Comment 8 for bug 1734653

I believe we are somehow converging.

Just a couple of additional remarks.

1 - Regarding the area associated to the broken edges:

If you look at the attached drawing, you'll see that we have actually 2 cases:
1) the 2 broken edges are associated to 2 individual cracks
2) the 2 broken edges are associated to the same fracture plane
In 1), I agree with you and the triangular surfaces are not coplanar
In 2), the triangular surfaces should be coplanar to account for the fact the the broken edges belong to the same fracture plane.
If we are concerned with the calculation of Vo, case 2) should prevail since we are dealing with pre-existing fracture planes rather than individual cracks but it would probably be easier (and more consistent?) to use the same logic for fracture planes and cracks so we should probably go for case 1). What do you guys think about that?

2 - Regarding the cubic law for computing the conductivity:

The cubic law is based on the concept of rectangular parallel plates of length L and width W (cf. drawing) and I don't know how we could adapt it to the case of triangular plates. Considering rectangular plates instead of triangular ones seems like something reasonable in our case. If you agree on this assumption, then we need to decide how to compute W. As a first guess, I would go for W equal to the distance between the middle of the broken edge (location of the crack) and the barycentre of the facet. What do you guys think about that?

Luc