Clean up particle-particle contact force class prior refactoring particle-wall contact force class in DEM #1237
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Description This PR adds a small section in the melting cavity example featuring the new constrain stasis feature from (Temperature-dependant solid domain constraints #1038). Documentation Parameter documentation was also updated : doc/source/parameters/cfd/constrain_stasis.rst Former-commit-id: e843eae
Description of the problem When calculate mass conservation in the post-processing was enabled (default) and the verbosity was set to verbose, the mass conservation table was displayed on terminal at the end of the simulation. If a lot of time iterations were made, this could cover a significant chunk of the terminal window and might not be pertinent to have it their when .dat files is also outputted by the simulation. Description of the solution In a similar way to "VOF Barycenter", the output of the "VOF Mass Conservation" is now displayed at every time iteration on the terminal. How Has This Been Tested? Expected tests have been updated output : applications_tests/lethe-fluid/gls_vof_1_isothermal_compressible_fluid.output applications_tests/lethe-fluid/gls_vof_2_isothermal_compressible_fluids.output applications_tests/lethe-fluid/gls_vof_dirichlet_boundary_condition.output applications_tests/lethe-fluid/gls_vof_periodic_boundary_condition.output applications_tests/lethe-fluid/time_dependent_boundaries_vof.output Mass monitoring has been added to the following test to improve robustness: applications_tests/lethe-fluid/heat_transfer_vof_phase_change_constrain_solid_domain.prm (.output) Future changes Change all set_precision in the make_table functions to set_scientific as default. Comments A new make_table_scalars_vectors function has been added in utilities.h/.cc. A new boolean argument in make_table functions has been added, namely display_scientific_notation to decide if the display of table values is displayed at a fixed decimal or in scientific notation. Former-commit-id: 82095ba
Description of the problem Some physical and interface property models require certain fields to compute property values and partial derivatives. When these fields were not defined a not so debug-friendly message was printed out. Description of the solution To help the debugging process, if an error were to occur, assertions were added to the property models in functions where the fields were required. How Has This Been Tested? By commenting lines in unit tests and code, the throw of the exception was verified in DEBUG mode. Comments Doxygen descriptions were also added to specific_heat_model.h. Former-commit-id: bc03310
Description of the problem The eigenvalue estimation for the smoother of the multigrid preconditioners was done through a PreconditionChebyshev. Description of the solution Due to recent changes in deal.II (dealii/dealii#16742), now we are able to do it using the PreconditionRelaxation directly. This allows us to remove the estimate omega function since it is done internally by the smoother if the relaxation parameter is set to 1. This allows also to remove the degree parameter that was only used in the case of Chebyshev. Former-commit-id: 30f1a30
Former-commit-id: d8e115f
Former-commit-id: b633f6f
since it is not used anyway according to https://www.dealii.org/developer/doxygen/deal.II/classMappingQ.html#a9b34a7f2ec00aa65e73402865ac2887d Former-commit-id: ab3a638
Former-commit-id: 3ea6a50
Former-commit-id: 190af0d
#1072) Description of the problem The recently implemented vof stasis constraint (Implement temperature-dependant solid domain contraints in VOF #1048) was not compatible with adaptive mesh refinements and checkpointing system. When refining the mesh or when restarting a simulation, the filtered phase fraction was reinitialized, but not filled with the appropriate values. Description of the solution The apply_phase_filter() function was added to post_mesh_adaptation() and read_checkpoint() functions. How Has This Been Tested? The stasis constraint was tested using the 2D LPBF benchmark : with restart and mesh adaptation; without restart and with mesh adaptation, and; with restart and without mesh adaptation. Former-commit-id: bb7cea3
Former-commit-id: 33d1886
Description of the problem The flow control subsection in the documentation for this example was not up to date. Description of the solution Update section in accordance to parameter file and new parameters for dynamic flow control. Former-commit-id: 3e3f3f3
Description of the problem No tutorial on how to use Pointwise the new meshing software used by the group. Pointwise tutorial is not that useful for our uses of it Description of the solution Add Pointwise tutorial Future work Add examples Co-authored-by: Bruno Blais <[email protected]> Former-commit-id: b47f7d2
Description of the problem For some cases we are running, it is relevant to monitor both the mass and the momentum in VOF simulations Description of the solution Add momentum monitoring to the mass monitoring. I don't think it's worth it to make multiple files, instead I appended the information to the existing file. How Has This Been Tested? Tested on some of our regular benchmark and the results make sense. Documentation Updated documentation accordingly Future changes The parameter monitor mass conservation is maybe not the most appropriate right now, but I don't want to break the prm file in this PR. We could have a parameter called monitor conservation instead, of maybe monitoring mass and momentum, but i 'd rather postpone this change because it's only esthetics Co-authored-by: hepap <[email protected]> Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: 73306f4
* fix lid driven cavity postprocessing * update doc Former-commit-id: 01be39f
Description of the problem The insertion box for the volume insertion was being defined by a large number of parameters (9 to be exact) Description of the solution The box is defined by one parameter and the order of direction by an other. Documentation The documentation has been updated accordingly. Comments The list insertion parameters are now parsed with "convert_string_to_vector", which is cleaner to look at. I have made (with the help of Amishga) a bash script to parse every prm file and change the old parameters into the new ones. We could do something with it... maybe? Former-commit-id: 219ca30
Description of the problem This is a follow up from Victor's comment from PR [Dem volume insertion box parameters #1074 The initial velocity and initial angular velocity are now define with one parameter each. Really quick PR. No example or application test (except for one) was using those parameters. Former-commit-id: 698b063
kinematic pressure: pressure/reference_state_density reduced pressure: pressure/critical_pressure The "kinematic pressure" in the documentation on incompressible Navier-Stokes equations was referred to as "reduced pressure"; this has been replaced. Former-commit-id: 734f8da
Description of the problem The naming scheme for the insertion files were not coherent with the rest of the code. (i.e., integrator_... , particle_wall_... , etc.). This was annoying when trying to find a specific insertion file. Now they are next to each other. Description of the solution The files are now named like this: insertion_method.cc or insertion_method.h How Has This Been Tested? Tests are running on my machine with these changes. Former-commit-id: 16425e9
Description of the problem The particle-wall nonlinear model wasn't following the same coulomb-limit as the particle-particle hertz-limit-overlap model Comments Tests outputs had to be updated since the position of the particles are slightly different. Former-commit-id: d1d9373
Former-commit-id: 4990749
Former-commit-id: 4e45598
Description of the problem A test seems to break on master version of deal.II Description of the solution Decrease tolerance to be more robust Former-commit-id: 9f84d35
Description of the problem On some clusters, it is recommend to specify OMP_NUM_THREADS = 1, otherwise Trilinos can start using multithreading. Description of the solution I added a line in the .dealii file with a small comment. Former-commit-id: 469bbe3
Description of the problem Using the Matrix Free solver, Joel cooked up a very cool 3d Taylor-Couette example that really showcases the power of the matrix free approach Description of the solution To keep this example for the future, we add it as an example test case. Co-authored-by: Pierre Laurentin <[email protected]> Co-authored-by: Laura Prieto Saavedra <[email protected]> Former-commit-id: 95dbbc5
Description of the problem The console output of the non-linear solver in the NS solver was showing the the norms of the newton update without distinction between the velocity and pressure contributions. The final residual norm of the linear solver was not outputted, only the tolerance. Description of the solution Add the distinction between the velocity and pressure contributions in the non-linear solver output. Add the final value of the linear residual when verbosity is at verbose. If it is at extra verbose, the solver output the residual for each iteration (already implemented but not documented). How Has This Been Tested? A unit test was added. vector_problem.cc Co-authored-by: Pierre Laurentin <[email protected]> Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: 4632e53
Description of the problem The third argument of the MPI_InitFinalize(...) call in all applications in Lethe was set to numbers::invalid_unsigned_int; if this is the case, then the number of threads is determined automatically according to the system and set via MultithreadInfo::set_thread_limit(...). Due to confusions this week in the lab, I though that it would be a good idea to set it explicitly to 1, which guarantees that it will start as many MPI processes per node as there are cores on each node avoiding multi-threading. For the main applications, there was a call MultithreadInfo::set_thread_limit(1) in the respective solve() function, however, according to deal.II, it only works if it is called before any threads are created, therefore, it is safer to call it at the beginning of main() instead. Description of the solution Removed MultithreadInfo::set_thread_limit(1) from the applications and set the third argument of MPI_InitFinalize(...) to 1 in all cases across the library. How Has This Been Tested? All the tests pass without any issue. Former-commit-id: 34ff2b6
Description of the problem The GMG preconditioners were defined within two functions in the matrix-free solver. This was preventing us from reusing certain parts of the multigrid. Description of the solution The GMG preconditioners now have their own class and objects. This also allows to use the inexact newton solver for the lethe-fluid-matrix-free application. In addition, some corrections to the local smoothing algorithm were introduced by eliminating the interface out matrix. How Has This Been Tested? All the existent matrix-free tests pass without any issue. Future changes -This is the first step towards reusing parts of multigrid to improve the performance of the matrix-free solver. Co-authored-by: Peter Munch <[email protected]> Co-authored-by: Bruno Blais <[email protected]> Former-commit-id: d7504a6
Description The previous implementation to calculate cell diameter used the measure function which assumed a linear mapping. Solution This PR adds a function that calculates the volume by integrating 1 over the cell, by summing JxW values returned by the FEValues object. Testing Tests that were outdated have been updated. Former-commit-id: 0430787
Remove include in ib that make it recompiled Move periodic offset as member variable Remove effective mass and radius as member variable Modify all force except DMT Put rolling torque calculation in function Clean up set pp force models with predefined template abuse First cleanup set pp force models Upgrade p-p contact forces doc and intermediate variables Former-commit-id: 794f832
Former-commit-id: a73797f
Former-commit-id: d2f58f4
Former-commit-id: 0ad538d
Former-commit-id: 28969ac
Former-commit-id: 5613050
Former-commit-id: 7998441
Former-commit-id: 5b7f3fe
This reverts commit 37963ad3a6e1f5472827af9bac67dcee105cbc8b. Former-commit-id: 93b3555
Former-commit-id: 3d6b0bc
Former-commit-id: 99fafe2
Former-commit-id: 0f0afbf
Former-commit-id: 8d88ce8
OGaboriault
force-pushed
the
dem_pp-forces-cleanup
branch
from
92075f5
to
8d88ce8
Compare
voferreira
approved these changes
Sep 26, 2024
| @@ -3,591 +3,243 @@ | |||
|
|
|||
| using namespace dealii; | |||
|
|
|||
| using namespace Parameters::Lagrangian; | |||
|
|
|||
|
|
|||
blaisb
force-pushed
the
dem_pp-forces-cleanup
branch
from
984cb53
to
dea765b
Compare
12 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
The particle-particle contact force class had some redundancies, so I managed to implement functions that are now called.
Some cleanup was been done here and there, but most of it is related to the call to the calculation of the contact force for a pair of particle to the proper model.
Before this refactoring, each (5) loop over the adjacent pair had the same series of
if constexpr (contact_model == ...), now each loop callexecute_contact_calculation()according to the contact type as a template.This clean up was done before applying the same architecture pattern for the particle-wall contact force class.
Testing
Tests are good.
Documentation
Miscellaneous (will be removed when merged)
I would like to have Gabo reviewing it before the merge since he has coded and may continue to code in the class and I want something that he agree on. However, I forbid him to do it during his mini-project hehe.
Checklist (will be removed when merged)
See this page for more information about the pull request process.
Code related list:
Pull request related list: