Spatially-varying grain bulk modulus within a region and Biot coefficient as a user-defined parameter (?)

[CusiniM]

In this PR:

• [x] Make grainBulkModulus a property varying cell by cell
• [x] Remove dependency of biotCoefficient on the grainBulkModulus and force it to be a user-defined parameter.

Personally, I don't particularly like the second point. Quoting this work: "The Biot tensor b is a function of Ce (drained elastic tangent moduli tensor of the solid skeleton) and Ks (bulk modulus of the solid mineral), indicating the portion of elastic deformation taken by the pore space out of the total elastic strain of the solid skeleton, which explains why b is paired to de." Thus, I think that the Biot coefficient (actually the Biot tensor) has a strong link to the choice of the solid model and it should be computed based on that. Thus, specifying it will force us to check for consistency.

Additionally, if we start introducing dependencies of the mechanical moduli on the state variables then, we will have to also update the Biot tensor accordingly which comes a lot more natural if it is not fixed.

If an heterogeneous distribution of the Biot coefficient is needed it should be easy enough to obtain it by specifying an heterogeneous grain bulk modulus.

I think there would be more value in modifying the code so that the Biot coefficient can be treated as an actual tensor as it should be.