this has been requested by [at]atgeirr.
Note: The FlowLinearSolverVerbosity, FlowNewtonMaxIterations and
FlowNewtonMinIterations parameters are still prefixed because they
clashes with parameters registered deeply within eWoms.
this has several advanges:
- a consistent and complete help message is now printed by passing the
-h or --help command line parameters. most notably this allows to
generically implement tab completion of parameters for bash
- the full list of runtime parameters can now be printed before the simulator
has been run.
- all runtime parameters understood by ebos can be specified
- no hacks to marry the two parameter systems anymore
- command parameters now follow the standard unix convention, i.e.,
`--param-name=value` instead of `param_name=value`
on the negative side, some parameters have been renamed and the syntax
has changed so calls to `flow` that specify parameters must adapted.
We introduced two runtime parameters for this: ilu_redblack and
ilu_reorder_spheres. If the last one is false, we try to preserve
the ordering within in the colors. Otherwise we try to achieve a D2
(alternative diagonal) ordering.
These versions are inspired by the ones used in SuperLU and the enums
to choose them have simuilar names, but without leading S (MILU_1-MILU_3).
The following variants are supported (chosen by the enum MILU_VARIANT):
ILU: plain ILU
MILU_1: lump diagonal with dropped row entries.
MILU_2: lump diagonal with the sum of the absolute values of the dropped row
entries.
MILU_3: if diagonal is positive add sum of dropped row entrires. Otherwise substract them.
MILU_4: if diagonal is positive add sum of dropped row entrires. Otherwise do nothing
Using the parameter ilu_milu=true|false (default=false) the user can now choose
to use the modified ILU0 decomposition. If selected values will
not be dropped for nonzero entries but added to the diagonal of U.
This approach will result in A*e = L*U*e for vector e with all entries
beging 1.
The approach is inspired by Geiger's system-amg but we use dune-istl
aggregation AMG for it. On the fine level all unknowns attached to a cell
form a matrix block and are treated fully coupled. To form the first
coarse level system we use only the pressure component to guide the aggregation
and neglect all other unknowns on the fine level. All other level are formed
in the usual way by scalar aggregation.
Currently,it has to be requested for flow_ebos manually by passing
"linear_solver_use_amg=true amg_blackoil_system=true" to it.
- restrict pressure changes. Set default to 1.0 (this also effects flow)
- change default number of linear iterations to 150
- tell stabilized newton the residual occilates even if it occilates in
only one phase (this also effects flow)
- avoid problems realated to division on small numbers
Tested on SPE9, norne and Model 2 with significant improvments.
- For some cases (for instance involving solvent flow) the reasoning for
only adding the pressure derivatives seems to fail. As getting the
sparsity pattern is non-trivial, in terms of work, the full sparsity
pattern is only added when specified by the parameter
"require_full_sparsity_pattern"
- For solvent runs "require_full_sparsity_pattern" defaults to true for
all other runs the default is to only extract the sparsity pattern from
the pressure derivatives.
the Iteration for a fixed number of cell variabled which is then used by the
NewtonBlackølInterationInterleaved class via a switch-case over the actually existing
numbers.
When running Norne with the interleaved solver sometimes exceptions
(diagonal matrix block is not invertible) occur for some rows in the
ILU decomposition. In a parallel run this means that some, but not all
processes will see the exceptions. This leads to a classic deadlock.
With this commit we catch the exception during the setup of the preconditioner,
determine with the other processes whether there were any exceptions, and
in this case all the processes will throw an exception.
Currently this limited to Dune::MatrixBlockError, but we could extend this.
It holds a reference to the sequential code. Previously this reference
point to a temporary object that was deleted upon exit of
constructPrecond.
With this commit use a unique_ptr to store the parallel
preconditioner. It also gets a custom deleter that will delete the
nested sequential iterator during destruction.
The current implementation avoids the using of formEllipticSystem() and
vercatCollapseJacs(), which take a significant amount of computing time
during the non-linear solutions.
