Ha. Spaceplane... or Submarine.

[cephalo]

So here was the plan: I have about 17 Kerbals, lots of tourists among them, on a return mission from Duna. Now that I just obtained all the space plane parts on the tech tree, I thought it would be cool to pick them out of Kerbin orbit with a new SSTO spaceliner and land them at the KSC.

I've designed quite a few successful rockets, some with outrageous payloads, so I thought I could design a simple spaceplane with the 16 Kerbal crew cabin.

I can't.

Wow, either my planes can't stay out of the water, or my engine power rips the plane apart. Forget about orbit, I'm not breaking 5000 meters with my Mk3 plane. All I need is a pilot seat, the Mk3 16 Kerbal cabin, and a docking port Sr. with some RCS control for docking with the mother ship. I can make small planes that fly, but Mk3 parts seem very large compared with the available wing parts.

Where's a really good guide for making Space planes? KER is great for making rockets, does it have a mode just for space planes so you know how much lift you need etc? How do you make strong, large wings that don't fall apart?

Edited July 9, 2015 by cephalo

[Sharpy]

Most guides for space planes won't help you with MK3 cabin. Small, lean, MK2 mostly. It's very hard to find an MK3 SSTO design around here.

Here's mine, feel free to steal the ideas. Since it had outrageously much fuel after the orbit, replacing some fuel tanks with the cabin should do the trick.

Of course landing it is a different matter. This one is SSTO, but not SSFO - it's intended for powered vertical landing (with landing struts extending from the four orange tanks backwards) and... it's bound to lose the wings in the process as they would extend farther back than the landing legs.

[cephalo]

Most guides for space planes won't help you with MK3 cabin. Small, lean, MK2 mostly. It's very hard to find an MK3 SSTO design around here.

Here's mine, feel free to steal the ideas. Since it had outrageously much fuel after the orbit, replacing some fuel tanks with the cabin should do the trick.

Of course landing it is a different matter. This one is SSTO, but not SSFO - it's intended for powered vertical landing (with landing struts extending from the four orange tanks backwards) and... it's bound to lose the wings in the process as they would extend farther back than the landing legs.

Wow that's a lot of engines. I have a question about that large wing surface. How did you get them not to fall apart? I used little pieces of struts on mine, but I don't know what that did to my flight characteristics.

[Sharpy]

Struts, and struts again. There's a strut to the hull roughly every two surfaces apart (meaning each unstrutted surface is adjacent to at least one strutted one.

And yep, you could do with a bit less Rapiers, but you need a plenty of whiplashes, a decent intermediate engine to bring you to orbital speed (Rhino here) and then Nukes for actual space flight.

You'd likely be better off giving up a space plane and going with vertical launch SSTOs.

[cephalo]

I got something to orbit! I needed a ton of wing surface, yet stock wings are too fragile to just extend out, so I solved the problem World War I style and made a biplane! It flies pretty well, and I got it to orbit with almost no fuel left, just enough to retro burn back into the atmosphere and maybe a powered landing somewhere that I didn't actually get to try.

I might be able to fly it better to save fuel, but it's tricky. Climbing becomes difficult around 11,000 meters, and not only am I impatient, but rapiers aren't very fuel efficient. That's when I've been triggering the Rhino to get to orbit, while there's still a big push through the atmo left to go.

I had some problems on the way down. Once I got into the lower atmosphere, my nose pitched up suddenly and snapped off the wings at mach 2. I think it's because I had the airbrakes deployed. I don't really know much about them except that they might be needed, but they are also control surfaces? Since all my airbrakes are on top of the plane, I think they pitched the nose up. It's either that, or its possible that my plane is tail heavy with no fuel. Not sure.

[Sharpy]

Nope. You used up enough fuel from the front tanks that your center of mass shifted behind your center fo lift. You need to move most of remaining fuel to front tanks for reentry if you have any - if you don't, drain your spaceplane of all fuel in SPH and move the wings so that the center of lift is at least flush with center of mass, if not a little behind. And of course you'll run into the launch problems of nosedive as you put all the fuel back in and the center of mass moves far forward...

Well, congrats, you've got half of the problem covered: you've made an SSTO. Now make it work as an SSFO as well for a complete success.

[Mic_n]

Well, congrats, you've got half of the problem covered: you've made an SSTO. Now make it work as an SSFO as well for a complete success.

Just for reference, the "From" is pretty much assumed. An SSTO that can't de-orbit and land is generally referred to as a "failure".

Edited July 7, 2015 by Mic_n

[cephalo]

Just for reference, the "From" is pretty much assumed. An SSTO that can't de-orbit and land is generally referred to as a "failure".

It's only a failure if you give up! I have two unrealized goals, how to get to orbit with enough fuel to do something other than reentry, and also how to fly and land something that looks like a collision between the space shuttle and a Wright Flyer. I'll get it eventually.

Right now I think I might need more engines. I'm only doing about 350 ms when I turn on the Rhino. That's kinda low isn't it?

[ajburges]

The thing you need to understand about Mk3 planes is the volume vs surface area scaling issue. Mk3 planes have awesome payload for their drag profile, but they need equally impressive aerodynamic surfaces for similar flight. Also, impressive payloads need impressive engines for the requisite thrust.

Airbrakes:

These are for two things commonly. Slowing down and increasing stability. They make reentry more controllable by varying your drag profile. If you don't use Kobra manuvers, they are needed to avoid overheating. If you locate them near the tail, they will also add flight stability in reentry. Be warned, 4 air brakes is not enough to control the speed of an empty Mk3 plane in a dive. Spam these bad boys if you use them.

Rocket power in atmo:

My standard flight profile uses any available dedicated anerobic engines to aid piercing the sound barrier. Your SSTO is typically heaviest on accent. Waffling at transonic speeds wastes more fuel mass then 6 seconds of rocket burn. Using discrete rocket engines also saves you the extra jet mass. Jets are dead weight above 20 km.

Engines:

Due to the payload, you need hefty engines. 1.25 m engines do not scale well to 3 m fuselages. I find a KER read of .6-.7 "vaccum" TWR is good for a Whiplash power plant, and a TWR of .6-.7 at takeoff weight for the anerobic mode should be sufficient after your jets do their thing.

Struts:

Struts (and fuel lines) are a tricky subject for space planes. They improve structural integrity (and fuel management), but are hell on your aerodynamics. I tend to attach engine pods to my wings and use a pair of struts per wing as tension cables/bars between the body and a pod. Keeps the wings just stable enough for flight and landing. I also manually transfer fuel/oxidizer from the pods to the fuselage.

My go-to Mk3 SSTO uses 8 Wiplashes and a mainsail. It can get a touch more than a freshly emptied orange tank to orbit. (It launches my research stations, automated tugs, and LKO fuel depots) It fits my current needs as I gather the science for large wings, RAPIERs, and/or Mammoth/Rino engines. I could likely replace 2 cargo bay segments with cabins and have more dV in orbit. The third cargo bay is required to store the 2 reaction wheels, solar panels, and batteries it needs. I also might eventually put more monoprop and a KER computer there. 100 units isn't much for a 60-100 t craft.

Edited July 7, 2015 by ajburges

[Mic_n]

It's only a failure if you give up! I have two unrealized goals, how to get to orbit with enough fuel to do something other than reentry, and also how to fly and land something that looks like a collision between the space shuttle and a Wright Flyer. I'll get it eventually.

Right now I think I might need more engines. I'm only doing about 350 ms when I turn on the Rhino. That's kinda low isn't it?

Wait, really? Yes, that's exceedingly broken.

My advice: start small. Its much easier to learn the concepts that way.

But basically... 350m/s is basically what people here tend to refer to as the transonic hump. The way the jet engines work it's a bit of a 'dip'.. once you get up past that the engines really start to put out a lot more power as they go faster, and you'll find your speed rapidly climbing up to 1000m/s and beyond, which is where you want to be before going to rockets. If you're using rapiers, that hump is quite prononunced since they really don't produce much thrust at all at low speeds. One thing you can to is manually flip some over to 'rocket mode' to get your speed up before switching back to airbreathing once you've pushed through that slump.

The idea is that you get as much of your orbital velocity as possible from your jet engines, which are vastly more efficient than rockets. You can also use some lift from the atmosphere to help turn that orbit and hence pull your Apoapsis up as well.

[cephalo]

Oh yeah. Thanks for the tip regarding pushing through the sound barrier with my rocket. That made a huge difference. After I shut it off, my rapiers were able to get up to 1000ms no problem. I then waited until they switched modes automatically to fire the Rhino again, and this time I made it into orbit with much more fuel, maybe enough for a rendezvous and de-orbit and landing. I could probably do better with a bit more practice.

I added a front canard to deal with the nose heaviness on takeoff, resulting from adjusting the wings to deal with tail heaviness on re-entry with little fuel. Now I get to try landing! Ugh. That's a whole nother can o worms. Hopefully it's just a matter of learning and not a matter of another design feedback loop.

[ajburges]

But basically... 350m/s is basically what people here tend to refer to as the transonic hump. The way the jet engines work it's a bit of a 'dip'.. once you get up past that the engines really start to put out a lot more power as they go faster, and you'll find your speed rapidly climbing up to 1000m/s and beyond, which is where you want to be before going to rockets. If you're using rapiers, that hump is quite prononunced since they really don't produce much thrust at all at low speeds. One thing you can to is manually flip some over to 'rocket mode' to get your speed up before switching back to airbreathing once you've pushed through that slump.

Actually, it's not a case that the jets dip so much as the drag spikes and jets don't get power boat until you are through most of that spike.

A for your plane, I can tell you now that's over designed. A crew bay, cabin and support is only a 12 t payload (conservatively). I can SSTO that with 4 Wiplashes and a Mainsail (possibly just a Skipper). A Rino and 12 RAPIERs has to be too much thrust for that mass. Reduce the weight (and fuel) of your anerobic mode, and you reduce all of your flight needs. You also have far too much wing. You only need enough wing to maintain a 10° AoA at 200 m/s. You won't be needing low speed lift until you lose most your weight. Anything more is a lot of extra drag for little gain.

I don't have experience with RAPIERs in 1.x yet, so I can't give good design tips for those (other than they need a dive or booster to go super sonic in a well designed plane), but try the following:

Make a Mk3 rocket sausage with cockpit, cabin, cargo bay (with solar panels, batteries, and reaction wheel), Mainsail, and fuel for 2 km/s dV (you loose 400 m/s adding jets and wings).

Aseble your tail: Delta wings or type B or C wings with control surfaces. Place as far back as you can.

Make your wing sub-assembly: structural A, long LF tank, bicouplers with jet engines/intakes, rear landing gear, and a Delta wing on the pod. Add control surfaces and structural wings to taste. (Especially in front of the structural A wing)

Shift the wings for CoM/CoL balance. Use two struts each to hold the pod relative to the body (ensure you make a triangle with each strut. Sides will be the wing, strut, and fuselage wall). Tweak out all the oxidizer from either the front or rear bicouplers.

Add the low mass decorations: gear, airbrakes, RCS.

You may need a second pair of engine pods (atach them to the fuselage with structural A wings and attach your wing assembly to those only anchor the outer pod with struts), but that is the only way this design would be deficient. This design with 4 pods is what I used for my previously mentioned cargo plane which takes detachable payloads larger than 16 t. On second thought, start with 4 engine pods (that's 8 Wiplashes). It may be less efficient, but my total payload is only above 22 t there.

[cephalo]

I landed it! Not perfect, almost as ugly as the plane itself, but nothing broken...

Ok, so its overdesigned. As my goals really are quite modest, it shouldn't take a monstrosity to get a 16 crew cabin to orbit for docking. I did learn a lot though.

[Mic_n]

Myself, I'd suggest a cluster of 4 whiplashes on the end of your fuselage (your payload + a mix of mostly LF + some O), and a couple of LFO pods in the wings carrying aerospikes and intakes, maybe with a couple of extra whiplashes in there if you need 'em.

Sometimes when you're designing you have a (perfectly natural) tendency to get into a bit of a loop... you don't have enough thrust, so you add more engines... now you run out of fuel too soon, so you add more fuel.. now you need some more engine to push it along.. maybe that's added some extra drag as well.. Often it's easier to just scrap it and start again from scratch, because removing stuff is actually a lot harder.

[cephalo]

One thing that surprised me with this design was the feasibility of a biplane in KSP. Aside from whatever extra lift you get, it actually does make these multi-part wings a great deal stronger, with not much downside. Before I added the top deck, a bit too much SAS oscillation would shake those bottom wings right off. With the top deck, everything is rock solid.

[SAI Peregrinus]

I cannot recommend B9 Procedural Wings enough. Though it does require FAR, that's not actually much harder than stock aero these days (I think it's a bit easier due to the analysis tools and voxelization). It eliminates the need for many-part wings, thereby eliminating the need for (very draggy) struts, and reducing part count. Your game will run faster, you'll have the FAR tools to tell you if the plane can work before you even leave the SPH, and it will look better.

[cephalo]

So I am redesigning again and came up with something much simpler, basically the sausage with two wings and some jets and a central rocket. Again though, I am making it to orbit on a razor thin, uncomfortable margin, so my plane wants to start growing again. I have a couple more topics to discuss.

I've tried the skipper, mainsail, and rhino, and the rhino just really likes to get the job done consistently, even though it needs a bit more plane than the others. It's also a suitable vacuum engine with better efficiency than the other two for any theoretical orbital maneuvers that might be done.

I have also tried using whiplashes instead of rapiers, and I'm finding a big trade-off here. Whiplashes are much better at low speeds and even though they lack the top speed of the rapiers, they get to the proper altitude for supersonic flight quickly and efficiently. They literally use 1/5th of the fuel to reach their top speed at altitude than the rapiers. By the time I'm at 10km and accelerating to supersonic, half of my liquid fuel is gone as the advantage of the rapiers is just coming into play. Either I need more rapiers than whiplashes, or I need to use rockets on the initial ascent for a while, or I just don't know how to fly with rapiers. How do you deal with the low initial thrust of the rapiers?

[Sharpy]

Well, I deal with it simply: I don't use Rapiers It seems for every headache they remove, they create two new ones.

BTW, they are there supposedly to remove the need for Rhino and the likes. The SSTO setup is either Rapier-Nuke, or Whiplash-[intermediate engine]-Nuke. Sometimes Whiplash-[intermediate engine] only, for things that don't plan to fly far, e.g. picking up stranded in LKO.

(well, save for Rocket SSTOs - these are [intermediate engine] only - usually Mainsail.)

Edited July 9, 2015 by Sharpy

[cephalo]

Yeah, I don't see rapiers being useful without some boost down low. At lunchtime I tried some early rhino assist, then level supersonic flight up to speed, then another final rhino assist, and this almost made it, and by that I mean it failed, but still. A little less early rhino might actually work.

I also noticed that your angle of attack affects your air intake quite a bit. If you change it too quickly, your rapiers will switch modes even below 20k. Since my plane generally flies a bit above prograde, I wonder if I could benefit from angling my intakes down a notch.

One more quick question, let's say you have one intermediate engine and one nuke, how do you provide symmetrical thrust?

Edited July 9, 2015 by cephalo

[ajburges]

Running RAPIERs without an additional booster is tricky. A quick jump to closed cycle leaves you to spool up the jets, and creeping past mach 2 in a dive wastes a lot of fuel. If you can do either of those effectively in RAPIERs alone, you have too much engine for every other aspect of the flight.

A reasonably popular trick in .90 was was a 50/50 split of RAPIERs and Whiplashes. It was good in .90 because it limited part counts, but didn't give too yuck anaerobic thrust. Today, there Whiplashes are there to push you into the RAPIER envelope.

As dedicated vaccum rocket engines go, you can't beat the Rino. Hybrid SSTO space planes can safely ignore Isp ASL. The lowest you could want to use rockets is above 7 km at which point you are 70% to your vacuum Isp. In the 2.5 m size, you compromise. You either take the Skipper for better Isp but less thrust, or the typically better Mainsail. While the Rino does have less than half the thrust of a Mammoth, only a tribute to Wackjob would need more than 2 MN of dedicated anaerobic thrust. If you really need more, you have RAPIERs. At the same time, you are within 3% of the most efficient chemical engines! That bell doesn't really look streamlined though. Might want to put it in a fairing.

If you want one intermediate stage and one nuke, consider a engine section that can be reconfigured in orbit with docking clamps. Once in orbit swap them out as needed. You could have the nuke carried as payload on ascent, or have a nuclear tug/engine on standby in orbit. You could even leave the intermediate stage in LKO to be picked up before reentry.

[Sharpy]

One more quick question, let's say you have one intermediate engine and one nuke, how do you provide symmetrical thrust?

Clipping.

http://forum.kerbalspaceprogram.com/threads/128057-Tiny-SSTO-plane?highlight=tiny+SSTO

This one clips together Whiplash, Nuke and a LF fuselage, through use of one cubic octagonal strut placed directly at the location of the node of the fuselage (but not *at* the node, just very precise surface attachment right where the node is.)

It may be considered cheating by some, and in certain cases, when done wrong, it may introduce explosive disassembly - generally a somewhat risky technique, but hey, if it gets the job done?

(also, cubic octagonal strut is really terrible from thermal standpoint - tiny thermal capacity, poor conduction - you may consider either adding some radiators to it or using some other part for the clipping trick.)

[Mic_n]

yup.. best way is to cheat, like that (which I've been known to do myself on occasion).. alt-F12 "allow part clipping in editor".

Otherwise, you don't do it symmetrically, and just work with the asymmetry.. I had a spaceplane with an "over and under" turbojet and LVT-45. By rights you can very carefully surface mount your engines and angle them so that they still push vaguely through your centre of gravity, that's one way. You can also make your craft quite long to minimise how much torque is applied, and add plenty of control authority - particularly SAS reaction wheels - to counteract it.

[cephalo]

I finally got something that is reasonable and comfortable. I'm ready for the real mission. I'll post pics this craft docking with the mother ship when I get to it.

It uses 10 rapiers and a rhino. My flight profile goes like this:

Take off at full power will all engines including the rhino.

Make a shallow climb, trying to trigger the rapier power loop right away.

Turn off the rhino once the rapiers can get supersonic. It's easy to get wrong, but you can turn the rhino back on if needed.

With this craft and a 5 - 10 percent AoA, the climb is fast enough, and the acceleration slow enough not to cause overheating.

Hit the rhino again at 26km or when the rapiers start slowing down.

I had enough fuel to dock with my Kerbin spacestation and land at KSC. On the runway this time! (After some quicksave re-loading of course. BTW, the KSC runway is outlined with an explosive minefield for some reason. Don't touch the edge!)

[Sharpy]

The runway is outlined with slopes, that are lethal at that speed. And has two lights sticking out at the end, so you'd better not start through the edge of the runway.

Personally, I always land next to it. Lots and lots of smooth grass.

[Mic_n]

It uses 10 rapiers and a rhino.

To each his own, but that's still massively overdoing it.

Ok, here's an example I just whipped up:

see Fat B-stard on KerbalX.com

It's a struggle getting up there, but it'll do it. 4 rapiers and a whiplash and it'll get that mk3 passenger bay to an 80x80km orbit with plenty of fuel to spare. CoM/CoL is pretty bad here, it really doesn't like lifting its nose much, but.. take off and nose up at about 25 degrees to altitude.. up to about 15km or so and you'll probably find yourself starting to drop as the atmosphere thins out. Go with it, nose down to about -5 and power through 400m/s before slowly bringing the nose back up to that 25 or so again with your speed climbing. Hit 1 to cut the whiplash and continue the climb on rapiers, then as your speed starts dropping hit 2 to flip them to rocket mode and close your intakes. Hold that 25 degrees to push your Ap up and away, it should be a few degrees above prograde at that point. Once you've got your Ap to 80km, cut your engines and drift to then circularise at the 80km mark. When you're ready to come back down, make your re-entry burn to bring your Ap down to 40-45km then turn back to prograde and again 30 odd degrees nose-up. Hit 3 to shutdown the rapiers, 2 to re-open the intakes, and 1 to re-activate your whiplash for a powered descent. airbrakes are there and should slow you reasonably well.

I've got a probe core and a few batteries clipped in up the front there, you'll probably see them getting warm on re-entry.. there's also a pair of spotlights up the back to help out if you need to land in the dark.

However, if it's just a matter of getting actual Kerbals to/from space.... command chairs in a mk2 cargo bay would make a great Budget Spacelines Option.

...and with a little fiddling there, the long bay can nicely fit three rows of 4 command seats on octo struts (though you could place them all individually if you want and forget the struts for extra weight saving). A short LF section in front, with short Mk1 adapters around the whole thing, a ram air on the front and rapier on the back.. couple of sleek wings on the sides and a trio of tail fins on the rapier itself. Clip a command pod and batteries in the fuel tanks, slap some solar panels on it (rapier has no alternator) and landing gear, and voila: superlight 12-man space transport that again can hit 80x80 and land again. That one also needs that power dive to push through the sound barrier with the rapiers... but it'll do it.

No craft file for that though Feel free to put that one together yourself.

Edited July 11, 2015 by Mic_n