stick-fixed vs stick-free stability
Jim,
You have to be careful about mixing stick-fixed stability and stick-free stability. Very different.
For stick fixed stability, the incidence angles don't come into play, except for some non-linear effects that are small. The RV-8 has plenty good stick-fixed stability at aft c.g. plenty good "static margin". The only effect of incidence is on the trim position of the elevator.
Stick free stability is effected by where the control surface naturally floats, i.e. zero hinge moment, and how that natural floating position changes with angle of attack. Where ever it floats to, it still exerts a force at the hinge (but no moment) that still influences the airplane moment balance. At the same time, the stabilizer is contributing to the moment balance. With large gradient of downwash on stabilizer with respect to angle of attack (d_epsilon/d_alpha), the stabilizer is not able to create a nose-down trimming moment if angle of attack increases. If the elevator float angle doesn't allow it to make any lift, then you will have no stick-free stability.
As a pilot, what you sense is that as the airplane slows down, you have to push forward more on the stick to keep it from continuing to slow down -- that is, the normal change in trim with speed is reversed.
In addition to stabilizer incidence, the other features that have a big effect on this are elevator camber, and elevator trim tab position. This is because the camber changes the float position as angle of attack and speed changes. As the airplane slows down, there is less dynamic pressure, so the moment due to camber gets smaller. With positive camber, or nose-up trim-tab setting, the float position of the elevator becomes less negative deflection as the plane slows down, and contributes less nose-up moment from it. So the stick force gradient will tend to be more positive (even though the stick force to trim may be forward pressure, it would take less forward pressure as you slow down). Conversely, if you have nose-down trim tab, and you need some back pressure to trim, the nose-down moment contribution from the elevator gets weaker as the plane slows down, so you have to push more.
This may seem counter-intuitive -- you are accustomed to having the trim tab trim the airplane, and the natural change in trim with speed that is caused by the stabilizer is dominant. This apparent reversal of trim change with speed only happens when the wing downwash greatly diminishes the trim contribution from the stabilizer. It is a 'feature' of very low aspect ratio wings.
At the face of it it sounds like what one would classically expect to happen in an airplane with the CG slightly behind the controls free neutral point. From what I've read the factory cg limits are based on fairly low static margins and given the manufacturing differences in a homebuilt aircraft it is quite possible to see how some aircraft could have lower/slightly negative margins when flown to the factory limits.
I find it interesting that in most of these discussions the tail/wing incidence angles keep working there way in. The control gradients are driven by the change in moments with angle of attack. These characteristics are unaffected by the incidence angle of the wing and tail (ie they don't affect the stability) as near as I can tell from looking at the equations, unless there is some non-linear effect that is escaping me.
Perhaps someone could enlighten me as to what I am missing.
Thanks
Jim