dumux issueshttps://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues2017-04-26T08:57:59Zhttps://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/338cleanup 3pwateroil2017-04-26T08:57:59ZTimo Kochtimokoch@math.uio.nocleanup 3pwateroilThe 3pwateroil model still uses some old idioms. For example
* [x] #337
* [x] not using range based for in some places (might have been all fixed short before the 2.10 release)
* [x] ... possibly more thingsThe 3pwateroil model still uses some old idioms. For example
* [x] #337
* [x] not using range based for in some places (might have been all fixed short before the 2.10 release)
* [x] ... possibly more things2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/366Dumux release 2.112017-04-26T08:57:59ZSina AckermannDumux release 2.11All major changes that should be included and are not yet committed should be announced.
Add a comment here describing the feature and areas affected by the change.All major changes that should be included and are not yet committed should be announced.
Add a comment here describing the feature and areas affected by the change.2.11Sina AckermannSina Ackermann2017-03-31https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/325FS#325 default to AMGBackend?2017-05-25T05:10:34ZBernd FlemischFS#325 default to AMGBackend?# Metadata
| Property | Value |
| -------- | ----- |
| Project | dumux |
| Category | General |
| Reported by | Bernd Flemisch (bernd@iws.uni-stuttgart.de) |
| Reported at | Jul 14, 2016 10:48 |
| Type | Feature Request ...# Metadata
| Property | Value |
| -------- | ----- |
| Project | dumux |
| Category | General |
| Reported by | Bernd Flemisch (bernd@iws.uni-stuttgart.de) |
| Reported at | Jul 14, 2016 10:48 |
| Type | Feature Request |
| Version | 2.9 |
# Description
Since we got rid of our self-written parallel linear solvers, the default linear solvers are sequential. This prevents us from doing a parallel run without having to modify the problem file and to recompile.
I would prefer to set the default linear solver to the AMGBackend. Not only is it parallel, is also showed the best performance in a recent independent study, at least averaged over a range of models, discretizations and grid sizes.2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/328FS#328 Use math functions from std:: and allow ADL2017-05-25T05:10:34ZChristoph GrüningerFS#328 Use math functions from std:: and allow ADL# Metadata
| Property | Value |
| -------- | ----- |
| Project | dumux |
| Category | General |
| Reported by | Christoph Grüninger (gruenich@iws.uni-stuttgart.de) |
| Reported at | Sep 9, 2016 11:46 |
| Type | DuMuX day...# Metadata
| Property | Value |
| -------- | ----- |
| Project | dumux |
| Category | General |
| Reported by | Christoph Grüninger (gruenich@iws.uni-stuttgart.de) |
| Reported at | Sep 9, 2016 11:46 |
| Type | DuMuX day |
| Version | 2.9 |
# Description
We use std::pow and pow for the power function. It is adviced to use using namespace std; and then just pow.
We should discuss an according change of all used math functions.
This task arose from a discussion in https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/merge_requests/1922.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/337get rid of matrixHeatFlux, rename heatCapacity2017-04-26T08:57:59ZBernd Flemischget rid of matrixHeatFlux, rename heatCapacityThe `3pwateroil` model still uses the method `matrixHeatFlux` from the spatial parameters. This should be made consistent with the other nonisothermal models, namely, by calculating the matrix heat flux in the flux variables using an eff...The `3pwateroil` model still uses the method `matrixHeatFlux` from the spatial parameters. This should be made consistent with the other nonisothermal models, namely, by calculating the matrix heat flux in the flux variables using an effective thermal conductivity relation.
For the same reasoning, one should rename `heatCapacity` to `solidHeatCapacity`.
The test of the `2pnc` model also still provides a function `matrixHeatFlux` in its spatial parameters. This should be removed.2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/342Implement classes facilitating a custom start.hh2017-04-26T08:57:59ZTimo Kochtimokoch@math.uio.noImplement classes facilitating a custom start.hhI often feel restricted by the way our `start.hh` and the `main.cc` file is structured. The idea is to make it simple to run a problem in dumux so the start function is predefined. However it's currently annoying to implement your own ma...I often feel restricted by the way our `start.hh` and the `main.cc` file is structured. The idea is to make it simple to run a problem in dumux so the start function is predefined. However it's currently annoying to implement your own main function since many of the steps in the standard `start.hh` method are just written out sequentially or in simple functions. It would be nice to make it more customizable by some OO programming. E.g. a simple class called the `DumuxMessageGenerator` and more important classes like a grid setup class or a parameter file / command line option reading class.
Simple examples would be:
1. I want to write a grid convergence test with a simple loop in main refining the grid and output the errors and rates.
2. I want to solve a problem with two grids.
3. I want to have objects or variables living outside the scope of a problem
These things are overly complicated now I think.2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/336New implementation of MillingtonQuirk can lead to different solution2017-04-26T08:57:59ZKatharina HeckNew implementation of MillingtonQuirk can lead to different solutionWhen using the new implementation for the Millington Quirk effective diffusivity the solution can be different. E.g. in co2plumeshape exercise in dumux - lecture. The old implementation used pow(sw phi, 7/3) the new implementation to avo...When using the new implementation for the Millington Quirk effective diffusivity the solution can be different. E.g. in co2plumeshape exercise in dumux - lecture. The old implementation used pow(sw phi, 7/3) the new implementation to avoid pow uses a cubicroot. See commit 7b864375056ec92d7b6250bc25b0b3ab0fa4329 for details.2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/349OPM 2016.102017-04-26T08:57:59ZMartin SchneiderOPM 2016.10-We should check if the new opm version is running with dumux.
-Furthermore, I get strange results when calling `gridView_().comm().rank()` with the cpgrid for sequential runs.-We should check if the new opm version is running with dumux.
-Furthermore, I get strange results when calling `gridView_().comm().rank()` with the cpgrid for sequential runs.2.11Bernd FlemischBernd Flemischhttps://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/334Sequential 2p impes restart+adaptive problem fails with current dune master2017-04-26T08:57:59ZTimo Kochtimokoch@math.uio.noSequential 2p impes restart+adaptive problem fails with current dune masterThrowing an `std::bad_alloc` error.Throwing an `std::bad_alloc` error.2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/343test_general is not adjusted to changes in start.hh2017-04-26T08:57:59ZChristoph Grüningertest_general is not adjusted to changes in start.hhtest_general is not adjusted to changes in start.hh. Please check all tests, tutorials and dumux-lecture programs, too.
```
Scanning dependencies of target test_general
[ 1%] Building CXX object test/common/generalproblem/CMakeFile...test_general is not adjusted to changes in start.hh. Please check all tests, tutorials and dumux-lecture programs, too.
```
Scanning dependencies of target test_general
[ 1%] Building CXX object test/common/generalproblem/CMakeFiles/test_general.dir/test_generalproblem2p.cc.o
/home/gruenich/dune/complete/dumux/test/common/generalproblem/test_generalproblem2p.cc:54:26: error: no member named 'readOptions_' in namespace 'Dumux'
std::string s(Dumux::readOptions_(argc, argv, paramTree));
~~~~~~~^
1 error generated.
```2.11Timo Kochtimokoch@math.uio.noTimo Kochtimokoch@math.uio.nohttps://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/335[test] mpnc test thermalnonequilibrium fails2017-04-26T08:57:59ZTimo Kochtimokoch@math.uio.no[test] mpnc test thermalnonequilibrium failsFails with the following error
```
Fuzzy comparison...
Traceback (most recent call last):
File "/temp/buildbot/corras/dumuxmasterdune24clang35/dumux/bin/testing/runtest.py", line 59, in
result = compare_vtk(args['files'][i*2], ...Fails with the following error
```
Fuzzy comparison...
Traceback (most recent call last):
File "/temp/buildbot/corras/dumuxmasterdune24clang35/dumux/bin/testing/runtest.py", line 59, in
result = compare_vtk(args['files'][i*2], args['files'][(i*2)+1], relative=args['relative'], absolute=args['absolute'], zeroValueThreshold=args['zeroThreshold'])
File "/temp/buildbot/corras/dumuxmasterdune24clang35/dumux/bin/testing/fuzzycomparevtu.py", line 43, in compare_vtk
root2 = ET.fromstring(open(vtk2).read())
IOError: [Errno 2] No such file or directory: '/temp/buildbot/corras/dumuxmasterdune24clang35/dumux/build-cmake/test/porousmediumflow/mpnc/implicit/combustion-00079.vtp'
2/95 Test #88: test_boxmpncthermalnonequil ............***Failed 29.65 sec
```
so the test didn't run until the number specified in the CMakeLists.txt file.2.11Kilian WeishauptKilian Weishaupthttps://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/339Thorough check that ALL tests use epsilons for floating point comparison in b...2017-04-26T08:57:59ZTimo Kochtimokoch@math.uio.noThorough check that ALL tests use epsilons for floating point comparison in boundary conditions and spatial paramsThis is the main reason for failing test. All tests should use a proper epsilon.
* This epsilon usually is made of a base epsilon `static constexpr eps_ = 1.5e-7;` times
a magnitude. In the case of grids this should be a characteristic c...This is the main reason for failing test. All tests should use a proper epsilon.
* This epsilon usually is made of a base epsilon `static constexpr eps_ = 1.5e-7;` times
a magnitude. In the case of grids this should be a characteristic cell size.
* If the grid is in the scale of 1m there is no adjustment necessary. However a runtime check would make it flexible for using any grid
* We could also introduce a preprocessor constant or global variable in a namespace for the epsilon
* We can also use more of Dune::Float_Cmp2.11https://git.iws.uni-stuttgart.de/dumux-repositories/dumux/-/issues/333viscosity of air implementation really the best available?2017-04-26T08:57:59ZHolger Classviscosity of air implementation really the best available?The current implementation of air viscosity is taken from the Book [Properties of Gases & Liquids]
and uses a method which can be applied to estimate viscosities for polar substances.
But air contains mainly non-polar molecules. Thus,...The current implementation of air viscosity is taken from the Book [Properties of Gases & Liquids]
and uses a method which can be applied to estimate viscosities for polar substances.
But air contains mainly non-polar molecules. Thus, I wonder if this relatively complex approach
is really justified or if it is really the most accurate available.
The ''old'' implementation in MUFTE is currently still used as simpleGasViscosity, but as far
as I remember, this was specific for dry air.
VDI-Wärmeatlas contains exact values for dry air viscosity. We could compare the current
implementations or implement a new one if necessary based on the tables there.2.11Holger ClassHolger Class