[Flightgear-devel] Bug in JSBSim FGPropeller.cpp
There is a bug in JSBSim causing multiengine aircraft with counterrotating props suffer from excessive yaw when engines are running. Significant aileron deflection is required to fly straight and level. I tracked down that issue and finally came to FGPropeller.cpp line 198: vH(eX) = Ixx*omega*Sense; If I understand correctly, this is some kind of force along the rotating axis of the propeller and it should not be multiplied with Sense (unless someone mounts a wrong propeller...) Making the line vH(eX) = Ixx*omega; removes the unwanted yaw and creates symmetrical forces and moments. The only aircrafts using counterrotating props are the P-38L and the SenecaII. All others should not be affected by this patch. Greetings, Torsten - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
On Tue 7 August 2007 17:57, Torsten Dreyer wrote: There is a bug in JSBSim causing multiengine aircraft with counterrotating props suffer from excessive yaw when engines are running. Significant aileron deflection is required to fly straight and level. I tracked down that issue and finally came to FGPropeller.cpp line 198: vH(eX) = Ixx*omega*Sense; If I understand correctly, this is some kind of force along the rotating axis of the propeller and it should not be multiplied with Sense (unless someone mounts a wrong propeller...) Making the line vH(eX) = Ixx*omega; removes the unwanted yaw and creates symmetrical forces and moments. The only aircrafts using counterrotating props are the P-38L and the SenecaII. All others should not be affected by this patch. Greetings, Torsten May be i don't understand your remark, and i could be wrong, but all aircraft propeller have not the same rotation direction , some are CCWon the same aircraft some are CW on others aircraft. Sure there is no problem with P-38L one propeller is given sense -1 /sense The other is given sense 1 /sense so, Sense could be -1 or 1 which is significant for the propeller effect. it is very important to keep in that calculation the Sense. Regards -- Gérard - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
On Tue 7 August 2007 18:20, gh.robin wrote: On Tue 7 August 2007 17:57, Torsten Dreyer wrote: There is a bug in JSBSim causing multiengine aircraft with counterrotating props suffer from excessive yaw when engines are running. Significant aileron deflection is required to fly straight and level. I tracked down that issue and finally came to FGPropeller.cpp line 198: vH(eX) = Ixx*omega*Sense; If I understand correctly, this is some kind of force along the rotating axis of the propeller and it should not be multiplied with Sense (unless someone mounts a wrong propeller...) Making the line vH(eX) = Ixx*omega; removes the unwanted yaw and creates symmetrical forces and moments. The only aircrafts using counterrotating props are the P-38L and the SenecaII. All others should not be affected by this patch. Greetings, Torsten May be i don't understand your remark, and i could be wrong, but all aircraft propeller have not the same rotation direction , some are CCWon the same aircraft some are CW on others aircraft. Sure there is no problem with P-38L one propeller is given sense -1 /sense The other is given sense 1 /sense so, Sense could be -1 or 1 which is significant for the propeller effect. it is very important to keep in that calculation the Sense. Regards In addition to, if these effect seems to be more than you espect it is possible to reduce the value with p_factor /p_factor with zero there is no effect. -- Gérard - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
There is a bug in JSBSim causing multiengine aircraft with counterrotating props suffer from excessive yaw when engines are running. Significant aileron deflection is required to fly straight and level. I tracked down that issue and finally came to FGPropeller.cpp line 198: vH(eX) = Ixx*omega*Sense; If I understand correctly, this is some kind of force along the rotating axis of the propeller and it should not be multiplied with Sense (unless someone mounts a wrong propeller...) Making the line vH(eX) = Ixx*omega; removes the unwanted yaw and creates symmetrical forces and moments. The only aircrafts using counterrotating props are the P-38L and the SenecaII. All others should not be affected by this patch. Greetings, Torsten If this is indeed a cause of a problem such as you describe, or a bug, I would be very surprised. Here's why: If you look at FGPropeller.cpp, you will see that the vector vH does not add to the moment of the aircraft unless there is already a rotational motion experienced by the aircraft. See the equation where vH is summed into the moment equation: vMn = fdmex-GetPropagate()-GetPQR()*vH + vTorque*Sense; If the rotational rates (returned by the function GetPQR()) are zero - that is, at steady state - the angular momentum (vH) parameter will add *nothing* to the total moment (vMn). The angular momentum introduces gyroscopic effects. If the effect is too strong, it could be that the Ixx value for the propeler/engine shaft is too large. In that case, the angular momentum calculated would be too large and would cause the aircraft to behave erratically when manuevering. Can you tell more about the conditions in your test, and more about the problem? Jon - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
If the rotational rates (returned by the function GetPQR()) are zero -
that is, at steady state - the angular momentum (vH) parameter will add
*nothing* to the total moment (vMn). The angular momentum introduces
gyroscopic effects.
If the effect is too strong, it could be that the Ixx value for the
propeler/engine shaft is too large. In that case, the angular momentum
calculated would be too large and would cause the aircraft to behave
erratically when manuevering.
Maybe I am wrong, but shouldnt these effects from the left engine compensate
those from the right engine if both are rotating in opposite directions?
Can you tell more about the conditions in your test, and more about the
problem?
I will try :-)
When you start the SenecaII and give full throttle on ground with parking
brakes set, the aircraft turns slowly to the left (because the brakes are to
weak - thats another story).
When you fly the aircraft with full throttle with wings level, a slight
aileron deflection to the right is needed (approx. 0.05) to keep the wings
level. When you idle the engines, no aileron is needed to keep the wings
level.
I put some debug code into FGPropulsion.cpp:
for (i=0; inumEngines; i++) {
cout computing engine i \n;
Engines[i]-Calculate();
vForces += Engines[i]-GetBodyForces(); // sum body frame forces
vMoments += Engines[i]-GetMoments(); // sum body frame moments
cout Forces: Engines[i]-GetBodyForces() \n;
cout Moments: Engines[i]-GetMoments() \n;
cout finished with engine i \n\n;
}
and FGPropeller.cpp:
vMn = fdmex-GetPropagate()-GetPQR()*vH + vTorque*Sense;
cout PQR: fdmex-GetPropagate()-GetPQR() \n;
cout vH(eX): vH(eX) \n;
cout vMn: vMn \n;
And the output for the Seneca is
computing engine 0
PQR: -0.00243528 , -0.000488289 , -0.0360579
vH(eX): 508.722
vMn: -376.628 , -18.3435 , 0.248403
Forces: 819.428 , 0 , 0
Moments: -376.628 , -529.63 , 4889.5
finished with engine 0
computing engine 1
PQR: -0.00243528 , -0.000488289 , -0.0360579
vH(eX): -508.722
vMn: -376.628 , 18.3435 , -0.248403
Forces: 819.428 , 0 , 0
Moments: -376.628 , -492.943 , -4889.5
finished with engine 1
and the P-38L says:
computing engine 0
PQR: 0.000489925 , 3.41744e-05 , -0.00822598
vH(eX): -3114.13
vMn: -6748.41 , 25.6167 , 0.106423
Moments: 3071.39 , 0 , 0
Moments: -6748.41 , -2090.55 , 25605.2
finished with engine 0
computing engine 1
PQR: 0.000489925 , 3.41744e-05 , -0.00822598
vH(eX): 3114.13
vMn: -6748.41 , -25.6167 , -0.106423
Moments: 3071.39 , 0 , 0
Moments: -6748.41 , -3126.87 , -25605.2
finished with engine 1
As you can see, the moments second values are different on the second value,
because PQR is not zero.
Both aircraft are sitting on ground, parking brake set, full throttle.
I hope this helps
Torsten
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now http://get.splunk.com/
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
May be i don't understand your remark, and i could be wrong, but all aircraft propeller have not the same rotation direction , some are CCWon the same aircraft some are CW on others aircraft. Correct - and some aircraft have both of it: one CW and one CCW. Sure there is no problem with P-38L one propeller is given sense -1 /sense The other is given sense 1 /sense so, Sense could be -1 or 1 which is significant for the propeller effect. it is very important to keep in that calculation the Sense. Yes - also true. In addition to, if these effect seems to be more than you espect it is possible to reduce the value with p_factor /p_factor with zero there is no effect. Since the engines of the P-38L and the Seneca are turning in opposite directions, the p-factor of both propellers are compensating each other. Thats why aircraft with counterrotating props are made: they are easier to fly! Try yourself: Get the P-38L in the air and fly full throttle straight and level, open the property-browser for controls/flight. I need a right aileron deflection to keep the wings level. If I reduce power to idle, the aircraft turns sharply to the right with that aileron setting. This should not happen on aircraft with counterrotating props. Torsten - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
On 8/7/07, Torsten Dreyer [EMAIL PROTECTED] wrote: Try yourself: Get the P-38L in the air and fly full throttle straight and level, open the property-browser for controls/flight. I need a right aileron deflection to keep the wings level. If I reduce power to idle, the aircraft turns sharply to the right with that aileron setting. This should not happen on aircraft with counterrotating props. I observe this, but I wonder if it isn't the way things should be (in spite of the counterrotating props). Maybe there's another effect at play that counterrotation doesn't solve entirely? I don't remember where now, but I thought I read something that satisfactorily explained the Seneca's behavior to me. Perhaps someone who has flown a counterrotating twin can weigh in. - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
Hans Fugal wrote: Perhaps someone who has flown a counterrotating twin can weigh in. Then Torsten is the right person here. Guess who does fly the real-life counterpart of the Seneca in FlightGear? Cheers, Ralf - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bug in JSBSim FGPropeller.cpp
explained the Seneca's behavior to me. Perhaps someone who has flown a counterrotating twin can weigh in. The Seneca POH says: Counter-rotation of the propellers provides balanced thrust during takeoff and climb and eliminates the `critical engine` in single engine flight. Not that much of information, but it is true: I get my feet out of the pedals after the gear has left the runway and you don't need your hands either to compensate any bank or yaw. - It's a good idea to be ready to jump in if one engine fails, of course ;-) Greetings, Torsten - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
