Dynamic roll during right sideways translation.

[tech_op2000]

Bug:
The Apache has an odd rolling characteristic when translating right. If right cyclic is applied it will begin a roll to the right that will accelerate until the helicopter is inverted. Even if cyclic is removed immediately. It does not exhibit this tendency in any other direction as far as my testing has revealed.

It feels like a bug, but I don't know if it could be just some odd physics caused by the tail rotor.

Steps to reproduce:
1. Enter a hover
2. begin translating to the right, about 20 knots is plenty.
3. add full right cyclic then immediately remove it
4. "Do a barrel roll Fox!" (Star Fox reference)
5. perform same action with a left translation.

Here is a short demonstration track:  (2.7.11.22211)
AH-64D Rolling Right.trk

[bradmick]

No.1 I would ask: Why are you slamming full left/right cyclic at a hover, particularly in a lateral hover.
No.2 I can reproduce this but it requires me to be at around 40 knots or so, and allow the sas to saturate on the yaw axis.

What is happening is the sas is trying to keep the nose aligned with whatever your previousl set heading was. I set the aircraft up perpindicular to the runway in the hot start instnat action mission and then initate a lateral hover. As the aircraft continues to build speed the SAS tries to keep the nose aligned, eventually (without pilot intervetion) the SAS saturates or just runs straight into its limit and effectively causes 'full left pedal' to be applied. Slamming cyclic full right and then left excacerbates this. This is something that is literally not done in the aircraft, and shouldn't with the sim also. I can altogether avoid this saturation event by pressing and holding the force trim interupted and commanding the SAS to center. 

 

Bottom line: Rapid cyclic movements and failing to periodically retrim the SAS in a lateral hover is causing the SAS to run out of authority and functionally apply 'full left pedal'. Since the thrust is increased on the tail rotor when this happens (and trust is to the right, and the high mounted tail rotor imparts a rolling moment to the aircraft) an aggressive right roll develops. Once the weight is over the CoG the aircraft will roll and you won't be able to recover it.

[tech_op2000]

2 hours ago, bradmick said:

No.1 I would ask: Why are you slamming full left/right cyclic at a hover, particularly in a lateral hover.

That would be because I have negative transfer of learning from acrobatic RC helicopters, which don't mind full control inputs and can handle all sorts of flight orientations much better than full scale helicopters. Also games with worse helicopter physics where a helicopter doesn't mind flying sideways at high speed (cough....Battlefield....Cough)

Generally, it was happening to me when I got surprised by enemy fire and tried to do an evasive maneuver to avoid AAA. I know, I know, The smart answer is to not get into that situation in the first place. The goal was to keep nose on target to engage with hellfire's or rockets without presenting a stationary target for return fire. But I do realize that is not a realistic tactic, better to turn in, fire, then turn out.
 

On a side note, RC Heli's do this really well.

 

2 hours ago, bradmick said:

Bottom line: Rapid cyclic movements and failing to periodically retrim the SAS in a lateral hover is causing the SAS to run out of authority and functionally apply 'full left pedal'. Since the thrust is increased on the tail rotor when this happens (and trust is to the right, and the high mounted tail rotor imparts a rolling moment to the aircraft) an aggressive right roll develops. Once the weight is over the CoG the aircraft will roll and you won't be able to recover it.

I went back in and did more after reading this. Including doing the maneuvers with SAS disengaged. It does sound like the SAS is partly to blame. It also seems the physics of trying to fly a heli sideways are also at fault, instead of any sort of sim bug.

I think I could explain it this way after my further tests: For a conventional helicopter such as the apache to translate to the right 90 degrees to its heading, it must provide increasing rudder power to overcome the rotor torque and the aerodynamic force of the relative wind attempting to weathervane the helicopter. Because the tail rotor is above the center of mass this induces a right rolling force that must be corrected with cyclic. The faster it translates right, the greater all these forces. 

If at this point, the pilot moves the cyclic aggressively right. It will remove the rolling force the cyclic was applying to stop the roll and even add some rolling force in the same direction as the tail. additionally, it will cause the helicopter to yaw to the right some. The SAS in an attempt to arrest the yaw rate increases tail rotor thrust causing more right rolling force. Once the roll begins, the amount of opposite cyclic that would be needed to arrest the roll would be massive. This is probably why I realized the first couple times I encountered it that I had more success just riding it out then recovering once the helicopter is back upright after the roll.

I could also reproduce it in the other direction by applying opposite rudder right as I initiate the roll, the rolling force of the adverse rudder helps it roll around much faster. In the same way, neutralizing the rudder at the start of the maneuver did make it roll about the same left or right. 

My takeaway from this is:
-It isn't a bug, it is instead a side effect of how a conventional helicopter counteracts torque and the aerodynamic forces pushing on a helicopter in sideways flight.
-Full size helicopters are not great at translating sideways at high speeds. Thus rapid sideways translations should be avoided.
-In helicopters where the rotor spins as it does in the apache, the effects of this will be much worse to the right than to the left just because of the physics involved.

Edited April 15, 2022 by tech_op2000