Board index FlightGear Development Aircraft Flight dynamics model

landing gear toe-in

Good sims require good FDMs (the "thing" that makes an aircraft behave like an aircraft).

landing gear toe-in

Postby OO ZVY » Tue Apr 06, 2021 5:06 am

A lot of FG tailwheel aircraft are not easy ( up to almost impossible ) to control on the ground.
Solution may be like for real airplanes ( I haven't tested yet ) main gear toe-in. ( out or maybe in ? )
What do you think ?
OOZVY

Image
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby danielHL » Tue Apr 06, 2021 1:20 pm

Interesting idea. I don't know if the physical models behind the gear elements in yasim and jsbsim are capable of simulating the resulting effects. JSBSim seems to have gear orientation tags, yasim doesn't (yet).
danielHL / D-FMPW
danielHL
 
Posts: 277
Joined: Fri May 02, 2014 6:23 pm
Callsign: D-FMPW
Version: next
OS: Linux

Re: landing gear toe-in

Postby danielHL » Tue Apr 06, 2021 1:41 pm

On second thought: you could rig a constant value into a control-input type "steer" to a yasim gear to simulate toe like a constant steering-command on the main gear (separate for left and right side, obviously). But I still lack the knowledge if this would do any good in terms of physics.
danielHL / D-FMPW
danielHL
 
Posts: 277
Joined: Fri May 02, 2014 6:23 pm
Callsign: D-FMPW
Version: next
OS: Linux

Re: landing gear toe-in

Postby danielHL » Tue Apr 06, 2021 4:19 pm

On a third note: I modified the DC3 (yasim) with a 0.1 antisymmetric steering input on the main gears. I don't know yet if this has any noticeable effect. Do you have any idea how to evaluate this? Do you have a specific aircraft and situation in mind where it would help?

Anyway, here is the code:
Code: Select all
diff --git a/Systems/dc-3-base.xml b/Systems/dc-3-base.xml
index 0eedad9..2ca5c7c 100644
--- a/Systems/dc-3-base.xml
+++ b/Systems/dc-3-base.xml
@@ -141,6 +141,11 @@
       <max-lb>2400</max-lb>
       <min-lb>0</min-lb>
     </weight>
+   
+    <gears>
+        <toe-left>-0.1</toe-left>
+        <toe-right>0.1</toe-right>
+    </gears>
 
     <!--
     <weight n="6">
diff --git a/Systems/dc-3-yasim.xml b/Systems/dc-3-yasim.xml
index fd0591c..de2b0af 100644
--- a/Systems/dc-3-yasim.xml
+++ b/Systems/dc-3-yasim.xml
@@ -364,6 +364,7 @@
         <control-input axis="/controls/gear/brake-left" control="BRAKE"/>
         <control-input axis="/controls/gear/brake-parking" control="BRAKE"/>
         <control-input axis="/controls/gear/gear-down" control="EXTEND"/>
+        <control-input axis="/sim/gears/toe-left" control="STEER" />
         <control-speed control="EXTEND" transition-time="12"/>
         <control-output control="EXTEND" prop="/gear/gear[0]/position-norm"/>
   </gear>
@@ -378,6 +379,7 @@
         <control-input axis="/controls/gear/brake-right" control="BRAKE"/>
         <control-input axis="/controls/gear/brake-parking" control="BRAKE"/>
         <control-input axis="/controls/gear/gear-down" control="EXTEND"/>
+        <control-input axis="/sim/gears/toe-right" control="STEER" />
         <control-speed control="EXTEND" transition-time="12"/>
         <control-output control="EXTEND" prop="/gear/gear[1]/position-norm"/>
   </gear>
danielHL / D-FMPW
danielHL
 
Posts: 277
Joined: Fri May 02, 2014 6:23 pm
Callsign: D-FMPW
Version: next
OS: Linux

Re: landing gear toe-in

Postby Ysop » Tue Apr 06, 2021 5:06 pm

yes, interesting! But...

- does this change the inherent stability problem, as centre of gravity is behind centre of effort?
- does this improve controllability for tail on ground and tail in air?

Also this can only be done to small single digit degrees, otherwise other problems like tire wear and brake effect.

Toe out might be more logic. When on the edge of control on an outside wheel it might be beneficial, when this point of contact moves further outside.
But when the ground loop monster is attacking, you need more than this. Moments are just too large.

Not sure, how this is done in real aircraft. If it would really improve things, everyone would have it. Maybe something for long discussions and you can feel an effect, if you really believe it.
User avatar
Ysop
 
Posts: 416
Joined: Thu Oct 25, 2018 9:06 pm
Version: 2020.3.1
OS: ubuntu 20.04

Re: landing gear toe-in

Postby OO ZVY » Wed Apr 07, 2021 4:39 am

Good morning,

Image

With toe-out, the airplane is caused to move
in such a direction that the tendency is to minimize the
centrifugal force applied to the CG by a swerve, and the
reason is that the left hand wheel has the greater load and
pulls away from the incipient swerve to the right. The
tendency for this layout is to pull the plane back to a
straight, stable course.

About one or two degrees ought
to represent a good compromise, for too much toe-out
would, in spite of affording a very stable landing roll, in-
troduce the disadvantages of excess tire wear and slight
drag on the take-off run.

see : https://pdfhall.com/landing-gears-toe-i ... b460c.html

I was thinking a solution like danielHL is proposing. does anyone already tried ?

Regarding real planes, If I am right this technique was used for the BF 109.

and more to think : toe in-out should changes with aircraft attitude during TO roll due to LG camber.

ZVY
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby Ysop » Wed Apr 07, 2021 6:44 pm

Thanks, now I get the idea behind it.

Maybe overthinking it: in the picture above with the aircraft moving forward at a slight beta, shouldn't the right wheel have more resistance and with that increase the yaw?

Not sure, if such has been implemented. And if it really has a noticeable effect. Control problem often come down to expecting taildraggers to be easy, lack of peripheral vision, lack of motion feedback and joysticks, which are not representing the movement needed in a real aircraft.

Which taildraggers do you find "almost impossible to control"?
User avatar
Ysop
 
Posts: 416
Joined: Thu Oct 25, 2018 9:06 pm
Version: 2020.3.1
OS: ubuntu 20.04

Re: landing gear toe-in

Postby OO ZVY » Wed Apr 07, 2021 8:06 pm

good evening,

In the picture, the plane experience a right turn so due to weight shift the weight is mainly on the left wheel. ( momentum due to centrifugal force and position of aircraft CG above the wheels )
For the same ( opposite ) reason, the weight on the right wheel is reduced ( the more yaw the less weight ) .
If we estimate a proportional relation between weight and friction, the friction on the right wheels is reduced while the one on the left increase.
Assuming extreme yaw, all weight will be on the left wheels with no weight on the right. ( no weight = no friction = no effect form the right wheel )
The toe-out induce a straight path at the picture angle (or may even generate a left yaw if more toe out ) to counter the right turn and stabilise the path. .

Regarding expectation of taildraggers to be 'easy', I have personal experience flying some real tailwheel plane. If more challenging than tricycle gear, they are not that unstable in smooth air and hopefully remains controllable.

Also, note that during take off roll, when lifting the tail, main gear camber and caster will induce a toe in-out change. This may explain why some FG aircraft are stable during the initial part of the take off and suddenly change to instability.

What do you think ?

:-)

ZVY
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby Ysop » Thu Apr 08, 2021 7:00 pm

Always nice to have input from pilots here, so very welcome :-)

My experience with a taildragger is taxing and flying, but not in the intermediate stages.

Just as an assessment, toe-out will not do the magic trick, but certainly there are improvements possible for taildraggers.
Of course it would be possible to introduce artificial stabilizing moments, but this approach is to be taken with care.

Maybe the rolling friction as function of weight on each wheel is not correctly represented, as there was a similar issue.
Also it should be easier to land on grass than on pavement.
To be investigated.

In general we have discussed quite a bit in the Piper J3 development thread.
User avatar
Ysop
 
Posts: 416
Joined: Thu Oct 25, 2018 9:06 pm
Version: 2020.3.1
OS: ubuntu 20.04

Re: landing gear toe-in

Postby OO ZVY » Sat Apr 10, 2021 6:42 pm



seems to work :

Zlin with toe in of 0.07 rad ( 4 deg ) very very instable.
Zlin with toe out of 0.07 rad ( 4 deg ) almost easy to control for TO and landing.

Will try with 2 deg more realistic next time.

Correction via YASIM a suggested by danielHL above.

let me know if you try it.

bgrs

OOZVY
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby OO ZVY » Sun Apr 11, 2021 2:19 am

Another very positive test with a more wicked aircraft : P40.
see controllability with a toe out :




ZVY
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby Ysop » Sun Apr 11, 2021 8:02 am

Good to hear!
Thanks for your research including video footage.
And thanks DanielHL for providing a YASIM solution.

Looks like, I've stumbled over something in JSBSim. So good hint :)
User avatar
Ysop
 
Posts: 416
Joined: Thu Oct 25, 2018 9:06 pm
Version: 2020.3.1
OS: ubuntu 20.04

Re: landing gear toe-in

Postby danielHL » Sun Apr 11, 2021 11:04 am

I think jsbsim has an orientation tag for the gears, that should also do the trick...

Thanks for testing this out! It's one little detail that is never often talked about and I wonder if anyone ever tried this in flightgear. By the way - could this also help with airplanes with tricycle gear? I'm thinking about the Citation II which also has some takeoff roll problems in crosswind...
danielHL / D-FMPW
danielHL
 
Posts: 277
Joined: Fri May 02, 2014 6:23 pm
Callsign: D-FMPW
Version: next
OS: Linux

Re: landing gear toe-in

Postby OO ZVY » Sun Apr 11, 2021 12:59 pm

Test with a ' heavy ' . Works perfectly.
User avatar
OO ZVY
 
Posts: 103
Joined: Sun Sep 11, 2011 7:13 pm
Callsign: OO-ZVY
OS: Ubuntu 18.04

Re: landing gear toe-in

Postby Ysop » Sun Apr 11, 2021 1:24 pm

The trick would be to model the airplane as it is and then everything should fall into place correctly. At least this is ONE philosophy.

There is a bit of discussion on controlability of JSBsim-Taildraggers. Maybe there is a bit room for improvement for those as well.

On a tricycle gear it would be a bit unnatural. But as a YASIM-"hack": Why not.
User avatar
Ysop
 
Posts: 416
Joined: Thu Oct 25, 2018 9:06 pm
Version: 2020.3.1
OS: ubuntu 20.04

Next

Return to Flight dynamics model

Who is online

Users browsing this forum: No registered users and 1 guest