Follow-up from "Restructuring of the TRM process: Split constitutive setting"
The following discussions from !4198 (merged) should be addressed:
-
@endJunction started a discussion: (+1 comment) I'd suggest to call this implementation from the one with member-access to avoid code duplication.
- Fixed in this MR: !4222 (merged).
-
@endJunction started a discussion: (+1 comment) A member is initialized here, but not in the same structures above (like CapillaryPressureData). I'd not explicitly initialize data if not needed (Eigen has different initialization strategies controlled at compile time), especially initialization to zero could be hiding some initialization issues later; nan would be better.
A comment could resolve the issue too...
-
@chleh started a discussion: Ordinary state tracking of chi(S_L) will be implemented in a follow-up MR.
-
@chleh started a discussion: RM test deformation dependent porosity swelling. (follow-up)
-
@TobiasMeisel started a discussion: Would:
BiotData BiotModel::eval(SpaceTimeData const& x_t, MediaData const& media_data) const
be possible? -
@TobiasMeisel started a discussion: (+1 comment) I guess
auto&& solution
is coming from non-copyable return type of integrateStress? Could mechancisBases be a shared resource? That would allow clients to decide and not to force how to bindsolution
-
@TobiasMeisel started a discussion: (+4 comments) init values?
-
@TobiasMeisel started a discussion: (+1 comment) yes, to the ToDo comment! It took some time to understand what is special about mutable transport_poro_data .
-
@TobiasMeisel started a discussion: (+1 comment) I would prefer any possible change from
eval(some_mutable_state&)
tonew_state = eval(const_prev_state)
. Maybe for now just adding a ToDo is also fine.