Markus Blatt
4527ce8ffd
Add access to the underlying information about the parallelization.
...
We need it serveral places and all of them seem to have access to
NewtonIterationBlackoilInterface. This makes it natural to give access
to it and prevent users from having to forward it manually at several
places in the simulator driver.
2015-02-12 10:41:43 +01:00
Markus Blatt
221565f038
Enable the use of parallel dune-istl solvers.
...
As with opm-core we use boost::any to provide additional
information about a parallel run. It is used to set a
ParallelISTLInformation object and and fill it with the
information obtained from a parallel Cpgrid.
Note that the simulator currently compiles sucessfully. Still,
we have to test the runs and do debugging.
2015-02-12 10:41:43 +01:00
Joakim Hove
02b682ea52
Merge pull request #297 from qilicun/support_minpvf
...
support MINPVF.
2015-02-12 09:34:35 +01:00
Liu Ming
3ddb930d1f
Adapt to api changes of opm-parser.
2015-02-12 16:16:07 +08:00
Atgeirr Flø Rasmussen
bed7bada40
Merge pull request #286 from dr-robertk/PR/add-support-for-restart
...
BlackoilOutputWriter: added support for backup and restore.
2015-02-11 14:05:24 +01:00
Robert K
0f20e187a6
typos and minor changes.
2015-02-11 13:26:57 +01:00
Robert K
61144a73af
Vector --> Container since we also read/write strings.
2015-02-11 13:18:07 +01:00
Robert K
2520aab0a1
added warning when restore file couldn't be opened.
2015-02-11 13:18:07 +01:00
Robert K
c6400fd30a
SimulatorFullyImplicitBlackoilOutput: warn when end of file was reached.
2015-02-11 13:18:07 +01:00
Robert K
a834e55f5a
BlackoilOutputWriter: added support for backup and restore.
...
This first implementation is mainly to support faster debugging.
2015-02-11 13:18:07 +01:00
Atgeirr Flø Rasmussen
2f632bc9c3
Merge pull request #299 from blattms/fix-dependency
...
[bugfix] Fixes dependency clause of #296
2015-02-11 12:45:00 +01:00
Markus Blatt
2be2dd5e41
[bugfix] Fixes dependency clause.
...
It caused dunecontrol to abort because of using a comma as a separator.
This commit removes the eroneous comma and adds a required version.
2015-02-11 12:35:40 +01:00
Atgeirr Flø Rasmussen
d99931ae0f
Merge pull request #298 from dr-robertk/PR/remove-warning
...
convergenceReduction: remove warning.
2015-02-10 14:44:16 +01:00
Robert K
0323624e3e
convergenceReduction: remove warning.
2015-02-10 13:08:16 +01:00
Liu Ming
6cd0c47b94
support MINPVF.
2015-02-09 11:27:10 +08:00
Atgeirr Flø Rasmussen
3b742e0093
Merge pull request #296 from blattms/require-dune-istl
...
Explicitly list dune-istl as a dependency.
2015-02-06 14:36:02 +01:00
Markus Blatt
ed36a439f8
Explicitly list dune-istl as a dependency.
...
This is in line with and closes issue #295
2015-02-06 11:28:42 +01:00
Atgeirr Flø Rasmussen
588615bd21
Merge pull request #285 from dr-robertk/PR/revision-of-output
...
Revision of black oil output by putting everything into a single class following the OutputWriter IF.
2015-02-05 15:20:35 +01:00
Robert K
2068b7ea16
Revision of black oil output. Put everything into a class following the OutputWriter
...
interface.
2015-02-05 14:39:47 +01:00
Atgeirr Flø Rasmussen
12248f5c14
Merge pull request #294 from totto82/default_bhp_wconhist
...
Hardcode bhp limit for WCONHIST/RESV to 1 atm.
2015-02-05 13:31:05 +01:00
Tor Harald Sandve
2f32afd36d
Hardcode bhp limit for WCONHIST/RESV to 1 atm.
...
TODO: Allow WELTARG to change the bhp limit
2015-02-05 09:41:37 +01:00
Atgeirr Flø Rasmussen
5b79ead6ba
Merge pull request #288 from andlaus/temperature_dependent_viscosity
...
add the glue code required for temperature dependent viscosity
2015-02-03 17:48:41 +01:00
Andreas Lauser
ff09e87787
add the glue code required for temperature dependent viscosity
2015-02-03 15:27:38 +01:00
Atgeirr Flø Rasmussen
3d8209abe4
Merge pull request #283 from blattms/restructure-newton-convergence-for-parallelization
...
Restructure newton convergence for parallelization
2015-01-30 15:38:25 +01:00
Markus Blatt
60c0f30388
Use const where appropriate and polish code formatting.
2015-01-30 15:26:55 +01:00
Markus Blatt
9d2d942396
Prevent copying of pore volume vector.
2015-01-30 15:26:33 +01:00
Markus Blatt
267bf39e4a
Use idx to access to access the phases.
...
In cases where not all phases are active, there no phase positions for
inactive phases and the computation may crash.
2015-01-30 15:23:12 +01:00
Atgeirr Flø Rasmussen
e6a053a392
Merge pull request #291 from blattms/fixes-issue-290
...
Fixes raisal of assertion in Eigen because of empty vector and matrices. Fixes #290
2015-01-30 15:12:48 +01:00
Atgeirr Flø Rasmussen
a8b715893f
Merge pull request #289 from totto82/vap_der
...
Include derivatives of vappars
2015-01-30 15:06:16 +01:00
Atgeirr Flø Rasmussen
69c1a660a8
Merge pull request #287 from qilicun/update_pv
...
Use original grid cell volume to compute pore volume.
2015-01-30 15:00:34 +01:00
Markus Blatt
7c8f992a2d
Fixes raisal of assertion in Eigen because of empty vector and matrices.
...
Apparently Eigen cannot handle empty containers during reserve correctly.
Therfore we check for the size of the vector and if it is zero simply create
empty jacobians.
Closes #290
2015-01-29 17:14:09 +01:00
Tor Harald Sandve
b57ddf1b2d
Include derivatives of vappars
...
The derivatives of the vappars are included in the Jacobian.
To avoid inf derivatives for vap<1, the oil saturation is restricted
from below by sqrt(epsilon).
2015-01-29 13:08:52 +01:00
Liu Ming
3ab3dc555a
Use original grid cell volume to compute pore volume.
2015-01-29 16:25:17 +08:00
Markus Blatt
d27e522ba4
docu fix (???) -> tempV.
2015-01-28 19:17:25 +01:00
Markus Blatt
3fcb99c27a
Fixes wrong indexing for residual_.material_balance_eq.
...
Curiously, this indexes differently from other stuff. It
uses the phase index and not the index within the phases
that are present.
2015-01-28 19:11:13 +01:00
Markus Blatt
cb9048d928
Only compute the sum of B for one phase and not over all phases.
2015-01-28 19:10:24 +01:00
Markus Blatt
f5ed31d128
Fixes rotated function parameters during function call.
...
There was a mixup in the order of the function parameters, both in
the declaration and in the actual function call. This commit resolves this.
2015-01-28 19:08:48 +01:00
Markus Blatt
0b18e6a15f
Improve documentation of convergenceReductions.
2015-01-28 16:54:28 +01:00
Markus Blatt
2209170c15
Use the already queried active phase index directly instead of looking it up each time.
2015-01-28 16:40:23 +01:00
Markus Blatt
622c8825a9
Correctly initialize the Eigen arrays.
...
Previously, we just called the default constructor. Of course this does not allocate
any memory and we experienced segmentation faults. With this patch we correctly
allocate the arrays by passing the number of row and columns to the constructor.
2015-01-28 16:38:10 +01:00
Atgeirr Flø Rasmussen
e3b0a92176
Merge pull request #284 from blattms/rename-residuals
...
Renames residuals() to computeResidualNorms()
2015-01-28 14:46:48 +01:00
Markus Blatt
20b6fb0853
Fixes typo: residuum -> residual
2015-01-27 14:05:38 +01:00
Markus Blatt
f820e1d441
Moves calculation of the total pore volume into convergenceReduction.
...
It is a global reduction in the parallel case and should therefore be
computed together with the other ones.
2015-01-27 12:49:55 +01:00
Markus Blatt
5e94e2ab08
Renames residuals() to computeResidualNorms()
...
It took me quite some time to understand the computations done
e.g. during the detection of oscillations, where the stuff returned
by residuals() is used as a vector of doubles. It turns out that
residuals() actually returns the norm of the residuals. To clarify this
we rename residuals() to computeResidualNorms() and residuals to
residual_norms. Having my dare devil day today, I even try to document the
method. (This documented method might feel kind of lonely between the others,
now;). Hopefully this saves others some time.
2015-01-27 12:04:46 +01:00
Markus Blatt
ddc2b820a7
Factors out the reduction part into a separate function.
...
Theses reductions are global and will require a different
behaviour in the parallel case.
2015-01-27 10:31:30 +01:00
Markus Blatt
0b81cd8f6f
Reduces code duplication and reorganizes code when checking convergence.
...
The computations made to check the convergence are the same for all existing
phases. Therefore this patch uses loops over the phase indices when cmputing
them,
In the convergence check there are several reductions (maxCoeff(), sum())
that will trigger communication in a parallel run. This patch seperates the
reductions from the other computations. The idea is to one reduction for the
reductions that need to done as global communication is expensive.
2015-01-27 10:31:30 +01:00
Bård Skaflestad
7dfcf7b0c0
Merge pull request #282 from atgeirr/fix-wellstate-init
...
Fix issues in WellStateFullyImplicitBlackoil::init().
2015-01-26 23:12:12 -06:00
Atgeirr Flø Rasmussen
ce3519dadb
Fix issues in init().
...
- Failed to set all rates to zero initially.
- Could set out-of-bounds index if number of controls changed.
2015-01-26 16:46:51 +01:00
Atgeirr Flø Rasmussen
56ef604773
Merge pull request #280 from dr-robertk/PR/restart-solver-when-residual-nan
...
Restart solver when residual is nan or to large.
2015-01-26 13:08:37 +01:00
Robert K
3416cba92a
added white space after comma.
2015-01-26 12:43:25 +01:00
