Excessive P-Factor on P47

[Kalashnikov63]

Respectfully, your technical team has severely over-modeled P-factor on the P-47.

As both a commercial and private aviator, having operated both reciprocating and jet aircraft, I can assure you that no aircraft, not even a P-47 with a 4-bladed paddle prop, can generate the p-factor I've observed on your model.  It is SO severely over-modeled, that it completely overwhelms the longitudinal thrust vector of the propellor.  P-factor is certainly  pronounced on an aircraft with such a large, 4-bladed propellor, and powerful engine as the P-47 has, but it should not violently, and irrecoverably, send the aircraft careening over to the left side of the runway, when AoA exceeds 4-5 degrees on takeoff.  I have never flown a P-47, but I have had the privilege of operating a P-51, and can say that your technical team has modelled its p-factor in a realistic manner.  Also, if anything, the 2800 Double Wasp and its derivatives, had tremendous torque, which is UNDER-modelled in the P-47.  If you wish more information, I would be happy to send it, or, in lieu of such, I respectfully request that your coders reduce the p-factor function output by at least 75%.

Thank you for your time.

Sincerely,

Dr. MDS, Ph.D., Applied Physics
NASA/JPL Cassini Mission Imaging Consultant.

[kablamoman]

Hello friend,

Have you ruled out the possibility that there is some other cause for the result that you're seeing? I do not find that P-factor is a significant contribution to left turning tendency in any of the warbirds in the sim. Nor should it be, really: Significant p-factor requires a higher relative airspeed than what you would typically see in the initial stages of the takeoff roll, coupled with a high angle of attack -- but once you've built up such speed, you're typically no longer at a sufficiently high angle of attack (assuming you are raising the tail) that the speed differential between the down-going and up-going (retreating) blades is enough to cause significant asymmetric disk loading.

In the initial takeoff roll, the effects from the helical propwash should and do dominate, but not to the extent that it should overwhelm directional control with rudder, assuming your tailwheel is straight and locked. Have you confirmed your tailwheel is locked? Perhaps there is an issue with the control mappings of your rudder axis? Sometimes there can be double binds from other devices. There are also special options for each warbird module for "auto-rudder" and "takeoff assistance" in the main menu options you can double-check to make sure they are disabled and aren't interfering.

As you pick up speed in the takeoff roll the next big left turning tendency would be from the gyroscopic precession when you lift the tail off the ground with forward stick. If this is the point at which you start to lose control and veer off the runway to the left, try easing the nose down a little more gradually so you have plenty of opportunity to react and correct with right rudder.

It sounds like you should have a firm grasp on pilot technique based on your background, so I've included advice regarding technique for others that may be curious, but sometimes the "simisms" catch the best of us, so I would encourage you to double-check the tailwheel lock, special options, and control bindings. Keep in mind that many folks fly the module every day without running into the problem you mention, many of them also pilots in real life, either for pleasure or as their day jobs.

Hope this helps!

Edited Saturday at 09:29 PM by kablamoman

[Nate--IRL--]

Hah - I only learned of the effects of Procession yesterday, and today I learn about P-Factor now too. Why does it all want me in the ditch on the left of the runway?  

Nate

[Kalashnikov63]

Hello,

   Thank you for taking the time to answer and share your thoughts.

  First of all, you're quiter correct, in that differnt rigs, hardware, etc, can yield differnt outcomes in performance.

>> 

Have you ruled out the possibility that there is some other cause for the result that you're seeing? I do not find that P-factor is a significant contribution to left turning tendency in any of the warbirds in the sim. Nor should it be, really: Significant p-factor requires a higher relative airspeed than what you would typically see in the initial stages of the takeoff roll, coupled with a high angle of attack -- but once you've built up such speed, you're typically no longer at a sufficiently high angle of attack (assuming you are raising the tail) that the speed differential between the down-going and up-going (retreating) blades is enough to cause significant asymmetric disk loading.  <<

I'm confused as to your statement; p-factor is virtually negligible at high IAS, and is not a de-facto issue while turning - it rears its ugly head on liftoff, during high power and high AoA situations.  It does not have a significant impact during takeoff roll, only immediately before, during, and after rotation.

In the initial takeoff roll, the effects from the helical propwash should and do dominate, but not to the extent that it should overwhelm directional control with rudder, assuming your tailwheel is straight and locked. Have you confirmed your tailwheel is locked? Perhaps there is an issue with the control mappings of your rudder axis? Sometimes there can be double binds from other devices. There are also special options for each warbird module for "auto-rudder" and "takeoff assistance" in the main menu options you can double-check to make sure they are disabled and aren't interfering.

The most severe instances of p-factor show themselves at the moment of rotation - as they should - which is after the tailwheel is off the ground.

As you pick up speed in the takeoff roll the next big left turning tendency would be from the gyroscopic procession when you lift the tail off the ground with forward stick. If this is the point at which you start to lose control and veer off the runway to the left, try easing the nose down a little more gradually so you have plenty of opportunity to react and correct with right rudder.

I'm quite certain that the effect at work is p-factor; it can be modulated by changing AoA, for better or for worse by increasing or decreasing.

It sounds like you should have a firm grasp on pilot technique based on your background, so I've included advice regarding technique for others that may be curious, but sometimes the "simisms" catch the best of us, so I would encourage you to double-check the tailwheel lock, special options, and control bindings. Keep in mind that many folks fly the module every day without running into the problem you mention, many of them also pilots in real life, either for pleasure or as their day jobs.

Hope this helps!

Any erudite contribution to a raised issue helps, so yes, it does!  

To compare, the p-factor from the P-51 is at least an order of magnitude less than that on the P-47 model; minimal effort is necessary to correct for it, even when attempting the most brazen, high AoA takeoffs, and aerial maneuvers.  I would not expect reduced p-factor on the P-51, as it also has a 4-bladed prop, driven by similar torsional forces.  

What truly perplexes me is the incredible UNDER-modeling of torque, especially on the P-47; with 20 crank pinions, the Double Wasp derivative installed in the P-47 was known to try to twist the airframe into a pretzel, when sudden increases in power were applied.  However, I cannot speak intelligently on this topic, as I have never flown a P-47.

[kablamoman]

4 hours ago, Kalashnikov63 said:

I'm quite certain that the effect at work is p-factor; it can be modulated by changing AoA, for better or for worse by increasing or decreasing.

Just to be clear, if you experience left yaw when you pitch forward to lift the tail in your takeoff roll this is gyroscopic precession. Pitching forward/lifting the tail would actually reduce the contribution of p-factor in this instance — resulting in a right turning tendency — as you are reducing the disparity of AoA and speed that the blades are experiencing between the left and right sides of the propeller disk compared to the more nose up three-point attitude.

I am curious, are you maintaining a three-point attitude for the entirety of your takeoff roll? If so, I would suggest you transition to a more nose-low takeoff attitude earlier in your roll and work on maintaining directional control through that transition by easing the nose down smoothly and deliberately, with an appropriate amount of right rudder to compensate. The aircraft should lift off on its own in this attitude, and there should be very little contribution from p-factor for the remainder of your takeoff.

[grafspee]

This topic was grilled so many times across different planes that i can not count them already. I will just repeat that feelings can't be entered in to code. They can fine tune code from pilot notes and tips but changing P-factor by 75% is like making a new FM from start. Not an easy task like moving p-factor slider which does not exist. If you poses documentation about P-47 flight tests which contradicts what is in game, sure you can drop it here and i'm sure that devs will take a look.

I remember that on some topic that clip was analyzed, that torque factor is near non at high power take off speeds that roll is induced by other factors and it is the last thing what happens to the plane.

 

Edited 19 hours ago by grafspee

[Kalashnikov63]

Just to be clear, if you experience left yaw when you pitch forward to lift the tail in your takeoff roll this is gyroscopic precession. Pitching forward/lifting the tail would actually reduce the contribution of p-factor in this instance — resulting in a right turning tendency — as you are reducing the disparity of AoA and speed that the blades are experiencing between the left and right sides of the propeller disk compared to the more nose up three-point attitude.

Absolutely correct.   The problem begins after the aircraft is situated at near zero AoA, and the landing gear lift off the runway; when I am pitching up slightly, past about 5 degrees AoA, which is what has me nearly convinced that it is p-factor.  I don't say "certain", because in the pure sciences we say that certainty is the enemy of progress...

I am curious, are you maintaining a three-point attitude for the entirety of your takeoff roll? If so, I would suggest you transition to a more nose-low takeoff attitude earlier in your roll and work on maintaining directional control through that transition by easing the nose down smoothly and deliberately, with an appropriate amount of right rudder to compensate. The aircraft should lift off on its own in this attitude, and there should be very little contribution from p-factor for the remainder of your takeoff.

No.  My takeoffs start as you're describing, then transition to a zero AoA segment, with the tailwheel aloft, a few meters of altitude are gained, and then I pitch up.  The exact same takeoff on the P-51 result in a normal takeoff; on the P-47, unless I walk on eggshells, I end up on the left side of the runway, in a ditch.

Respectfully (to the coders), unless there is something amiss with my configuration (a distinct possibility), P-factor should be detuned.

[Kalashnikov63]

Hello,

This topic was grilled so many times across different planes that i can not count them already. I will just repeat that feelings can't be entered in to code. They can fine tune code from pilot notes and tips but changing P-factor by 75% is like making a new FM from start. Not an easy task like moving p-factor slider which does not exist.

I do not know what language the coders are using - my guess is that it is a "C" derivative, with possible binary data fields that read into defined functions.  Defining functions is one of the quickest way of modeling a mathematically defined behavior, and functions can either be recursive, or interdependent, again, dependent on the language used to program the sim.  I would think that p-factor behavior would be best modelled through a defined function, with arguments derived from both local and global variables to return what we observe.  Again, I have no knowledge of how they are programming the sim, so I am ill-prepared to pass judgement on the degree of difficulty involved in modifying that effect.

If you poses documentation about P-47 flight tests which contradicts what is in game, sure you can drop it here and i'm sure that devs will take a look.

Not on the P-47; I have never flown one.  On the P-51, a far more ubiquitous aircraft, with many still flown today, I do.  But I am inclined to opine that if ED's coders wanted my input, they would ask for it, especially after I offered it.  It would make little sense to scan dozens of pages of documents to post on a discussion board, that no one asked for. 

I remember that on some topic that clip was analyzed, that torque factor is near non at high power take off speeds that roll is induced by other factors and it is the last thing what

Torque is a function of a great number of things that require the use of partial differential equations to properly model, but in a nutshell, it is a result of the engine trying to turn, against the inertia and drag of the propellor, which is resisting the torsional forces created by the engine.  What many miss, is that torque effects are most pronounced at the time when the balance between twisting force (torque) of the engine, and resistance thereof by the propellor CHANGES.  If the forces are in dynamic equilibrium, there will be no torque effects, but during power/IAS/drag changes, the struggle between the engine trying to spin the propellor, and the latter resisting, results in a change in total angular momentum, and a rolling moment.

Good point.

[grafspee]

22 minutes ago, Kalashnikov63 said:

 

Torque is a function of a great number of things that require the use of partial differential equations to properly model, but in a nutshell, it is a result of the engine trying to turn, against the inertia and drag of the propellor, which is resisting the torsional forces created by the engine.  What many miss, is that torque effects are most pronounced at the time when the balance between twisting force (torque) of the engine, and resistance thereof by the propellor CHANGES.  If the forces are in dynamic equilibrium, there will be no torque effects, but during power/IAS/drag changes, the struggle between the engine trying to spin the propellor, and the latter resisting, results in a change in total angular momentum, and a rolling moment.

Good point.

Watch that video and you will see, if you willing to ofc, you can see that pilot did not apply any inputs to counter roll and as soon plane left carrier there was no roll, if any significant direct torque was present there plane should immediately start to roll but it didn't. First what happened was swing to the left due to massive p-factor and then plane started to pitch up as well and still wings were level. Then when left wing stalled abrupt roll was induced but not from torque not from p-factor but from stall.

Edited 9 hours ago by grafspee

[Kalashnikov63]

Hello,

I remember that on some topic that clip was analyzed, that torque factor is near non at high power take off speeds that roll is induced by other factors and it is the last thing what happens to the plane.

Respectfully, I do not understand the relevance of the clip.  As the text indicates, the pilot committed several errors during liftoff.  They didn't point out excessive AoA, that, despite the fact that earlier Corsairs had a 3-bladed prop which reduced p-factor, to me, this is not just a control surface issue, it was a failure to control both the torque and the p-factor of that particular aircraft.

4 minutes ago, grafspee said:

Then when left wing stalled abrupt roll was induced but not from torque not from p-factor but from stall.

Which was caused by failure to compensate for both torque and p-factor, and incorrect pilot inputs.  Stalls are a result of incorrect operation of an aircraft; they don't appear from naught.

[grafspee]

8 minutes ago, Kalashnikov63 said:

Hello,

I remember that on some topic that clip was analyzed, that torque factor is near non at high power take off speeds that roll is induced by other factors and it is the last thing what happens to the plane.

Respectfully, I do not understand the relevance of the clip.  As the text indicates, the pilot committed several errors during liftoff.  They didn't point out excessive AoA, that, despite the fact that earlier Corsairs had a 3-bladed prop which reduced p-factor, to me, this is not just a control surface issue, it was a failure to control both the torque and the p-factor of that particular aircraft.

Point is to see what is going with a plane when you do not steer it correctly and you can see roll develops in time.

1 blade or 2 or 3 or 4 it does not matter really. What matters is the power of the engine. Same rpm = same power = same torque.

 

8 minutes ago, Kalashnikov63 said:

Hello,

I remember that on some topic that clip was analyzed, that torque factor is near non at high power take off speeds that roll is induced by other factors and it is the last thing what happens to the plane.

Respectfully, I do not understand the relevance of the clip.  As the text indicates, the pilot committed several errors during liftoff.  They didn't point out excessive AoA, that, despite the fact that earlier Corsairs had a 3-bladed prop which reduced p-factor, to me, this is not just a control surface issue, it was a failure to control both the torque and the p-factor of that particular aircraft.

Which was caused by failure to compensate for both torque and p-factor, and incorrect pilot inputs.  Stalls are a result of incorrect operation of an aircraft; they don't appear from naught.

That roll was not from torque but from single wing stall.

Edited 9 hours ago by grafspee

[Kalashnikov63]

3 minutes ago, grafspee said:

1 blade or 2 or 3 or 4 it does not matter really. What matters is the power of the engine. Same rpm = same power = same torque.

Respectfully, that is incorrect.  Three-bladed propellers are significantly less prone to the effects of p-factor than 4-bladed propellers.

[grafspee]

7 minutes ago, Kalashnikov63 said:

Respectfully, that is incorrect.  Three-bladed propellers are significantly less prone to the effects of p-factor than 4-bladed propellers.

I think you are mixing P-factor with torque then. I was talking about torque.

Edited 9 hours ago by grafspee

[kablamoman]

1 hour ago, Kalashnikov63 said:

Just to be clear, if you experience left yaw when you pitch forward to lift the tail in your takeoff roll this is gyroscopic precession. Pitching forward/lifting the tail would actually reduce the contribution of p-factor in this instance — resulting in a right turning tendency — as you are reducing the disparity of AoA and speed that the blades are experiencing between the left and right sides of the propeller disk compared to the more nose up three-point attitude.

Absolutely correct.   The problem begins after the aircraft is situated at near zero AoA, and the landing gear lift off the runway; when I am pitching up slightly, past about 5 degrees AoA, which is what has me nearly convinced that it is p-factor.  I don't say "certain", because in the pure sciences we say that certainty is the enemy of progress...

I am curious, are you maintaining a three-point attitude for the entirety of your takeoff roll? If so, I would suggest you transition to a more nose-low takeoff attitude earlier in your roll and work on maintaining directional control through that transition by easing the nose down smoothly and deliberately, with an appropriate amount of right rudder to compensate. The aircraft should lift off on its own in this attitude, and there should be very little contribution from p-factor for the remainder of your takeoff.

No.  My takeoffs start as you're describing, then transition to a zero AoA segment, with the tailwheel aloft, a few meters of altitude are gained, and then I pitch up.  The exact same takeoff on the P-51 result in a normal takeoff; on the P-47, unless I walk on eggshells, I end up on the left side of the runway, in a ditch.

Respectfully (to the coders), unless there is something amiss with my configuration (a distinct possibility), P-factor should be detuned.

Then I misunderstood! I read your original post and took it to mean you were veering to the left prior to becoming airborne.

Sorry about that!

[grafspee]

3 hours ago, Kalashnikov63 said:

No.  My takeoffs start as you're describing, then transition to a zero AoA segment, with the tailwheel aloft, a few meters of altitude are gained, and then I pitch up.  The exact same takeoff on the P-51 result in a normal takeoff; on the P-47, unless I walk on eggshells, I end up on the left side of the runway, in a ditch.

Respectfully (to the coders), unless there is something amiss with my configuration (a distinct possibility), P-factor should be detuned.

I will test that, i miss understand that too. Maybe they changed something since i have flown P-47.

I did test it right now and P-47 fly as always at take off it barely needs any inputs and after lift off if i gentle pull stick to rise nose nothing abrupt happens unless i pull hard then i stall very quickly.

My guess is you take off too early and stalls it after take off this is only explanation i can think off ofc if you have your controls properly setup. If you are doing non turbo take off P-47 needs a lot more run way then P-51 and it climbs not as good as P-51.

Edited 7 hours ago by grafspee