Spaceplane Tanker design assistance

[Jarin]

So I've built my first successful SSTO tanker! Two flights under its belt now (first did not return so well, but some adjustments to CoL fixed that nicely for the second), and I'm pretty proud of it! Thanks to @bewing @AeroGav and @mk1980 for the general spaceplane advice.

As a first design, it's obviously going to be less than optimized, so now I'm hoping for some advice in refining it.

craft file here

Current outstanding issues, aside from just general efficency... it wobbles like hell, especially when low on fuel. I suspect this is due to the wing design? Also, I think I'm getting extra drag, with each wing segment adding its own drag rather than being obscured. I had them separately attached to the center for stability, but maybe I should have the wing components attached front to back to eachother? Fuel percentage to orbit varies widely on how well the ascent went. 

Flight profile (also open to suggestions): run out the end of the runway and probably end up drifting down towards the water as you accelerate. Build speed at sea level 'till 500m/s or greater, then start to climb. Make certain your TWR and thrust are going up, otherwise level out and accelerate more before climbing. Shallow climb, aim for 1200/1300m/s before the engines start to drop off. Kick in the Skipper before flameout. Once the Whiplashes die, hit 5 to close the intakes and let the rapiers autoswitch. Push your trajectory up to 40-50km then hit 2 to kill the rapiers and follow orbital prograde to desired AP with the skipper. When delivering fuel, keep the very front nosecone full (minus a bit of ox if you want) and empty the rest.

Re-entry: PE to 30km over KSC (if using trajectories mod, just put the X on KSC). Standard 15-20 degree inclination to slow. Main thing to watch is overheating the RCS, but it's not too troublesome. Feel free to invert and pull "up" towards the ground to avoid overshooting KSC. I've was screaming over the mountains at 1700m/s and still had no trouble bringing it down with inversion and airbrakes. Final approach at around 100m/s, then pop the brakes and settle lightly to the runway at 50 or less.

 

I think the wings are my biggest issue here, but I don't think I can reduce them in size, as I'm already pushing my limit on getting into the air when fully-loaded. Other designs attempted: all whiplashes and all rapiers. The first couldn't push the AP out of the atmosphere on the skipper alone, and the second gently nosedived into the ocean for lack of thrust. 

Edited June 10, 2016 by Jarin

[bewing]

Wings never obscure each other, no matter how you place them. So that's not fixable. Other than that, your report sounds very much like the typical spaceplane stuff that all my spaceplanes do. Poodles are a bit more efficient and lighter than skippers -- did you try a poodle?

Other than that, I think there's a possibility that your lower tailplane is obstructing the output from some of your radial engines, but I'm not sure about that. If that's true, you might get a lot of additional thrust by moving it.

Did you open the aeroGUI overlay to get an idea of the vectors in flight? Especially with the rapiers running, I think.

 

Edited June 10, 2016 by bewing

[mk1980]

hmm doesn't look too bad.

some general notes:

-instead of the inverted mk->1.25m bicoupler adapters with the ramp intakes, you could use the normal (or the double length) mk2->1.25m adapter with one chock cone. 4 shock cones in total should be enough to feed the 8 jet engines. i think that would also reduce drag somewhat (?)

- instead of the skipper, you can use an mk2-> 1.25m tri or quad coupler and stick aerospikes on it. 3 aerospikes produce a bit less thrust than a skipper, but they have better fuel efficiency. not a big deal though. an alternative might be to not use rapiers at all and put a rhino engine at the back. it's heavy, but it has huge thrust and is (almost) as efficient as the terrier & poodle vacuum engines. probably overkill for that plane (don't need that much thrust), but worth a thought if you try to make something even bigger some day.

i'll load up the craft file in KSP and take a closer look. maybe i can give you a tip regarding the wings if i see them in action.

 

EDIT:

i think the angled wing geometry (outer sections of the wings) messes up the lift. i removed the outer sections and just added a delta wing and a structural wing and it seemed to work quite well.

probably should have stopped at that point, but then i continued and changed more stuff.

removed the RCS thrusters and the MP tank and added vernor engines instead. i think they have lower drag (?) and better heat resistance and they are more powerful (and they consume LF/Ox fuel instead of MP)

i also removed the whiplash and skipper engines and ran the plane with 8 rapiers (sorry, couldn't resist)

a small (useful!) change would be to move the solar panels into the cargo bay. you can also set the "toggle cargo bay" in the same action group as the "toggle panels", so it's still just 1 key press. they are better protected in the bay and (i think) produce less thrust.

then i also changed the "X" shaped tail fins to a more coventional setup (2 planar fins moved a bit upwards so they don't occlude the engines) 

and i replaced the winglets (canards) with 2 sets of the .. i think they are called "high performance canards" (?). i angled them slightly upwards (with the rotate tool +SHIFT key for finer rotation . i think it rotates in 5° steps) so they produce some automatic pitch without pressing the key

and then your plane didn't reall look like your plane any more

some images of the plane "in action". it's not exactly what you were asking for, but i got carried away a bit. i hope the proposed change to the wing geometry will at least be useful

 

Edited June 10, 2016 by mk1980

[Jarin]

2 hours ago, bewing said:

Did you open the aeroGUI overlay to get an idea of the vectors in flight? Especially with the rapiers running, I think.

I looked at it, but I wasn't sure what to make of it. I didn't see any obvious big red flags on the tailplanes, but I don't know how occlusion is indicated.

Anyway, since fighting for lift seemed to be one of my problems, I added a topwing, along with the previous suggestions. A poodle lacked the necessary thrust, but the quad-aerospike handles like a dream. I separated the x-tailplane to clear the thrust path. I don't notice much difference, but hey, I'm not burning my tail off anymore.

The topwing and new intakes seem to have helped greatly. Now, aside from still being a bit of a pig getting moving, the flight profile after 500m/s is just point the nose 10 degrees up and let the plane fly itself to whiplash flameout.

Only major problem I'm noticing with these edits is that the off-centeredness of the main thrust is more obvious now that my primary engines don't vector. There's a serious (though not uncorrectable) tendency to nose up for orbital burns.

Edit: Also, I think I may have bodged up the tail a bit. Getting some weirdness I'll have to fiddle with. Included a shot of the aero overlay to see if someone can spot an obvious problem.

 

1 hour ago, mk1980 said:

-instead of the inverted mk->1.25m bicoupler adapters with the ramp intakes, you could use the normal (or the double length) mk2->1.25m adapter with one chock cone. 4 shock cones in total should be enough to feed the 8 jet engines. i think that would also reduce drag somewhat (?)

Not sure if it had much effect on drag, but it gives me more fuel, and looks a lot nicer, so net win regardless.

 

Edit: Further discovery on testing... does NOT re-enter well still loaded with fuel. 

Edited June 10, 2016 by Jarin

[Jarin]

On 6/10/2016 at 11:48 AM, mk1980 said:

EDIT:

i think the angled wing geometry (outer sections of the wings) messes up the lift. i removed the outer sections and just added a delta wing and a structural wing and it seemed to work quite well.

probably should have stopped at that point, but then i continued and changed more stuff.

removed the RCS thrusters and the MP tank and added vernor engines instead. i think they have lower drag (?) and better heat resistance and they are more powerful (and they consume LF/Ox fuel instead of MP)

i also removed the whiplash and skipper engines and ran the plane with 8 rapiers (sorry, couldn't resist)

a small (useful!) change would be to move the solar panels into the cargo bay. you can also set the "toggle cargo bay" in the same action group as the "toggle panels", so it's still just 1 key press. they are better protected in the bay and (i think) produce less thrust.

then i also changed the "X" shaped tail fins to a more coventional setup (2 planar fins moved a bit upwards so they don't occlude the engines) 

and i replaced the winglets (canards) with 2 sets of the .. i think they are called "high performance canards" (?). i angled them slightly upwards (with the rotate tool +SHIFT key for finer rotation . i think it rotates in 5° steps) so they produce some automatic pitch without pressing the key

and then your plane didn't reall look like your plane any more

some images of the plane "in action". it's not exactly what you were asking for, but i got carried away a bit. i hope the proposed change to the wing geometry will at least be useful

I think my tweaks get more fuel to orbit than your redesign, but it's still a bit of a bear to rendezvous. The flat wing edit helps with the stability though. I'll definitely try the double-forward-up-tilted-canards to see if that makes the climb easier. 

Honestly, what I really need to do is rejigger the whole thing to have a few nukes somewhere onboard for orbital operations. Built a new utility SSTO that's one-stop-to-anywhere. ISRU onboard and just needs a bit of fuel from my orbital dock to reach Minmus. If I ditch the ISRU for more onboard, it's nonstop minmus-and-back for 6 kerbals. Four rapiers and three nukes.

In long-run-config below. The front long mk2 gets switched for an inverted bay with fuel equipment. It also needs a second set of small gear, with the long gear retracting so the drills can touch the ground.

Used your recommended shock-cone-on-long-adapter here too, and it works great. Obviously, if it can carry ISRU gear, that front tank can be a cargo bay with whatever is wanted for basic orbital operations.

Craft here

 

[bewing]

In my testing on my own spaceplanes, I actually found that the advanced canards are significantly less efficient than the AV-R8s that you were using initially. Also in my designs, I find that during launch I need downward canard as much as I need upwards -- so pre-twisting them may be a mistake.

And the standard canards out at the ends of the wings ... I tried that myself once because it seems like a good place for another control surface, but it turns out to be a big mistake. They have no leverage to turn the plane (always too darn close to CoM), so they only add drag when they actuate.

 

[Jarin]

7 hours ago, bewing said:

And the standard canards out at the ends of the wings ... I tried that myself once because it seems like a good place for another control surface, but it turns out to be a big mistake. They have no leverage to turn the plane (always too darn close to CoM), so they only add drag when they actuate.

Those are my roll control, and they seem to work okay for that. I'm sure it's unnecessary, but I tend to restrict control surfaces to a primary axis, so those elevators next to the engines are just for pitch. So the canards out there serve three purposes, of roll, a touch of extra lift, and appearance.