I work with a Ball on Plate scenario and I observed, that If I set the rolling_friction_coefficient = 0.0 that there is also no
static friction = friction_coefficient*static_friction_scale*normal force.
I use the ScaledBoxFast friction model for my ball and my plate. I used
compliance = 0.0001
static_friction_scale = 1.1
friction_coefficient = 0.8
I adjusted the same parameters for both materials.
In the experiment, my Ball directly starts to roll if my plate is inclined (as if there is no static friction).
This behaviour is actually not how I would expect it.
Perhaps I'm misunderstanding the problem, but a ball should roll freely even if it has static friction.
Rolling friction (or more specifically, friction in the primary and secondary angular directions) is the only thing that will resist rolling down a slope. So if that rolling friction is not active, I would expect the ball to roll freely.
Static friction only resists sliding at the contact surface. It increases the friction coefficient while there is no sliding at the contact. Basically, using the Coulomb friction model:
Kinetic friction coefficient (μ_k) = friction_coefficient
Static friction coefficient (μ_s) = friction_coefficient * static_friction_scale
So if you have a box sliding down a plane, the amount of force resisting the sliding will be higher before the box moves, and lower once the box is sliding. But for a ball rolling, the static friction will resist sliding at the contact point, but will not prevent the ball from rolling.
If I've misunderstood the problem, let me know.