Physical av #735
Merged
Physical av #735
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
MTCam
added
enhancement
Draft
8 tasks
Date: Fri Nov 4 04:35:10 2022 -0500
Update CI scripts to use driver scripts to drive tests.
There are no remaining changes to Kaushik's fork: inducer/arraycontext@main...kaushikcfd:arraycontext:main
* Add option to turn off inviscid terms in CNS operator. * Deflake8 * partial update * tweak mengaldo BC code 1) separate individual aspects of BCs (adiabatic, slip, etc.) into their own component classes 2) use *_plus for values that will be used in conjunction with *_minus values in numerical flux functions, and *_bc for values that will be used to compute flux directly * use grad_temperature_bc instead of grad_temperature_plus * remove duplicated inviscid/viscous flux methods * make inviscid/viscous flux internal variables * change grad_momentum_bc to grad_velocity_bc in _SlipBoundaryComponent and use it in SymmetryBoundary * deprecate SymmetryBoundary * add comment about energy for inviscid * only use _ldg_bnd_flux_for_grad in mengaldo BCs * rename state_plus to state_plus_inviscid and explain why * don't pass callbacks to base class constructor unless they're expected to be used * add FIXME * add some docstrings * simplify cv_plus/cv_bc construction by using cv.replace * add replace_fluid_state function and use it to further simplify state construction in BCs * tweak some docstrings * add note * restore SymmetryBoundary docstring to avoid unused reference error in doc build * Reset CI * Deawkwardize doc wording. * Make clear that lewis must be an array * Add option to turn off inviscid terms in CNS operator. (#856) Co-authored-by: tulioricci <[email protected]> * Try to fix doc err. * Adding variable spec. diff to PowerLawTransp (#772) Co-authored-by: Mike Campbell <[email protected]> Co-authored-by: Matthias Diener <[email protected]> * Partial update * Refactor Mengaldo: Isothermal, and Adiabatic for clarity * Update doublemach to provide its own boundary. This AV is deprecated, preparing for removal * Matts changes update 1. * Add Matts replace fluid state helper * Use smoothness consistent * Switch to mengaldo bcs * Update doc with boundary interface change. * Undo test changes * Random cleanup * Remove duplicate func. * Update docs * Add another Poinsot ref, fix doc err? * Extricate AV-specific junk from fluid boundaries. * Extricate AV from fluid boundaries. * Use production MANIFEST.in, try to fix gmsh ish with doublemach. * Update lingering bcs * Update per final review comments * Update outdated comment --------- Co-authored-by: Matthew Smith <[email protected]> Co-authored-by: tulioricci <[email protected]> Co-authored-by: Tulio <[email protected]> Co-authored-by: Matthias Diener <[email protected]>
lukeolson
reviewed
Apr 4, 2023
mirgecom/eos.py
Outdated
| # MJA, it doesn't appear that we can have a None field embedded inside DV, | ||
| # make a dummy smoothness in this case | ||
| if smoothness_mu is None: | ||
| smoothness_mu = 0. * cv.mass |
There was a problem hiding this comment.
Are there any (compile) risks with initializing smoothness_mu from cv.mass? Should we use zeros or zeros_like directly from the array context?
There was a problem hiding this comment.
This pattern is used pervasively. While we prefer the zeros_like facility (which was created exactly for this type of thing), this pattern is not (or has not presented itself as) a compilation issue.
There was a problem hiding this comment.
Looks better -- there are several places where this is used with smoothness_*
There was a problem hiding this comment.
I only found a couple places in eos. I suppose where they are not passed (i.e. for inviscid-only and runs with AV off), that setting them to the same zeros is OK. e1ad33f
lukeolson
approved these changes
Apr 5, 2023
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.
Production feeder for physical artificial viscosity.
Add a new transport class that implements artificial viscosity as an additive term to the physical viscosity.
This requires an additional parameter included with the fluid state, the smoothness. The smoothness is derived from the discretization and needs to be available to the transport class to compute the new viscosity term. We have added the smoothness field to dv.
Adds a new example which doubles as an integrated test to illustrate how to setup and use the new transport class and tests that the solution works and stays in the acceptable numerical range with physical AV enabled.
based on work in #717
Questions for the review @lukeolson :