Board index FlightGear Support Flying

Taildraggers  Topic is solved

Controlling your aircraft, using the autopilot etc.

Re: Taildraggers

Postby D-ECHO » Wed Aug 04, 2021 6:43 pm

Hi dilibert,

I've revisited the Cessna 140, adding references to the values used in the FDM and improving it a bit. But even though I changed the empty mass CG a bit now in accordance with the flight manual, it was already in the correct range before. Thus, while moving the CG to the nose will definitely improve the described problem, it's not a fix as it makes the aircraft more unrealistic (and also worsens the flight model in many other critical areas like pitch tendency, stability in the air, stall behaviour).

EDIT: You were faster dilibert ;)
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby D-ECHO » Wed Aug 04, 2021 6:45 pm

In addition to the above, as I can perfectly control the Cessna 140 on takeoff here, I'm pretty sure the issue in this case is your input way, i.e. only "full brake" or "no brake" when using the keyboard and also the lack of rudder pedals.

I'm sorry I can't offer you a better help, but you wouldn't want to control a real aircraft with this binary approach. In the same way that your suggestion of making the brake keys apply the brakes gradually depending on "press duration", a better and more direct way of controlling the rudder could imo fix your problems here.
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby dilbert » Wed Aug 04, 2021 8:40 pm

Thanks, DECHO. Sticking with the http://mirrors.ibiblio.org/flightgear/f ... raft-2020/ version of your C140 for now. It's a little touchier on the runway than the real bird, but manageable.
Moving CG forward in it would also encourage nose ups, so I've abandoned that approach.

Regarding the C18: I've been flying three versions: that of http://mirrors.ibiblio.org/flightgear/f ... raft-2020/, that of https://sourceforge.net/p/flightgear/fg ... /Aircraft/, and
the one I modified.

Barringer's https://sourceforge.net/p/flightgear/fg ... /Aircraft/ version works the best of the three, so I'm sticking with it and eliminating the one I modified.

It's still a bit touchy, but flyable.

Don't use keyboard. Use either a logitech Extreme 3D Pro, or a Saitech Mad Katz V1 stick. Rudder response does seem too lackadaisical in responding to unanticipated horizontal rotation. Perhaps,
my I3 machines might be the reason :?:

Thanks again and kind regards.
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby wlbragg » Thu Aug 05, 2021 5:07 pm

Absolutely if you have any issue with frame rate it's going to compound it. I know on the Cub pressing the tail wheel down with the elevator helps considerably, but it too is a balancing act.
Kansas and Ohio/Midwest scenery development.
KEQA, 3AU, KRCP Airport Layout
Intel i7/GeForce RTX 2070/Max-Q
User avatar
wlbragg
 
Posts: 7588
Joined: Sun Aug 26, 2012 12:31 am
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/RTX 2070

Re: Taildraggers

Postby dilbert » Thu Aug 05, 2021 6:15 pm

Have done multiple tests on the C18. I'm flying the trunk version.

It would travel a ways down the runway and at about 40 knots, or so, make a radical, rapid, 90 degree turn and leave the runway.

After start, I then set the autopilot Heading Bug to runway heading and clicked Heading Control; after which I set Altitude Control at 1000, and clicked Pitch/Altitude Control. I then added full throttle and disconnected the stick from the computer. The plane meandered left and right, but the autopilot maintained the runway.

Thereafter, I did the same, but left the stick plugged in, but without using it, with identical result.

I ran the test again, this time keeping Heading Control engaged, but with Pitch/Altitude Control unchecked. On this run the plane meandered, keeping the runway, as before; but at 48 KNOTS it made a radical 90 turn to the right, which the autopilot heading control was unable to arrest.

It seemed that the problem might be happening when the tailwheel left the ground. So I ran multiple tests, watching the tailwheel from the left rear in chase view. The wheel remained compressed from ground contact right up and until and after the radical turn, which pretty well ruled out tailwheel lift as a trigger.



In summary, I think:

It's not caused by pilot overcontrol.
It's not caused by the stick
It's not caused by normal meandering on the runway, which the auto pilot has no trouble correcting.
It can happen left or right; and seems to have no relation to prop effect.
Once it occurs, the autopilot is helpless in arresting it.
It does not appear to be triggered by tailwheel lift.
With Heading Control alone engaged in the autopilot, it happens consistently at 48 KNOTS.
It normally takes about 20 seconds into the run, but reducing throttle increases the time reaching 48 KNOTS, when it seems to happen.
With both Heading Control and Pitch/Altitude Control engaged, it doesn't happen.

What triggers it remains a mystery.

Kind Regards
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby D-ECHO » Thu Aug 05, 2021 8:34 pm

Could you please look up the current wind when your issue happens. To do this, use the property browser ( press "/" ), navigate to /environment/ and scroll down to wind-from-heading-deg and most importantly wind-speed-kt.

When testing with the Beechcraft C18 S (YASim), I don't encounter any such issue.
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby dilbert » Thu Aug 05, 2021 9:58 pm

Currently am on runway 7 at KPNA with Wind-from-east-fps= 5.1905--. Wind-from-heading-deg='160.164212 (double). Don't quite understand the'160.16--, but that's the reading.

Set autopilot Heading to 70 and clicked Heading Control. Added power and released brake. Plane meandered left, then auto pilot corrected it to the right, and at about same point as before,
plane went hard right and left the runway.
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby dilbert » Thu Aug 05, 2021 11:36 pm

In the past half hour I tested the PT22, the CAB G.Y> 30,the Arsenal vg33, the DH1-c, and Vivian Meaza's Spitfire. They all did pretty much the same thing, usually getting wingtips in the bargain.
I also tested the PT17 and the L19. Both made similar radical departures, but the autopilot was eventually able to steer them back.
So the problem seems other than plane related.

I tested the C140 last. It veered left, corrected back right; then radically spun right, ending up180 degrees to its original heading, facing the runway threshold. :(
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby D-ECHO » Fri Aug 06, 2021 7:05 am

wind-from-east-fps is only the eastern wind component, so total wind speed would be very helpful.

I'm also wondering how the autopilot is able to keep the heading, as far as I know it only uses the ailerons to compensate, or do you have Auto Coordination enabled?
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby dilbert » Fri Aug 06, 2021 11:54 am

Wind speed was 5 knots

More information:

I ran the exact same test this morning using my A8 laptop and Ubuntu64
I used the C18
It did the same thing: spun out.
Autopilot stopped spinout at about 90 degrees and gradually returned heading to correct direction.
Then the program crashed back to the Terminal.
Last message on Terminal: 69.92[ALRT}:flight This aircraft uses version-'YASIM_VERSION_CURRENT'(3)
and beneath that: YASim warning: possible convergence problem

At the moment, still have the terminal displayed, if there's a question about earlier entries--
Kind Regards
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby D-ECHO » Fri Aug 06, 2021 12:16 pm

Are you absolutely sure that the wind speed was 5 kts? I'm not trying to bug you, but if the eastern wind component is about 5 kts and the main wind direction is from 160 deg, then the total wind has to be well above 5 kts, also being crosswind ( runway heading 070, wind direction 160 ) which could cause the problems.

If I didn't miss it, could you please say whether you use auto coordination?

> This aircraft uses version-'YASIM_VERSION_CURRENT'(3)
is nothing to worry about, whereas

> YASim warning: possible convergence problem
as far as I remember (it's been some time that I used YASim) suggessts that the FDM is not well tuned and that the "solution" (i.e. flight behaviour) YASim generates will not be realistic.
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby wkitty42 » Fri Aug 06, 2021 12:49 pm

D-ECHO wrote in Fri Aug 06, 2021 12:16 pm:> YASim warning: possible convergence problem
as far as I remember (it's been some time that I used YASim) suggessts that the FDM is not well tuned and that the "solution" (i.e. flight behaviour) YASim generates will not be realistic.

this
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9148
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: Taildraggers

Postby dilbert » Fri Aug 06, 2021 1:23 pm

Current readings:
wind-from-down-fps='0.0034239 (double)
wind-from-east-fps='7.3191--- (double)
wind-from-heading-deg'120.0292812 (double)
wind-from-north-fps='-4.2306---(double)
wind-speed-kt='3.00878-- (double)

These aren't much different from yesterday, other than wind-from-heading-deg, which changed from 160 to 120; and they probably haven't changed much since my run earlier this morning.

Just flew again: wind speed is 3 kts.
Ran C18 on Windows. Plane didn't wander much starting down the runway; but it suddenly made about a 90 turn to the left, a ways into the run...

Have Dentist appointment shortly, but will be able to test more later. :)

kind regards
Last edited by dilbert on Fri Aug 06, 2021 1:45 pm, edited 1 time in total.
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

Re: Taildraggers

Postby D-ECHO » Fri Aug 06, 2021 1:34 pm

Ah sorry I missed that the wind-from-X is given in fps, not kts. Now it makes sense :)

Hope your dentist appointment goes well :D
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: Taildraggers

Postby dilbert » Sat Aug 07, 2021 12:58 pm

Have tested 11 additional tail dragger models.

At this point I think the following:

1 There is a general, programmed ground looping routine in YASIM.

2 Rate of turn is determined by aircraft weight and distance of CG from front wheels.

3 Initiation is a condition of individual aircraft speed (48 for the C18, 30 for the Jenny, 75 for the P40, etc.) .

4. In my opinion. It's somewhat overdone and unrealistic for most planes.

There are a few planes where dealing with it has been relatively successful, the L19 and the Arsenal VG33 being good examples.

With some planes, the tendency has been reduced by moving the CG close to the main wheels, but at the expense of nose ups, the Percival Mew Gull being a good example.

With the De Havilland DH-89 the tendency has been all but eliminated, but only at the expense of its having exceptional flight stability.

I attribute the good behavior of the L19 and the Arsenal VG33 to two things:

1.The CGs in both are only about 1.5 behind the wheels, as compared to 3 in the C18. (Enough to reduce the turning moment; but not enough to easily cause nose ups.)

2.The ground loop routine is triggered to initiate at relatively high (and fairly close to flying) speed for each (50 for the L19 and 104 for the Arsenal), which gives their vertical fins and rudders better ability to resist it.

Kind Regards :)
Last edited by dilbert on Sat Aug 07, 2021 2:38 pm, edited 1 time in total.
dilbert
 
Posts: 774
Joined: Wed Jan 07, 2015 9:36 pm
Location: KJKA, Alabama, USA
Callsign: db1
Version: 220.3.19
OS: W10,Ubuntu64

PreviousNext

Return to Flying

Who is online

Users browsing this forum: No registered users and 6 guests