Need Help with Autogyros

[Geschosskopf]

I've always wanted to make an autogyro in KSP and I've recently seen pics from folks who have done so successfully.  But I'm running into some serious fundamental issues with free-spinning rotors that I'd like some help with.

The problem I have is that the free-spinning rotors have no angular momentum.  Thus, when the direction of the airflow changes relative to the blades, the blade speed changes rapidly, including instantly switching direction, which is bad ;).  The rotor disk has to be pitched with its leading edge slightly down relative to the airflow to make the rotors spin the correct direction at the proper speed.  But maneuvering the autogyro causes this angle to change.  How do you get around this problem?

For instance, my autogyro has the rotor plane tilted down towards the front of the fuselage about 15^.  The autogyro rolls down the runway and the rotors start spinning.  At about 40m/s, the rotors make enough lift for flight but won't pull the thing off the ground, so I have to pitch up to take off.  This reduces the angle of the rotor disk to the airstream, which causes the rotors instantly to slow down, reducing lift significantly, and the thing falls back to the runway.  Pitching up more than 15^ in fact causes rotor reversal.  If I stay on the ground until 90m/s, the autogyro as a whole has enough momentum to get enough altitude that it can level off (which puts the rotor disk back in the preferred 15^ down pitch) before losing all lift.  And then I can fly along at a very low altitude in a straight line.  If I try to climb, it's the same problem with pitch angle as taking off.  If I try to turn, the rotation of the autogyro changes the orientation of the rotor disk to the airstream and again slow, stops, or reverses the blades, or creates asymmetric lift and a nasty roll.

So how do successful autogyros avoid these issues?  Thanks.

[RocketSquid]

Now, take all of this with a grain of salt, seeing as KSP aerodynamics aren’t totally accurate, but real autogyros tend to have their rotor planes tilted back, with the blades pitched only a couple degrees from the rotor plane. The air is meant to go up through the blade as with a wing, rather than down as with a helicopter blade. In most designs I’ve seen, the backwards tilt of the rotor plane is mostly accomplished by the front of the autogyro being higher than the back. From what I can tell, one end of the blade (the one moving upwards and into the wind) is driven and provides more lift, and the other (the one moving downwards and with the wind) is the driving one, since it’s stalled and thus more affected by drag than by lift. 

[Geschosskopf]

3 hours ago, RocketSquid said:

Now, take all of this with a grain of salt, seeing as KSP aerodynamics aren’t totally accurate, but real autogyros tend to have their rotor planes tilted back, with the blades pitched only a couple degrees from the rotor plane.

Quite so.  but airflow from below has only made my rotors spin backwards, which forces the thing into the ground.  I'll tweak around with it some more but I'm not entirely hopeful

[bewing]

On an autogyro, the rotors are supposed to spin backwards. But because of drag, they spin backwards slower than the vertical updraft of air through them -- generating a net upward force and a force that accelerates their forward rotation.

 

 

Edited June 4, 2019 by bewing

[Geschosskopf]

51 minutes ago, bewing said:

On an autogyro, the rotors are supposed to spin backwards. But because of drag, they spin backwards slower than the vertical updraft of air through them -- generating a net upward force and a force that accelerates their forward rotation.

I understand this is how things work in real life.  But the KSP atmosphere is something different. In KSP, all wing surfaces are symmetrical so must have a positive angle of attack to generate lift.  But blades only auto-rotate with a positive angle of attack if the relative airflow is from slightly above.  If the relative airflow is from below, the blades immediately reverse direction and start rotating with a negative angle of attack, causing downforce.   That's the problem I'm trying to overcome.

 

 

[Geschosskopf]

Well, after a huge amount of test pilot fatalities, learning about real autogyros somewhat, and of course the good advice of @RocketSquid and @bewing I finally got working autogyros.  Thanks for the help!  In the interests of posterity, here are some of the important things I learned...

1.  Rotor Mast Tilt and Blade Pitch Angles

Yes, the mast must be canted backwards slightly.  I found an angle of 10^ worked best.  But how to solve this causing the blades to spin backwards and creating negative lift?  By counter-intuitively giving the blades a -5^ (nose-down) pitch.  Then they move through the air forwards and create positive lift to fly with.  This was the critical break-through to get things into the air and able, if controllable, to stay there.  This is when I started killing test pilots, as now they could get high enough to crash badly :).

2.  Controls

Real autogyros tilt the rotor disk for pitch and roll control so I tried that, using stacked hinges under the rotors and tied to the control axes.  This MIGHT possibly work if you use a joystick but I use the keyboard and it didn't work at all.  Part of the problem is that for some reason, the axis controls aren't self-centering with the robot parts like they are for control surfaces.  Instead, they stay where you left them when you release the key.  This made control response way too slow, and trying to remember how many taps I'd used in each direction too complicated, for controllable flight with the keyboard.  Also, the "medium" in-line hinges (the biggest currently available) aren't strong enough for the aerodynamic forces, which bodes ill for trying this with a joystick.  Thus, after many crashes, I gave up on this and equipped the autogyro with lots of reaction wheels and aerodynamic control surfaces.  Ultralights don't need ailerons (torque alone suffices if you have 2 reaction wheels) but anything bigger needs ailerons.

3.  Rotor Design

Real autogyros have a single rotor with 2 blades, and these blades are on free-moving teeter hinges.  The idea is, the advancing blade will tilt up on its hinge to balance its lift with the retreating blade.  So I tried this but it doesn't work in KSP.  Even with stupidly large amounts of hinge travel, the lift would never balance and the craft would leave the ground, immediately roll over, and crash.  So, I went with the blades rigidly attached to the hub.

However, having a single rotor left the fatal lift imbalance in place, so I countered this by having stacked, co-axial, counter-rotating rotors.  As both are unpowered, the only difference is that the blades face in opposite directions.  Then I experimented with the size and number of blades.  Because there's nothing to force the blades to always spin at the same speed, the relative positions of the blades will vary during flight.  The more blades you have, the more any such lift imbalance averages out.  HOWEVER, more blades make more lift, which limits forward speed before you get an uncontrollable pitch-up.  So in the end, I settled for 3 blades per rotor as the best compromise.  YMMV.

NOTE:  the rotor blades are made of control surfaces.  Be sure to disable ALL their control axes.  Bad things happen if you don't :).  For larger autogyros, using wing parts for rotors might work, in which case you wouldn't have to worry about this.

4.  Horizontal Stabilizier Downforce

Most real modern autogyros have a fixed horizontal stabilizer that creates downforce on the tail.  This is believed to cure  a fatal problem various tail-less designs had/have.  I found this feature absolutely essential for maintaining basic stability.  Autogyros in KSP have a strong tendency to pitch up uncontrollably even at moderate speeds without it.  So, put a 5^ up angle (so it's like down elevator) on your horizontal tail surface, whether it's fixed or flying.

5.  Separation of CoM and CoL

Autogyros require the CoL (rotor hub) to be noticeably farther behind the CoM than conventional airplanes.  This is also to keep the nose from pitching up at low and moderate speeds.  Apparently the CoL marker shown in the editor doesn't take rotor rotation into account.  The actual CoL seems to be ahead of the marker and gets farther ahead as forward speed (and thus, lift) increases, eventually getting in front of the CoM.  So you have to put the rotor back a ways from the CoM to give yourself a decent range of speed before uncontrollable pitch-up occurs.

Because of this moving CoL, all autogyros with a fixed rotor disk angle have a maximum speed beyond which their controls can no longer keep the nose down.  This means you don 't need (or can't use) as much engine as you might want.  It MIGHT be possible to put the rotor mast on a robotic part that would allow trimming it for higher speed by leaning it slightly forward from its normal position of 10^ back.  But as mentioned above, the inline hinges aren't up to the aerodynamic forces so I'm not sure this would work.

6.  Miscellaneous Stuff

• In level-ish flight, you need to be gentle with the pitch inputs.  Just like with real authogyros, rapid pitch changes does strange things to the rotor's speed and thus your lift.  Making turns of up to about 45^ are best done with a gentle bank, lots of rudder, and elevator used to maintain nose attitude.  However, you can turn very rapidly and tightly by banking 80-90^ and yanking the thing around with lots of elevator, using the rudder gently to maintain nose attitude.
• Steep, power-off descents up to about 45^ dive angle work fine.  It's almost like parachuting.  You gain little if any speed, especially if you spiral down using the rudder to increase your side area to the wind for added drag.  Just be careful not to get TOO slow doing this or you won't be able to pull out.
• Turn off the rotor brakes.  Applying the wheel brakes also brakes the rotors and their sudden stoppage imparts a lot of torque to the craft, which can roll it over.

7.  Some Results:

Here is the Autogyro Mk1 Mod 37, the initial test bed which learned all the bloody lessons above.  It started out with a fixed rotor tilted forward and didn't fly much if at all.  Then it got hinges on the mast and the blades, which were eliminated over time, and had numerous rotor and control surface configurations.  In this form, it ultimately reached a point where it could be flown, although it was squirrelly and required constant hands-on flying to go approximately in the desired direction.  Landing was only rarely successful--usually, 1 or more parts would hit the ground and break, and landing on the desired point was quite difficult.

So once I reached this point, I started over with something designed with the proper arrangement of parts from the start, and also downsized a bit as I really wanted something smaller.  That led to the Mk2, which flew beautifully from the start.  But it had the smaller, spherical fuel tanks which provided insufficient range for a round trip to Airfield Island.  Thus, the only change with the Mk2 Mod 1 was the larger fuel tanks.

This thing is actually a joy to fly although it flies slightly less well with 2 Kerbals compared to 1.  Slower and not QUITE as laterally stable, which is strange given the drag is symmetrical with 2.  But it's still good enough to upload.  Interested parties may find it over in the Spacecraft Exchange, along with all its vital stats and pilot's notes.

Thanks again for the help!

 

 

Edited June 5, 2019 by Geschosskopf

[RocketSquid]

Glad I could be a help. My autogyros are still plagued with all manner of problems, unfortunately. They keep flipping. 

[Geschosskopf]

10 hours ago, RocketSquid said:

Glad I could be a help. My autogyros are still plagued with all manner of problems, unfortunately. They keep flipping. 

Do they flip about the pitch or roll axis?  In my experience, solving these problems goes like this:

Roll Axis:

• Cause:  Asymmetric lift, which itself can have several causes.
• This always happens with a single rotor.  So if you have only 1, add a 2nd that turns in the opposite direction.  Be sure it's otherwise an exact copy of the 1st (same blades, same hub) as the different rotary parts seem to free-wheel at different speeds.
• If this is happening with counter-rotating rotors,  the problem is likely caused by the rotors going at different speeds due to changes in airflow from maneuvering.  If the autogyro is constantly twitching around the roll axis in straight-ish flight at low-moderate speeds before flipping out, you probably need more rotor blades to balance out the lift more evenly.  But adding blades increases lift, which will require adjusting the issues relating to pitch instability.  If this happens in a turn, then probably either the rotor mast isn't angled back enough (this decreases the sensitivity to airflow changes) or you let the nose rise or fall too much too fast during the turn.  Even with co-ax rotors, 1 will feel the effects more than the other because of their difference in height above the autogyro's CoM.  Anyway, 1 rotor will get enough slower than the other (or even stop or reverse direction instantly in extreme cases), leaving the other creating way unbalanced lift as if you had only 1 rotor.

Pitch Axis:

• This comes from exceeding the autogyro's speed limit.  If you go fast enough, the CoL will move in front of the CoM and your nose will yank up very quickly, causing a radical change in airflow direction and doing bad things to your lift, which then often causes the roll issues noted above and you just tumble out of the sky.  So, if this is happening well below 100m/s you probably need to increase the autogyro's speed limit.  If this is happening above 100m/s (= 194 knots), then you have too much power because autogyros shouldn't be going that fast anyway 
• To increase an autogyro's speed limit, you can do 3 things:
  • Move the rotor mast farther behind the CoM
  • Reduce the blade area, either by reducing the number of blades or using smaller blades.  But this reduces the lift and thus the size/payload.
  • If you don't have one already, add a horizontal stabilizer providing 5^ up force (down elevator) on the tail.  If you already have such a stabilizer, make it bigger or move it farther aft.
• If you have too much power, either switch to a smaller engine or limit the thrust of the engine you have.

Edited June 6, 2019 by Geschosskopf

[RocketSquid]

19 minutes ago, Geschosskopf said:

Do they flip about the pitch or roll axis?  In my experience, solving these problems goes like this:

Roll Axis:

• Cause:  Asymmetric lift, which itself can have several causes.
• This always happens with a single rotor.  So if you have only 1, add a 2nd that turns in the opposite direction.  Be sure it's otherwise an exact copy of the 1st (same blades, same hub) as the different rotary parts seem to free-wheel at different speeds.
• If this is happening with counter-rotating rotors,  the problem is likely caused by the rotors going at different speeds due to changes in airflow from maneuvering.  If the autogyro is constantly twitching around the roll axis in straight-ish flight at low-moderate speeds before flipping out, you probably need more rotor blades to balance out the lift more evenly.  But adding blades increases lift, which will require adjusting the issues relating to pitch instability.  If this happens in a turn, then probably either the rotor mast isn't angled back enough (this decreases the sensitivity to airflow changes) or you let the nose rise or fall too much too fast during the turn.  Even with co-ax rotors, 1 will feel the effects more than the other because of their difference in height above the autogyro's CoM.  Anyway, 1 rotor will get enough slower than the other (or even stop or reverse direction instantly in extreme cases), leaving the other creating way unbalanced lift as if you had only 1 rotor.

Pitch Axis:

• This comes from exceeding the autogyro's speed limit.  If you go fast enough, the CoL will move in front of the CoM and your nose will yank up very quickly, causing a radical change in airflow direction and doing bad things to your lift, which then often causes the roll issues noted above and you just tumble out of the sky.  So, if this is happening well below 100m/s you probably need to increase the autogyro's speed limit.  If this is happening above 100m/s (= 194 knots), then you have too much power because autogyros shouldn't be going that fast anyway 
• To increase an autogyro's speed limit, you can do 3 things:
  • Move the rotor mast farther behind the CoM
  • Reduce the blade area, either by reducing the number of blades or using smaller blades.  But this reduces the lift and thus the size/payload.
  • If you don't have one already, add a horizontal stabilizer providing 5^ up force (down elevator) on the tail.  If you already have such a stabilizer, make it bigger or move it farther aft.
• If you have too much power, either switch to a smaller engine or limit the thrust of the engine you have.

I think I’m just going to have to limit the thrust. One Wheesley is too much thrust, but a Juno on either side isn’t enough. The two-Juno design tends to take off, fly for a bit, then land (quite safely most of the time, it has that going for it) and roll on the ground until it has time for another takeoff.

[Geschosskopf]

2 hours ago, RocketSquid said:

I think I’m just going to have to limit the thrust. One Wheesley is too much thrust, but a Juno on either side isn’t enough. The two-Juno design tends to take off, fly for a bit, then land (quite safely most of the time, it has that going for it) and roll on the ground until it has time for another takeoff.

My Mk1 testbed above had 1 Wheesley and that was too much, also.  It would fly (in its hard-to-control manner) at up to about 1/2 throttle and 90m/s, but would pitch up uncontrollably with more power.  Rather than bother limiting the thrust of the part, I just put a post-it note on the dashboard saying not to exceed 50% throttle ;).  That said, the Wheesley has the advantage of a thrust reverser.  This is a huge help bleeding off airspeed in a short distance, facilitating short landings.  That's why I kept that engine on the thing.

[RocketSquid]

41 minutes ago, Geschosskopf said:

My Mk1 testbed above had 1 Wheesley and that was too much, also.  It would fly (in its hard-to-control manner) at up to about 1/2 throttle and 90m/s, but would pitch up uncontrollably with more power.  Rather than bother limiting the thrust of the part, I just put a post-it note on the dashboard saying not to exceed 50% throttle ;).  That said, the Wheesley has the advantage of a thrust reverser.  This is a huge help bleeding off airspeed in a short distance, facilitating short landings.  That's why I kept that engine on the thing.

Next mission: build an autogyro so big it uses a Goliath