Spaceplane Assistance Requested

[capi3101]

Evening all.

Tonight I launched my first successful SSTO spaceplane, the mighty Auk I:

Javascript is disabled. View full album

Pretty much a no-frills spaceplane, though it does have an ejectable command section for emergencies, a probe core for unmanned flight and a reaction wheel for better overall control (which I'm not convinced is necessary/useful for the design I have). She made orbit and I even managed to land it back at KSC intact (at night no less and in a save game without my ground beacons) afterwards. She had achieved orbital velocity around 30,000 and still had some minimal thrust capability on the jet around 50k.

I do have a problem with the design, and it's one that I need some help overcoming. I can't get this thing to take off prior to reaching the end of the runway. I mean, she'll accelerate the whole way but I can hold down the S key to no effect until after the runway drops out from beneath me. She flies great after that point. So how do I set it up so she'll take off sooner?

Also, what's up with my center of lift and center of thrust indicators? By which I mean why would the center of thrust be pointed slightly downwards?

Also, anybody got any ideas on "where I should go from here"? I mean, I what do y'all typically do with your spaceplanes?

I'm also finally beginning to understand why folks think the terms "spaceplane" and SSTO are synonymous - given what I've learned so far, it'd be difficult (though not impossible, I'd wager) to build a spaceplane that couldn't achieve orbit.

Edited January 7, 2014 by capi3101

[Rhomphaia]

The center of thrust issue is because the seperatrons for ejecting your command pod are counted as well.

To assist take off you could try adjusting the landing gear so that the nose is pointed up slightly on the runway.

[Pursuedtank]

Easy fix for the taking off issue, either twist the gear 180 degrees or move them fowards so they are just behind the center of mass This way they will be pivoted on rather than trying to lift the whole craft up on the back.

[bsalis]

Some options to get off the runway, which you can try some or all of...

* Move the rear landing gear forward, closer to the CoM

* Move the control surfaces further from the CoM

* Raise the rear landing gear to have the craft pitched up when landed (as Pursuedtank suggested)

* More control surfaces to lifting surface ratio

[capi3101]

The center of thrust issue is because the seperatrons for ejecting your command pod are counted as well.

D'OH!!!!

Well, there's a novice mistake on my part. Thanks.

I imagine that's why the center of lift indicator also has a slight tilt to it as well.

Easy fix for the taking off issue, either twist the gear 180 degrees or move them fowards so they are just behind the center of mass. This way they will be pivoted on rather than trying to lift the whole craft up on the back.

Okay...what exactly would twisting the gear 180 degrees do? I see where that would put the wheels themselves in closer proximity to the center of mass...is that the idea?

Some options to get off the runway, which you can try some or all of...

* Move the rear landing gear forward, closer to the CoM

* Move the control surfaces further from the CoM

* Raise the rear landing gear to have the craft pitched up when landed (as Pursuedtank suggested)

* More control surfaces to lifting surface ratio

Well...the elevons are probably about as far from the center of mass as I can comfortably get them. I suppose I could adjust the position of the stabilizers, but they're more of a rudder than anything.

"Control surfaces -to- lifting surface ratio". That's a new one; could someone help explain it? How much lift do you really need anyway? I haven't found much in terms of guidelines on that topic.

Alrighty: summing up the suggestions so far --

I'm at work now; won't be able to try these changes out until after I get home later tonight (probably around 02Z)

Also, would there be any benefit to putting more distance between the back wheels? I know that a wider axle track makes a rover less likely to flip - can the same be said for aircraft? Not that big of an issue; just curious...

[FuzzyLlama]

A wider base can help if you are often wobbling side to side on the .runway, but this is indicative of another problem similar to the one that caused your plane to not generate lift on the runway.

Another trick that effectively increases your angle of attack when sitting on the runway is angling your wings. Hold shift while using q and e to rotate wing parts. Angling wings so that their leading edges are higher than the back of the wings will do two things: increase the amount of lift generated and rotate the direction of your lift vector.

[capi3101]

Yeah, this thing has pulled to the side during take off before, usually to disastrous effect (the "too low to eject" kind). That's not a consistent problem, though.

[Armchair Rocket Scientist]

Okay...what exactly would twisting the gear 180 degrees do? I see where that would put the wheels themselves in closer proximity to the center of mass...is that the idea?

Partly. When you try to pitch up, your control surfaces are applying a downward force on your aircraft's tail, making the craft pivot around the axis of the wheels. The center of mass is on the other side of the wheels, exerting a downward force on the aircraft's nose. To pitch up, the torque from the control surfaces must exceed the torque from the aircraft's weight. Moving the wheels closer to the CoM and farther from the control surfaces makes the torque from the weight smaller and the torque from the control surfaces bigger, making it easier to lift the nose. However, if it's TOO close to the CoM, there is a risk of tailstrikes.

Well...the elevons are probably about as far from the center of mass as I can comfortably get them. I suppose I could adjust the position of the stabilizers, but they're more of a rudder than anything.

You still have some room to move the wings back until the back of the root edge is flush with the back of the rocket fuel tanks. You can also replace your large control surfaces with 3 small ones, giving you extra pitch authority.

I replicated your design with these modifications, but they move the CoL too far back. The plane takes off with ease, but has trouble keeping the nose up. Fortunately, a pair of structural wings mounted just forward of the others will improve that.

Also, would there be any benefit to putting more distance between the back wheels? I know that a wider axle track makes a rover less likely to flip - can the same be said for aircraft? Not that big of an issue; just curious...

This is true, but your wheels look far enough apart that if you aren't flipping out and crashing on the runway you should be fine.

[draeath]

To avoid tailstrikes you could cram a third gear up near the edge of the engine on the tail, along the centerline. It wouldn't normally have contact but when it did, it would be as an alternative to dragging your tail across the pavement like a dog!

Might even help with landings too.

[numerobis]

My recent experience with tail gear is that you survive the tail strike then follow it up by a very hard landing up front; in my case my plane bounced in an unfortunate direction. Not really recommended.

[Markus Reese]

Ahh, I love flying. So will address a couple of your problems ^.^

1. Pulling off to the side is usually caused by slight roll when you have lift, this is a bit relate to your trouble nosing up. Easiest solution is to widen the stance of the landing gears. I do this on single fuselage craft by using that /__| style connector at the wing with the gears on the end of that. Strutted to stabilize.

2. The pitch up is really all about getting the gears forward more. Mount them onto those tail wings a bit more forward than current, the plane should pitch up easier.

3. Tail slap. This is usually caused by trying to lift off at too low of air speed and a bit due to kerbal "lift" physics. Once you are able to get that pitch up, let the SAS do some of the work and hold the nose pitch upwards. If you have enough lift, once you get to speed, the ground will just start to fall away...... Pretty much all of my planes have alot more tail overhang (from COG) with the same clearance as yours does. Just let it lift... oooooooooommmmm -v()v-

4. With the COG that far back, how balanced does it stay as you burn off your aviation fuel? Guess if you get into the liquid fuel early, should stay stable?

[bsalis]

MOVE ALL CONTROL SURFACES OUTWARDS?

Close but not quite right. You need to move them back to give you better Pitch authority (to help get off the runway). Moving them outward will give you better Roll authority. You get more torque the further the control surface is from the CoM because it increases the moment arm (Force x Moment arm = Torque)

"Control surfaces -to- lifting surface ratio". That's a new one; could someone help explain it?

Simply more control surfaces give you more force, so more torque (again, Force x Moment arm = Torque). Another way of thinking about this is that the control surfaces shift your center of lift (CoL). So from my experience, I found that the more lifting surfaces you have, the more control surfaces you need to have to be able to shift the CoL. The higher the control surfaces to lifting surfaces ratio, the further the CoL moves. Essentially, you pitch up off the runway when control surfaces shift the CoL in front of the CoM.

... oh which reminds me of another tip to get off the runway.

* Move the CoM and CoL closer to each other.

Edited January 8, 2014 by bsalis

[DeMatt]

Could try adding a pair of canards (or AV-R8's, if you don't like the actual canards) to the nose, so they'll lift it when you pull up rather than just your elevons pushing the tail down.

[capi3101]

Success - I just had to move the rear gear forward and out on the FL-T400s, and flip them around. For good measure I went ahead and replaced the large elevons with two smaller ones per wing. Took off about halfway down the runway, after my speed had gotten up to 100 m/s. Worked like a charm; got into orbit, circled around a couple of times and came back down. Overshot the KSC runway on entry but I was able to come back around and land; first time landing from an ocean approach (runway 27).

I'm now trying to figure out ways to build a spaceplane that can do moar stuff - which for the moment means "transport more than one Kerbal at a time". I started in last night on the design of the mighty Auk II, a seven-Kerbal shuttle capable of round-trip interplanetary travel (Duna, Eve, Dres). As you might imagine, it's a pretty massive craft and I've discovered that I haven't got enough intakes on the thing (flameouts start around 16,000). I don't have screenies of the ugly monstrosity yet; it's got five turbojet engines for the ascent, two ram intakes and fourteen radial intakes. I had been under the impression that you want about three ram intakes per turbojet and the numbers on the wiki suggest the radial intakes are capable of providing more air, but that seems to not be the case. What am I doing wrong here?

The Auk II really is gawdawful ugly; I let it fly off the end of the runway because it seems to be the best way to avoid tail-strikes with it. Has some other issues too - namely the ejection system is currently incapable of safely returning the crew to Kerbin in the event of an emergency and it isn't capable of recovery from flat spins. It does fly at least.

[capi3101]

Question for y'all - is there a slip/skid indicator in the game anywhere? I've just today learned about the notion of coordinated turning with aileron and rudder; I had been turning by rolling and then pitching - behavior I've since learned can induce spins (and which probably explains why I've had a tough time with flying around in atmo to this point).

Is there a guideline for how much rudder needs to be applied -vs- banking angle or anything like that?

[Soda Popinski]

Question for y'all - is there a slip/skid indicator in the game anywhere? I've just today learned about the notion of coordinated turning with aileron and rudder; I had been turning by rolling and then pitching - behavior I've since learned can induce spins (and which probably explains why I've had a tough time with flying around in atmo to this point).

Is there a guideline for how much rudder needs to be applied -vs- banking angle or anything like that?

Wouldn't be surprised if SteamGauges had turn and slip. I might install this mod if I get more into planes.

[draeath]

KSP's stock aerodynamics aren't all that complicated, just FYI.

Unless you're running FAR, you don't really need to care about coordinating your turns.

[capi3101]

Well, I was curious to see if coordinating the turn would make changing my heading any easier. Using rudder alone doesn't do the trick (precession effects, I'll wager) but neither does banking.

[capi3101]

Next question then: about how much lift do you need per tonne of mass? I couldn't get this thing to climb for love or money...

I'm pretty sure I've got enough thrust there, and she'll pitch up okay but she still goes in the drink every time.

EDIT: I'm pretty sure all five of the Turbofans are lighting up, but lemme make sure about the two inboard. It would explain a great many things if I had only three operational engines instead of five, wouldn't it?

Edited January 13, 2014 by capi3101

[LethalDose]

First, I'd agree moving the rear gear forward, it can make a huge difference.

You may also want to consider using 2 turbojets and one rocket motor, instead of two rockets and one jet. This will give you more acceleration down the runway, and therefore, more lift when you nose up.

You can also fire the rocket(s) on the tarmac to get more speed during take-off. I've had to do that with my larger SSTOs to get enough velocity to get airborne, and even then I go off the end of the runway sometimes. But that's okay, because then you can just nose-up and let the thrust get your ascent started.

Also, an adv. SAS unit near your CoM may help force the nose up on the runway, and provide more

Finally, what on Kerbin are you doing mounting 24-77's on that thing!!! You will need WAAAAAAAAAAAAAAAAAAAAAY more thrust than that to get into orbit. I usually use LVT-30's (215 kN of thrust and good ISP in the upper atmosphere).

Also: Is this a career mode game? That craft has a bizarre tech mix...

[numerobis]

I'm working on a little script.

I type:

planeDesigner.forces( ((jets.turbojet, 5), (jets.ramAirIntake, 18), (lift.largeControlSurface, 4), (lift.sweptWing, 4), (lift.deltaWing, 2), (lift.avt1, 2), (lift.wingConnector, 2), (lift.canard, 2)), 61.5, 0, 20, 50, 70)

Where 0 is the flight path (flat), 25 is the angle of attack, 85 is the speed, 70 is the altitude of the runway. In other words, we're on the runway at just about our top speed, and we roll back to 25 degrees.

And it replies:

lift and drag: (-471.39,  386.04) kN
thrust:        ( 520.41,  242.67) kN
gravity:       (       , -603.17) kN
net:           (  49.02,   25.54) kN

So that's liftoff! Barely. As you climb, the flight path grows, which shrinks your angle of attack. You took off by pitching to 25 degrees; if you climb 100 meters (up to 170m), get your flight path up to 5 degrees, and keep the same pitch, you now have the following:

lift and drag: (-488.94,  298.93) kN
thrust:        ( 520.41,  242.67) kN
gravity:       (       , -602.97) kN
net:           (  31.46,  -61.37) kN

Pitch up higher, to 30 degrees, and you'll be able to maintain a bit steeper than 5 degree flight path, but not much.

So you should be able to get off the ground and stay there. By the time you're at 1km, pitch 30 degrees, you're flying 93 m/s at a 4.5 degree flight path angle. But this is pretty marginal: much higher and your lift drops off, but your thrust can't yet pick up the slack because (a) you're slow, so the turbojet puts out little thrust and ( you're slow, so there's no centrifugal effect to speak of. You'll end up burning a lot of fuel and slowly climbing and speeding up as your mass falls. Eventually you'll break the wall.

My suggestion: either more lift, or more thrust. Lift is cheaper to provide. Four extra swept wings would make a big difference. Even better, get the small control surface, and place some on the wing connector. Angle them to be flat with the wing, then shift-w (or shift-s) twice to angle it ten degrees, so that when you're flying level, it's at a 10-degree angle of attack. Disable all the control inputs (yaw/pitch/roll), and now it's just extra wingspan. At liftoff, each one will be providing 24 kN of lift, so 10 of them is equivalent to a T30 firing under your belly -- but a T30 that only weighs 100kg and doesn't use fuel.

[capi3101]

First, I'd agree moving the rear gear forward, it can make a huge difference.

You may also want to consider using 2 turbojets and one rocket motor, instead of two rockets and one jet. This will give you more acceleration down the runway, and therefore, more lift when you nose up.

You can also fire the rocket(s) on the tarmac to get more speed during take-off. I've had to do that with my larger SSTOs to get enough velocity to get airborne, and even then I go off the end of the runway sometimes. But that's okay, because then you can just nose-up and let the thrust get your ascent started.

Also, an adv. SAS unit near your CoM may help force the nose up on the runway, and provide more

Finally, what on Kerbin are you doing mounting 24-77's on that thing!!! You will need WAAAAAAAAAAAAAAAAAAAAAY more thrust than that to get into orbit. I usually use LVT-30's (215 kN of thrust and good ISP in the upper atmosphere).

Also: Is this a career mode game? That craft has a bizarre tech mix...

You're talking about the Auk I, right? I got my takeoff issue solved with that one. I figured I'd have more spaceplane questions to ask, though, which is why I left the thread listed as "unanswered". Thank you anyway.

Regarding the Auk I, though, I used a single turbofan because it did the job; it got me up to orbital velocity before flaming out in the mesosphere (around 35,000 or so). I avoided the double turbofan so I wouldn't have to worry about flat spins, and with the jet doing the bulk of the work all I needed was a little extra thrust for keeping the Ap where I wanted it, circularization and deorbiting; the twin 48-7S handled that easily. They don't even have to light until I'm well up into the atmosphere. The design has a reaction wheel - I use those instead of Inilne Stabilizers (ASAS) because of the lower mass - and that does help with banking a great deal. Doing this sandbox; seemed to be the best way to do things since I'm just experimenting.

The Auk III now, that's an attempt to do something useful. It's just not doing anything useful yet.

[capi3101]

I'm working on a little script.

I type:

planeDesigner.forces( ((jets.turbojet, 5), (jets.ramAirIntake, 18), (lift.largeControlSurface, 4), (lift.sweptWing, 4), (lift.deltaWing, 2), (lift.avt1, 2), (lift.wingConnector, 2), (lift.canard, 2)), 61.5, 0, 20, 50, 70)

Where 0 is the flight path (flat), 25 is the angle of attack, 85 is the speed, 70 is the altitude of the runway. In other words, we're on the runway at just about our top speed, and we roll back to 25 degrees.

And it replies:

lift and drag: (-471.39,  386.04) kN
thrust:        ( 520.41,  242.67) kN
gravity:       (       , -603.17) kN
net:           (  49.02,   25.54) kN

So that's liftoff! Barely. As you climb, the flight path grows, which shrinks your angle of attack. You took off by pitching to 25 degrees; if you climb 100 meters (up to 170m), get your flight path up to 5 degrees, and keep the same pitch, you now have the following:

lift and drag: (-488.94,  298.93) kN
thrust:        ( 520.41,  242.67) kN
gravity:       (       , -602.97) kN
net:           (  31.46,  -61.37) kN

Pitch up higher, to 30 degrees, and you'll be able to maintain a bit steeper than 5 degree flight path, but not much.

So you should be able to get off the ground and stay there. By the time you're at 1km, pitch 30 degrees, you're flying 93 m/s at a 4.5 degree flight path angle. But this is pretty marginal: much higher and your lift drops off, but your thrust can't yet pick up the slack because (a) you're slow, so the turbojet puts out little thrust and ( you're slow, so there's no centrifugal effect to speak of. You'll end up burning a lot of fuel and slowly climbing and speeding up as your mass falls. Eventually you'll break the wall.

My suggestion: either more lift, or more thrust. Lift is cheaper to provide. Four extra swept wings would make a big difference. Even better, get the small control surface, and place some on the wing connector. Angle them to be flat with the wing, then shift-w (or shift-s) twice to angle it ten degrees, so that when you're flying level, it's at a 10-degree angle of attack. Disable all the control inputs (yaw/pitch/roll), and now it's just extra wingspan. At liftoff, each one will be providing 24 kN of lift, so 10 of them is equivalent to a T30 firing under your belly -- but a T30 that only weighs 100kg and doesn't use fuel.

Would be very much interested in seeing your code...Java, right? I use Python myself when the coding mood strikes me.

Caught the meaning behind most of the parameters...you've got the lifting parts and quantities, followed by mass...and the next four numbers I'm not sure about.

If I get the gist of what your code has calculated, though, I should have sufficient thrust for takeoff as things are (though just barely). That makes me suspect the two inboard engines even more...

I will make the changes you've suggested to the Auk III design and let you know how it goes.

[numerobis]

It's python; the KSP-scripts that I've been using forever (slowly adding little things here and there; your post made me put some of them together).

"The next four numbers" are what I describe in the line that follows the code.

The takeaway message I'm getting is that your plane needs more lift. I tend to use an engine every 15t to 20t; you've got one for every 12t. Any lack of lift can always be made up by adding thrust, of course, but lift can be added with a lot less mass.

[capi3101]

It's python; the KSP-scripts that I've been using forever (slowly adding little things here and there; your post made me put some of them together).

Ah! Ka-ching!!!

"The next four numbers" are what I describe in the line that follows the code.

Okay...that bit confused me; what was written in the line didn't match what was there in the code:

I'm working on a little script.

I type:

planeDesigner.forces( ((jets.turbojet, 5), (jets.ramAirIntake, 18), (lift.largeControlSurface, 4), (lift.sweptWing, 4), (lift.deltaWing, 2), (lift.avt1, 2), (lift.wingConnector, 2), (lift.canard, 2)), 61.5, [B]0, 20, 50, 70[/B])

Where 0 is the flight path (flat), 25 is the angle of attack, 85 is the speed, 70 is the altitude of the runway. In other words, we're on the runway at just about our top speed, and we roll back to 25 degrees.

You have 0, 20, 50, 70 instead of 0, 25, 85, 70.

The takeaway message I'm getting is that your plane needs more lift. I tend to use an engine every 15t to 20t; you've got one for every 12t. Any lack of lift can always be made up by adding thrust, of course, but lift can be added with a lot less mass.

Gotcha.