Zeiss Ikon

Real world rocket engines are always run slightly fuel-rich for two reasons. One is to ensure that the exhaust is rich in carbon monoxide, and poor in carbon dioxide; this is done because CO, with its lighter molecular weight, gives better exhaust velocity (= Isp) than the heavier molecules of CO2 (as well as about double the gas volume). The Isp levels of most of the engines in the game are inconsistent with a CO2-heavy exhaust, but right in line with a fuel-rich combustion giving mostly CO (in addition to the water vapor). The other reason is that a fuel rich combustion is far easier on the materials of the combustion chamber and nozzle -- metals in general are subject to oxidation, because they're fully reduced. As a result, a reducing environment tends to leave the metal mostly unchanged, while an oxidizing environment tends to rapidly eat away at the metal. You see much the same thing in oxy-acetylene welding; you'll prefer your flame to be slightly on the reducing side of neutral, rather than slightly oxidizing, for most welding processes, and this is why. The F1 engines in the Saturn V first stage were actually "film cooled" -- they used an injection system that put a film of kerosene vapor along the combustion chamber and nozzle wall. Without enough oxygen present to fully combust the hydrocrabons, this film left the chamber without reacting, and carried away heat from the chamber and nozzle in the process. Look at video of a Saturn V startup and you can see the blackish layer partially obscuring the bright core of the exhaust -- that's the cooling film. This method has been used in many hydrocarbon fueled engines, because it's cheap, fairly simple to implement, and works pretty well.

My recollection from the Eve SSTO discussion is that there are a few peaks that reach as high as 7.5 km -- which strongly implies little land above 5 km. As for 3000K oxygen, fluoride passivation is a marvelous process. But as you note, that doesn't help much for a hydrogen atmosphere. And running at lower temperature to ensure material compatibility cuts deeply into the efficiency of a Brayton cycle (gas turbine/turbojet). I think we're back to a closed-circuit "steam" system driving a propeller (or ducted fan) for loiter and to get high enough (in Jool and Eve atmosphers) for rockets to work at departure. Or just not trying to land on Eve or Jool...

If you have an NTR that can also be run as a thermal turbine/jet, any atmosphere that can be compressed would work as reaction mass -- Jool's (presumed) hydrogen-helium, Eve's (presumed) carbon dioxide/nitrogen, or Duna's (presumed) thin carbon dioxide/nitrogen. You'd need a variable geometry compressor and nozzle to have a single engine work in all three locations, but that's not a hard problem (compared to actually getting the ship there in the first place). If Jool's atmosphere is, in fact, mostly hydrogen, any oxidizer ought to work with it -- and oxidizer isn't significantly heavier than (presumed) kerosene to run a conventional jet on Laythe. The big issue you'd have with this nuclear thermal jet is Eve -- if the atmosphere is mostly CO2, it might well be compressing to a liquid at the surface (that's about 800 psi at room temperature, though much above room temperature you wind up supercritical before it properly liquefies -- so you don't get a well defined surface like the "water" on Eve). Compress that, anywhere near "sea level", and you'll find it liquefying in the compressor; you'd have to preheat it in order to compress it, then heat it still more to get it to provide turbine energy (to run the compressor stage) and give a pitiful little bit of jet thrust. I'm afraid that on Eve, at least down low, a propeller is likely to be the most efficient way to move something in the air -- and then you'll want wings to hold it up, so TWR less than one won't keep you grounded. A Rankine cycle turbine (steam turbine, though it could run on anything that has a convenient boiling point) running off your nuclear thermal source is probably the best way to do that -- and that engine will work anywhere there's atmosphere to push and a way to get rid of waste heat.

I very strongly suspect this will result in a significant increase in system load, not to mention Squad having to (more or less) replicate a Computed Flight Dynamics system (which can be pretty complex code -- because it's pretty complex math).

(Science game) This time, the engineers found and fixed the problem with staging in the Munar Explorer series (or so they claimed), and in the process built the first example of Munar Lander. As previously noted, both Munar Explorer and its outgrowth, Rescue Alpha, had exhibited staging issues with the parallel staged boosters. Instead of pumping fuel from booster to booster and thence to the core, they were running all tanks dry simultaneously, remaining attached even though the radial separators had fired, and essentially operating as a "big, dumb booster". This was enough for a Munar orbit mission, and sufficient to rescue Sigemy after he'd flown his tanks dry doing a Munar flyby with a ship that wasn't really up to it -- but it was clearly going to be unacceptable for an actual Mun or Minmus landing mission. Much coffee later, a nameless intern was caught in the VAB with a tape measure. After convincing security to call Gene before hauling her off the base in chains, she was allowed to report to the Design Committee with her findings: the strap-on boosters were being compressed too much at mounting, causing the radial decouplers to fail without sensors reporting the failure. The excess compression also blocked the fuel lines between boosters and the core, preventing all the fuel transfers from taking place (fortunately, the transfer pumps weren't powerful enough to burst the transfer pipes, else the resulting fuel/oxidizer venting would likely have resulted in a spectacular fireball well before the first staging event should have occurred). With this knowledge in hand, the engineers and mechanics in the VAB assembled a new Munar Explorer booster, and could easily see the corrected spacing between boosters and core. Based on that visible correction, they upgraded the upper stage into a two-stage lander -- using the only landing legs available, which carry a caution to avoid hard landings, and splitting the tank previously used in the final stage (the one that orbited the Mun and returned to Kerbin) into separated descent and ascent stages, with another Terrier engine between. They incorporated RCS, in part because they could, in part because there was reason to believe it could be used to make short hops between points on the Munar surface without burning return fuel, and in part because it would give a last chance to peel off velocity before reentry, improving the likelihood of externally mounted experiments surviving the landing. The first launch of the new Munar Lander was a resounding success, with the parallel staging working exactly as the Design Committee had always insisted it should. Maltrey was chosen for the first Munar landing attempt, in part because of his outstanding work in rescuing Sigemy, and in part because he was the only pilot other than Jeb willing to fly the landing after seeing the so-called landing struts. At the time of this report, Maltrey was in parking LKO, with (reportedly) enough fuel remaining in the transfer stage for transMunar insertion and Munar capture. He has four goo canisters, and four each barometers and thermometers, and hopes to not only plant the first flag on the Mun, but also retrieve enough science from the surface to keep R&D busy for weeks.

Of course, the other concern with a poorly licensed mod is wondering what else the author might have screwed up -- or whether there might even be a security risk to your computer. As I understand it, Squad is very, very generous with licensing for mods -- so if someone just can't be bothered to go through the process, you have to wonder whether it's plain, ordinary incompetence (which is, after all, a very Kerbal quality) or malicious intent. Not saying anything specific to that mod, mind you -- just that KSP mods (in my deeply UNinformed opinion) seems as if they could embed "arbitrary code" on your system, which is the usual definition of "payload" in the software security world.

That's what Vernor engines are for...

When you have to embed the booster nozzles in the floor to add bits to the command pod because you can't shift view any higher within the VAB.

Today marked the first test flight of the new Munar Explorer, built with the latest available technology (for the moment) and virtually certain to be able to orbit either Mun or Minmus. Might even carry a lander, if the engineers at KSC can ever get parallel (asparagus) staging figured out. As it is, the fuel transfer from booster to booster and booster to core doesn't work, and the boosters won't stage off the core, even though the decouplers fire as intended. More tests coming up. However: even with the staging issues, Munar Explorer is more than capable of sending a Mk. 1 Command Pod and a large enough service stage for capture and return into transMunar trajectory. Sherny Kerman was up in the rotation for this mission, and despite his dismay at the staging issues, handled the flight like a pro, eventually lowering both his periapsis and apoapsis below 11 km and collecting all the science from "In orbit above the Mun" that there was to collect before burning back to Kerbin. He set up a 35 km apoapsis for reentry, and had significant fuel remaining in the service stage when he dumped it just as reentry heating started. Unfortunately, during reentry, things got a bit hotter than on previous missions, and while the command pod itself and its parachute were fine, all the instruments except one thermometer burned off the exterior of the command pod (they've survived from a previous Munar flyby, suggesting reentry profile was wrong). This wasn't much help in gaining enough data to turn into the technology to rescue Sigemy (who'd been stranded in his Orbiter Alpha Type 2 after running his tanks dry trying to make a Munar flyby ahead of schedule), so a simpler mission was set up. Engineers quickly upgraded the Swiveler II, an early "spam in the can" vessel, with extended tanks to produce Swiveler III, and Jeb himself strapped into the command pod for a hop to the polar regions. With careful piloting, Swiveler III has enough dV to fly to any point on Kerbin. As a result, getting to the North Pole was a cinch -- Jeb let the vessel go far higher than needed, topping out above 400 km, and used the time in flight to take the requested observations -- but then made a discovery about suborbital reentry: even from a mere 400 km, and well below orbital velocity, plunging steeply into the atmosphere will subject the capsule and pilot to excessive stresses. Not only did all the experiments (except one thermometer, again) burn off the exterior of the pod, but Jeb came within an eyelash of blacking out from the G load as the capsule plowed into dense air while still at high supersonic speed. Landing was safe, however, and a small amount of science was retrieved. Finally, the decision was made to attempt something entirely new: Sherny made an EVA from a command pod still on the launch pad, collected surface samples and reports, and returned those data, finally providing enough knowledge to let R&D produce the final, critical component needed to rescue Sigemy: Reaction Control Systems. A variation of Munar Explorer was dubbed Rescue Alpha. With the science experiments deleted, and what the engineers swore was a fully rebuilt and inspected parallel staging setup, a second command pod was added to provide a seat for a single rescued Kerbal, provided he could fly his suit jets well enough to board via the crew hatch. Maltrey won the drawing to pilot the rescue mission (Jeb swore up and down his name had been left out of the helmet, even though Gene found it after going through the rest of the pilot slips). Unfortunately, the staging still wasn't right; all six boosters remained attached and failed to transfer fuel as designed, but again, even with the booster running in "big, dumb" mode, there was enough dV for the mission, which was to do another thing never done before: rendezvous with a ship already in orbit, and "park" close enough at low enough velocity for a Kerbal to cross in EVA, using the never-tested suit jets. The mission was complicated by the high apoapsis and periapsis (hence long period -- close to that of Mun, in fact) of Sigemy's orbit, combined with a significant orbital inclination, but Maltrey had spent tens of minutes on the simulator (which suspiciously resembles a video game), and went from closest approach of 400+ km to parked at less than 5 m separation with aplomb. Then the cowboy, Sigemy, made it look easy as he jetted across from his vessel (the only thing still working being the reaction wheels to let him stabilize) to the second command pod's hatch on Rescue Alpha. Once inside, there was minor concern whether there would be enough dV to return to Kerbin, but Maltry quickly put those concerns to rest; not even waiting for an apsis (apparently unwilling to spend a week or more to adjust that long, long orbit), he did a radial burn that brought him to his chosen reentry periapsis, and by doing so saved enough fuel to be able to decelerate by a couple hundred m/s just before atmospheric interface. The two capsules (the lower festooned with RCS tanks, the upper with parachute canisters) proved to be stable during reentry, and fortunately the "anything on the right planet" landing put the pair on land, so they didn't have to find out if the stack would tip over, or leave Sigemy under water until the recovery crew could get him out.

I've only been playing for a couple months -- so far, I haven't yet landed anywhere beyond Minmus, and only made a single Duna flyby (in sandbox, yet, though that was direct launch -- one booster, nothing assembled or refueled in orbit, and without a helpful assist from Duna's gravity well, that vessel would never have come home). Come to that, I've never found the controls to transfer either crew or resources, hence never refueled a vessel in flight. Never used a significant mod, either -- the only one I've installed so far is Better Burn time, which is nice but, for orbital maneuver's, not really essential (I expect it'll look a lot better the first time I need to get down to Mun on a too-small fuel tank and need a near-suicide burn to make it). Lots more stuff to look forward to -- building and flying spaceplanes and other SSTO vessels, assembling ships in orbit for missions bigger than one Kerbal to fly by Duna, ISRU refueling, nuclear and ion engines. Most of the game is still ahead of me.

This would be good.

I'll get on the bandwagon for this one -- for landing on moons as well as for landing boosters back at KSC. If you build a lander with 1.25 m parts, it's really not very stable on landing even with a Terrier engine; if you mount a Swivel or Reliant, you'll find the legs won't even touch the ground before the engine bell. If we had longer legs, more or less in SpaceX style, we'd have a more "spread" stance on a surface, even if we don't get much additional height. Otherwise, I'm going to have to come up with a way to launch a lander with radial tanks on a 1.25 m launcher, just so I have a place to put the legs to give enough stability for a landing without perfectly zero horizontal velocity (especially likely before I get RCS in my Science save). And yes, so far I'm almost mod free; the only mod I've installed to date is Better Burn Time. Nothing against mods, just trying to keep the load on my computer under control (apparently people who ride their old machines until they drop shouldn't try to play games at all).

Honestly, I doubt there's much to choose between methane and kerosene. Methane gives higher Isp, kerosene lets you use smaller, lighter tanks (methane falls between kerosene and hydrogen in both respects). The advantage of methane for SpaceX is that sending a tank of hydrogen (no oxidizer, just H2) to Mars will let them make both methane fuel and liquid oxygen to fuel an ascent vehicle and/or return transfer -- that's ISRU in the real world, or approximately the Zubrin method of going to/from Mars. I honestly have no idea how they'll keep hydrogen liquid for a year and a half while it gets to Mars, never mind in the Martian atmosphere after they land it and wait for the solar panels to come on line to power the conversion unit (much easier to keep the LOX and methane cold, as they're both more than 100 C warmer than LH2). But yes, at least for hydrocarbons, the lighter the fuel (in terms of molecular weight) the more of the exhaust is (light) water vapor and the less (heavier) carbon monoxide and (heavier still) carbon dioxide -- and the lighter the average molecular weight of the exhaust, at a given chamber pressure, the higher the exhaust velocity (hence Isp). That's part of why the NERVA engine (designed in the 1960s, but never flown, at least to date) had such high Isp: it was to use liquid hydrogen, only, at very high temperature, so had the lightest possible exhaust molecular weight and lots of energy in that exhaust.

The high density of kerosene lets you make the tank smaller, which improves your mass ratio (smaller tank = lighter tank). It's still very unclear whether NASA gains anything in terms of payload to LEO by burning LH2/LO2 in their SSME derivative engine, even at their very high pressures. And SpaceX's new Raptor runs at almost SSME pressure, but on Methane/LOX -- about the same size and weight as a Merlin, but with (IIRC) four times the thrust and higher Isp. Liquid methane, however, is denser than liquid hydrogen, so they'll be able to get the same total impulse with about half the tank volume.

SpaceX Merlin engines run on Kerosene and LOX, and as their booster demonstrated yesterday, they can be ignited at least three times on a single flight. I get the impression that they can be reignited more times than that; the igniter for an engine like that is basically a spark plug in a blowtorch.

(Science game started recently) Yesterday, Sigemy took his turn in the command pod, scheduled for another orbital mission. Unfortunately, Sigerny turns out be a cowboy; he got the Orbiter Alpha Type 2 in to orbit with about 80% fuel remaining in the service stage, and decided to violate orders; he was sure the vessel had enough dV remaining to make it to Mun. Ground control was screaming at him as the insertion burn ran down, but it was too late; he was committed to go to Mun. He made two tiny correction burns along the way, lowering his Mun periapsis and adjusting his Munar inclination. Just before he went out of comm range of Kerbin, he reported back that he had plenty of fuel for the Munar periapsis burn, but wouldn't have enough to adjust his Kerbin periapsis afterward. He wouldn't be departing into Kerbol orbit, in other words, but he'd be stuck in a high Kerbin orbit (4489 x 27359 km) until a rescue could be mounted. Good news is, he comes into comm range for a couple days every periapsis; bad news is, the technology to rescue him doesn't yet exist. He'll be out there for a good while. Doing what he could with what was available, Maltrey stepped into the command pod of an untested and newly upgraded vessel, Munar Orbiter -- perhaps a bit optimistically named, but that was the intent of the upgrade -- and set off to bring back as much science as possible, to speed the day when rescue would be technologically possible. Launching with the now-standard collection of four goo canisters, and three each barometers and thermometers, he reached orbit with much more dV than Sigemy had had, and promptly set up a Munar transfer. Not drawn in by the optimism of those who'd named his vessel, he didn't attempt to make orbit around Mun, but had plenty of dV to ensure a prompt return to Kerbin. Due to a misunderstanding, he rode through a high Kerbin periapsis and another orbit to just beyond the Mun's path, where he adjusted his periapsis for reentry. His service stage had more than 20% fuel remaining when he kicked it loose, just as reentry heating started, suggesting that with careful planning his vessel might, in fact, be capable of entering a Munar orbit and still returning to Kerbin -- but with all the data he brought back, R&D was able to crank out another round of upgrades rending Munar Orbiter obsolete before the charred paint on the command pod was fully cooled. Working deep into the night, engineers and the VAB crew put together a magnificent creation, with six strut-braced boosters feeding fuel into the core, a second core stage the same size as the booster core, and a third core stage with three fourths the tankage and a high efficiency engine. Dubbed Munar Explorer, there is no doubt in any mind that this ship can orbit both Mun and Minimus, and there are whispers that it may be able to carry a lander to one or both. Once R&D invents landing legs, that is.

I tried a pad test of fuel cells, but couldn't really tell what they were doing. I should probably rerun that test and take the step of draining the command pod's battery before putting the testbed on the pad.

Well, I just went and looked, and apparently it's not an option. So, if you have a tank you want to contribute to your entire vessel, don't put a heat shield behind it. I've had fuel cells that seemed to work when mounted on the command pod, however, with the pod's heat shield between them and the service or transfer stage's tank(s). Maybe I'm confused at how fuel cells work -- the in-game documents aren't very clear on that, as with many other things.

No, I was asking if SpaceX was using Mechjeb to fly their booster landings.

Well, it does if it's monopropellant. I haven't tried a heat shield separating Lf/O tanks from the rest of the ship, but I seem to recall that they have a crossfeed toggle, too (defaults off, just like a docking clamp or stack decoupler).

Didn't get to see it this time, but the computer driving those boosters seems to have suicide burns completely nailed. Is that Mechjeb?

What I'd rather see with fuel cells is the ability to dedicate a small Lf/O tank to fuel cell operation -- so if I burn my tanks dry with (say) a Kerbin return insertion burn, I don't have to get home on just the installed batteries (with no more way to recharge them, if I've installed fuel cells instead of solar panels). Flow priority won't do this; no matter how low the priority, a running engine will still suck down all the fuel. Careful setup of crossfeed enable/disable might do the job, but it'd be easier to just have a dedicated tank. An FL-T100 or toroidal would be plenty for a longish mission, generally, if I could be sure the maneuver engines wouldn't suck down the fuel.

Nice. One each at the leading and trailing Lagrange positions (L4 and L5), and one at (if the game handled the physics using real n-body calculation) the opposition point (L3). The latter should be a tiny bit further from Duna than Ike is, but the difference might be too small to notice. The L4 and L5 are stable enough you wouldn't even need station keeping thrusters (in real life the corresponding Jupiter-Sun points actually have families of asteroids captured from the main belt), but you'd probably need to make provision to refuel the L3 station in proper physics; it's only metastable.

So, basically similar to Europa or Callisto in our system?

I don't know -- a lot of modern phones, especially the top end of the current generation, have similar computing power to a five year old desktop computer. I think my new Google Pixel actually outperforms my Core2Quad desktop machine; I'm certain it's faster than my Core2Duo laptop. The screen on a phone is a little small, but if you have a stylus, you can touch with enough precision to do the work. My tablet isn't anything like that fast -- but it was $100 a year ago, bought mainly to have an Android device with a keyboard (which plugs into one of the USB ports), and it's still competitive with my laptop (tablet has quad core 1.7 GHz, which compares well with a Core2Duo at 2.1 GHz). Tl;Dr I think modern phones and laptops have plenty of performance to handle the building side of this, and might even allow launching and flight if you turn down the physics delta-time slider. And for Android, at least, it's just a little more work than a cross-compile, since Android is essentially Linux Lite for ARM processors and phone/tablet peripheral hardware. If this were available for Android, it might be the first Android app I pay for.