My first plane that kind of works, but I would still like to make it a bit more stable, any tips?

[Hadva2]

I used tail fins for control because Elevon 1 control surfaces are way too small.

 

 

[Aegolius13]

Did you put the Elevons on your swept wings?  They may not have worked because that's very close to your center of mass/lift (can't remember which one matters).  Control surfaces are like levers, so being further out helps.  You can get improve leverage by placing control surfaces at the front of the plane (canards, which you already have), and/or at the very back (horizontal tail fins).  Placing smaller fins at the back of the fuselage might work better than those giant wing mounted fins.

Have you set your front canards to "inverted"?  I've seen conflicting things, but I think they will go in the wrong direction by default if they're in front of the center of mass.

Edited October 8, 2016 by Aegolius13

[AeroGav]

There is only so much you can do with the game's physics engine, all craft tend to wander a bit due to errors in the way forces are transmitted from part to part causing phantom rolling motions.    Sometimes they fly straight,  later the same airplane always pulls left, restart the game, and it might roll right.   Just have to live with it i guess.

Pitch stability looks ok.    

For roll,  you can try the following - 

• apply dihedral to outboard wing segments
• attach wings to top of fuselage so most of mass hangs below - gives a pendulum effect (this will raise your centre of mass however, and cause a slight pitch up tendency when power is applied due to engine being below CoM) 
• improve yaw stability, where possible

Yaw stability is often overlooked.   I've heard that it's best to have tail surfaces not much higher or lower than centre of mass if possible, stops the rudder acting like an aileron and causing a secondary roll effect when applied.   For this reason twin tailbooms that are very long and extend far behind the aircraft are better than one very tall one nearer the CG.     But watch for tailstrikes...

Pic - notice the dihedral on outboard sections for roll stability, though i haven't bothered to shift the wings up to a shoulder mount in this case.

Notice the twin tailbooms with some rudder above and below the centre of gravity.  I couldn't shift the engine any further back or i'd have CG issues when empty.

Pic - on this one, i went for 6 small rudders instead of one mega tall one.

Finally, a tip for roll control is to apply  decent sized ailerons/elevons all along the trailing edge of the wing, even if you have no need for fast roll rates.  Then right click and "limit authority" of the aileron surfaces.   Low deflection angles give the lift you need raise a wing, without much drag, which causes the adverse yaw.  As the aileron gets closer to fully deflected/stalling angle 30 degrees,   you get a lot more yaw-inducing drag for not much extra lift.  That is why large, but only lightly deflected ailerons are better for people without rudder pedals  (lol - if anyone can even get a joystick working in this game i'd love to hear it!)

Edited October 6, 2016 by AeroGav

[ForScience6686]

Design it more like a IRL plane.  Unless you have some purpose for that design.  You have very little roll authority, and why use girders for engine attachment, just use mko fuel tanks.  As for the insanely swept wings, I think they are hindering performance more than anything.  You're running on junos so you're not running fast enough to even need swept wings.  Maybe raise it to the top of the fuselage and angle it up slightly.  

[AeroGav]

22 minutes ago, ForScience6686 said:

  You have very little roll authority,  

agree!

Quote

and why use girders for engine attachment, just use mko fuel tanks.  

Suspect that whilst light, they are v draggy

Quote

As for the insanely swept wings, I think they are hindering performance more than anything.  You're running on junos so you're not running fast enough to even need swept wings.  .

Stock aerodynamics do not take account of sweep angle or wing shape at all.  The only thing it effect is centre of lift and centre of mass, otherwise it's a cosmetic effect.

[AeroGav]

2 hours ago, Aegolius13 said:

 

Have yousety your front canards to "inverted"?  I've seen conflicting things, but I think they will go in the wrong direction by default if they're in front of the center of mass.

Canards sometimes attach back to front, especially if you hook them up to something the game normally expects to find on the rear of an aircraft (tail cone flipped 180 degrees being used as a radome).  This can cause control surface inversion.  However, the solution is to set the control authority to a minus number, eg. flip it from +100% to -100% , if you find it moving in the opposite direction to intended.

It happens to elevons too.   I attached elevons to the leading edge of the wing to use them as elevators in this design.   The game normally expects the pivot point to be in front of the bit that goes up or down, because i'd flipped them 180 degrees to attach them to the leading edge, i had to use the -100% authority trick to get them going the right way.

[Plusck]

4 hours ago, Hadva2 said:

I used tail fins for control because Elevon 1 control surfaces are way too small.

They shouldn't be for that size of aircraft.

You probably have a hard time getting airborne because you have wheels far behind the CoM. When taking off, your plane will rotate around its CoM. It can't "rotate" very well if the fulcrum (the hindmost set of wheels) is placed a long way behind it.

All control surfaces use odd (but logical, I guess) rules to determine how far they move in which direction to control which axis of movement. So to roll, they want to be placed far out in a line running through the CoM. To pitch or yaw, far forward or aft of the CoM. If they're angled compared to "forwards", this makes the game only "sort of" use them for each control direction, depending on their precise angle. Change the angle a touch and they may well reverse direction to control (the game considers that they're applying forces the other way, even if it's not true as long as you're flying straight).

So to get the best control, just follow the simple rules implied by that last paragraph and you won't need large control surfaces.

Also, you're using canards. Canards are really not great for fine control, although the computer'll do fine using them if you're activating SAS to fly. However, if you ditch the hindmost set of wheels and add a couple of elevon 1s to the back of the plane (on the rearmost edge of your wing, for example), it would do fine.

[ForScience6686]

41 minutes ago, AeroGav said:

agree!

Suspect that whilst light, they are v draggy

Stock aerodynamics do not take account of sweep angle or wing shape at all.  The only thing it effect is centre of lift and centre of mass, otherwise it's a cosmetic effect.

However, with his wings trailing so far back, as he pitches up drag will increase at the rear which should point the nose back down, right?  I see what you're saying for level flight, but the drag would still be modeled, correct?  I don't have any examples to go off of since I've never encountered a design like this.

[AeroGav]

1 minute ago, ForScience6686 said:

However, with his wings trailing so far back, as he pitches up drag will increase at the rear which should point the nose back down, right?  I see what you're saying for level flight, but the drag would still be modeled, correct?  I don't have any examples to go off of since I've never encountered a design like this.

Those outboard sections, swept well to the aft, will generate lift far behind his CG when trying to pitch up,  this will end up forcing the nose back down again.  From a stability standpoint , that's a good thing , but it's countered by the front/inboard sections being too far forward.

He could have achieved the same centre of lift by using a straight wing and moving it's attachment point to the fuselage aft a bit. From a performance standpoint it'd be identical and it'd be easier to build and modify.

[Plusck]

25 minutes ago, AeroGav said:

It happens to elevons too.   I attached elevons to the leading edge of the wing to use them as elevators in this design.   The game normally expects the pivot point to be in front of the bit that goes up or down, because i'd flipped them 180 degrees to attach them to the leading edge, i had to use the -100% authority trick to get them going the right way.

I suspect that this is an example of what I mentioned earier: "logical" alignment of control surfaces that are not quite what they should be. I wouldn't want to accuse anyone of shoddy coding... but if you take the centreline of your frontmost elevons (perpendicular to their axis of movement) and draw a straight line through the craft, they pass in front of the CoM, not behind. I think that they'd operate normally if you rotated them a touch forward, because that line would pass behind the CoM... and I think that that is what is telling the game whether they should be normal or inverted, even though it is obvious to a human what they should be doing.

Edited October 6, 2016 by Plusck

[ForScience6686]

1 hour ago, AeroGav said:

Those outboard sections, swept well to the aft, will generate lift far behind his CG when trying to pitch up,  this will end up forcing the nose back down again.  From a stability standpoint , that's a good thing , but it's countered by the front/inboard sections being too far forward.

He could have achieved the same centre of lift by using a straight wing and moving it's attachment point to the fuselage aft a bit. From a performance standpoint it'd be identical and it'd be easier to build and modify.

I disagree on performance.  The plane is fighting itself whenever it pitches.  Especially with those engines, he must be losing a lot of speed when maneuvering.  Standard wing setups would be much more beneficial.

[Hadva2]

21 hours ago, Aegolius13 said:

Did you put the Elevons on your swept wings?  They may not have worked because that's very close to your center of mass/lift (can't remember which one matters).  Control surfaces are like levers, do being further out helps.  You can get improve leverage by placing control surfaces at the front of the plane (canards, which you already have), and/or at the very back (horizontal tail fins).  Placing smaller fins at the back of the fuselage might work better than those giant wing mounted fins.

Have yousety your front canards to "inverted"?  I've seen conflicting things, but I think they will go in the wrong direction by default if they're in front of the center of mass.

In my case canards seem to be working correctly, and changing the deploy direction in my case doesn't do anything anyways, any idea why? And also thanks, putting control surfaces further back helped alot.

21 hours ago, AeroGav said:

For roll,  you can try the following - 

• apply dihedral to outboard wing segments
• attach wings to top of fuselage so most of mass hangs below - gives a pendulum effect (this will raise your centre of mass however, and cause a slight pitch up tendency when power is applied due to engine being below CoM) 
• improve yaw stability, where possible

Yaw stability is often overlooked.   I've heard that it's best to have tail surfaces not much higher or lower than centre of mass if possible, stops the rudder acting like an aileron and causing a secondary roll effect when applied.   For this reason twin tailbooms that are very long and extend far behind the aircraft are better than one very tall one nearer the CG.     But watch for tailstrikes...

Is this half decent? Also thanks.

 

And sorry to both of you for late replies, I have been rather busy lately.

[Kryten 2X4B 523P]

If you havent already, take a look at this thread in the Tutorials forum, its really really useful in explaining all the concepts you need to build a plane successfully. Ironically, building early planes can actually be harder using the first aircraft tier parts, but one you get the hang of it its great.

 

[Hadva2]

Wow, insanely helpful, alot of work must have gone into that, thanks.

[ForScience6686]

You really like those girders eh?  I'd say it's improving.  But your elevons are angled quite oddly on the tail fins, and I'd guess they aren't doing much to help.  The forward elevons have hardly any lever, so they won't be much help either, move them forward.  I'm still not getting what your plan is for this craft though.  I would remove those girders.

[Aegolius13]

I believe the invert option on the canards only affects when they are "deployed", not when used for normal pitch/roll control.  Deploying is a separate command which can be set to an action group, like to use as flaps.  

[Hadva2]

11 hours ago, ForScience6686 said:

You really like those girders eh?  I'd say it's improving.  But your elevons are angled quite oddly on the tail fins, and I'd guess they aren't doing much to help.  The forward elevons have hardly any lever, so they won't be much help either, move them forward.  I'm still not getting what your plan is for this craft though.  I would remove those girders.

Uh, without girders ? 

And are the canards forward enough? If they go more forward they will just bend, should I use girders?

[ForScience6686]

51 minutes ago, Hadva2 said:

Uh, without girders ? 

And are the canards forward enough? If they go more forward they will just bend, should I use girders?

I wouldn't use girders, I haven't seen many used IRL.  You have size mk0 liquid fuel tank for the juno engine.  Those girders are adding a lot of drag and just aren't necessary.  Your task find are better, but get rid of the elevons, with the fins angled like that you have pitch and yaw control.   At this point you don't need the canard, I would get rid of it.  As for roll control, put control surfaces on the wings.  You could also add flaps to improve low speed performance.

What's the purpose of this plane?  That will help you decide what characteristics it should have.

[Hadva2]

1 hour ago, ForScience6686 said:

I wouldn't use girders, I haven't seen many used IRL.  You have size mk0 liquid fuel tank for the juno engine.  Those girders are adding a lot of drag and just aren't necessary.  Your task find are better, but get rid of the elevons, with the fins angled like that you have pitch and yaw control.   At this point you don't need the canard, I would get rid of it.  As for roll control, put control surfaces on the wings.  You could also add flaps to improve low speed performance.

What's the purpose of this plane?  That will help you decide what characteristics it should have.

Oh, so the way the tail fins are in that picture is OK?  And what are flaps?

Also the purpose of this plane is just doing survey missions.

[Plusck]

5 hours ago, Hadva2 said:

And are the canards forward enough? If they go more forward they will just bend, should I use girders?

From this, it sounds like you haven't really explored You just need to move them around with the offset (key 2) and rotation (key 3) tools in the SPH/VAB.

Canards are generally not useful unless you have a weight-distribution problem or re-entry-attitude problems. If you insist on keeping them, you can easily offset them forwards and rotate them to face the right way using those two tools (turning off snap for the rotation, or pressing shift while rotating, to get the angle right). Likewise with the elevons at the back.

You have far more wing than necessary, so this plane will take off and land at very low speeds. However, you still don't have any roll authority (tiny elevons near the wingtips would be fine) and the rear wheels look to be far too far behind your CoM, so rotating on the runway for takeoff and gentle landings will be hard (though, admittedly, this is a "weight distribution" problem that canards can help with...).

 

edit: actually realised my first phrase was stupid, since you've obviously been using rotation.

Another thing I just noticed (playing around with light planes myself in 1.1.3): those wheels are really only good for extremely light and slow aircraft. If those three Mk1 sections have fuel in them, the wheels probably won't survive landing.

For example, this plane weighs only 3t, but it shows a tiny amount of wheel stress just being on the runway, more just by being on the grass, and needs pitch to be held up after landing to be able to brake in a straight line (if you drop the controls, the nose wheel gets stressed and stops steering straight). I had a slightly heavier one (with a half-full fuel tank in the middle) and it just crashed all the time on landing.

Spoiler

 

Edited October 8, 2016 by Plusck

[Kryten 2X4B 523P]

A question if I may...

Are those Mk1 fuel tanks youre using for fuselage full or empty?

If theyre full thats an insane amount of fuel for that plane, and the mass of those alone is probably pushing those landing gear to breaking point if youre only using a pair of LY-01s and an LY-05.

A single juno engined plane in the 3.5 ton range using a single mk0 fuel tank (50 units) will take the plane most of the way to the north pole on a one way trip. If youre dead set on using a mk1 fuel tank, only use 1 of them, empty it out and mount a couple of mk0 fuel tanks radially, with the junos and air intakes mounted on the Mk0s. Then use a tail connector A and mount the tail fin / rear control surfaces on that.

Also, as a rough guide you want about 1.5 lift per ton of aircraft mass for an early plane (probably scales up tho) - so if youre plane weighs 3.5 tons without the wings, you want about 5.5 lift in total from the wings and control surfaces. Each wing part or control surface has an amount of lift if you right click on the part in the category screen it will tell you how much lift it has.

The more lift you have, the lower the speed you will need for takeoff, the less lift you have, more speed needed.

Something like this will see you through the early, nearby survey missions - the one on the left only comes in at 3.2 tons, has about 25 parts (from the first aircraft tier and below) and uses a single Juno engine.

 

 

[Hadva2]

5 hours ago, Plusck said:

From this, it sounds like you haven't really explored You just need to move them around with the offset (key 2) and rotation (key 3) tools in the SPH/VAB.

Canards are generally not useful unless you have a weight-distribution problem or re-entry-attitude problems. If you insist on keeping them, you can easily offset them forwards and rotate them to face the right way using those two tools (turning off snap for the rotation, or pressing shift while rotating, to get the angle right). Likewise with the elevons at the back.

You have far more wing than necessary, so this plane will take off and land at very low speeds. However, you still don't have any roll authority (tiny elevons near the wingtips would be fine) and the rear wheels look to be far too far behind your CoM, so rotating on the runway for takeoff and gentle landings will be hard (though, admittedly, this is a "weight distribution" problem that canards can help with...).

 

edit: actually realised my first phrase was stupid, since you've obviously been using rotation.

Another thing I just noticed (playing around with light planes myself in 1.1.3): those wheels are really only good for extremely light and slow aircraft. If those three Mk1 sections have fuel in them, the wheels probably won't survive landing.

For example, this plane weighs only 3t, but it shows a tiny amount of wheel stress just being on the runway, more just by being on the grass, and needs pitch to be held up after landing to be able to brake in a straight line (if you drop the controls, the nose wheel gets stressed and stops steering straight). I had a slightly heavier one (with a half-full fuel tank in the middle) and it just crashed all the time on landing.

  Hide contents

 

Thanks for everything, I think I finally got it. But also something off-topic, if you may, what use do wings on rockets have? I guess they help with control but to what rate?

 

2 hours ago, Kryten 2X4B 523P said:

A question if I may...

Are those Mk1 fuel tanks youre using for fuselage full or empty?

If theyre full thats an insane amount of fuel for that plane, and the mass of those alone is probably pushing those landing gear to breaking point if youre only using a pair of LY-01s and an LY-05.

A single juno engined plane in the 3.5 ton range using a single mk0 fuel tank (50 units) will take the plane most of the way to the north pole on a one way trip. If youre dead set on using a mk1 fuel tank, only use 1 of them, empty it out and mount a couple of mk0 fuel tanks radially, with the junos and air intakes mounted on the Mk0s. Then use a tail connector A and mount the tail fin / rear control surfaces on that.

Also, as a rough guide you want about 1.5 lift per ton of aircraft mass for an early plane (probably scales up tho) - so if youre plane weighs 3.5 tons without the wings, you want about 5.5 lift in total from the wings and control surfaces. Each wing part or control surface has an amount of lift if you right click on the part in the category screen it will tell you how much lift it has.

The more lift you have, the lower the speed you will need for takeoff, the less lift you have, more speed needed.

Something like this will see you through the early, nearby survey missions - the one on the left only comes in at 3.2 tons, has about 25 parts (from the first aircraft tier and below) and uses a single Juno engine.

 

 

Thanks! I think I finally kind of understand planes.

[AeroGav]

7 hours ago, Hadva2 said:

Uh, without girders ? 

And are the canards forward enough? If they go more forward they will just bend, should I use girders?

That's a minor problem that comes when mounting to a cone.   After attaching the part, press 3 to get into Rotation mode,  keep it on the 15 degree step mode rather than fine adjust, and press F key to toggle into "Rotation - Absolute" rather than "Rotation - Local".   Now you will be able to correct the forward elevons so they point straight outward.  This is also VERY important when mounting landing gear, to avoid problems where the plane skids off the side of the runway and flips before you can even get takeoff speed.

Re:  Vertical Stabilizer.

I find myself hardly ever using the rudder control, the actively steered part of it isn't that important.   What does matter in terms of passive yaw stability,  is the "relative wing area" of the part being used for vertical stabilizer, and how far behind the CG it is.      The  Tail fin is an "all moving" part so it has relatively low "relative wing area" for it's mass.    That;s why I often use wing strake parts mounted vertically as my fin and just stick a small elevon on the back as an afterthought.

https://kerbalx.com/AeroGav/simple-bird

   

For the reasons i gave in my previous posts, plenty of aileron surface is a good thing because it' gives the option of turning down control authority to reduce the amount of yaw generated away from the direction you're trying to turn the plane.

Similar for pitch control surfaces  -  because we keep the centre  of lift behind CG for stability,  we need constant nose up input at all times just to stay flying.    If the pitch control surface is small and highly stressed, at a large deflection angle just doing that, then it will be a major source of drag.

Finally, drag.    Apart from all those box girders people keep complaining about,  I strongly suspect you have no kind of tail cone (or nose cone flipped 180 degrees)  at the back end of your plane.   This creates enormous flat plate drag, just as much as it would if your plane had no nose.

Try pressing F12 in flight to see the drag forces you are generating.  Red lines = the drag problems you need to fix !

[Plusck]

3 hours ago, Hadva2 said:

But also something off-topic, if you may, what use do wings on rockets have? I guess they help with control but to what rate?

I use absolute minimal control surfaces on rockets. They are useful, but only from ground to about 15km, so at least 3/4 of the ascent to orbit doesn't need them at all and they are dead weight. On Eve, wing sufaces matter but the atmosphere is so dense that to get the same effect, you need a tenth of the wing surface. Again, therefore, only the very smallest surfaces are actually useful.

So from Kerbin, large fins at the bottom of boosters can be helpful: they keep everything facing the right way, and you ditch them at the first opportunity.

[ForScience6686]

On 10/8/2016 at 7:24 AM, Hadva2 said:

Oh, so the way the tail fins are in that picture is OK?  And what are flaps?

Also the purpose of this plane is just doing survey missions.

The tail fins are fine except for the elevon running through them.  As for flaps, you would stick a pair of elevons on the main wing at the rear close to the fuselage.  Deselect all control authority and set it's toggle function to an action group, or even the gear action group.  It helps to slow you down and provide more lift for landing purposes.

I think you're just about to a proper plane for your surveys.  But I'm not seeing any science gear.