Addition regarding the second question: As far as i understand, if a bearing load is applied to a complete rotational inner surface, it will always be applied such that the signs of the distributed load vectors are opposed to those of the surface normal vectors (meaning that the load is always compressive). The question is whether this behavior is maintained when the load is applied to an outer surface (which would correspond to figure c) ) or is "flipped" (meaning the load would become tensile as in figure d) )
In the meantime, I have noticed that bearing loads can also be applied to circular surface segments, which raises two further related questions:
3) Will the angular coordinates of a segment be internally rescaled to a full circle (meaning that if a bearing load is applied to a half circle, it would act on only one quarter of the complete circle)?
4) For full circular surfaces, there seems to be no possibility to change the load direction from "compressive" to "tensile" or vice versa since a change of sign will simply cause the load to act in the same manner, but on the diametrically opposing side of the surface. What happens if the surface is just a segment of a full circle? Will the parts of the load arc that are acting outside the segment be set to 0 (cf. figure e) ) or will they change direction (cf. figure f) (which would enable a change of the load direction)?
In any case, I would recommend to cover these questions in the documentation and implement an option to change load direction.