Board index FlightGear Development Aircraft

JSBSim Piper J-3 Cub  Topic is solved

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: JSBSim Piper J-3 Cub

Postby wlbragg » Sat Mar 07, 2020 2:46 pm

Help yourself! :)
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: JSBSim Piper J-3 Cub

Postby D-ECHO » Sat Mar 07, 2020 3:08 pm

Thank you :)
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: JSBSim Piper J-3 Cub

Postby dany93 » Thu May 07, 2020 6:07 pm

Hi,

I have a weird behavior with the (default) tailwheel version from FGAddon, downloaded this day. At runway threshold, brake parking and brakes OFF, it yaws around one wheel then the other one, whatever my controls on brakes. Impossible to have it rolling.
Tricycle and Amphibious work OK.
The issue is only with the tailwheel version.

Tested at three airports (KSFO, KJFK and LFBO)
With three different versions of FG (2019.2.0 from Jan 2020, from April 2020 and 2020.2.0 from 06 and 07 May 2020).
With Autostart and, If I remember well, manual start ("S") too.

My "old" version of the J3Cub (github e580b7322, 2019-07-28) works well.

Or do I miss something?

Dany

Linux Mint 18
SimGear 1d89a76 Thu, 7 May 2020 11:47:25 +0100
FGData acd3737 Wed, 6 May 2020 15:41:18 +0100
FlightGear da666ae Thu, 7 May 2020 11:13:12 +0100
dany93
 
Posts: 861
Joined: Mon Sep 07, 2009 4:43 pm
Location: France (Paris region)
Version: 2020.4.0
OS: Linux Mint 20.3

Re: JSBSim Piper J-3 Cub

Postby wlbragg » Thu May 07, 2020 6:52 pm

Hey Dany,

Yeah, I think I need to back that change out and remove the change to the range of the castor wheel section.
See:https://github.com/wlbragg/J3Cub/commit/5d674bd6db8d23738f974007ee47adc5f96087de
It needs some work. I don't know what approach to take. I kind of left it hanging. I don't know if it is something that needs to be taken up with JSBSim folks on sensitivity of that part of the code. Or if we can make some kind of dampening logic to tone down the speed of the castor or make it non-linear?
I'd like to make it work so we can turn on a dime like in real life.

Your not saying that it used to work and now it doesn't are you?

It's been awhile since I tried it and it was really touchy. To the point of unusable with anything but a peddle rudder control system, which I have. I added it fo FGADDON so people would test and I could get some feedback and opinion.
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: JSBSim Piper J-3 Cub

Postby dany93 » Thu May 07, 2020 11:21 pm

wlbragg wrote in Thu May 07, 2020 6:52 pm:Your not saying that it used to work and now it doesn't are you?

I'm saying that it used to worked in the past (my "old" github J3Cub still works well) and it no longer works now.
it was really touchy. To the point of unusable with anything but a peddle rudder control system

Rolling at very low speed, when the tail wheel weighs on the ground and still has a steering effect, my "old" J3Cub is easy to control, not touchy. It becomes touchy during a short time between this ground step and the one when the rudder becomes efficient. I don't know for the real J3Cub, but taildraggers are well known as being touchy on the ground. This is due to the CG being behind the main wheel contacts (differently from a tricycle).
Pulling the yoke helps during the touchy part (I think that it is to be done), but the risk is to take off at too low airspeed.
dany93
 
Posts: 861
Joined: Mon Sep 07, 2009 4:43 pm
Location: France (Paris region)
Version: 2020.4.0
OS: Linux Mint 20.3

Re: JSBSim Piper J-3 Cub

Postby dany93 » Fri May 08, 2020 11:12 am

The difficulty to steer a tailwheeler is rather at rolling on the ground after landing, but it can also been observed on these videos at acceleration before takeoff.
(P51D)
https://www.youtube.com/watch?v=kZThXV1x6RA
at 0:45

https://www.youtube.com/watch?v=Ys7QqNTASWk
at 0:25

https://www.youtube.com/watch?v=eOXxUApaaWo
at 0:17
Observe the feet movements when he progressively pushes the throttle.


I think that the best is doing operations progressively, firstly at low throttle, holding the yoke pulled, up to a velocity not too high (to avoid takeoff at low airspeed), then release the yoke progressively.

The issue with caster wheel in JSBSim is that, in JSBSim, the wheel orientation is controlled by the velocity vector of the ground relative to the aircraft. With some velocity, this is correct and like reality for a caster wheel. But a real caster wheel has its axis shifted from the rotation axis, which makes it self-orientable even at zero speed, only by the force exerted. With JSBSim, this self-orientation by the force does not exist. So, because the JSBSim caster wheel tends to orientate near +/-90° at rest (probably due to a last slight yaw before stopping), having the aircraft rolling is very hard or nearly impossible as long as it does not reach a low velocity. Which can be made easier by setting a weak static and dynamic friction coefficient.

For the DC3, Fritz's idea (workaround) was to set a very weak static and dynamic friction coefficient of 0.05. Rather like a "ball".

Anyway, better to have an aircraft which is touchy to control at rolling at the beginning of acceleration (which may be not so far from reality) than an aircraft which is impossible to have it rolling.
dany93
 
Posts: 861
Joined: Mon Sep 07, 2009 4:43 pm
Location: France (Paris region)
Version: 2020.4.0
OS: Linux Mint 20.3

Re: JSBSim Piper J-3 Cub

Postby wlbragg » Fri May 08, 2020 5:16 pm

Pulling the yoke helps during the touchy part

Yes, I found that out, makes it a little more tolerable.

I think that the best is doing operations progressively, firstly at low throttle, holding the yoke pulled, up to a velocity not too high (to avoid takeoff at low airspeed), then release the yoke progressively.

Absolutely takes way more technique and practice. Even with rudder peddle it requires concentration.

If you want to try to work on this, please do. I think the castor to 360 deg is only in fgaddon right now. I think I left the git repo alone. Actually I should reverse that and make the git repo the touchy one and leave fgaddon easier until we can fine tune this.

When I get a chance I can try some of your suggestions.
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: JSBSim Piper J-3 Cub

Postby dany93 » Sat May 09, 2020 2:38 pm

wlbragg wrote in Fri May 08, 2020 5:16 pm:If you want to try to work on this, please do. I think the castor to 360 deg is only in fgaddon right now. I think I left the git repo alone

Honestly, I don't know what to improve. Maybe for someone who knows the real one.
For me, the steering at very low speed (taxiing) is easily doable.
I'm not sure that the critical step is a matter of caster wheel tuning. During this step, there is less or no weight on the tail wheel.


Actually I should reverse that and leave fgaddon easier until we can fine tune this.


I agree 100%. FGAddon is for all users, the aircraft must work.
Yes, it's touchy, but I'm not sure it is more difficult than reality (I don't know for the real one but see my videos above). With tailwheelers,You have to struggle with the rudder on the ground.

Having full throttle is not necessarily worse. It makes the critical step shorter and the rudder is air blown, more efficient. With some practice and anticipation (2-3 attempts), I even can keep in line and do good takeoffs by starting from at rest full thrust, brakes applied (although this is not recommended with a tailwheeler if your propeller is clean... :lol: ).

and make the git repo the touchy one


I don't see the interest. Kind of incitement for people to improve it ? I think the "easier" one is a better start. If users are fine with it, no problem.

Otherwise, this discussion might be the best place.
dany93
 
Posts: 861
Joined: Mon Sep 07, 2009 4:43 pm
Location: France (Paris region)
Version: 2020.4.0
OS: Linux Mint 20.3

Re: JSBSim Piper J-3 Cub

Postby Ysop » Mon Jun 01, 2020 9:10 am

This still gets me thinking, so I took the aircraft for a spin - and the spin(s) escalated a bit.

Looks like the tailwheel is free spinning without any self-centering. Once the aircraft circles on the ground there is no stop, even with rudder centered.

From pictures it can be seen, that the contact point of the wheel to the ground is aft of the rotation axis and there is also centering by springs.

Is it possible it make max_steer unit depending on side component of the ground reaction or yaw moment plus part of side force?

Linear approach could be used in first assumption. This would simulate the effect of the spring and make life before airborne a bit easier. The JSBSim behavior as pointed out by dany would be significantly smaller then.
And it would be of help for a similar problem on a nosewheel with partly steering by differential braking.

Small experiment with the max steer angle: When limiting it to 90, to limit the problem, pedal effect is surprisingly reversed.
User avatar
Ysop
 
Posts: 1349
Joined: Thu Oct 25, 2018 10:06 pm
Version: 2020.3.18
OS: ubuntu 22.04

Re: JSBSim Piper J-3 Cub

Postby dany93 » Mon Jun 01, 2020 11:30 am

To make it clear (just in case), the previous github version (prior to 5d674bd6) with
Code: Select all
        <contact type="BOGEY" name="TAIL">
            <location unit="IN">
                <x> 203.00 </x>
                <y>   0.00 </y>
                <z> -36.20 </z>
            </location>
            <static_friction>  0.60 </static_friction>
            <dynamic_friction> 0.50 </dynamic_friction>
            <rolling_friction> 0.02 </rolling_friction>
            <spring_coeff  unit="LBS/FT">   250.00 </spring_coeff>
            <damping_coeff unit="LBS/FT/SEC"> 150.00 </damping_coeff>
            <steerability> STEERABLE </steerability>
            <max_steer unit="DEG">-10</max_steer> <!-- instead of 360 -->
            <brake_group> NONE </brake_group>
            <retractable> 0 </retractable>
        </contact>

was correct and controllable IMO.

Dany
dany93
 
Posts: 861
Joined: Mon Sep 07, 2009 4:43 pm
Location: France (Paris region)
Version: 2020.4.0
OS: Linux Mint 20.3

Re: JSBSim Piper J-3 Cub

Postby D-ECHO » Mon Jun 01, 2020 11:40 am

Just two informations regarding this based on previous experience:

*max_steer set to 360 means a free castering wheel with no steering (not sure if that was clear)
*The problem with the steering being reversed is probably due to the tailwheel configuration. I used to override the relevant property with an own function, but just setting it to -10 like in dany's post is also a neat way
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: JSBSim Piper J-3 Cub

Postby wlbragg » Mon Jun 01, 2020 3:15 pm

The problem setting it to 10 is it limits the tightness of the turn. That is not realistic for this aircraft. Yes it's controllable.

<max_steer unit="DEG">-10</max_steer> <!-- instead of 360 -->

So what does -90 do?
Last edited by wlbragg on Mon Jun 01, 2020 3:28 pm, edited 1 time in total.
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: JSBSim Piper J-3 Cub

Postby D-ECHO » Mon Jun 01, 2020 3:26 pm

How is the tail wheel controlled in real life?
D-ECHO
 
Posts: 2460
Joined: Sat May 09, 2015 1:31 pm
Pronouns: Bea (she/her)
Version: next

Re: JSBSim Piper J-3 Cub

Postby wkitty42 » Mon Jun 01, 2020 3:35 pm

if it is a castor wheel, it isn't controlled directly... the left and right brakes should control the direction of the craft until such time as there is plenty of air flow past the rudder for it to have authority... even then, the wheel brakes would still have some authority while they are on the ground taking lift into account...
"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: JSBSim Piper J-3 Cub

Postby wlbragg » Mon Jun 01, 2020 3:52 pm

As far as I know it is 360, but it doesn't behave this way.
I changed it to -90 and tested. I think that is better. With slight back pressure it is controllable and I can still turn tight. That was the main reason for attempting 360.
I pushed it to fgaddon.
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

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 13 guests