Plane Failure

[RocketSquid]

Okay, so I'm building a plane designed to be highly maneuverable and to land on aircraft carrier decks. However, it exhibits ridiculous levels of yaw-roll coupling, and seems completely unable to land. Regardless of the landing speed or angle, the wheels fail, destroying most of the plane even in best case scenarios. I have had sub- 80 m/s landings that destroy the whole plane. The plane uses one "Wheesly" basic jet engine and 2 "Panther" engines. Even when I moved the wings down on the fuselage and angled them downwards (anhedral low wing), it still remained uncontrollable.

[Snark]

5 minutes ago, RocketSquid said:

Okay, so I'm building a plane designed to be highly maneuverable and to land on aircraft carrier decks. However, it exhibits ridiculous levels of yaw-roll coupling, and seems completely unable to land. Regardless of the landing speed or angle, the wheels fail, destroying most of the plane even in best case scenarios. I have had sub- 80 m/s landings that destroy the whole plane. The plane uses one "Wheesly" basic jet engine and 2 "Panther" engines. Even when I moved the wings down on the fuselage and angled them downwards (anhedral low wing), it still remained uncontrollable.

A screenshot would be helpful for diagnosis... otherwise it's difficult to answer "I have a craft, why doesn't it work" questions. 

6 minutes ago, RocketSquid said:

ridiculous levels of yaw-roll coupling

One possible explanation-- did you remember to disable roll authority on your tailfin?

Beyond that suggestion, really need to see a screenshot.

[FullMetalMachinist]

2 hours ago, RocketSquid said:

Even when I moved the wings down on the fuselage and angled them downwards (anhedral low wing), it still remained uncontrollable.

I could be mistaken (it's been known to happen) but I thought that if you had low wings, you wanted dihedral (upwards) angle for greater stability. 

[Nich]

Is yaw introduced by roll?  Or roll introduced by yaw? My first guess agrees with Snark

[Snark]

24 minutes ago, Nich said:

Is yaw introduced by roll?  Or roll introduced by yaw? My first guess agrees with Snark

Well, a classic problem with airplanes is leaving roll authority enabled on the tailfin (it's easy to do, that's the default behavior.)

That's a situation in which yaw induces roll, and roll induces yaw, and the tailfin ends up having less ability to keep you stable on the yaw axis, which is the reason for having it there in the first place.

Turning off everything but yaw authority on the tailfin solves all three of those problems.

Just how bad the problems are depends on the placement of the fin.  The "higher" it is (i.e. farther outward from the craft's central axis), the worse this problem will be.

[RocketSquid]

1 hour ago, FullMetalMachinist said:

I could be mistaken (it's been known to happen) but I thought that if you had low wings, you wanted dihedral (upwards) angle for greater stability. 

I had them angled downwards for greater maneuverability, like fighter jets IRL.

[FullMetalMachinist]

6 minutes ago, RocketSquid said:

I had them angled downwards for greater maneuverability, like fighter jets IRL.

Well, yes, but you said in the OP:

4 hours ago, RocketSquid said:

angled them downwards (anhedral low wing), it still remained uncontrollable.

So I figured you would want something that would increase stability/controlability. And for what it's worth, most modern fighters have wings that go straight out, neither anhedral nor dihedral. 

[GDJ]

Just a guess from the descriptions but it sounds like your COL is too far below the COM. You'll need to alter your plane's wing layout to reduce the distance between the COM and COL, preferably so the COL blue sphere is touching or within the yellow COM sphere. Also examine your payload and accessories mounted on the fuselage. If you have external tanks, mount them as low as feasible to drop the COM closer to your COL.

Downward angled wings don't necessarily cause this to happen, but it doesn't help. I found that having downward angled wings will work best if the wings root are mounted higher on the fuselage, but that might negate any advantage you're looking for. I would try to mount the wings centered on the fuselage, dead level (no dihedral angles).

Start simple first, make the plane work properly, then make small changes.

Edited March 8, 2016 by GDJ

[RocketSquid]

5 hours ago, GDJ said:

Just a guess from the descriptions but it sounds like your COL is too far below the COM. You'll need to alter your plane's wing layout to reduce the distance between the COM and COL, preferably so the COL blue sphere is touching or within the yellow COM sphere. Also examine your payload and accessories mounted on the fuselage. If you have external tanks, mount them as low as feasible to drop the COM closer to your COL.

Downward angled wings don't necessarily cause this to happen, but it doesn't help. I found that having downward angled wings will work best if the wings root are mounted higher on the fuselage, but that might negate any advantage you're looking for. I would try to mount the wings centered on the fuselage, dead level (no dihedral angles).

Start simple first, make the plane work properly, then make small changes.

It was initially built with the wings dead center, perfectly level, but it experienced the dutch roll problem. To try an fix it, I moved the wings downward. When that didn't work, I also angled them downwards.

Anyways, the tailfin changes helped a lot, but now I don't have enough roll authority.

Darn plane.

[Snark]

1 hour ago, RocketSquid said:

Anyways, the tailfin changes helped a lot, but now I don't have enough roll authority.

Glad to hear it helped! However, the moral of the story is that you never had enough roll authority, you just thought you did.    You pretty much never want your tailfin "helping" you with roll; it's a cure worse than the disease.

Still waiting for a screenshot of this plane, so it's still hard to give specific advice.  But roll authority tends to be very easy to get.  All you have to do is put your ailerons as far out towards the tips of the wings as possible (i.e. not close in to the centerline of the plane).  That will give them more of a lever arm to work with and they'll roll you hard.

(When I build planes myself, I usually have the opposite problem-- too much roll, too "twitchy"-- so I need to move them farther in.  But then, I'm typically not trying to build fighter jets.)

[RocketSquid]

48 minutes ago, Snark said:

Still waiting for a screenshot of this plane, so it's still hard to give specific advice.  But roll authority tends to be very easy to get.  All you have to do is put your ailerons as far out towards the tips of the wings as possible (i.e. not close in to the centerline of the plane).  That will give them more of a lever arm to work with and they'll roll you hard.

Okay, finally got the screenshot to upload.

Also, the elevons are Elevon 3's, as you can probably tell.

 

[Snark]

32 minutes ago, RocketSquid said:

Okay, finally got the screenshot to upload.

Whoa.  Mean-lookin' little critter ya got there. 

Okay, yeah, that plane is seriously going to have roll/yaw instability issues with two tailfins placed up high like that, and so far forward, if they don't turn off their roll authority.

Is there any way you can slide them farther to the rear?  The farther back you can make them, the more effective they will be (both in stabilizing you, and in giving you yaw authority).

Also, do you really need two of them?  I assume that for tight turns you'll be doing roll-then-pitch-up, rather than trying to yaw, right?  Dual-tailfin planes have been known to have control problems in KSP.  What if you have a single tailfin in the center, and then move those airbrakes out to the sides?

If you're looking for maneuverability (and more roll authority), is it worth considering putting some canards on the front?

If you can use the translate widget to slide the tailfin (or fins) a little bit farther in the ventral direction (i.e. toward the SPH floor, so they don't stick up so high above the plane), you'll probably have better luck with your stability-- you'll keep the same yaw authority while getting less roll interference.

[Nich]

Yes you have way too much Yaw authority.  In addition you have pitch and roll going to your elevon 3's.  This will do some strange things when you are trying to roll and pitch at the same time.  Namely if your pulling up and rolling left the craft will yaw right because the right side has more drag (2 up vs 1 up 1 down on the left)  This is further confounded by the fact that your yaw stability has a very small yaw arm but a large roll arm and the yaw directly cancels your roll.  Are you using the weasly to reverse thrust?  It is a lot of dead weight above mach 1.  Sepratrons, veneers, or parachutes might be better.  Also setting the landing gear so the wing creates down force will make brakes more effective.  Finally circular intakes are subsonic intakes I do not believe they work well about mach 1.

Here is an example.  The precoolers provided me with more then enough air.  Landing distance was very short Tops speed was respectable.  It probably had too much pitch authority as I would lose a lot of speed in a 15g 15 degree AOA bank.  Roll was better then anything I had flown in a LONG time lol.  It was a little unstable at high speed flight so I would probably change the yaw surfaces to pure vertical as it doesn't need any more pitch control.  Any yes the third pic I hit the 90 at 45m/s and stopped where you see me.  I am guessing less then 50m.  Make sure you turn off the brakes on the front wheel and get the backs close to CG.

 

Edited March 9, 2016 by Nich

[GDJ]

I pretty much do the same thing as the others, and I also use the same parts, but set them up a bit differently as far as layout:

Here's my plane (inspired from a Northrup T-38). You can see where I adjusted the COM, COL, and COT. The COL is a bit lower than the COM, but only by a small margin, and it's behind the COM for stability.

From the rear you can see that the COT is dead entered with the COM. I find that works very well with light aircraft that have plenty of wing surfaces to lift the plane. Larger heavier planes can have the COT below the COM to help keep the nose up. Some aircraft like the Fairchild A-10 "Warthog" have the engines up high, but the thrust line points downwards to help with pitch control. Otherwise that plane would want to pitch down constantly.

Here is the front wing Elevon settings. The inner larger ones only handle roll. The outer Elevon's do nothing, but they can be adjusted to alter the wing characteristics. Having the large Elevon's handle the roll commands still allow the plane to have plenty of roll authority, but it's dampened by the wings mass. Easier to handle this way. No pitch authority on the front wings here.

The rear stabilizers only handle the pitch authority. Since these Elevon's are far away from the COM and COL, they can apply alot of force with a smaller Elevon. Less drag, still easy to handle. Having pitch authority on the front wings will be largely ineffective on this plane because the Elevon's would be trying to lift the entire plane vertically and adding alot of unwanted drag. This is useful for take-offs and landings where the airspeed is quite low, but for normal flight it's better to not have this option selected.



The Vertical Stabilizer has a single large Elevon to handle yaw, and it handles yaw only. This Elevon is large despite being at the rear of the plane. There's only one, so the Elevon has to be large to match the two smaller Elevon 2's in the back. This has alot of yaw authority, but it snaps back to neutral fast and easy. It will cause the fuselage to roll a bit under use because it's a non-symmetrical layout in terms of lift surfaces, but the front wings can correct that easily.

This plane will fly dead level at 10000 metres at 670 m/s with this layout. It's powered by a pair of modified J-20 Juno engines with a output of 70 kN each. The plane weighs in as equipped at 4.9 tonnes. Light, quick, fast, and 15+G turns are not a problem.

Edited March 9, 2016 by GDJ

[pincushionman]

For your landing problems, I think you have a lot of wing, a lot of engine, and a lot of speed, and you need far less of all three. 80 m/s is darn near 180 mph. How slow can you go before you stall or lose control? You want a low stall speed, but once you get down you want to stick to the ground, not float.

[GDJ]

12 minutes ago, pincushionman said:

For your landing problems, I think you have a lot of wing, a lot of engine, and a lot of speed, and you need far less of all three. 80 m/s is darn near 180 mph. How slow can you go before you stall or lose control? You want a low stall speed, but once you get down you want to stick to the ground, not float.

60 m/s would be better, but 80 is not totally unrealistic. I've landed faster, but it's certainly touchier at higher speeds. Air Brakes and/or drogue chutes may help in this case. Once he touches ground and has the brakes applied he'll slow down fairly fast and stick, especially if the nose of the plane is a bit pitched down (one or two degrees) once all the wheels are down.

[RocketSquid]

So, I will add canards, and will try offsetting the tailfins to improve maneuverablility. Meanwhile, I'll add some drogues for landing.

[Val]

18 hours ago, RocketSquid said:

Okay, finally got the screenshot to upload.

Also, the elevons are Elevon 3's, as you can probably tell.

 

I recreated this craft, except with a different tailfin arrangement.

The reason it's so hard to land is because CoL is too far behind CoM. It has to go really fast to and give full pitch up to keep from lawn darting into the ground.