gchristopher

Totally agreed. (Or with omitting it entirely.) I just don't know how to draw a line between "reasonable" air intake usage and "airhogging." 3 intakes per engine is often considered a minimum for a useful spaceplane. Is 5-6 per engine excessive? This one is 10 per engine, which is definitely excessive. Do you try to not clip them and at what point is covering a wing with intakes considered too much? Either way, this has been fun! Thank you! It's really interesting to try to figure out just how much travel you can get out of a single launch.

Here's another attempt in progress! [table=width: 900, align: left] [tr] [td] SPH view[/td] [td] Takeoff![/td] [td] Air intake abuse gets us to a 68.6km semi-major axis using only jets.[/td] [td] Mun landing #1[/td] [/tr] [tr] [td] Docking, lazily letting Mechjeb align both craft.[/td] [td] Attaching KAS struts. (Still very weak and floppy)[/td] [td] Kerbin landing #1.[/td] [td] Back to LKO.[/td] [/tr] [tr] [td] Mun landing #2.[/td] [td] Kerbin landing #2: overshot the runway and had to turn around and land from the east.[/td] [td] Back in orbit.[/td] [td] Mun landing #3.[/td] [/tr] [tr] [td] Kerbin landing #3, didn't do quite as well at aiming at the runway.[/td] [td] Mun landing #4.[/td] [td] Kerbin landing #4. Descent was mostly an uncontrolled tumble! TAC fuel balancer saved it in the final moments by restoring a flyable COM![/td] [td] Mun landing #5.[/td] [/tr] [tr] [td] Kerbin Landing #5: almost overshot the runway again. A dramatic S turn brought it in.[/td] [td] Mun Landing #6.[/td] [td] Kerbin Landing #6.[/td] [td]Low-angle approaches to steep terrain were typical.[/td] [/tr] [tr] [td]Landings were not always picture-perfect.[/td] [td]Mun Landing #7: toggling landing legs works to right the lander again![/td] [td]Kerbin Landing #7[/td] [/tr] [/table] This attempt is using all stock physics. Mods are mostly for automation (MechJeb and TAC fuel balancer), reducing part count (Procedural Wings and welded/multiplied air intakes), and KAS struts to stabilize the lander on the launch vehicle, instead of fiddling with multiple docking ports. The vehicle is a horrendously air intake abusing spaceplane. Its stats are: Launch Mass, 97.922t, Cost: 171,258.6√, 1 LV-N on the lander, 12 TurboJets, 120(!) Ram Air Intakes, LiquidFuel: 9540, Oxidizer: 2508, Monoprop: 115 Part count is 141, but would increase by ~100 without procedural wings and by another 216 without welded air intakes. (But performance should be nearly identical, other than a sad CPU.) Here's the resource log at each runway stop on Kerbin: [table] [tr] [td][/td] [td]LiquidFuel[/td] [td]Oxidizer[/td] [td]Mass[/td] [td]Usage: (LF/Oxy)[/td] [/tr] [tr] [td]Launch[/td] [td]9540[/td] [td]2508[/td] [td]97.922t[/td] [td][/td] [/tr] [tr] [td]Landing #1[/td] [td]8064[/td] [td]2121[/td] [td]88.602t[/td] [td](1476/387)[/td] [/tr] [tr] [td]Landing #2[/td] [td]6870[/td] [td]1759[/td] [td]80.817[/td] [td](1194/362)[/td] [/tr] [tr] [td]Landing #3[/td] [td]5806[/td] [td]1453[/td] [td]73.948[/td] [td](1064/306)[/td] [/tr] [tr] [td]Landing #4[/td] [td]4863[/td] [td]1149[/td] [td]67.714[/td] [td](943/304)[/td] [/tr] [tr] [td]Landing #5[/td] [td]4033[/td] [td]848[/td] [td]62.059[/td] [td](830/301)[/td] [/tr] [tr] [td]Landing #6[/td] [td]3228[/td] [td]557[/td] [td]56.523[/td] [td](805/291)[/td] [/tr] [tr] [td]Landing #7[/td] [td]2432[/td] [td]225[/td] [td]50.524[/td] [td](796/332)[/td] [/tr] [/table] At this rate, it looks like another two round trips to the Mun should be possible before running out of oxidizer just short of a third trip, but I didn't think through COM very carefully, with all the engines and intakes loaded on the back of the craft, so it might become unflyable before then. (Still, if it makes 7 trips, that's better than the 6 I thought I'd get at the start.) I think the score at this incomplete stage is: Mun multiplier: 2.5, Trips: 5, Start Mass: 97.922, End Mass: 62.059 (Mult*(Tr-1)^2*100) = 4000 + ((Em/Sm)*50000) = 31688 (capped at 4000) Um, bonuses would be Junk hoarder: 200 (not cumulative with No Junk Left Behind?), Manning the Guns: 100? So, the score so far is 8300. As time permits, let's see if it can make any more trips! This entire thing is an exercise in ridiculous airhogging, which is pretty darn exploit-ey. Maybe the score deserves an asterisk and a frowny face? edit: Trip 7 is complete, for a score of 18,400. And that's as far as we're getting, because the oxidizer won't last for another round trip!

Another few categories (all of which have popular plugins) to consider would be: - Capabilities to move/change/reconnect parts, such as Kerbal Attachment System and Infernal Robotics. - Rendering/beautification/texture improvements. - Depth of gameplay with expanded contracts/biome/science options. All three of those might rate above weapons.

From a previous challenge: Sped Boat Mark I. All stock. Tops out around 320 m/s. Not useful for actually going anywhere. All the relevant design features are in the screenshot.

The more I thought about it, the more dissatisfied I became with the scoring system. Basically this is a CPU challenge, not an efficiency challenge, since your score doubles if you slap twice the parts on your vehicle. So, how about if we look at the entries with a scoring that weights efficiency higher? What if we count USABLE fuel (9:11 LF:oxy ratio) instead of just total fuel? The results are interesting. [table=width: 700] [tr] [td]Name[/td][td]LF[/td][td]Oxy[/td][td]Total[/td][td]Usable Fuel[/td][td]Launch Cost[/td][td]Recovered[/td][td]Net Cost[/td][td]Bulky Score[/td][td]Cost/Fuel[/td][td]Cost/Ton[/td][td]LOG10 Score[/td][td]Usable LOG10 Score[/td] [/tr] [tr][td]Vector 2[/td][td]0[/td][td]37,932[/td][td]37,932[/td][td]0[/td][td]517,374[/td][td]506,026[/td][td]11,348[/td][td]126,792[/td][td]0.3[/td][td]59.83[/td][td]15.31[/td][td]0[/td][/tr] [tr][td]DundraL (FAR)[/td][td]12,761[/td][td]78,745[/td][td]91,506[/td][td]28,358[/td][td]927,640[/td][td]871,132[/td][td]56,508[/td][td]148,180[/td][td]0.62[/td][td]123.51[/td][td]8.03[/td][td]2.23[/td][/tr] [tr][td]Wanderfound (FAR)[/td][td]3,728[/td][td]4,186[/td][td]7,914[/td][td]7,612[/td][td]149,322[/td][td]143,322[/td][td]6,000[/td][td]10,439[/td][td]0.76[/td][td]151.63[/td][td]5.14[/td][td]4.92[/td][/tr] [tr][td]Pecan[/td][td]2,880[/td][td]3,520[/td][td]6,400[/td][td]6,400[/td][td]159,458[/td][td]132,189[/td][td]27,269[/td][td]1,502[/td][td]4.26[/td][td]852.16[/td][td]0.89[/td][td]0.89[/td][/tr] [tr][td]Vector 1[/td][td]218,435[/td][td]266,976[/td][td]485,411[/td][td]485,411[/td][td]8,029,533[/td][td]0[/td][td]8,029,533[/td][td]29,345[/td][td]16.54[/td][td]3308.34[/td][td]0.34[/td][td]0.34[/td][/tr] [tr][td]Bothersome[/td][td]10,354[/td][td]12,654[/td][td]23,008[/td][td]23,007[/td][td]355,906[/td][td]0[/td][td]355,906[/td][td]1,487[/td][td]15.47[/td][td]3093.76[/td][td]0.28[/td][td]0.28[/td][/tr] [tr][td]s1l3nt_c0y0t3[/td][td]2,880[/td][td]3,520[/td][td]6,400[/td][td]6,400[/td][td]365,306[/td][td]277,469[/td][td]87,838[/td][td]466[/td][td]13.72[/td][td]2744.93[/td][td]0.28[/td][td]0.28[/td][/tr] [/table] I refer to the current score as "bulky score." I favor LOG10(payload) / (cost / payload) as a way to reward larger payloads while still giving greater weight to efficiency. Also, fuel or oxidizer in excess of the 9:11 ratio either way is generally unusable in space. (especially oxidizer) Disregarding unbalanced fuel amounts is another good idea for a measure of the usefulness of an efficient lifter. Vector's first entry is the most egregious example of how scoring so heavily on payload mass is bad. It would be the third highest bulky score despite having the worst cost efficiency of anything posted so far! Vector's second entry is twice as cost-efficient as the nearest competitor and delivers a massive amount of fuel, so a challenge with "Economic" in the title really should rate it first. But interestingly, Wanderfound's entry does the best job of delivering a useful fuel balance efficiently, so you could make an argument that it's the most practical "economic" entry so far!

Not quite? If you just want to cut part count and improve aesthetics, I think it'd be: @PART[*]:HAS[@MODULE[ModuleResourceIntake]] { @mass *= 10 @MODULE[ModuleResourceIntake] { @area *= 10 } } Fixed the area increase. Wouldn't increasing the mass give you the same performance as if you'd stacked 10 intakes at that spot?

Actually, I think I was wrong to agree with this. I'm sorry! Cost is intrinsically related linearly to payload, since the game doesn't really have any economies of scale beyond very small values. The formula really should produce only 2x a score for a payload that's twice as large. You use twice the lifting parts and spend twice the fuel. I'm sorry! The formula should only be linear in payload size, as Bothersome correctly points out.

Yep, it needs a better scoring formula. Maybe if you still want to reward bigger payloads without it being as dominant a factor, something like: P = Payload, fuel + oxidizer, in tons C = Net Cost after recovery of any launchers Score: P/C * Log10(P) Then a 100 ton payload is worth only twice as much as a 10 ton payload? If the intent wasn't to make a size challenge, Score = P/C with a minimum size, or size categories, is fine.

Suggestion: Turn this into a piloting challenge instead of a building challenge. Since there's a lot of subjective rules, (what is "recognisably a plane", what altitude is "low-altitude soup", where exactly to draw the line on "abusive airhogging"), winning this challenge will mostly be about who can bend the rules enough and still pass your interpretation. Sounds like a headache for all involved. How about, if you want a good challenge, make everyone fly the exact same plane and see who can get to orbit the fastest, as a test of piloting skill? If you want a build component, maybe have people suggest/build craft for the challenge, then after picking a plane, start the timed challenge with everyone flying the chosen plane.

I'm surprised we haven't seen more reusable testing. I can think of four launch profiles that might be compared: (stock only) Ballistic rocket ascent Ballistic jet ascent, switching to rockets (without much lift from wings) "Reasonable" spaceplane ascent, switching to rockets in atmosphere "Airhog" ascent, switching to rockets exoatmospheric with a high semimajor orbital axis There's room for interpretation between some of those categories. You can fly a wingless jet-rocket (profile 2) quite a long ways, picking up orbital speed at the 20-30km altitude and raising your semimajor axis before switching to rockets. The only difference between a reasonable spaceplane and an airhog is the number of intakes, and there's no way to firmly set a limit there. To me, the difference is that a normal spaceplane doesn't achieve BOTH a high apoapsis and high semimajor axis, where an airhog with 1 intake/ton can easily leave the atmosphere with a 40-50km semimajor axis, and be less than 50dv from orbit just using jets. Maybe also compare a vertical landing profile vs. a runway landing. (Or with and without parachutes?) I don't think the weight categories will end up mattering too much, since a 100 ton lifter can be made just by doubling all the parts in a 50 ton lifter, etc. Has anyone tried recovering a pure rocket in .24 yet? With only two results in this thread, there's not a ton to report yet: [table=width: 500, class: grid] [tr] [td][/td] [td]Runway[/td] [td]Rocket Landing[/td] [td]Parachute[/td] [/tr] [tr] [td]Ballistic rocket[/td] [td][/td] [td][/td] [td][/td] [/tr] [tr] [td]Ballistic jet[/td] [td]~300 √/t (Tsynique)[/td] [td][/td] [td][/td] [/tr] [tr] [td]Spaceplane[/td] [td][/td] [td][/td] [td][/td] [/tr] [tr] [td]Airhog Plane[/td] [td]~30 √/t (me)[/td] [td][/td] [td][/td] [/tr] [/table]

I tested an all stock (plus Mechjeb) recoverable spaceplane lifter that uses intake spam (without debug clipping) to lift a 22.9t payload for 27.8 √/ton cost. Note that this is NOT using FAR, which should probably have its own category. Stats: Launch Weight: 47.572t Payload Weight: 22.9t (Only a 48% payload fraction, airhog planes max out around 70%) Launch Cost minus payload: √204,734 Recovery Value: √204,097 Fuel: LiquidFuel: 920 starting / 139 landed Oxidizer: 440 starting / 376 landed. The flight profile was to climb steeply to ~20km altitude, level off and use jets to accelerate to an orbital speed of about 2,340km/s at ~35km altitude, then maintain that speed using intake spam until about 63km, altitude, so apoapsis was 63.5km and periapsis was 35.4km when switching to the LV1-N engines for the rest of the ascent. Getting into orbit only burned 30 oxidizer. The craft file is here. 27.8 √/ton is pretty cheap and this isn't even a particularly good airhog spaceplane. I didn't put any effort into tuning it for efficiency, such as the relatively low payload fraction. I hope that we'll see better cost performance from less cheesy approaches, but it looks like intake spam might also lead to better cost in addition to the biggest payload fractions.

FAR should be a separate category for both recovered and non-recovered lifters. You can't combine FAR and non-FAR recovered craft into the same category; they're playing with dramatically different physics. Do the weight categories matter that much? Higher weight classes should generally just be multiples of the best result from a lower weight class, right? For stock recovered vehicles, I've been wondering whether a rocket-style launch, normal spaceplane ascent or airhog-style spaceplane would offer the cheapest per ton result. I'm starting to suspect that intake abuse might give the best cost per ton in addition to the highest payload fraction.

How about judging by semi-major axis ( (apoapsis + periapsis)/2 ) instead of just apoapsis? On the plus side, that'd be a lot more useful for a space program than just a really big apoapsis. On the down side, it might be dominated by airhog designs.

That was one HECK of a precise landing. Yikes.

I ended up doing a LOT of RCS tug work for the Affordable Space Program challenge. (Especially stock.) Here are my observations. I'd really like to see what people think makes one good, more than just seeing craft designs. Things I Learned from Building a Rhombicubeoctohedral Space Station: General: Use Docking Port Alignment Indicator. It's wonderful. Skinny RCS Tugs are better at fitting into tight or oddly shaped spaces. The claw is far stronger than docking ports and the sanest choice for moving potentially high-mass parts Consider making the orbital maneuver tug and your RCS tug as two separate craft. You need a mod like RCS Build Aid to have any real way of knowing if your RCS is balanced. Lighting: Lights facing away from the tug to illuminate a target are MUCH LESS USEFUL than lights mounted on the target to illuminate itself. A light facing away from you has to be correctly aimed to usefully see what you're aiming at (like setting a docking port at a target.) In the dark this is really hard to do, because everything is dark until you are aimed exactly right. So instead, put lights on each craft aimed at illuminating itself! Then every light is always aimed, and everything is always lit. Effective Translation: [*]Translation with a big heavy weight on the end of your tug is challenging. Things that help are: Use MechJeb RCS Balancer. Stock capabilities are seriously deficient if you naively use a tug to align a large module. Make the tug long so the RCS thrusters can be spaced farther apart. The longer lever allows for more torque from RCS. Put most of the tug mass on the end farther away from the target. This helps keep the center of mass inside the bounds of the RCS ports. Put RCS thrusters on the actual station module! (Reasonable translation with stock RCS is basically impossible without this.)

Oh, true! In practice, the tasks of toggling air intakes, sequentially shutting down jets, and reducing throttle was tedious, annoying, and ate up all my action keys. That's once case where the Mechjeb utilities to manage intakes and clamp the throttle do a lot to increase the fun. That'd be extremely difficult to do by hand for large planes. Thanks!

Lots of good advice, especially the "just fly a plane around and have fun" advice. There're a LOT of ways that a SSTO spaceplane can be successful. Even stupid spaceplanes can work. I've built plenty. For me, there were three major areas of learning: - Getting it off the runway. (Other posters/links have already covered the major points: structural stability, enough thrust, enough lift, relative COL/COM position, and having the COM close to the rear wheels.) - Getting to space: I have a couple thoughts here that others haven't covered. - Landing: for me, this mostly consisted of flying slowly enough, which meant having enough wings. Getting to Space: (These are all stock game observations. Mods such as FAR do change everything.) Once you're in the air, there are a lot of flight patterns that will get you to space. All have some variation on using air-breathing engines until you run out of air, then use rockets. (Or switch RAPIERS) There's lots of good advice already referenced on how to switch your engines off to avoid spins, and how to manage air intakes. Early on, it's probably easiest to just bring up some extra rocket thrust and fuel, and just switch over to a familiar rocket ascent once the air runs out. This is straightforward, and doesn't require any fancy piloting. All the jets and wings need to do is get you up and out of the first 30k of thick atmosphere. If you start focusing on efficiency, then how you fly the plane matters more. Looking at Kasuha's flight reports taught me the most about how this works. Don't worry about things like this until WAY later, but for the sake of discussion, here's what I've learned. Typically there's an ascent at lower altitudes (<20k) where all you're doing is trying to climb out of the thick soupy lower atmosphere. Then you begin a careful balance of speed vs. altitude. Jet engine max thrust depends solely on speed. As you go faster, you have less thrust. This took me a while to really start understanding the implications. If you start going too fast, too low, your jet engine thrust might not be enough to continue an ascent, and your altitude will drop until you slow down enough to gain more thrust and give the ascent another try. If you're going too slow, too high, you run out of air for your intakes and you don't make good use of your jets. As far as I can tell, a really good jet ascent gains speed and altitude smoothly up to 25k-30k (more on that in a second) such that you can power up to as high a velocity as possible (2,300 m/s is great) during the region when you have enough atmosphere to feed your air intakes, but not so much that the drag stops your acceleration. This takes trial and error and varies a lot by plane. This is where the number of air intakes start mattering more than any other concern. ("air hogging" is the term, mostly used derisively, because it's unrealistic, but I'm just taking the game as it is, and leaving opinion out of it.) Once you're up high, the limiting factor is only how much air you can get to power the jets. If you have enough air intakes (up to 1 per ton!) then you can keep just enough thrust going from the jets to keep up with the tiny amount of atmospheric drag. All you're trying to do here is maintain that velocity you built up earlier, while slowly creeping up in altitude. This is the really weird thing. At 30K altitude, moving at 2,300 m/s won't put you in orbit. But if you can eke out just enough jet thrust to maintain that speed, then by the time you're pushing 40-50k altitude, the same orbital velocity will give you a apoapsis that's ABOVE the atmosphere! So just zip along on a tiny bit of jet thrust and you'll eventually skip up and out of the atmosphere on the far side of your orbit! (Which is probably crazy and unrealistic) This is how people get the ultra-efficient payload fractions, because this way, the jet does almost all the work. You might need only 20 m/s of rocket thrust in space to circularize your orbit in this case. Even if you can't get all the way out of the atmosphere on jet thrust alone, you still have done most of the work and you'll be going fast enough in thin enough atmosphere to be able to rely on the low-thrust, high-efficiency nuclear engines. The best I've done is having 70% of the launch mass be payload this way. Apologies, I suspect this is an advanced topic, but I haven't read much discussion of it. I wonder if I've summarized the idea well?

Anything from the Affordable Space Program Challenge entry. Composite SSTO's that delivered awkwardly-shaped payloads, combined in orbit to return together, then were refueled and loaded with new payloads without recovering.

I have multi-docked 5 at a time in 23.5. Really good alignment during construction and docking is required.

Sped Boat Mark I. All stock. Not useful for actually going anywhere. It didn't carry any payload. 319.4 m/s.

It could join Unladen Swallow in this challenge.

That's a huge mass to have in a payload bay. Pretty awesome, Hodo! How finicky were those ginormous loading gantries to use?

Sped Boat Mark I. All stock. Tops out around 320 m/s. Not useful for actually going anywhere. All the relevant design features are in the screenshot.

Even without intake/wing reuse, you can get better than 66% payload fraction on a spaceplane with ridiculous airhogging (~1 intake per ton). Intakes are pretty much the only factor in max lifting ability. But the time investment would be pretty extreme. That's a lot of time creeping up in altitude while maintaining velocity while keeping your thrust 150 mm/s over atmospheric drag. Part count might hurt more than tonnage. I've lifted 100 tons payload on a 150-ton spaceplane with 160 ram intakes, but that's one tiny strut per intake, so the air alone was 320 parts.

This hydrofoil topped out at 145 m/s while carrying a rover as cargo. It was dropped off by a VTOL for an amphibious challenge. Here's all I know, stolen from other, better, boats: - only have low drag aerodynamic parts underwater, the bounding boxes determine buoyancy. - don't have any part come into new contact with the water while running (I use physics-ignoring small hardpoints for the risers) I had to fiddle with thrust, CoM, and angle of small control surfaces under the water before it'd stay level at speed. It doesn't corner well. It flies very poorly. Nosing up even slightly will launch 2000m in the air to its death. 200 m/s should be very doable with a little more thrust/drag.