~fluidity-core/fluidity/darcy_weak

~fluidity-core/fluidity/darcy_weak_bcs » Changes from revision 4110

From Revision 4087 to 4068

Rev	Summary	Authors	Date
4087	Add an option have_dual_pressure. Add an option have_dual_pressure. Correct a couple of bugs for the dual model. Only use the block pressure matrix when really required as it is bugged in 2d.	Brendan Tollit	11 years ago
4086	update makefile dependencies update makefile dependencies	Brendan Tollit	11 years ago
4085	after each explict solve of saturation for a subcycle the pa... after each explict solve of saturation for a subcycle the parallel halos need updating. Makefile.dependencies requires updating	Brendan Tollit	11 years ago
4084	Add Jackson2Phase and opposite 2 phase relperm correlation. Add Jackson2Phase and opposite 2 phase relperm correlation. Adds a scaling coefficient argument to the functionality	Brendan Tollit	11 years ago
4083	Add functionality to Sparse_Tools_Petsc.F90 to add a scalar ... Add functionality to Sparse_Tools_Petsc.F90 to add a scalar field to the diagonal of a specified block of a petsc_csr_matrix Add to darcy impes the lambda*cv_mass matrix in the pressure equation and the saturation sources for dual model. These terms assume pressure is control volume wise, or that a lumped mass matrix is used. A full consideration of pressure as FE wise requires functionality not there in femtools.	Brendan Tollit	11 years ago
4082	Change the pressure assemble to use the petsc_csr_matrix. Change the pressure assemble to use the petsc_csr_matrix. Adjust femtools to include a petsc_solve routine that will take the petsc_csr_matrix and a scalar_field_pointer array. Similar adjustments for the dirichlet BC's. Include in Petsc_Tools functions to copy to and from a petsc vector and a scalar_field_pointer array. Seems to work fine for 1d and 3d. For 2d although the answer may be correct petsc keeps complaing about trying to insert a number into a row-column it doesnt know about. This can start to happen part way through into the simulation (after a few timesteps). This slows it down alot. Looks like a bug, just cannot see if this is in the Darcy code, femtools or petsc.	Brendan Tollit	11 years ago
4081	Add a new variable to darcy_impes_type called pressure_matri... Add a new variable to darcy_impes_type called pressure_matrix that is type petsc_csr_matrix. This is allocated only for the prime di and takes account if it is a dual simulation by making it a /2,2/ block. Currently not used.	Brendan Tollit	11 years ago
4080	Break the assemble and solve routine up into three parts. Break the assemble and solve routine up into three parts. Part one is before the pressure solve, part two is the pressure solve, and part three is after. Set up calls in the main to calc the diagnostics for di_dual if required.	Brendan Tollit	11 years ago
4079	Point the di%transmi in the same place the other pressure of... Point the di%transmi in the same place the other pressure of porous media are pointed, not in the initialise routine as this only takes one di.	Brendan Tollit	11 years ago
4078	added a pointer for transmissibility in the dual porous medi... added a pointer for transmissibility in the dual porous media in the main file	Peyman Mostaghimi	11 years ago
4077	Similar to the last commit add a subroutine to calc the pres... Similar to the last commit add a subroutine to calc the pressure local surface matrix and rhs, set up a bit for the dual model also.	Brendan Tollit	11 years ago
4076	added a subroutine for pressure volume element local assembl... added a subroutine for pressure volume element local assemble	Peyman Mostaghimi	11 years ago
4075	Adding transmissbility stuff in assemble file Adding transmissbility stuff in assemble file	Peyman Mostaghimi	11 years ago
4074	Fix some if blocks in Darcy_IMPES.F90 Fix some if blocks in Darcy_IMPES.F90 Merge trunk into branch test cases still pass.	Brendan Tollit	11 years ago
4073	Change TransmissibilityLambda to TransmissibilityLambdaDual Change TransmissibilityLambda to TransmissibilityLambdaDual	Brendan Tollit	11 years ago
4072	Add transmissibility_lambda_dual as a field to the di type Add transmissibility_lambda_dual as a field to the di type	Brendan Tollit	11 years ago
4071	Correct small spelling mistake in error message and field Correct small spelling mistake in error message and field name in extract for di initialise	Brendan Tollit	11 years ago
4070	Added transmissibility_lambda_dual to the main file and rena... Added transmissibility_lambda_dual to the main file and rename the transmissibility option in the schema to transmissibility_lambda_dual	Peyman Mostaghimi	11 years ago
4069	Conflicts resolved Conflicts resolved	Peyman Mostaghimi	11 years ago
4068	Fix some space style in Darcy_IMPES_Assemble Fix some space style in Darcy_IMPES_Assemble In Populate_State allocate the necessary fields for dual perm/poro Like prime per/poro fields this inserts the fields in state(1) and then alias in all other states to these. This requires a different name for the dual perm and porosity such that the schema has changed. I could not see anyother way around this. This will need to be accounted for in the darcy initialise routine which can easily be done with a logical argument which will set a post append character to either "Dual" or "".	Brendan Tollit	11 years ago