Large spaceplanes?

[godefroi]

I've now gotten two different space planes to orbit, and my experience shows that only very small spaceplanes are successful. I have a feeling that Whackjob will disagree with me on this point, but I'm unable to prove any differently.

My first spaceplane had three RAPIER engines, and three of the long 1.25m tanks. A couple delta wings, canards, a couple AV-R8 tail fins, and that's it. It barely gets into a 75x75 orbit. My second is a design I copied from KSP-TV, which has one turbojet and a couple 24-77 engines. A pair of swept wings, canards, AV-R8 tail fins, and that's it. One long 1.25m tank, and a mk2-mk1 adapter for jet fuel.

I've never been successful when attempting to build larger spaceplanes. What am I missing? What are the general tips for large cargo-capable spaceplanes?

Edited March 12, 2014 by godefroi

[Claw]

How large are you talking. There's a pretty big spectrum between 3x FL-T800s and a 400t behemoth.

Quite often, if you're having problems getting into a descent orbit with a small plane it might be due to how you're flying it. Also, you're probably going to get a lot of negative reviews about rapier engines. They have a place in designs, but most people dislike them.

I would say three RAPIERs on a plane that (sounds) that small was probably excessive. So I might venture a guess that you used up a lot of liquid fuel flying just getting to altitude and might have been speeding a bit. There are a ton of threads up here about rules of thumb for wings/engines/etc. Also, you can probably browse around the spacecraft exchange for some ideas.

Anyway, some basic guidelines...

1x TurboJet per 12-15 tons

Around 0.5 to 0.75 points of lift rating per ton (but no more than 1.0 lift rating per ton), basically more wings if you're flying lower TWR

About 1x MK-1 fuselage worth of liquid fuel per TurboJet

3-10 ram intakes per TurboJet, depending on how "air hoggy" you want to be.

Plus however much liquid fuel/oxidizer & rocket engines to accomplish your space goals.

Most space planes that fit into this spectrum do a descent job of getting into orbit if you fly near terminal velocity during your climb. Around 25km you'll want to be level-ish and really want to ramp up your speed since that's where TurboJets and air breathing RAPIERs give the most thrust.

These are really basic rules of thumb to start off with. Small space planes (10-20t range) typically require very little delta V to get to a descent orbit. Larger space planes will typically require a bit more, mostly because you'll run into asymmetric thrust problems with the engines.

Edited March 9, 2014 by Claw

[capi3101]

Claw's beat me to it as far as the general basic guidelines - so I guess I'll post some screenies to prove that yes, it can be done, and hopefully to provide some inspiration:

...those are two different planes; apparently I don't have any other screenies of the Auk VIII (aside from the one of it landing successfully at the Runway). The bottom two are pics of the Auk VII, a slightly lighter but still viable heavy spaceplane.

Claw helped me get both of them into space so I'd definitely take his advice if I were you.

Edited March 9, 2014 by capi3101

[godefroi]

Excellent advice, thank you very much.

Anyway, some basic guidelines...

1x TurboJet per 12-15 tons

Around 0.5 to 0.75 points of lift rating per ton (but no more than 1.0 lift rating per ton), basically more wings if you're flying lower TWR

About 1x MK-1 fuselage worth of liquid fuel per TurboJet

3-10 ram intakes per TurboJet, depending on how "air hoggy" you want to be.

Any tips for getting more ram intakes? I have a really hard time getting more than one per engine, especially when there's only one engine, because the cockpit is immediately forward of the engine...

I read somewhere that one should avoid anything but the ram intakes, but I don't know if that's accurate.

[godefroi]

Also, wings are an issue... the base parts pack seems very anemic on wings of any size, and the procedural wings, while great and all, have never, not once, produced a flyable spaceplane for me.

[celem]

You can produce some very large spaceplanes with ease if you drop the idea of jet engines altogether. Just make it vaguely aerodynamic and strap it until its got something like 2 TWR. Treat it like a horizontal rocket launch, you just need to get the nose up and burn for it. This tactic yields some truly horrid payload fractions.

Cobbling stock together to make lift surfaces can be done but in general sucks. The procedural wings are great but unpredictable given that KSP doesnt work anything like reality where aircraft are concerned, so they do work, but the shapes may be a little counter-intuitive. You need FAR for it to even start making sense to be honest.

[Claw]

Excellent advice, thank you very much.

Any tips for getting more ram intakes? I have a really hard time getting more than one per engine, especially when there's only one engine, because the cockpit is immediately forward of the engine...

I read somewhere that one should avoid anything but the ram intakes, but I don't know if that's accurate.

You're welcome. Hopefully it helps you out.

Here's one idea for getting more intakes without it being an eyesore. You use cubic struts turned on their side. Then you can attach intakes to the cubic struts. In the example below, I attached the struts to an I-Beam in order to make moving them around easier.

Some people think cubic struts are "cheating" so you can also use dual, tri, and quad adapters on the front end of fuselage sections.

[Claw]

You need FAR for it to even start making sense to be honest.

FAR certainly changes the aerodynamics of KSP. I would also argue that once you learn how to work within stock KSP it acts just fine. It certainly lets you get away with more than you would be able to with FAR which is why you can end up with some pretty crazy designs.

If you want to use FAR, then great. If not, don't feel forced to.

[capi3101]

As far as the wings are concerned, I've learned that you can't really be afraid to make bi-planes or tri-planes, and to keep it to no more than two wings attached end to end (see my screenie of the Auk VIII above). You strut the two wings together, strut the near wing to the fuselage, and run one strut from the fuselage to the end of the far wing. When you stack them up into bi-plane and tri-plane configurations, you run a strut between near each wing layer near the far end (that way the wings flex the same and don't collide with one another during flight, which as with most things in the game is generally a Bad Thing).

Ram Intakes are what you want to stick to; they bring in the largest amount of air. Air is good - the more you can bring in, the higher the operational ceiling of your jets; the higher the operational ceiling of your jets, the later you have to light your rockets; the later you have to light your rockets, the more delta-V they'll have once you're finally in space.

[celem]

FAR certainly changes the aerodynamics of KSP. I would also argue that once you learn how to work within stock KSP it acts just fine. It certainly lets you get away with more than you would be able to with FAR which is why you can end up with some pretty crazy designs.

If you want to use FAR, then great. If not, don't feel forced to.

Yes, I certainly didnt want to imply that you need FAR. Stock planes are still fun and stock aerodynamics work fine once you understand them. I meant more that if you want to design something youve seen in reality or that might work on Earth then it will probably work better under FAR. When designing for stock KSP you always have to remember that it is KSP, and you cant say it wont work until you try and prove/disprove it.

I also like and am stealing the cubic strut on its side trick, that looks pretty nice, good tip. I tended to use adapters so far, but its a bit ugly (and not so FAR friendly for that matter)

[Claw]

Yes, I certainly didnt want to imply that you need FAR. Stock planes are still fun and stock aerodynamics work fine once you understand them. I meant more that if you want to design something youve seen in reality or that might work on Earth then it will probably work better under FAR. When designing for stock KSP you always have to remember that it is KSP, and you cant say it wont work until you try and prove/disprove it.

I also like and am stealing the cubic strut on its side trick, that looks pretty nice, good tip. I tended to use adapters so far, but its a bit ugly (and not so FAR friendly for that matter)

Yeah, sorry. I wasn't trying to single you out. Sometimes people feel forced into using mods, so I was just meaning to leave it a matter of choice. I haven't personally used FAR so I have no opinion one way or the other about it.

So that being said, I don't know how that ram intake stick reacts under FAR. I would like to think it's a reasonable way of doing things, but I don't know how the drag would work out. Nor do I know if intakes stacked like that work in FAR or if they block each other out. Someday I will download it, but my attempts to use mods has generally resulted in an increased amount of crashes. So I mostly use stock now.

(I hope the inlet stack serves you well. )

Edited March 11, 2014 by Claw

[Claw]

Also, wings are an issue... the base parts pack seems very anemic on wings of any size, and the procedural wings, while great and all, have never, not once, produced a flyable spaceplane for me.

I must missed this post previously. I haven't used procedural wings (or B9, another aerospace add on), but when I hear people talk about either mod they are usually also discussing FAR. I'm not sure how reliable procedural wings is in stock KSP. If I recall correctly, there are people who successful use B9 with stock KSP aero.

Wing selection is pretty limited in stock KSP parts. Realize you can also use control surfaces, many of which have a lift rating of 0.7. They're a bit easier to hide than delta wings but still provide a good deal of lift. If you use them, make sure to disable the control surface using tweakables, unless you want them on. Just watch out because if you get too many control surfaces fighting each other, it makes your craft jittery (that's a SAS problem).

[Hodo]

Also, wings are an issue... the base parts pack seems very anemic on wings of any size, and the procedural wings, while great and all, have never, not once, produced a flyable spaceplane for me.

I can't tell you much about the stock KSP aerodynamic model because I got rid of it after an hour of playing in the stock KSP. It made no sense to my slightly educated mind. So I installed FAR so aerodynamics work, mostly right, instead of like flying through slush.

PWings actually work great, I use them all the time on vast majority of my spaceplanes, and I have built a few. I have even started building space planes in the Real Solar System plugin with an Earth sized Kerbin, while they aren't as good as my stock sized Kerbin space planes they do work.

The thing is with the stock KSP, you really just want to take everything you know about aerodynamics and what looks like it should fly and throw it out the window. In KSP you can get away with putting a barn door on the front of an aircraft and if you put enough engines behind it it will fly just fine. If you put enough reaction wheels and control surfaces, you can create a working UFO with no jet engines.

I know a lot of people don't like the inline intake nacelle, but it is actually pretty decent at adding a second intake on an airframe without slapping two RAM intakes together.

[godefroi]

Huh. Air-hogging makes a BIG difference. I put 10 intakes on a little single-turbojet plane, and flew it straight to ~45km and ~1.5km/sec before flameout. The navball had switched to "orbit" mode before flameout, and my Ap was at ~63km. I would've hardly needed rockets at all to hit a 75km orbit. Nice!

[Claw]

Huh. Air-hogging makes a BIG difference. I put 10 intakes on a little single-turbojet plane, and flew it straight to ~45km and ~1.5km/sec before flameout. The navball had switched to "orbit" mode before flameout, and my Ap was at ~63km. I would've hardly needed rockets at all to hit a 75km orbit. Nice!

I'm glad you found something that helps you out.

If your plane is small (around 10 tons or less) and you have ten intakes, you can probably get an even higher AP than what you posted. I usually aim to be close to 2km/sec at 30-32km. (Actually, the last "hard" airspeed I aim for is 1000m/sec at 24km.)

Also, if you didn't already know, after flameout you can throttle the engine down and it will re-light. So you can continue to run the TurboJet for a while (just keep slowly throttling down).

With small space planes and the right ascent, I found that three intakes can let me run the engine (at reduced throttle) to over 60km. An AP over 200km is achievable.

Note that this is why people consider air-hogging as cheating.

[KerikBalm]

In addition to air hogging.... do you start to reduce throttle at high altitudes?

Your engine will flame out at a lower altitude if it is at 100% throttle, than if it is at 50% throttle. So keep an eye on your air intake resource, both the buffer remaining, and the consumption rate - as you get higher, throttle back, and keep throttling back until you are no longer raising your apoapsis/gaining speed. Then turn off the jets, close the intakes, and fire the rockets (or with rapiers, switch modes, close intakes)

I don't air hog, or part clip, and put up with stock aerodynamics (it doesn't make real world sense, but once you get used to it, it is at least consistent and sense can be made of it).

This one barely gets 41 tons to orbit, but it does:

But I wasn't satisfied with its stability on reentry, its payload mass capacity, and its payload size restrictions. It was also rather inefficient (using a mk1-2 command pod... what was I thinking using that 4 ton behemoth?)

So, I set about making a new heavy lifter. First successful test flight released a 54 ton payload, and had over 30 tons of fuel left (29.5 tons of liquid fuel and oxidizer in the right ratio)... so in theory, at least an 80 ton payload capacity

It takes a lot of struts to hold together. I wish we had "rockomax" size wings to reduce part count directly and indirectly (don't need to many struts to hold many small parts together).

[mellojoe]

Huh. Air-hogging makes a BIG difference. I put 10 intakes on a little single-turbojet plane, and flew it straight to ~45km and ~1.5km/sec before flameout. The navball had switched to "orbit" mode before flameout, and my Ap was at ~63km. I would've hardly needed rockets at all to hit a 75km orbit. Nice!

Finding ways to get as many air intakes on your planes as possible makes a world of difference.

Also, if you use multiple engines, one will always flame out first. And it will always be the same engine that flames out first. Once you learn which engine is the trouble-maker, you can just babysit that one engine and know that the rest will stay in line. Just right click on the trouble-maker-engine and watch the thrust. When it rapidly starts to drop, then you know it is about to flame-out. Drop the throttle a bit. Keep an eye on thrust numbers. Again, when you see it rapidly dropping, drop throttle a bit. Babysitting just one engine means you can keep as much power on jets as possible.

The order in which you put your engines on your plane also dictates which engine will be the trouble-maker. Typically, the last engine put on will be the first to flame out. Knowing this, I will try to use an odd number of engines (ie: 3) and put the middle one on last. This way, the trouble maker is the one in the center and not one on the side. When that first flame out hits, you won't end up in a flat spin if its the middle engine. This isn't a perfect system, and sometimes the last engine ISN'T the first to flame-out, but more often than not I find this to be true.

[Jouni]

In addition to air hogging.... do you start to reduce throttle at high altitudes?

Your engine will flame out at a lower altitude if it is at 100% throttle, than if it is at 50% throttle. So keep an eye on your air intake resource, both the buffer remaining, and the consumption rate - as you get higher, throttle back, and keep throttling back until you are no longer raising your apoapsis/gaining speed. Then turn off the jets, close the intakes, and fire the rockets (or with rapiers, switch modes, close intakes)

Is reducing throttle relevant anymore? I assumed that 0.23 does it automatically, and flameouts occur only after the intakes don't provide enough air to run all engines at any throttle at all. After that point, the only thing you can do to keep going is to shut down some of the engines, so that the remaining engines can use more air.

[capi3101]

^^^

No, reducing your throttle still works. 0.23 does not do it automatically, though, so you do still have to pay attention. Me, I do it by watching my intake air amount; I really start to give it attention when it gets down to 0.05 IntakeAir per engine. When the plane begins to pull to one side, I throttle back a bit and keep on doing that until A) I'm no longer accelerating or I'm at 1/3 thrust. That's when I fire up the rockets.

[Claw]

Is reducing throttle relevant anymore? I assumed that 0.23 does it automatically, and flameouts occur only after the intakes don't provide enough air to run all engines at any throttle at all. After that point, the only thing you can do to keep going is to shut down some of the engines, so that the remaining engines can use more air.

My understanding is that 0.23 changed a bit of the way intakes work, but it doesn't do any throttle modulating at all. What does happen is that an engine that isn't getting enough air will start to "roll back." Basically at a certain level of intake air, the thrust output will gradually decrease with a continued reduction of intake air. The engine will roll back a certain amount, but eventually not enough air is available and that engine will flame out. You can reduce the air that engine requires by reducing the throttle and, hence, keep it running a bit longer.

This becomes more obvious when flying multi-engine designs. At some point one of the engines won't have enough intake air because one engine gobbles up all it needs, then the second engine doesn't have its full requirement. That's why multi-engine designs will begin to yaw before an engine actually flames out.

So in some regard, KSP does "auto-throttle" in a very limited sense. The engine isn't "on/off," but the effect isn't quite the same as actually reducing throttles. At high altitude, you can sometimes run on less than one notch of throttle and it's enough thrust to overcome aerodynamic drag as you leave the atmosphere.

If I'm feeling aggressive about it with multi-engines, I watch the trim axes for when the SAS inputs yaw, then begin throttle reductions. For single engine I've done enough profiles to know about how long after air intake goes to 0.0 to start reducing throttles.

Like Capi, I keep running TurboJets till I'm no longer accelerating. Unlike Capi, that might be less than 1/3 thrust. Also, you can get away with starting the rockets but still flying your TurboJet profile for a while. Acceleration due to the rockets will keep the intake air higher, but the usefulness of that kind of profile depends on the craft.

[godefroi]

You've inspired me to create a one (or maybe 2!) big orange tanker to make refueling runs to my fuel depot. Excellent information here!

[capi3101]

^^^

Start with one - building a tanker is tricky enough work as it is. Make sure you test it thoroughly in the VAB for its flight characteristics (notably where the Center of Mass winds up once you empty that tank). My most successful tanker plane, the Auk VIII, can only deliver about 60-70% of its payload per run; the reason is because if I deliver every drop, the CoL goes ahead of the CoM, and I lose control once I start getting deep into the atmo. Mind you that this was a craft that was designed with a ballasting system already.

[Jouni]

No, reducing your throttle still works. 0.23 does not do it automatically, though, so you do still have to pay attention. Me, I do it by watching my intake air amount; I really start to give it attention when it gets down to 0.05 IntakeAir per engine. When the plane begins to pull to one side, I throttle back a bit and keep on doing that until A) I'm no longer accelerating or I'm at 1/3 thrust. That's when I fire up the rockets.

I had to try this. I normally use a light one-kerbal spaceplane with one turbojet and one ram intake for testing purposes. At 100% throttle, its service ceiling is around 33 km with a good ascent path. By manually reducing throttle after 32 km, I was able to get to over 43.6 km, but it took a really long time. The thrust was something like 4-5 kN just before flameout.