fluids: Use standard boundary integrals for primitive

fluids: Make STG QFs primitive compatible

fluids: Make stg components dependent on primitive

fluids: Make newtonian boundary QFs primitive compatible

fluids: Change newt->primitive to is_primitive

fluids: Add missing \n to blasius warning message

Fluids - Newtonian with Primitive variables (#1011)

* Fluids - Initial commit for Newtonian primitive variables

* Fluids - include A0 (dU/dY) in the output

* Fluids - added ICs for IG in primi

Fluids - Newtonian with Primitive variables (#1011)

* Fluids - Initial commit for Newtonian primitive variables

* Fluids - include A0 (dU/dY) in the output

* Fluids - added ICs for IG in primitive variables

* Fluids - added Jacobean QFunction for primitive variables

* Fluids - added RHS QFunction for primitive variables

* Fluids - fixed compilation errors and warnings

* Fluids - added ICs for density_current with primitive variables

* Fluids - In/OutFlow BCs for channel in primitive variables

* Adding the missing parts after rebasing onto main

* Fluids - Use correct component names for primitive variables

* Fluids - Primitive variables supported only with implicit scheme

* Fluids - drop in/outflow for channel flow and call Exact_Channel_Prim() in Exact_Channel()

* Fluids - Set solver's QFunction data in an if-else statement

* Fluids - style

* Fluids - add a comment to explain why the the gravity body force is excluded from the potential energy.

* Fluids - Exact_Channel return State

* Fluids - density_current: some style and cleanup

* Fluids - DC: refactor & cleanup

* Fluids - Singel QFunction for prim&cons

* Fluids - Use absolute temperature

* Fluids - DC: Fix pressure

show more ...

tests: Shrink fluids test run time

doc: State output_freq=-1 -> output final state only

fluids: Force final output except if output_freq=0

fluids: Skip I/O if output_freq == 0 (#1027)

Merge pull request #1018 from CEED/jrwrigh/stg_precompute_ektot

examples/fluids: skip legacy preallocation (we use COO)

When using -snes_fd_color or running a model that doesn't provide a
Jacobian (in which case coloring will be used automatically), we still
nee

examples/fluids: skip legacy preallocation (we use COO)

When using -snes_fd_color or running a model that doesn't provide a
Jacobian (in which case coloring will be used automatically), we still
need to preallocate.

show more ...

examples/fluids: explicitly handle MFFD

examples/fluids: use shell for Amat, block diagonal aij* for Pmat

This adds two independent options for tuning solves.

-amat_type shell: causes Amat and Pmat to be different matrices, with shell
us

examples/fluids: use shell for Amat, block diagonal aij* for Pmat

This adds two independent options for tuning solves.

-amat_type shell: causes Amat and Pmat to be different matrices, with shell
used for Amat. It's technically possible to use other matrix formats
here.

-pmat_pbdiagonal: causes Pmat to be assembled as point-block diagonal.
Note that this usually only makes sense in addition to -amat_type shell
because otherwise your KSP is only working with the point-block diagonal.

show more ...

fluids: Store 1/Ektot

fluids: Implement stgdata for ektot

fluids: Add STG Preprocessing and Calc QFs

fluids: Abstract out spectrum calculation

examples/fluids: use DMGetCellCoordinateDM to handle periodicity

Note that without -dm_sparse_localize 0 (default=1), the cell DM will
only have cells with localized coordinates (and there is no vec

examples/fluids: use DMGetCellCoordinateDM to handle periodicity

Note that without -dm_sparse_localize 0 (default=1), the cell DM will
only have cells with localized coordinates (and there is no vector
representation that contains both).

show more ...

Merge pull request #1009 from CEED/jrwrigh/dirichlet_with_libceed

Fluids - Use libCEED to compute Dirichlet boundary conditions

examples/fluids: fix spelling in output

fix(fluids): Create and use basis_xc_sur for dirichlet bcs

fluids: Minor improvements for dirichlet and stg

fluids: Set all outputs in STG QF