeing in direction with the ground wind (which is not always the case).

As I recall, the Merlin engine can be throttled to something below 50% (which is pretty deep for a real-world rocket engine), but even burning only the single center engine, with the nearly-empty tanks they have at landing (for performance reasons, they'd like to land just before the engine flames out due to lack of propellant, but not let the engine actually flame out because a mixture imbalance during a burn can damage stuff), that's still too much thrust to genuinely hover. On the bright side, they have a computer flying the beast, and a bunch of failed attempts (which yielded a lot of data on what they could and couldn't get away with) preceded the current long string of successful landings, so they can manage with this. The other side of the hoverslam is that, as noted, it's the lowest fuel consumption landing possible -- which means less fuel need be reserved from the booster's main job of accelerating the second stage and payload toward orbit. The very fact they can reserve fuel while launching commercial payloads is a testament to the performance of the Falcon 9.

Energy conservation ought to give you the AP (just like it lets you back out bullet velocity from the height reached by a ballistic pendulum). The kinetic energy the ships have after their boost gets converted back into potential energy. It'll be a system of equations, though, because you also need to know what the velocity will be at AP to know how much KE stays in that form. KEP + PEP = KEA + PEA. Remember to calculate your PE from the center of Kerbin (gravitational zero point) rather than from ground level, though, or your height will be off by the PE of Kerbin's radius. And of course, you have to calculate PE with an inverse square law applied to the value for g. Might be better to just launch a craft (via hyperdit?), let KSP tell you its Ap height, double that, and subtract Kerbin's diameter.

(1.6.1 RSS/RO/RP-1) No screen shots at the moment, but I succeeded in building an airplane that can (in KRASH simulation) launch from the runway and reach the speed of sound. It takes two of the Derwent V jet engines (still the only jet I have in this career) to get the 10+ T craft up to flying speed, and it still needs the entire runway, but once off the ground and cleaned up, it can climb at a reasonable rate at about 150 m/s. Once up to test altitude, the jets are ejected (to be recovered with a parachute on each nacelle) to remove their drag and the XLR-11 alcohol/LOX engine is started. This engine can push the craft to and through 350 m/s in level flight at 10 km altitude, perhaps faster if I have the patience to climb higher before starting the six minute burn (XLR-11, technologically new, also throttles to 50% thrust and can be ignited a total of four times in a single flight, so reserving a little fuel might give the option for a go-around on a rejected landing). Once the rocket fuel is burned off (and the kerosene for the jets dumped), the airplane is basically an empty fuel tank with wings, landing gear, and a cockpit; it ought to have a ridiculously low stall speed, so if I can manage my terminal energy well, ought to land back on the runway without major issues. In all previous RO careers I've played, I've done the "Crewed Speed Record: 350 m/s" and "Speed of Sound" contracts with a rocket, generally effectively an A-4 with a cockpit (and varying amounts of pilot protection) and sometimes with stubby wings. If I have the patience to wait through construction time for Research 2, it'll be the first time I've flown a runway-launched airplane for that achievement. If the RO and Kerbal Construction Time devs happen to be around, I'd like to suggest that the very first craft, airplane or rocket, launched in a new career be exempted from construction time -- reflecting the fact that its completion might have signaled the beginning of the actual space program we're playing. Build an X-1 equivalent or a WAC Corporal, and it ought to be ready to fly at the beginning of 1951 (given these were both 1940s achievements -- 1947 and 1948, respectively, as I recall).

@RizzoTheRat It's hard to say if your method is any more efficient than the one I gave (stop thrust when AP is high enough), but mine is certainly easier to fly. I'll admit, I've been influenced by playing Realism Overhaul. With that mod set (RSS, Real Fuels, FAR, a bunch of historical engines parts packs, most of which don't throttle, have limited burn time, and have only a single ignition) you burn all the way from ground to circularization with minimal staging delays (because you have no control without thrust), and tailor your thrust profile by controlling the mass, hence TWR, of each stage (plus its payload, which includes all later stages). Historically, a coast phase like the one SpaceX uses with the Falcon 9 (second stage shuts down when apogee reaches desired height, then restarts near apogee for final insertion) just wasn't practical until engines with significant throttling and restart capability existed. After manually piloting a rocket to orbit in RO once, I'm generally happy to let MechJeb do the heavy lifting. I'm pretty sure MechJeb is doing something similar to "chasing AP" -- based on the way i see, for instance, an early Atlas-alike stage-and-a-half rocket pitch up and down at various places in its launch -- but Positive Guidance is much more sophisticated than just that -- it actually pitches below horizon when it gets close to desired altitude and is short of velocity, which avoids raising AP as it raises PE to produce a circular orbit without burning at the existing AP or PE.

This is also true in real life. There are many very good fixed-wing pilots who just can't learn to hover, because it requires a fourth continuous control input (throttle, collective, two cyclic axes, and rudder, vs. throttle, roll, pitch, and yaw). One way I've seen it referred to is that "Every airplane out there wants to fly. A helicopter does not."

That layout (aside from the push-pull propulsion) reminds me a bit of the Bumblebee biplane that once held (and may still hold) the real-world record for smallest flying, crewed airplane (by dimensions) -- had a span less than two meters, and just about two meters long; just about 1.5 m tall on its fixed gear. Teeny little heavily staggered biplane, reportedly a major PITA to fly.

Liquid air locks are used routinely in some settings. A "moon pool" in an oceanographic ship (used for launching ROVs and occasionally crewed deep submersibles) is one sort; the swim-in entrances in saturation dive habitats are another (there was a good depiction of the latter in several scenes in The Abyss). In both cases, pressure at the water surface is the same as pressure at the same height outside the structure -- but there is pressure present, so the liquid (water, in every working case I know of) doesn't just boil off or evaporate.

According to the linked article, SED had lower power consumption than LCD -- at least as it existed in the noughties. That seems likely to be because of the inefficiencies of lighting the entire backplane (dark pixels and light) with a cold cathode fluorescent array. OLEDs are much more efficient, in that they only draw light emission power for lit pixels, and in proportion to brightness, at what would have been, in first generation, about the same conversion efficiency as fluorescent, and likely is significantly better now. They also have most of the other advantages SED and FED had over LCD -- viewing angle, brightness ratio, and response time. There's a good reason almost all top line smart phones use OLED -- and second- or lower line don't. OLED works better than LCD, but LCD costs less to build. OLED also doesn't need to be encased between layers of conductive-coated glass; it's behind the flexible display technology that created foldable phones (the next generation of those will work. Really!) and roll-down TV screens. A lot of top-end computer monitors are OLED as well. Effectively, LCD beat SED with a VHS-over-Beta victory of being good enough that first was best, and now is getting its lunch eaten by OLED because that's so much better.

Assuming you're launching a rocket, vertically, you need to do a gravity turn. This amounts to tipping toward the east (D key, if you built the rocket with the default pod position in the VAB) by a few degrees when your rocket gets between 50 and 100 m/s, then following the prograde marker until apoapsis is around 80 km. Throttle down to zero and make a circularizing maneuver at your AP marker, and when that's done, shazam! you're in orbit. The amount of tilt needed at the start of the gravity turn is dependent mainly on your TWR -- if it's 2 you'll need a lot more tilt than if it's 1.2, because you'll gain speed along your heading faster, and therefore have less time to build up horizontal velocity that tips the prograde marker toward the horizon. This method hasn't changed significantly since at least 1.2 (I first played in 1.2.2). Earlier (before 1.0?) there was an alternate method required by the "soup" of the atmosphere, to fly to 10 km before turning over, then immediately yaw to around 45 degrees and follow prograde (from there on, it's pretty much the same) -- but the atmosphere was changed in 1.0 or 1.1 to have less drag at sea level and to gradually taper off, more like a real atmosphere, so the gravity turn works better now. The one exception is if you're trying to launch something like a complete space station all at once -- something big, fragile, and draggy may still get to orbit best by launching vertically to clear most of the atmosphere, then turning over to the east to build horizontal velocity after you have enough vertical speed to coast completely above the air. For spaceplanes, the technique is completely different from what I've given here, but since I've never flown a horizontal launch spaceplane to orbit, I'll let those who have advise on how to do that.

No discernible CoM shift from full to empty tank (CoM marker is inside the tank), steering on the main gear was off from the beginning. After fixing the gear, I was able to complete the design mission without problems, despite using the "worst" swept wings in the game. And I slightly understated performance -- while it's only good for about 200 m/s at sea level, it'll do that on 70% cruise throttle at 7000 m. I was able to visit six biomes (For Science!) near the Brownsville, TX launch site (in Realism Overhaul) and return to base on the 1200 liters of fuel I provided, though I should add at least another couple hundred in the stored version -- I landed (on the grass near the launch center) intact, with literally about a minute of full power fuel remaining. As expected, the airplane slows down much better when bingo fuel (that's close to 1 T of fuel weight burned off); I landed at around 50 m/s compared to a full-tank dirty stall speed around 65 m/s. This is actually a nice handling airplane now that I can get it off the runway. Trimmed for current speed, there's no significant change between MechJeb autopilot and "free" flight without MechJeb or SAS. Long term no-SAS flight is marred only by a slow tip-off into a bank one direction or the other; likely another degree of dihedral would fix that. I do need to reduce the built-in wing incidence by a couple degrees; at 200 m/s at altitude, I'm flying with almost five degrees nose down to hold altitude. Glides well, too, when clean; zero pitch coincides with trim for about 100 m/s as currently set up, which ought to be close to minimum sink or maximum L/D. Drag brakes would be welcome; the twin rudders are set up for air brake function, which changes pitch trim, but I don't have that tech unlocked yet (all parts in this aircraft are in the starting tech node for 1.6.1 Realism Overhaul).

Sigh... My CG does shift from a full tank to empty. Down, a few centimeters. No discernible fore or aft movement at all. I designed it with the CG in the center of the fuel tank. If the airplane is stable with a full tank, it'll be stable when empty. And it's heavy enough at just under 4 T that the weight of a single Kerbal more or less (two seat cockpit, may or may not carry a scientist on the science mission it's built for) won't change the CG location enough to notice (test flights were done with only Jeb aboard, so if anything it'll be more stable with the second seat filled). I may try building another version with the other swept wings, but as you note, this should land better with depleted tanks, because after burning most of the 1200 liters of kerosene in that tank under the engine, it'll be 600-800 kg lighter (roundly 20% of its weight) . I'll also give another test flight with, say, 10% fuel and see how much more I can slow it down when it's lighter. The main reason I don't have confidence in my ability to land it is because I can't seem to land anything in game. I blame keyboard controls. Thirty-some years ago, I was a pretty good pilot with an R/C transmitter...

I beg to differ. My stock career, which ran from 1.3.1 to 1.6.1 (abandoned when I installed Realism Overhaul) had an experiment with asteroid ISRU that was stabilized to the rock with three KLAW attachments on the same craft. Took me several attempts to get all three to catch, but once I got the third one on, everything was absolutely stable. This doesn't work very well for pushing the rock, mind you -- any engines on the three-point attached craft will have only their gimbal range to compensate for misalignment with the composite CoM -- but if you want to ensure that your drill stays in position to actually extract ore from the rock, it works quite well -- or did in, IIRC, 1.4.1, anyway.

Signal lag is an option with RemoteTech mod. G force limits are an option in the stock game (that defaults "on" in Moderate and Hard modes) -- for parts as well as crew. RO covers unthrottleable engines with limited ignitions via RealFuels and parts configs. Reentry blackout, as noted in another reply, is an option in the stock game (only makes a difference if "coms required for control" is active). Principia "fixes" gravity, to the limits of what a PC can process in real time, at least (and includes mods to fix issues with the stock solar system -- like making one of Jool's moons retrograde so they won't eject one another). There are mods for decaying orbits, even beyond RSS extending the Earth's atmosphere to 140 km. Persistent Rotation mod carries angular momentum through a time warp. Most of the rest is necessary to make the game one that can run in a single-threaded program on a real-world Windows/Mac/Linux box. I agree, Squad did a lot of stuff right -- and top of the list was making the game open for modifications. Out of the box, it's great fun and will keep a new player busy and learning for weeks or months. Add mods and it can satisfy someone who's been learning about space science for fifty years -- and still run on a computer that old guy can afford.

(1.6.1 RO/RSS/RP-1/Principia) I finally (with a little help from @bewing and @Dale Christopher) got my Research 1 airplane into a flyable state. It wasn't a "flight" problem, as such; rather, it wouldn't stay straight on the runway. After straightening out the main gear (I'd been trying different toe angles, which would have worked for a real airplane on gravel or grass), adjusting the nose gear (shortened, to reduce wheelbarrow effect, and friction reduced to 0.3), and moving the CoM aft and main gear forward by similar tiny amounts, it can take off -- and flies as well as I expected. Only good for about 200 m/s in level flight, but then it's a single 1940s tech engine. The irony here is (even after tooling the two procedural tanks) this will take close to a game year to build, and it'll get flown once, to get science from half a dozen biomes near the Brownsville, TX launch site (can reach Shores, Water, Grassland, Tropics, Mountains, and Desert, and return, in about an hour of flight time, most of which can be compressed with physics warp). But now it can go into my stash of good craft for RO, so I don't have to build it again every time I start a new career (which happens some in RO due to going broke or other forms of dissatisfaction).

@Dale Christopher If you look at the top of this thread, you'll see my wheels weren't far behind the CoM -- the nose actually bounced a bit when the plane loaded onto the runway. I moved the CoM back a smidge, and moved the gear forward a bit, and now the nose bounces even more when the plane loads. Doesn't look like much changed -- just lowered the engine and moved it aft a short distance (20-30 cm at most). The other changes are hard to see or just plain invisible. But now, I can take off, and it's a nice flier once in the air and trimmed. Here's the updated craft file. Thanks for the help, @bewing, I'd forgotten the FAQ had information specific to runway problems.

@Dale Christopher Thanks for the suggestions. @bewing mostly had me covered. Lowering nose wheel friction and following the FAQ suggestion (and method) of removing all toe (in or out) on the main gear, combined with a couple minor tweaks (moved the engine down and back a little, which moved the CG aft, moved the main gear forward, as you suggest, which lightens the load on the nose gear and eases rotation, shortened the nose gear to reduce wheelbarrowing) and it now tracks pretty straight and corrects predictably. I also made some further adjustments to wing incidence and decalage based on trim requirements. As with any longitudinally stable aircraft, it requires trim changes for different speeds, though the top mounted engine stands in for the down thrust commonly used in models. Also as expected, once off the ground, it flies very much like an R/C advanced trainer -- and makes me wish for a joystick. I had underestimated the wing loading; it actually needs almost 70 m/s to rotate, and with near-full tank, dirty stall (gear down, 30 flaps) is close to 65 m/s. IOW, it takes off more like a fighter than a transport or trainer, probably not surprising considering the resemblance to an F-86. I may mount parachutes; there is no way I'll ever be able to land this, while flying with the keyboard. I'm on my laptop now, but I'll come back later and show a photo of the corrected version, and upload the working craft file.

I'll go look at the FAQ again -- been a while since I read through it. For an R/C model, the solution to this would be a few degrees of toe in on the main gear (acts as a brake on the outside of a veer or the "heavy" wing), but I've never flow tricycle (the three powered R/C models I've owned and flown were all conventional gear, which steers a bit differently). I'll try reducing the nose wheel friction to half or less and see if that helps.

Yep, we have Real Solar System and Realism Overhaul (which latter includes RealFuels and RealPlume to make the engines more realistic), and some of us use Principia to handle n-body physics and allow things like chaotic orbits, perturbation, and so forth. And no, it doesn't really need a particularly "beefy" computer. I have one that's quite a bit less performance than a top line gaming machine (AMD FX8350 with 4.1 GHz turbo, nVidia GTX750, some kind of SSD, and 16 GB RAM) and it actually performs better with all the mods for 1.6.1 RSS/RO/Principia than it does in stock 1.8.1. I have to presume this is because the mods I'm using are more efficient code than the stock game. I just came from launching a rocket with a WAC Corporal engine (pressure fed aniline/furfuryl fuel, red fuming nitric acid oxidizer) boosted with a V-2 analog (RD-100 burning 75% ethanol/water with liquid oxygen and high test peroxide to drive the pumps). The booster has 70 second rated burn time and a single ignition; the upper stage 47 seconds rated burn and single ignition (plus I had to manage ullage, and spin up the rocket at the same time, so the upper stage would ignite and wouldn't tumble at 32+ km staging altitude). The main difference is that KSP is a video game -- Realism Overhaul is a simulation. Landing on the Mun in KSP is a challenge. Landing (softly) on the Moon in RO is, well, a "moonshot" level project, or at least a simulation of one.

(1.6.1, RSS/RO/RP-1 -- lots of mods, of course) I'll confess. I hate KSP airplanes. And this one demonstrates why. Simple design, looks like a cross between a Mig 15 and a really early German jet (HE-111 cockpit and engine on top will do that). If this were an R/C airplane, I could throttle up, wait a few seconds to build speed, rotate, and then trim as I climb out before testing with rolls, loops, and so forth. In Realism Overhaul, however, I can't get it far enough down the runway to get anywhere close to flight speed (should be around 40 m/s with flaps down, maybe less). I've built and even designed model airplanes -- my center of mass is ahead of my center of lift (maybe too far ahead; in an R/C bird I'd probably find it to be nose heavy), ahead of the main gear. Steering is disabled on the main gear. I've got the wing set at a small incidence angle, and the horizontal tail a couple degrees less (I expect to need nose up trim on early test flights). But every time I start my takeoff roll, the slightest deviation either direction compounds. I've tried setting the gear parallel to centerline, one degree toe out, and up to three degrees toe in. Same result. Here's the craft file. Any help?

There isn't a way to change parts buying mode in an existing save (at least not without save file editing, which I've never tried). In a "new game" there's a setting in the difficulty settings to "require individual parts purchases" (actually, I think it's "don't require individual parts purchases" that needs turned off), that will require buying parts after each node is researched. It's present in the stock game, at least in 1.8.1, and defaulted "on" in RP-1. Mind you, this is funds purchasing. The closest I know of to researching parts individually is Community Tech Tree, which breaks things down a good bit compared to the stock tree.

Ferrofluids, however, have similar problems to mercury, in that they consist of magnetic micro- or nano-particles suspended in a liquid such as oil -- the magnetic particles increase the density of the oil (though it's still less dense than mercury). Also, the oil will evaporate in a vacuum, though this process is slowed by the magnetic particles (reduced surface area and adsorption bonding). That said, you could actually get away with some external pressure -- after all, your spacecraft are pressurized internally. There's no actual compression on the hull until external pressure exceeds the internal pressure. A ferrofluid twice as dense as water could be thirty meters deep on the Moon before it reaches that pressure. Even mercury could be more than five meters deep in 1/6 G before it reaches 1 atm. Now, getting that much mercury to the Moon is a problem left as an exercise.

Not sure where people get the idea that ethanol would be good in this application. Sure, it has a low freezing point, but it offsets that with higher vapor pressure -- that is, it evaporates. There isn't a low density liquid with low vapor pressure, sorry to say.

I've made an SSTO, once. It looked remarkably like a rocket booster. Reusable, due to landing legs and parachutes -- lots of parachutes. I have trouble even getting an airplane to fly (admittedly, I've only tried in Realism Overhaul, which uses Ferram Aerospace Research to "upgrade" the stock aerodynamics -- and I only have the keyboard to fly with). I admire anything that will both fly reasonably well, and do so over a wide enough speed range to take off from the runway, fly into orbit, reenter, and land (never mind on the runway).