An action constraint is just a way of applying an animation (or action, as it’s called in Blender. Throughout this post, wherever you see the word animation you can substitute it with the word action).
Normally, for example, you’ll create an animation by inserting keyframes for the various bones (for objects with an armature) and for the objects (that don’t have an armature). Then you hit ANIM and ooh and aah.
An action constraint just allows you to use that animation as part of another, larger animation.
One animation may be of the wing folding.
Another might be of the wing swiveling back, up, and forward to grab air, then thrusting down to push air.
When you use an action constraint, you have a little bone off to the side somewhere which indicates how far through that animation you are by its rotation angle. For example, angle 0.0 could be wing in normal position, and angle 90.0 could be completely folded. Angle 45.0 would then be halfway through the ‘folding’ animation. Another little bone off to the side somewhere could be similarly used for the ‘flapping’ animation.
An important thing is to keep the ‘little bones off to the side’ (action constraint targets) in an orthogonal view before rotating them. Otherwise rotating will have less than the desired effect.
To answer your post, yes, you can pose the bones any way you like in 3D space, just as in any other animation.
[ ] each action has a good (descriptive) name
[ ] the values for an action constraint allow full range of motion
- min and max are set to something like [-90, 90] or [0, 90] or [90, 0] (as appropriate)
- start and end are set to the first and last frame of the named action
- rotation axis is set to the appropriate axis. For example, if you want to
look at the action target bone from the front view when manipulating it,
set the axis to Z.
[x] all bones with an action constraint targeting the same bone have the same values
[ ] you only rotate the action target bones from front/side/top (as appropriate) view.
Hope this helps.