OilfieldToolsNet/opm-simulators
4
0
Fork 0
mirror of https://github.com/OPM/opm-simulators.git synced 2025-02-25 18:55:30 -06:00
Code Issues Packages Projects Releases Wiki Activity
5,198 Commits 24 Branches 102 Tags 90 MiB
07e0d71906
Commit Graph

2613 Commits

Author SHA1 Message Date
Robert Kloefkorn
07e0d71906 [feature][flow] Add a common executable for all flow variants, i.e. 
flow_ebos, flow_ebos_polymer, flow_ebos_solvent, flow_ebos_2p.
 2017-08-23 16:21:15 +02:00
Atgeirr Flø Rasmussen
58eea94460 Merge pull request #1247 from totto82/removeState-PR 
Remove reservoirState from BlackoilModelEbos
 2017-08-22 20:47:16 +02:00
Tor Harald Sandve
6146190844 Fixes comments from PR 
Don't sum ghost cells in the RateConverter
Some cleaning and comments.
 2017-08-18 08:46:44 +02:00
Atgeirr Flø Rasmussen
8998207d7e Merge pull request #1253 from blattms/print-bubble-dew-point-problems-only-on-0 
Output bubble/dew point computation problems only on rank 0.
 2017-08-17 22:25:53 +02:00
Tor Harald Sandve
6bca2ea69c Fix rebase 2017-08-17 11:34:33 +02:00
Tor Harald Sandve
2bf8d48272 Remove redundant local state in ParallelDebugOutput 2017-08-17 11:13:00 +02:00
Tor Harald Sandve
593546da4b Fix 2p case in relativeChange(...) in BlackoilModelEbos 2017-08-17 11:13:00 +02:00
Tor Harald Sandve
ce84a59b29 Remove reservoirState from BlackoilModelEbos 
1) Use the solution variable directly in RelativeChange(...)
2) Add a method in the RateConverter that takes the simulator instead of the state.
3) Pass the reservoir pressure directly to the well initialization.
4) Move convertInput(...) to SimulatorFullyImplicitBlackoilEbos.hpp.
This code is only used to convert the initial reservoir state.
5) Modify  updateState(...). The solution variable is updated directly and adaptPrimaryVariable(...)
from ewoms is used to switch primary variables. An epsilon is passed to adaptPrimaryVarible(...) after a switch
of primary variables to make it harder to immediately switch back.

The following code used by flow_ebos still uses the reservoirState
1) the initialization
2) restart
3) output of the initial state
4) the step methods in AdaptiveTimeStepping and NonlinearSolver.
The reservoirState is not used by this methods, so after the initial step, an empty reservoirState is passed around in the code.
 2017-08-17 11:13:00 +02:00
Atgeirr Flø Rasmussen
e3ceecdfa4 Silence unused variable warnings. 2017-08-15 11:17:30 +02:00
Markus Blatt
7b3b99f692 Output bubble/dew point computation problems only on rank 0. 
We do this by switching to output the global index of the cells.
In a first step the problematic cell indices are gather on process 0.
Then they are logged there.

This should prevent spurious ouput at the end of PRT and DEBUG files after the
simulation time is printed. This happened previously for some parallel
runs of model 2. Unfortunately, it seems these problems do not appear any
more for the current master. At least I could not reproduce them.
 2017-08-14 21:34:54 +02:00
Atgeirr Flø Rasmussen
c6e729b1bf Merge pull request #1252 from atgeirr/debug-reordering-solver 
Fixing bugs and improving the reordering solver
 2017-08-10 14:09:58 +02:00
Atgeirr Flø Rasmussen
67823c5893 Disable debugging output of detailed cell data. 2017-08-10 13:09:22 +02:00
Atgeirr Flø Rasmussen
784bcf6892 Experiment with relaxing single cell iterations. 2017-08-10 12:31:04 +02:00
Atgeirr Flø Rasmussen
9cf6832140 Ignore limits to rs and rv changes. 
As implemented with a relative limit, even with 1e9 default limit it
would still be impossible to get away from a zero value. It is
possible that the limits may return later, implemented in a less
buggy way, however for now they do not seem necessary.
 2017-08-10 11:25:47 +02:00
Kai Bao
c59aa9127e making rate_converter to be reference to the one in Simulator 
keeping the const property in the Well Model.
 2017-08-10 11:20:09 +02:00
Kai Bao
cf902e1a8b Simulator Blackoil and StandardWellsDense uses same rate_converter 
fixing the running of group control.
 2017-08-10 10:25:39 +02:00
Atgeirr Flø Rasmussen
0a440d98ce Merge pull request #1238 from totto82/removeConvertResults2 
Ask ebos to formulate the equation by surface volumes directly
 2017-08-10 10:01:32 +02:00
Atgeirr Flø Rasmussen
1e0facec4a Fix updateState() saturation update. 2017-08-09 15:47:39 +02:00
Atgeirr Flø Rasmussen
608a674858 Make dh_sat behave like in coupled transport solver. 2017-08-07 16:17:57 +02:00
Kai Bao
338828fdce fixing the running of the vfp related for flow_ebos 
the output of the THP is still not yet correct.
 2017-08-07 11:39:11 +02:00
Atgeirr Flø Rasmussen
c39341006d Bugfix: follow opm-material convention for capillary pressure 2017-08-04 11:39:51 +02:00
Atgeirr Flø Rasmussen
e9e4416428 Make computeCellState() a const method. 2017-08-04 11:39:09 +02:00
Atgeirr Flø Rasmussen
12bd67e248 Create oil-in-place data if overall efficiency is requested by output. 2017-08-01 16:40:34 +02:00
Atgeirr Flø Rasmussen
babf06eade Merge pull request #1243 from blattms/write-initial-oip 
Write the initial OIP values if requested.
 2017-08-01 16:15:46 +02:00
Kai Bao
ab1b57b994 fixing the if condition for 2p + dummy phase 
from if (numComp < numEq) to if (numComp < numWellEq)
 2017-07-31 12:43:50 +02:00
Markus Blatt
900fec427d Cleaned up bogus documentation and whitespace 2017-07-20 12:37:01 +02:00
Markus Blatt
c06e5b7cf8 Removed boolean parameter indicating initial write to writeTimeStep. 
Instead we use the timer to determine whether this is the initial write.
 2017-07-20 12:12:41 +02:00
Markus Blatt
1939453adb Write the initial OIP values if requested. 
This is when writing the first time step as is this is
the first step when OIP is available.
 2017-07-19 22:06:08 +02:00
Markus Blatt
542f181f7e Narrow down the possible Matrix types for ParallelOverlappingILU0 
It now has to be a BCRSMatrix, but the block type is flexible and
needs to this way. flow_legacy uses MatrixBlock, and flow_ebos uses
FieldMatrix.
 2017-07-03 13:50:21 +02:00
Markus Blatt
1bb0968283 Work around unstable matrix inversion in DUNE 2.[34] 
The versions are missing the specialized code for inverting
a 3x3 matrix that makes the algorithms quite a bit more stable.
With this patch we fall back to using our own MatrixBlock that does
not suffer from this deficiency.
 2017-07-03 10:20:23 +02:00
Tor Harald Sandve
082e00d4ec Ask ebos to formulate the equation by surface volumes directly 2017-06-29 13:58:20 +02:00
Robert Kloefkorn
ce7f3f46a1 [feature] Added two-phase executable for numEq = 2. 2017-06-29 12:56:37 +02:00
Atgeirr Flø Rasmussen
a980993289 Make reordering solver work with iterated implicit. 
Also tighten tolerances. They were too slack, giving significant
differences from non-reordering implementation towards the end of
SPE1CASE2 for example.
 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
9b2d696b86 Tweak tolerance. 
Note that a more proper dimensionless residual check is still a good idea.
 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
c40ef4d9d0 Modifications to catch up with master branch. 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
4f9a9359ee Refine convergence check, retain max change info. 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
6167fff62a WIP: change what is logged, keep track of max change. 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
ab735b33e1 WIP: add sequential model for debugging. 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
0f8cd2c212 Bugfix: use correct sign for z gradient. 2017-06-28 07:47:17 +02:00
Atgeirr Flø Rasmussen
e653067951 Adapt new interface with timers. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
b5b8fa14de Bugfix: use correct pv multiplier for initial and current state. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
fb6dd3adf8 Add well flux terms. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
2c20bb6cde Bugfix: take absolute value of residuals for comparison. 
Also tighten hardcoded tolerance for testing, and update a comment.
 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
87477f7a44 Bugfix: update output reservoir state. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
79fed93b76 Fix saturation bug. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
f12c781f3e Fix bugs. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
e83b8cd0ac Finished updateState(). 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
0a3c65707d More work in progress... 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
56853a0272 Avoid derivative pollution. 2017-06-28 07:47:16 +02:00
Atgeirr Flø Rasmussen
5b21cdf54f Still working on reordering solver. 2017-06-28 07:47:16 +02:00