Trouble finding spaceplane SSTO guides

[Jarin]

Hey, I've been looking through here and the spacecraft exchange, but I'm having trouble finding a general guide on SSTO plane assembly. I can build your basic passenger ferry easily enough, but I can't for the life of me scale up to craft delivering larger payloads (long Mk2, Mk3, etc). Would especially appreciate guide or examples for slightly lower-tech planes, for careers that haven't hit the top for the RAPIER yet (pre-whiplash would be even cooler).

Edited June 2, 2016 by Jarin

[bewing]

The MK2 and MK3 parts have far too much drag to SSTO with panthers. A Whiplash or two can get a small MK2 SSTO through the airbreathing part of its ascent.

An MK3 design needs a couple Rapiers, period. Maybe a couple Whiplashes, too.

Don't use more jet engines that you absolutely need to.

After you get past the basic aerodynamic stuff, it just comes down to a matter of minimizing weight, minimizing drag, adding enough wing, adding moar fuel, and adding moar rockets.

Don't use struts. Don't use fuel ducts. Do manual fuel transfers when needed.

Canard wings are more efficient than negative control surfaces ... you would be wise to minimize the number of negative control surfaces, because they add weight and drag. (<- this opinion is not universally shared.)

 

[Jarin]

1. Okay, Panther SSTO limited to MK1 parts then? Or is it just not reasonable to build Panther SSTOs for actual use (as opposed to "hey, I managed something on a razor margin because I could" like mk0 SSTO)? 

2. "absolutely need to" - guidelines here? A side-note... I've never once pulled off the "dive for speed then re-ascend" maneuver that I've seen mentioned.

3. "enough wing" - I can never figure out what wing area I need for a given mass of spaceplane. Any rules of thumb? Also any pointers on expanding wing space past a single Big-S? It doesn't have an obvious way to tile multiple like the regular wing parts do...

4. Struts, more an addendum to 3, really. Larger planes need more wing. How do you add more wing without it wobbling around if you're not strutting the wings? 

5. Negative control surfaces? You mean elevators behind the center of gravity, right? Why do control surfaces up front add more drag than control surfaces in back?

[AeroGav]

21 minutes ago, Jarin said:

 

3. "enough wing" - I can never figure out what wing area I need for a given mass of spaceplane. Any rules of thumb? Also any pointers on expanding wing space past a single Big-S? It doesn't have an obvious way to tile multiple like the regular wing parts do...

4. Struts, more an addendum to 3, really. Larger planes need more wing. How do you add more wing without it wobbling around if you're not strutting the wings? 

Enough wing - open up the aero data GUI during an ascent, alt f12 then physics tab to show the option.  Watch your AoA and lift drag ratio.  Tune your AoA for best lift:drag ratio.    If you find yourself having to pitch up to AoA that give significantly worse lift:drag than optimal to avoid getting too fast too low and overheating (or exceeding the best speed of your airbreathing engines while still deep in atmosphere) you need more wing.

re: tiling big S wings

It can be done, though it might be better to install B9 Aerospace procedural wings since it gives freedom to pursue the aesthetic you want. It's not your fault the game didn't include stuff that's right size.

Struts - try to avoid attaching wings to wings or too deep in the "tree structure", try to keep them higher up the node.   eg. attach each wing to it's own fuselage piece and offset it.  If you need to attach wings to wings for outer sections, putting heavy masses like engine pods on the join can stop this joint flexing up when lift is created.

Again though procedural wings are your friend.   I also recommend GPOSpeedFuelpump,  since manual transfers become unmanageable in larger craft. Eg. you can transfer fuel from multiple wing tanks to a single rocket engine feeder tank in the fuselage, but if you have multiple pod mounted rocket engines you can't run simultaneous transfers 

[bewing]

1. Yes, panther SSTO = MK1. But MK1 has no cargo, so they are passenger SSTOs only.

2. An empty MK3 Cargo SSTO (full of fuel, but with no cargo) will mass about 75 tons. To push that through the air, you will need 250kN at takeoff, 400kN for acceleration and atmospheric climb, tapering to 120kN at 40km+ altitude. And those engines have to get you to 1200 m/s in airbreathing mode, or you won't make it. 15 tons of those 75 are budgeted for engines. You can go a little over budget if it gets you a big bonus. If you want to really carry some stuff, though, then double the mass and quantity of everything.

Re the sidenote: to make that maneuver work, you have to really know your engine. All those jets have a high-thrust zone that you can see when you have them really pushing their limits. An afterburning panther at 410m/s is just getting started. A Whiplash or Rapier at 370 m/s is just about to start really ticking. Below those speeds, the engines feel pathetic. If you are stuck just below a magic speed, then climb, push the nose over into a very shallow dive (maybe 1 degree down) and let the plane do its thing. If you cross the magic number, then the engine will come to life. And you need to leave it be, to get really into the zone. If you pull the nose up manually, you will lose it again (lots of drag = -50 m/s). The nose will rise by itself over time, and you have to let it happen.

3. Yeah, that's a tricky one. If your nose is pointing way above prograde, that is draggy and bad. If you see your nose doing that during ascent, play around to see if you can add a lifting surface somewhere. The BigS is a problem, yes. All you really have available there (if you care about looks) is the BigS strakes. I have seen people clip multiple BigS wings exactly into each other. You can put rectangular wing parts against the fuselage and attach the BigS wing to the outside of that. A lot of workarounds to adding wing involve clipping. But basically, once you've built an almost-functioning SSTO -- just try adding a little more wing to it and see what happens. If it gets you to orbit with more fuel, then try adding a little more.

4. No solution is perfect.  A little wobbling may be the price you have to pay to get your payload to orbit. And the struts and fuel ducts issue is a bug in 1.1.2 that will hopefully be fixed very soon. Along with the bug that (unless you use Claw's patch) SSTOs currently have zero body lift.

5. Elevons work by pushing the back end of your plane down to make you go up. Your lifting surfaces have to work even harder to counteract that extra downward force, beyond just countering gravity. A canard makes you go up by lifting the front end = more efficiency.

[Blaarkies]

Follow that guide. It looks long, but it is a fun read. Experiment with each page and you will master atmospheric flight in no time.

If you still have trouble with the space part, continue on with this:

The thing is, pure stock spaceplanes are somewhat limited by the part type to be small planes...going bigger than MK2 makes things get complicated real quickly. B9 parts+B9 proc wings makes this a lot more fun though

[Jarin]

First guide, I have the basics of aircraft down pretty well. Not having much trouble with that, just the specifics of engineering larger craft without losing too much efficiency to reach orbit. Second guide is... badly out of date for current advice, I'm afraid. Case in point: "MOAR INTAKES" and the "soup". Neither of which has really been a thing since pre-1.0. =/

 

[mk1980]

i don't have a guide, but here's a few images of SSTO planes I made. most of them are from an actual career

note that some of the designs are a bit overkill. i only learned in the process that you don't really need that many air intakes, so some of them could have been built a bit more efficient easily (simply by using less intakes)

panther based SSTOs are somewhat viable (even the mk3 versions), but you actually get similar results by making them pure rocket planes with no air breathers (as seen in one of the images), so the usefulness of the panther is debatable. in essence, if you want to make a panther SSTO, it's mostly a rocket anyway since the panthers only get you to about mach 2.5 at 10km or something before you have to switch over.

the whiplash is an entirely different thing. they are very capable of doing most of the typical SSTO plane stuff (see the "orange tank" lifter on the last image)

 

[Jarin]

10 hours ago, mk1980 said:

(see the "orange tank" lifter on the last image)

Awesome, any chance I could see a breakout of the wing structure on the fuel tanker and that last concept? Just to see what's under the Mk2 engine nacelles.

[mk1980]

13 minutes ago, Jarin said:

Awesome, any chance I could see a breakout of the wing structure on the fuel tanker and that last concept? Just to see what's under the Mk2 engine nacelles.

removed the engines and took screenshots in the SPH editor

the big bird uses mostly wing connector type B (i think) and large delta wings / structural wings type B on the outside. the small angled pieces on the tips of the wings are small delta wings. the fins at the back are also structural wings type B and the wings on the front are type A (i think), partially clipped into the fuselage (only for aesthetics)

control surfaces are elevons type 3 (i think)

 

the low tech fuel plane uses 2 wing connectors and a delta wing and the angled tips are structural wings type A. i think there's also a small wing segment (wing connector D or something?) partially clipped into the fuselage at the back. the canards are 2 pairs of standard canards and the elevons are type 1 i think (probably should have used bigger ones)

 

not sure if the angled wing tips make much sense. i guess they would work better if they were perpendicular (?)

 

overall, my designs aren't as elegant or efficient as some of the other examples that were shown. basically, i substitute knowledge about aerodynamics with brute force and ignorance  also, i tend to pack more engines then the bare minimum necessary. it's nice to lift an orange tank to orbit with just 4 rapiers or something, but spending 20 minutes on the ascent gets old really fast. i try to aim for a thrust: weight ratio of ~0.6-0.7 at the runway, so the planes can actually break the sound barrier without a dive and can continously climb at a 15-20° degree angle and reach space almost as fast (in real time) as a rocket. savin a few minutes of playtime every flight is more valuable than saving a few hundred funds worth of fuel (to me, anyway)

[AeroGav]

Something that's worth discussing is the whole cargo SSTO concept.   Now, if you use your space plane to launch a probe, and the probe explores space, then you aren't dragging air breathing engines, landing gear, wings etc. any higher in Kerbin's gravity well than low orbit, so that's a saving.

But , mk3 cargo bays are incredibly heavy,  very draggy, as are mk3 nose and tail sections/adapters/cockpits.  These days I just prefer to have the space plane itself explore the Kerbol system. It already has wings (useful on laythe and duna), NERV engines, strong landing legs, and large capacity for storing liquid fuel.      Multiple Big S wing sections can enable a design to be both efficient at high altitude flight,  landing at a safe speed on Duna, and have plenty of LF capacity in its wings.

Refuel your spaceplane after reaching kerbin orbit either by docking another spaceplane/fueller rocket with it,  or fly it up to your Minmus In Situ Resource Mining station.  If you are using IRSU,  it doesn't really matter that your space plane is carrying uneccessary mass on an interplanetary journey, since you can mine over 2000lf per day with one small drill and converter.

The only time you need rockets is for Tylo (a lightweight, chemical rocket fuelled shuttle required with high TWR) and Moho (astronomical delta v requirement means very pared back ship).  In both cases these are likely to be minmalist vessels that would fit in a mk2 cargo bay however. 

Duna lander 

 

[Jarin]

3 hours ago, mk1980 said:

basically, i substitute knowledge about aerodynamics with brute force and ignorance

Sounds like my kind of design philosophy. What's your part-tree look like for those wings? Is that basically just 4 separate wings that can wobble up and down independently, or did you take the drag hit to strut them together?

[mk1980]

can't look it up at the moment (no KSP on this computer), but IIRC the inner wing connectors are all attached to the main fuselage (mk3 cargo bay) and the 2nd column is attached to the inner column, the mk2 engine nacelle assemblies are attached to the 3rd "row" (the one that ends in the angled small delta wing) and there is 1 strut from between the outer an inner engine nacelle and 1 strut between the inner nacelle and the cargo bay. there may also be 1 strut between the big delta wing segment and the 2nd wing connector of the "main row". i think i added one because the deltawing would flex upwards a bit and so there is a gap in the wing during flight that just looked ugly.

overall i tried to keep the strut count low. i think the whole plane has like 8 struts (2 of which are inside the cargo bay so the cargo fuel tank doesn't wobble). seemed like an acceptable tradeoff. i suppose it could be avoided with mods (procedural wings or reinforcement mods or something) but i generally (try to) solve my design problems with stock parts if possible.

[Jarin]

Speaking of mods, are there any out there that fix/adjust the crazy drag that struts and fuel lines have right now?

[FullMetalMachinist]

2 hours ago, Jarin said:

Speaking of mods, are there any out there that fix/adjust the crazy drag that struts and fuel lines have right now?

None that I know of, but 1.1.3 is due in the next week or two. I've got my fingers crossed that that's one of the fixes. 

[KerikBalm]

On 6/2/2016 at 3:30 AM, Jarin said:

Would especially appreciate guide or examples for slightly lower-tech planes, for careers that haven't hit the top for the RAPIER yet (pre-whiplash would be even cooler).

Pre-Whiplash is not good at all for carrying cargo.

With whiplashes, for larger planes, I recommend a KR-2L. as its got great vacuum Isp, and great TWR.

If you want to get >35% payload fraction, it gets a bit harder... but for 25-35% payload fraction with turboramjets or rapiers, its really just enough engine and wing to be able to go supersonic in a shallow climb... which should get you >700 m/s by 10km, and well over 1,000 by 20 km (ideally, over 1,150 m/s for turboramjets, and over 1,300 m/s for rapiers). Carry about 1,200- 1,500 dV for your rockets, and you should be good.

Also, minimize drag as much as possible, avoid struts, fuel lines, blunt ends (you're familiar with the "rapier-spike" idea?), etc.

[Jarin]

On 6/3/2016 at 2:56 PM, KerikBalm said:

(you're familiar with the "rapier-spike" idea?)

I... think so? Is that the reverse nosecone slapped onto the back of the rapier? Does that actually help drag in-game (real physics I can see how it would, but KSP)? Seems like it would just be adding weight.

[KerikBalm]

It does reduce drag in KSP. All uncapped nodes are draggy, putting a cone on it does reduce drag, and the mass penalty for the small cones is quite small.

Keep in mind though, that without using the part offset tool, or a gimbal limit, the nosecoen is likely to get cooked by the engine exhaust.

[Mk3 Maniac]

Here is a very small start. SCALE UP, multiply the fuel and engines from how much nerves you have, (I tried it), but what do I know I am a MK3 maniac. (and can't describe stuff).

[Wanderfound]

On 6/2/2016 at 1:49 PM, Jarin said:

1. Okay, Panther SSTO limited to MK1 parts then? Or is it just not reasonable to build Panther SSTOs for actual use (as opposed to "hey, I managed something on a razor margin because I could" like mk0 SSTO)? 

2. "absolutely need to" - guidelines here? A side-note... I've never once pulled off the "dive for speed then re-ascend" maneuver that I've seen mentioned.

3. "enough wing" - I can never figure out what wing area I need for a given mass of spaceplane. Any rules of thumb? Also any pointers on expanding wing space past a single Big-S? It doesn't have an obvious way to tile multiple like the regular wing parts do...

4. Struts, more an addendum to 3, really. Larger planes need more wing. How do you add more wing without it wobbling around if you're not strutting the wings? 

5. Negative control surfaces? You mean elevators behind the center of gravity, right? Why do control surfaces up front add more drag than control surfaces in back?

1) Mk2 Panthers are possible, but their payload capacity is lousy. Usually not worth the hassle. OTOH, Mk1 Panther SSTOs aren't good for much except crew transfer either.

2) Usually, if you can get off the runway before the end, you've got enough engine. Low TWR ships (e.g. a 1 RAPIER/2 nuke Mk2) might need to level off while breaking through transonic. Higher TWR ships can keep climbing the whole way.

3) See: 

Certain styles of build benefit from large wings, but they aren't compulsory. If you can get your takeoff speed down to 100m/s or below, you're probably okay. Building with wing incidence helps to get maximum lift with minimum fuselage drag; important if you want decent fuel efficiency from a larger ship.

To demonstrate what you can do with just a few wing pieces:

Able to lift up to 60 ton

 

3,700m/s unrefuelled; enough for a Mun return.

 

Eeloo...via Minmus, Duna, Dres, Laythe and Pol.

 

Full science, scanning and ISRU kit aboard, 4km/s in the tanks.

/

 

4) Usually, you shouldn't need struts. You certainly shouldn't need visible struts. Keep your wings as simple as you can; use fewer, bigger parts. Autostrut is available if needed, but usually isn't.

To avoid needing struts, you need to think about the tree structure of your craft. If you're putting a long stack of fuel tanks on the side of the fuselage, don't hang them off the nose; support the centre of the stack. Etc. If you do have to strut, you can usually arrange things so that it's just a single pair that can be hidden inside a wing or somesuch (saves aesthetics, doesn't reduce drag).

5) Control surfaces at the back work by pushing the tail down, canards up front work by lifting the nose up. While canards are deflected for climb, the lift of the canard surface is increased; while rear elevators are deflected for climb, the lift of the control surface is decreased (usually into negative lift). However, the effect is minimal; most of your lift should be coming from the wings, not the control surfaces.

And there are other tradeoffs; canards are draggier and more stall prone; as overall AoA increases, the local AoA on an elevator decreases, moving it away from stall. With canards, the reverse happens; they gain power as AoA increases (giving canard planes a tendency to be agile but twitchy) and may stall out, losing lift and creating drag.

 

Edited March 29, 2017 by Wanderfound

[Jarin]

8 hours ago, Wanderfound said:

stuff

Pretty solid info. On the other hand, I asked that question eight months ago in a previous game version, and have run an entire spaceplane career in the interim.  

Edit: Ah, m3m bumped the thread, I see.

Edited March 29, 2017 by Jarin