Heli torque linked to collective.

[StuLes]

Heli related stuff.
Engine torque is related to the use of the pedals. This is mostly working correctly.
But, it is also somehow linked to the collective, which is not correct. IE during normal operation if I raise collective, the rotor drag is increased and the engine power is increased, and I need to add more pedal to compensate.
Now, if I remove the engine (engine failure), torque is quickly removed, but the accompanying reaction is not forthcoming. ie there is no torque reaction to the loss of torque. Instead the antitorque reaction only occures with the lowering of the collective. THis is wrong. Subsequent raising and lowering of the collective during the ‘auto’ has an incorrect torque reaction. There is no engine, and hence not torque.
Anyone have any insight into this model behaviour?

Is this a 'settngs' issue, or an MS issue?
Thanks
Stu

 

[=rk=]

Is this a 'settngs' issue, or an MS issue?

I think you maybe left out some alternatives. The collective changes the blade angle of attack and I think you're mistaken about there being "no torque," because in your scenario there is no engine torque. As long as the blades are spinning, there is a sliding scale quantity of rotational torque and if you don't believe so, just reach up there and try to stop them with your hand. In fact, this is the torque which makes autorotation possible. Also and FYI, the pedals are connected to the tail rotor and are used to control yaw. There is no direct connection to the engine and any effect on engine torque is the consequence of vertically elevating the helicopter via collective.

 

[StuLes]

Hi Rick
I appreciate you entering into discussion about this.
The sprag clutch on a smaller heli, or freewheeling unit on a larger heli, isolate the rotor from the gearbox/eng combo. So, during the auto, the rotor is just spinning (freewheeling) with no way to transfer torque back to the body of the heli. as this is one way bearing that can trasfer torque to the rotor, but not recieve torque from the rotor. Change pitch, and the rotor will speed up or slow down. Gyrocopters are in perpetual auto rotation, and have no tail rotor, as one is not needed. They have a two way bearing, ie no torque to or from the rotor.
So I might have just answered my own question. The freewheeling unit on the MS model is not freewheeling, and hence feeding back torque where it shouldn't be doing so.
It works on piston heli's but for some reason, not on turbine heli's.
Regarrds
Stu

 

[=rk=]

Glad to think I helped. Imagine trying to manipulate MSFS physics to edit the helicopter tail rotor model into representing a biaxial rotor servo flap controlled aircraft with hand and foot controls, which is set up to mimic traditional helicopters. It's pretty hard to get it to to rotate evenly in both directions without initiating some other unanticipated consequence, let's just say my version does an even better job of mimicking traditional helis, especially when you turn off assists.

Same thing with the landing gear algorithm, the idea that anyone at Asobo has ever adjusted an actual shock absorber and then translated that to aircraft suspension settings demands an extremely powerful graphics card to simulate.

 

[StuLes]

"grand prix legends' was a sim driving game in the 90's that had adjustable spring rates, spring travel, bumpstops, damping and rebound damping all modelled in and tunable. So I guess it just needs the priority and desire to model it in.

With regard to the freewheeling/sprag clutch not working, what is the proceedure for getting MS to address that item?

This setting is a dud.
This one does not work as it should ie the rotor can turn the engine, which it shouldn't be able to do. Changing values does nothing and can be verified on 'debug eng' page as you cannot get the torque to go to zero.
clutch_maximum_torque_up = 0.0

This setting is good.
This one, the engine can turn the rotor, which it does. Can be verified by reducing to zero, and the engine can no longer drive the rotor. ie torque goes to zero.
clutch_maximum_torque_down = 1000.0

Stu

 

[=rk=]

Ok at the turn of the century just before my son was born I was tuning cars in "Gran Turismo." decades earlier I had made these same tunes to my own cars for actual track racing. It seemed incredibly accurate, wheel camber was modeled, drift patterns between fwd, 4wd and rwd and along came this tuner/racer that was winning all the races and all he did to tune was that he simply jacked all his rear ends as high as possible and cranked the front suspension down to no travel. Basically he made a wedge with his cars and the problem was that the "aerodynamic model" had over exerted itself against the physics model. I know I raced and such tunes are undrivable in RL - but they work in simulation.

With regard to the freewheeling/sprag clutch not working, what is the proceedure for getting MS to address that item?

I don't believe there is a direct procedure, there could possibly be a technique. To explain I have to go back to my understanding of suspensions, because I never flew helicopters as you have. The suspension model is like "taffy." It sort of holds its shape and if it's shape is necessary, you can support that with settings, but if you push it or stress it, it falls apart or melts. In an actual "automobile," you can remove the shock absorber, you can place a stiffer spring, etc. In the suspension model, every individual setting affects every individual aspect to some varying degree. Therefore if you change damping in the front, the mains will also change to some degree.

I imagine the engine tuning model is also leaking into itself similarly. My situation was exacerbated because I had some geometry intersecting the ground that I had to set up like suspension to get the sim to accept it, so I had two "wheels" very close together that were literally fighting each other. My technique and this is what I recommend you consider, is that instead of perfecting each leg of the suspension as I would in RL, I went around each contact point and made a small adjustment, then went to the next, etc. Each pass I made successfully, I moved further away from the "tuning" style and closer to the "wedge" style, but it worked for me.

Another thing to consider is nomenclature. I do not understand the term "clutch_maximum_torque" in terms of "up" or "down." Torque as an absolute, no problem, "the clutch holds until the torque limit is passed, at which point the clutch slips." But up torque and down torque? Up as in pounds per square foot, or up as in elevation? Different torques engaging and disengaging? It's my suggestion that Asobo faked it too. My final idea about the clutch situation is to again refer to actual clutches. In fact, they never actually fully disengage. What happens is that the drive force gets offloaded from the friction pads onto a much more slippery part of the clutch mechanism. The drive never actually gets freewheeled, but there is far less torque getting through the clutch to the driveshaft. It's almost unnoticeable, unless one is turning the entire drivetrain by hand and I am suggesting that if you massage your settings in, you may be able to coax this clutch_maximum_torque_up into cooperating.

 

[StuLes]

I decided to try to get the community of 'heli developers' and realism enthusiasts to upvote the solution. Please 'vote'

Sprag clutch for Helicopters

Currently, helicopter developers are hindered seriously by the lack of a ‘Sprag clutch’. Current ‘workarounds’ are unsatisfactory. What is a "sprag clutch’. A sprag clutch is a one-way It resembles a roller bearing but, instead of cylindrical rollers, non-revolving asymmetric figure-eight...

devsupport.flightsimulator.com

 

[=rk=]

It kind of looks like you maybe got through. This was on this months Twitch Developer Stream. If you weren't in attendance, they told us, "yes, it will be much better, our actual helicopter pilot testers have informed us so."