How do you go about designing/improving a spaceplane?

[Nephrahim]

Bit of an in-depth question, but I'm just wondering...  how do you guys go about improving/designing your space planes?  I can MAKE them, but making them GOOD seems like such a byzantine process.

I've been playing a new campaign mode, and got a contract for a 3 person tourist trip to the mun and back.  Decides it would be fun to use a spaceplane for this, so went about designing this beast:

I'm fairly happy with it, although it has some pretty big problems.  It tends to veer side to side during takeoff, not sure why.  The more important part, though, is that I can't manage to get the Delta V to the mun and back.  This one can land on the mun, but that's about it.  I don't know how to fix this, though.  With a rocket it's easy, add more Delta-V, weather that means bulking up the stages or adding a new one.  With the spaceplane though there's such a complicated relationship with lift, drag, thrust, and Fuel.  I add more fuel, and I can't go fast enough, so I add more engines, but then I end up not having enough fuel again.  Is there any way to tell what I'm doing other then spending a few minutes trying to get to the mun and back?  It takes a while with these NERV engines.

So yes, if anyone has any tips for THIS plane, that's fine, but I'm more interested in learning how to fish, y'know?  So far all my space planes seem improvised and not nearly as well designed as my rockets.

[Rocket In My Pocket]

Mostly trial and error lol. Design it-Fly it-Revise it-Fly it-Revise it-Fly it-#$@*(#!@^&-Start over-Design it-Fly it.

This mod is a huge help as well.

The other thing to remember is that spaceplanes/SSTO's simply aren't the best answer to every problem. (A spaceplane doesn't need to be an SSTO either, you can always add drop tanks or boosters.) Sometimes you need to step back and ask "Would this be easier with a rocket?" Personally I only use them for LKO missions, so I'm gonna leave the more detailed advice to our resident SSTO enthusiasts/experts, who will likely be showing up soon. Best of luck!

Edited December 2, 2018 by Rocket In My Pocket

[Nephrahim]

Thanks, I'll look at it.  Oh, and I am FULLY aware I could save tons of time just launching a rocket and getting it over with, but at this point it's the principle of the matter.

[Rocket In My Pocket]

1 hour ago, Nephrahim said:

Thanks, I'll look at it.  Oh, and I am FULLY aware I could save tons of time just launching a rocket and getting it over with, but at this point it's the principle of the matter.

Fair enough.

Well, generally the advice you are going to get is "less is more." Start taking everything off that you can afford to lose.

I'm no expert in SSTO's as I said, but that looks like an awful lot of engines to me, and NERV's are heavy.

[Cavscout74]

One thing to help with the veering to the side on takeoff, you can try reducing the friction control setting on your nose gear.  It defaults to auto, but you can change it to manual & move the slide that appears down a little under one, so your nose gear has less friction than your main gear.  

You also have to pay close attention to drag.  Just from the screenshot, it looks like you have a pair of stack bi-couplers.  Those are pretty bad about drag.  Judging by your screenshot, I'd say (other than the bi-couplers) you have too many engines (and associated mass). 

Here is a design of mine - with a short Mk 2 fuselage section, the rest 1.25m.  It uses a pair of rapiers & a single NERV and after reaching an 85km circular orbit with a pilot & 4 passengers, typically has around 2000 m/s dV - plenty to orbit Mun or Minmus & return.

https://kerbalx.com/cavscout74/CS-08S-Lightning

 

Also, @AeroGav posted two nice tutorials on here that helped me out.  There are other good tutorials here also, but I happen to have these bookmarked:

 

Edited December 2, 2018 by Cavscout74

[Fearless Son]

I am going to echo @Cavscout74 on the point about drag.  Drag is a huge limitation on spaceplanes, drastically increasing their fuel requirements to reach orbit.  Further, every extra bit of mass you add just to get it into orbit becomes more dead weight (figuratively speaking) once you get there.  With a multi-stage launch (like most rockets) this isn't a problem since you eject the mass that's no longer useful, but with an SSTO spaceplane that's not the case.

So a "less is more" approach is really important here.  The tyranny of rocketry applies everywhere, but everything goes double for SSTOs since the first stage is also the final stage.  If you can strip it down, do it.  You really don't have the luxury of making something big and fancy for this kind of mission profile.

[bewing]

6 hours ago, Cavscout74 said:

Here is a design of mine

Lovely design. One trick you may not know, though -- it looks like you have a klaw on the nose? If you "arm" the klaw, you drag goes down by a lot. So that's a smart thing to do on ascent.

9 hours ago, Nephrahim said:

So yes, if anyone has any tips for THIS plane

Yeah, so, general tips for cutting stuff:

Get rid of the struts and use autostruts instead.
Avoid using MK2 parts unless you need them or know exactly what you are doing with them -- MK2 parts have unusually high drag and drag is the main thing that kills you.
Don't add a lot of unused/unusable kerbal seats. They weigh tonnes. Weight is the second thing that kills you.
Reduce the number of engines to the bare minimum that works. Ditto about weight.
You have some secondary nacelles that are "incomplete" -- they end in nothing. That is somewhat bad for drag. Either put a whiplash on the end, or move the fuel tanks around enough to get rid of that extra nacelle.
The 3 tailfin design is cute, but "cute" adds mass and you don't need it. All you need is one fixed airfoil that doesn't move -- fixed airfoils weigh less.
Two nacelles with a shock cone on the tip of each is enough air intakes. Each additional air intake adds mass and drag.

 

[Cavscout74]

1 minute ago, bewing said:

Lovely design. One trick you may not know, though -- it looks like you have a klaw on the nose? If you "arm" the klaw, you drag goes down by a lot. So that's a smart thing to do on ascent

Thanks.  I did not know that trick.  The claw version was a one time variant for a LKO rescue mission, normally it's a shielded docking port.  I know they aren't the greatest for drag, but it was simpler than the inline port.  I may have to check it out further with the klaw though.

[Nephrahim]

1 hour ago, bewing said:

Lovely design. One trick you may not know, though -- it looks like you have a klaw on the nose? If you "arm" the klaw, you drag goes down by a lot. So that's a smart thing to do on ascent.

Yeah, so, general tips for cutting stuff:

Get rid of the struts and use autostruts instead.
Avoid using MK2 parts unless you need them or know exactly what you are doing with them -- MK2 parts have unusually high drag and drag is the main thing that kills you.
 Don't add a lot of unused/unusable kerbal seats. They weigh tonnes. Weight is the second thing that kills you.
Reduce the number of engines to the bare minimum that works. Ditto about weight.
You have some secondary nacelles that are "incomplete" -- they end in nothing. That is somewhat bad for drag. Either put a whiplash on the end, or move the fuel tanks around enough to get rid of that extra nacelle.
 The 3 tailfin design is cute, but "cute" adds mass and you don't need it. All you need is one fixed airfoil that doesn't move -- fixed airfoils weigh less.
Two nacelles with a shock cone on the tip of each is enough air intakes. Each additional air intake adds mass and drag.

 

I've seen people talking about the MK2 parts being less efficient...  but that's compared to the MK1, not 3, right?  I can't imagine it'd be easier to go into space with a n MK3 part designed ship?

I need to get at least 3 (I wanted 4 but maybe I'll have to scale back.) people into orbit with this.  The MK2 design seemed the easiest way to do that, 2 cockpits.  I've seen SSTOs designed with science labs, so it didn't occur to me I'd have so much trouble with THIS.

Oh, and I'm glad you find the tailfins cute, but they weren't for being cute   It was trying to get rid of the veering from side to side at takeoff.

Sigh....  guess I'll go back to the board with what you guys are telling me.  Maybe some MK1 parts and a MK1 to 2 to get the 3 seats instead of 4 (I hate losing the pilot seat for the RPing, but I'll try anything to get it working.)

Oh, one last thing, I've GOTTEN this into orbit with over 2000 Delta V left....  or so Kerbal Engineer would have me believe.  I assume it's just not calculating the Delta-V right?  Because I tend to use twice as much as it says I should.

[bewing]

2 hours ago, Nephrahim said:

I've seen people talking about the MK2 parts being less efficient...  but that's compared to the MK1, not 3, right?  I can't imagine it'd be easier to go into space with an MK3 part designed ship?

Well, you've heard of the square-cube law? MK1 is about going to space with a sleek, cheap, elegant design. The advantage of MK3 is more. An MK3 design has plenty of room for more engines, and more fuel, and more lifting surfaces, and more cargo, and just more. So yes, in fact, with the extra fuel and engines and heat resistance -- it is easiest to go to space with an MK3 design. If you are in career mode, it costs a fortune.

The advantage of MK2 parts is that they are pretty.  Otherwise, they suck.

2 hours ago, Nephrahim said:

I need to get at least 3 (I wanted 4 but maybe I'll have to scale back.) people into orbit with this.  The MK2 design seemed the easiest way to do that, 2 cockpits.

Yeah -- it's not a bad solution. Depending on what kind of probe cores you have, or whether you have piloting covered some other way, crew cabins can generally a better way of packing kerbals around.

2 hours ago, Nephrahim said:

Oh, and I'm glad you find the tailfins cute, but they weren't for being cute   It was trying to get rid of the veering from side to side at takeoff.

 

Yeah, veering on takeoff is a very common problem for planes. There are a handful of causes and a handful of fixes. It's discussed in the FAQ thread on this forum. As said above, my favorite fix is manually reducing the friction on the front wheel.

 

[Foxster]

15 hours ago, Nephrahim said:

I add more fuel, and I can't go fast enough, so I add more engines, but then I end up not having enough fuel again. 

As others have said, this is basically the problem. It is a matter of doing almost the opposite: take stuff away and reduce drag. 

You want a mk1-only (or smaller) craft with the minimum number of stacks. Each stack should be tapered at both ends (except where you have engines and even that is often surmountable). Don't use stack adapters - far too draggy. Don't use mk2 parts. Use mk1 crew cabins for the passengers. Don't have bits stuck on the outside - if you must carry extra bits then consider a small service bay. Use low-drag wings and minimise the number and size of elevons. Take away air intakes until the engines stall and then add one back. To help make sure parts like wheels are on straight (which can often lead to runway drift) use the Hanger Grid mod, highlight the part and hit M and N keys to align part with axis.    

Edited December 3, 2018 by Foxster

[Fearless Son]

On 12/2/2018 at 10:52 PM, bewing said:

Well, you've heard of the square-cube law? MK1 is about going to space with a sleek, cheap, elegant design. The advantage of MK3 is more. An MK3 design has plenty of room for more engines, and more fuel, and more lifting surfaces, and more cargo, and just more. So yes, in fact, with the extra fuel and engines and heat resistance -- it is easiest to go to space with an MK3 design. If you are in career mode, it costs a fortune.

So would you say that the MK2 was a kind of "worst of both worlds" thing?  All the disadvantages of the MK1 and MK3 put together, without the advantage of either alone?

Edited December 4, 2018 by Fearless Son

[bewing]

3 hours ago, Fearless Son said:

So would you say that the MK2 was a kind of "worst of both worlds" thing?  All the disadvantages of the MK1 and MK3 put together, without the advantage of either alone?

Well, that's on ascent. Which is kinda the most important phase for a spaceplane. Horrible amounts of drag are a bonus on descent. So, MK2 looks awesome and is easy to land. But yeah, none of the benefits of MK1 or MK3.

[Guest]

17 hours ago, Fearless Son said:

So would you say that the MK2 was a kind of "worst of both worlds" thing?  All the disadvantages of the MK1 and MK3 put together, without the advantage of either alone?

Mk2 has its uses -- the small cargo bay is just right for a science payload + support gadgets, and the specially-shaped tanks with the piece that has two engine nodes give a lot of design flexibility; the Mk2 crew cabin is also great for medium-capacity spaceliners. And the two-seat inline cockpit is very nice for some uses as well.

But yeah the problem is drag. I almost never use Mk2 parts near the nose, and try to keep the number of Mk2 parts low. They also have some very weird CoP effects -- if you've got, say, a Mk3 plane with a Mk2 nose, it is quite likely to turn somersaults upon re-entry; the nose will shift the CoP forward much more than you'd expect.

[Hotel26]

On 12/3/2018 at 12:49 AM, Nephrahim said:

It tends to veer side to side during takeoff, not sure why.

I recognize this symptom and the key part is not 'veer' [to the side] but 'side to side'.  When this happens to me, it's always because I haven't -- and rectifiable by going back into the SPH and -- use[d] the rotation gizmo in Absolute mode to absolutely confirm that the yaw and roll axes of the wheels are perfectly orthogonal to the runway.  Click the gear with the rotation tool in the snap-to-grid mode and deliberately cock them off center-line and then reset them to center-line.  Wheels aligned forward down the runway [not pigeon-toed] and wheels aligned vertically [not bow-legged].  It's the latter misalignment in particular that sends the craft wandering back and forth over the runway center-line...

Edited December 5, 2018 by Hotel26