A KSP Design Log
Sun 16 March 2025
I had only intended to post this as a reply to my own "toot" on the Fediverse, but it actually went over my instance's post limit of 5,000 characters, so here we go! 😆
On Friday night, I posted a toot saying that I was intending to play Kerbal Space Program that night (which I haven't played in a while). My idea was to make a close-to-stock Learjet (like the one I flew in MS Flight Simulator many years ago) in KSP, but then add parts to it to make it space-faring, but 100% recoverable, and absolutely indistinguishable (both visually and from a BOM perspective) once the space "bits" had been jettisoned.
So, I was up until 2am Saturday morning doing this. Staying up way past my bedtime seems to be a KSP tradition for me, going all the way back to early 2016 😅. It explains why I don't play the game very often. It is so engrossing to me!!
While I had intended to use the (circular cross-section) mk1 parts for the Learjet for optimum visual realism, I had to use mk2 parts, because the mk1 passenger cabin only seats two. That's why. It had nothing to do with the fuel capacity or heat resistance of mk2 parts. 😁
I did disable the mk2 cockpit's reaction wheels (Learjets don't have reaction wheels!), and the Learjet itself was pretty decently realistic, sporting a little fuel adapter/tail piece (no oxidizer) and angled pointy nose cone in the back, and two engine nacelles sporting each a mk1 Liquid Fuel Fuselage, a standard mk1 Circular Intake, and a "J-33 'Wheesley'" engine for comparable performance.
I still ran out of electricity during the orbital portion of the flight, because keeping the lights on drains power, but the power came back as soon as the "Wheesley" engines came back online within the atmosphere. I may add a solar panel for the next iteration, maybe placing them on the orbital engine pods for realism (since, again, Learjets don't usually have solar panels).
The Center of Gravity was a little fiddly and tail-heavy, but I blame KSP for not putting any fuel capacity inside the belly of the passenger compartments like real airplanes have. (Or in any of the smaller wings, for that matter). I might add a few mk0 fuel tanks with mk0 nose cones attached radially to the bottom of the passenger compartments and translated internally, but I'm not sure if I want to go that far. That's a design tip I gleaned from "Cupcake Landers" on YouTube, who used an array of mk0 fuel tanks tucked into the base of a plane consisting of just two cargo ramps, wings and engines.
I then added two detachable RCS/OMS (space engines) pods consisting of about 12-16 of mk0-size "Oscar-B" fuel tanks each connected to two pylons in the back, arranged in a double stack so that I had room for four 48-7S "Spark" mk0-sized engines, total. Each stack/cylinder of the RCS/OMS engine pods was capped with a mk0 nose cone. The pods didn't provide a LOT of thrust (about 0.3 TWR), but it was fine for orbital maneuvering, and also provided basic attitude control during orbital maneuvering burns. I added a couple mk0-sized FL-R20 RCS fuel (monopropellant) tanks to each pod (four total), and four directional "Place Anywhere 7 Linear RCS Ports" to each pod. Since the RCS thrusters used for roll control were attached pretty far from the rotational axis of the craft, they gave a bit too much roll authority and caused the RCS to oscillate a lot, but I didn't want to put any thrusters on the Learjet itself, because Learjets don't have RCS thrusters! I also removed the monopropellant/RCS fuel from the mk2 cockpit for the same reason. I kept learning that less (power) is more when it comes to RCS, at least for vessels of this size.
The RCS/OMS rocket engine pods also had a "Mk16 Parachute" on the front of each cylinder (four total), and I tested jettisoning the rocket engine pods while in low-altitude flight powered by the two "Wheesley" engines. The parachutes deployed just fine at subsonic speeds (which is all the "Wheesleys" are capable of, lol), and landed safely, with no part loss. Circling around the parts in the Learjet to remain in loading/rendering distance to ensure their safe recovery was a bit of a pain (they had to be jettisoned at 400m altitude, as 200m proved fatal to the parts' recovery), so I just landed, taxied to the parts to inspect them up-close, then took off again.
The launch booster is still in early alpha. I was mainly testing the RCS/OMS and reentry, so I just created a huge, somewhat overpowered orbital booster with three SSME-clone "S3 KS-25 'Vector' Liquid Fuel Engines", a "Probodobodyne HECS" probe core, and a "PB-NUK Radioisotope Thermoelectric Generator" (RTG) for electricity during non-powered flight. The orbital booster had no reaction wheels or RCS, but relied on the Vectors' overkill 15° gimbals for attitude control and re-orienting, and had a wee bit more than enough fuel to push the Learjet into orbit and perform its own retrograde burn. Of course, the vector engines required adjustments to the maximum deflection angle during latter parts of the flight, because overzealous gimballing led to a lot of painful oscillations, as per norm for that engine with the not-exactly-NASA-spec RCS algorithm in KSP.
I did lose a single vector engine due to heating from a butt-first reentry (despite using four rather large wings to try to keep the heat-resistant nose pointing forward), but I think this can be remedied by fuel balancing for the next launch: preferring to keep the front two fuel tanks full and to drain the rest first in order to keep the center of gravity further forward. The single mk1-sized heat shield at the very front (just behind a mk1 nose cone) should be more than enough to handle reentry heating if I can keep the nose pointed forward, but I think I will scrap this alpha orbital booster and replace it with a horizontal takeoff winged booster, as the TWR requirement for horizontal takeoff is only 0.3-0.5 TWR, as opposed to 1.3-1.5 for vertical. That allows for a much more efficient, less brutalist craft. Nevertheless, the entire booster was recovered in a water landing with parachutes and no engine power, save the loss of the one Vector. I used the trick of placing extra parachutes on the base of the dorsal side of the booster to prevent it from landing on its butt and then tipping over (forcing a bit of a belly-flop landing, which is safe enough in KSP).
The orbital booster is connected to the Learjet orbiter by an engine pylon that has been mounted to the bottom of the Learjet and translated to visually mate with the tail. The (next) root part of the booster, a mk1 rocket fuel tank is then translated to visually mate with the back of the pylon. Using a pylon like this allows both the very front of the booster and the tail of the Learjet to remain aerodynamically efficient after separation, vs. using any kind of stack separator. The overall craft retains structural stability (despite KSP's noodley physics) through the use of autostruts on several parts, particularly the cockpit (set to autostrut to the current heaviest part), and several parts on the periphery of the craft, particularly on the engine pods and the booster.
Surprisingly, the Learjet handled reentry just fine, even without any real heat shielding. Using more heat-resistant mk2 parts for the bulk of the plane helped, but I also did allow myself one minor cheat: using another trick learned from "Cupcake Landers," I placed a single "Place Anywhere 7" Linear RCS Port at the very tip of the nose of the cockpit, and each engine intake. They were turned inward to look more like decorative parts, and had their own RCS toggle turned off, so they weren't ever used for attitude control, but just as "heatsinks" to keep the more vulnerable parts safe. I don't think the RCS port "balls" on the engine intakes were needed after all, but I know from experience that having one on the nose of a plane helps with reentry heating, especially if using mk1 parts (which I did not do, as they're far too vulnerable to reentry heating, even with the "RCS Ball" trick).
The RCS was a bit fiddly during reentry (since the RCS pods were only located on the rear extremities of the orbiter, on the spacefaring engine pods), but it managed to maintain a slight angle of attack of about 20° during reentry, which was enough to induce enough drag and keep the plane from getting hot during reentry, even from a 500km apoapsis. I reentered with a modest 50km periapsis, which required an extra aerobraking pass, but kept the safety margins comfortably high.
The next time I play (I never know when that will be), I look forward to designing a proper horizontal-takeoff-and-landing booster to bring the Learjet into orbit more efficiently, without requiring bombastic Space Shuttle Main Engines. I might also add an optional kick stage to make a Mün landing possible. 😄
100 Days to Offload 2025 - Day 15
Category: Entertainment Tagged: 100DaysToOffload ADHD Entertainment Gaming Hobbies Non-religious post Non-technical post Science Fiction