It is explained here that adverse yaw is caused by the difference in induced drag due to the difference in lift being generated by each wing.
I've also seen a couple of websites (including Wikipedia) that argue that this explanation is wrong. They explain the cause of adverse yaw as being the change in the direction of the lift vector during a roll, or what some refer to as the "twisted lift" concept. They usually have a diagram like the following:
The twisted lift explanation discounts a difference in induced drag due to a change in the amount of lift. They contend that lift and drag remain the same but they twist in different directions. The descending wing has a slightly higher local AoA and the rising wing slightly lower. Since lift is perpendicular to the local airflow the lift vector rocks forward on one wing and aftward on the other.
This explanation seems flawed to me.
- First of all, in the diagram they show the force on each wing being the same. If that were the case the aircraft wouldn't roll at all, it would ONLY yaw.
- Second, the vector of lift being perpendicular to the airflow is correct, but that is really just an arbitrary division of the total aerodynamic force. The yaw axis of the plane is not dependent on airflow. It would be dependent on the portion of the force vector parallel to the wing chord, aka the axial force. As shown in another question, depending on a number of factors the force vector in relationship to the wing does not necessarily move forward with an increase in AoA.
- I can't come up with solid numbers, but trying to figure this out for normal flight speeds it would require a very fast roll to even make a 1° difference in local AoA. I ran numbers for a 3° per sec roll at 120 KTAS on a 36ft wingspan (C172) and came up with a change in AoA of about .27°, and that Is only at the wingtip. It would decrease the closer it got to the fuselage. It doesn't seem like it would be enough to cause a significant change in vector and a resulting yaw moment. Or at least it would be secondary in magnitude to the difference in the drag due to reduction of lift.
Since the changes in lift are being created either by changing the effective camber of the wings (ailerons) or by altering the trailing edge flow (spoilerons) rather than a change in AoA I don't think it's quite so simple to predict that the direction of the aerodynamic vectors will necessarily move in opposite directions as this theory assumes.
If this explanation is correct can someone show me what I'm missing?