Need advice : launching a spaceplane from the top of a rocket

[TaoDue]

Hi everyone!

I am trying to launch a spaceplane to low Kerbin orbit using a conventional launcher. The spaceplane is mounted on top of the rocket, where the usual payload would be. This is actually my attempt to recreate ESA's Hermes shuttle :

I've successfully built one before the aerodynamic model was patched but now (KSP v1.0.4) I haven't been able to build one that flies without flipping over... I understand that flipping is caused by the CoL being in front of the CoM so the rocket is naturally unstable, but I haven't been able to find a workaround to counter that phenomenon... I tried to attach more fins to the bottom of the rocket or/and add tons of SAS but nothing works. The only way I can reach space is to fly it straight up until it leaves the atmosphere which is of course very inneficient... I don't know if I miss something with the design or if I don't fly it properly.

Did anyone succeed with this design in KSP V1? Any advice would be much appreciated

Thank you in advance!

Edited August 1, 2015 by TaoDue

[cantab]

Adding larger fins at the bottom should help - you want the rear fins to be enough to more than compensate for the unhelpful wings on the shuttle. On a launcher your strongest control is often the engine gimbals so I advise lighting one or more gimballing engines at launch, don't launch on SRBs alone even if they have the thrust. The usual advice of never straying too far from the prograde marker applies too.

[Alshain]

I tried doing a DreamChaser replica (which is really a more modern attempt at Hermes, unlike the Hermes the Dreamchaser has been constructed and flown in atmosphere) but it is very difficult to maintain control with those lifting surfaces on top. I had to build massive control fins on the bottom of the rocket to bring the CoL back down.

Edited July 31, 2015 by Alshain

[xtoro]

My kerbal way of fixing a similar hardtofly rocket was tons of fins, but it was way too many. I switched out the fins for a few airbrakes. As I found out, there was only a few times where it was difficult to keep from flipping. So on my ascent, if I found the rocket started to wander, I would put the brakes on just for a few seconds to get it to correct itself and then release. Airbrakes were on the bottom where my fins were originally. Kept the part count down and i think overall it caused less drag for the entire ascent.

[Jouni]

Top-mounted shuttles are hard in KSP, because you're trying to launch a huge plane on top of a tiny rocket. The real Hermes would have been a tiny plane on top of a huge rocket.

Unless you build something really ridiculous, the rocket won't be aerodynamically stable. Even if you add two sets of giant wings to the boosters, the rocket will do its best to flip over. Active guidance with thrust vectoring and control surfaces is your best bet (and remember to disable the control surfaces on the plane itself).

[Foxster]

You'll have to use a fairing...

This. I've done it and this is the best way to do it.

[Lighthawk]

Depending on the design of your spaceplane, you could also try sandwiching the plane between two rockets using radial decouplers.

[Geschosskopf]

Another option is to hang the spaceplane BELOW the rocket. In this way, the spaceplane acts as the rocket's tail feathers and all the mass of the rocket is at the top, so it will fly straight. At least given sufficient torque and/or RCS, plus some small (compared to the spaceplane) fins on the nose. This works quite well in 1.0.4.

In the above pic, the rocket lifter is an SSTO heaving 2 identical planes AND a full orange tank nuclear tug into LKO. At which point (as shown on the right), TAC Self-Destruct got rid of the lifter.

If you just want to do a single plane, you can hang it nose-down below the rocket then dock a transfer stage to the plane's rear end once in LKO. Because this gives a bit of asymmetrical lift, however, it requires the rocket to have more attitude control muscle than when launching a pair of planes.

It's a good idea to disable all the control surfaces on the plane(s) before launching. But you must remember to turn them on again before reentry

Edited July 31, 2015 by Geschosskopf

[Sharpy]

You'll have to use a fairing, or neutralize the lift of the wings some other way. The solutions would be VERY kerbal...

Launch two planes, belly to belly.

[ajburges]

CoP is ahead of CoM. This makes craft dynamically unstable. You can maintain course with either passive correction or active control. Given how wacky SAS can be, I recommend passive stability over control authority.

There are 3 good ways to add passive stability to a rocket.

> More lift in back. Wings get greater drag at higher AoA. Winglets are slipstream compensating when behind CoM

> More drag in back. Less elegant than the first solution, but less sensitive to different atmospheric pressure. Also comes in handy to recover modules.

> Fairing over forward wings. Occlude the aspect moving your CoP up so much.

[Slam_Jones]

Basically, you need more wing area at the bottom of the rocket than the top. Personally, I'd probably add Wing Strakes in 6-symettry, maybe adding elevons if they don't provide enough stability.

[Skorpychan]

Stick it lower down on the rocket, so the CoL isn't so far from the CoM.

Asymmetric lift? Stick two on it.

Weight? Use a bigger lifter.

Alternatively, fairing.

[LtWhiskers]

Nothing wrong with a launch straight up to 45km, then slowly turning to circularize. If it works and doesn't bankrupt you I say go for it. Also struts connecting top of plane to top of rocket may help.

[ajburges]

Nothing wrong with a launch straight up to 45km, then slowly turning to circularize. If it works and doesn't bankrupt you I say go for it. Also struts connecting top of plane to top of rocket may help.

Yuck!

A shuttle final stage is aerodynamic enough that you shouldn't need to abandon a gravity turn. Using air brakes below 24 km to keep CoP rearwards is probably more efficient than that.

[GoSlash27]

I was able to do a simple LKO orbit and return to the runway like this:

It was mostly a matter of countering the lift of the wings with lift in the tail and keeping a low t/w ratio. The main LF booster had way more fuel than it actually needed, so it could've been tidied up a bit.

It's an unfinished design (no solar panels, airbrakes, or RCS thrusters) but if you want a copy of the craft file just give me a holler.

Best,

-Slashy

Edited August 1, 2015 by GoSlash27

[Wanderfound]

Option 1: lots of gimbal and a very vertical launch. Don't pitch over until you're in the upper atmosphere.

Option 2:

[GoSlash27]

Wanderfound,

If you take a look one post up, that design doesn't use gimbals and follows a normal gravity turn on launch.

Best,

-Slashy

[Wanderfound]

Wanderfound,

If you take a look one post up, that design doesn't use gimbals and follows a normal gravity turn on launch.

Best,

-Slashy

I wasn't intending it to be an exclusive list. But isn't that a Mainsail (or Skipper?) as the main engine? Both of those have gimbal.

As you demonstrate, the "very small payload wings balanced by substantial tailfins" approach can work, especially if you're careful to keep close to prograde during the ascent. It does tend to be rather fiddly and restrict your payload size, though.

[TaoDue]

First, thank you all because it helped a lot!

I followed some of your advices and I managed to build a spacecraft that is stable enough. I could follow a conventional trajectory with a gravity turn at 10 000 m and reach LKO!

Basically I used a combination of your suggestions :

- Less tall rocket

- Air brakes at the bottom (these compensate deviation when SAS is on!)

- Large wing strakes at the bottom

- Limit TWR

Good job on that guys! I'll try the other designs that some of you proposed as well

Thank you!

[panzer1b]

Back in 0.90 and before, i often launched planes and or capital ships with alot of wing armor by having a rocket above the entire vessel. Ofc now you really have 2 options (if if HAS to be placed atop the rocket), either spam alot, and i mean ALOT of find and stabilizers or even wings on the very bottom of the rocket to overpower the wings on your plane or whatever it is you want atop the rocket, or place the craft inside a fairing. If its very large and wont fit in a fairing, there is a tweak you can make to the fairings to allow a larger radius, at the cost of making thge rocket look 100% rediculous with super tiny engine with massive cylinder thats over 10 times the engine cluster's radius, (ive done this when i made a cylon basestar and refused to assemble it in orbit, or when i was launching an entire refinery complex to dres and i really didnt want to assemble such a large thing on site).

That said, if whatever you want to lift will not work with fairings for whatever reason, and it is excessively draggy, then your ONLY option is having a puller rocket that is above the payload.

[chris206]

My spaceplane that I am launching is a beast. On its own power it can achieve orbit and land 3x. I have spaceplane mods and a nuclear expansion mod so my plane only uses liquid fuel.

My goal with launching it from a rocket is that I wanted to escape kerbin soi so I can maneuver to duna and fly there AND return to kerbin.

This is what I came up with

dont know why I cant upload pics so heres a link

https://goo.gl/photos/6Bk35oju4E7o5GaX7

Edited September 2, 2015 by chris206

[Nothalogh]

Even the USAF designs for Dyna-Soar had big honkin stabilizers on the Titan booster

[ForScience6686]

This is exactly what fairing are good for. I use this method on my rescue craft and only fins are needed to stabilize. The air break method sounds good but may induce too much drag in a craft that constantly imparts lift on the rocket and relies too much on them.

[RoboRay]

My solution...

[RoboRay]

Dear god, that actually flies?!?? Nice work!

It flies great!

Here's some larger variations that don't throw away expensive engines...

Album: http://imgur.com/a/DXk9r#0

Album: http://imgur.com/a/9QbIy#0