Add validation case val2a #189

lin-yang-ly · 2024-10-14T23:48:08Z

(Ref. #188)
(Ref. #12)

lin-yang-ly · 2024-10-14T23:49:47Z

@simopier @chaibhave This is the draft PR for validation case 2a. Just let me know if you have any suggestions or comments for the model I built. That would be helpful for me to build the correct model. Thank you!

moosebuild · 2024-10-15T00:03:15Z

Job Documentation, step Sync to remote on 4755cf1 wanted to post the following:

View the site here

This comment will be updated on new commits.

moosebuild · 2024-10-15T00:07:11Z

Job Coverage, step Generate coverage on 4755cf1 wanted to post the following:

Coverage

Coverage did not change

Full coverage report

This comment will be updated on new commits.

simopier

Thank you @lin-yang-ly.

A couple preliminary comments.

Plus, don't forget to list val-2a in the V&V list in index.md.

test/tests/val-2a/tests

test/tests/val-2a/comparison_val-2a.py

test/tests/val-2a/gold/val-2a_TMAP4_out.e

test/tests/val-2a/gold/val-2a_TMAP7_out.e

test/tests/val-2a/val-2a.i

test/tests/val-2a/comparison_val-2a.py

lin-yang-ly · 2024-10-15T17:16:46Z

@simopier I just updated the PR with your comments. Just let me know if you have more suggestions!

lin-yang-ly · 2024-10-17T23:56:07Z

@simopier @chaibhave I just upload the updated model. The result is finally close to the experimental result. The two figures below are the permeation flux on the right side based on data from TMAP4 and TMAP7, respectively. The result based on TMAP4 is close to experiment, except for the small bumps around 9000s and 15000s. However, the result based on TMAP7 is far away from experiment. I think it is because lots of data from TMAP7 are wrong.

Based on current result, I think

I can continue to study why there is small bumps around 9000s and 15000s.
In fact, the time of implantation start and stop in TMAP4 is wrong as well. Thus, I am thinking about only compare with experiment in one model only using correct data from TMAP4 or TMAP7, instead of comparing in two models with error data everywhere. Because our main purpose in this case is validating the experiment result instead of verifying TMAP4 or TMAP7.

Just let me know if you have any suggestions about that. Thank you all a lot for helping me figure this out!

lin-yang-ly · 2024-10-21T14:28:09Z

Better Results for TMAP4:

and TMAP7:

chaibhave · 2024-10-21T14:45:01Z

Better Results for TMAP4:

and TMAP7:

Looking good! It would be helpful if you add what changes made the results better when you show the updated figures.

lin-yang-ly · 2024-10-21T14:58:02Z

Better Results for TMAP4:
and TMAP7:

Looking good! It would be helpful if you add what changes made the results better when you show the updated figures.

Thank you for the suggestion! I used the recombination flux instead of diffusion flux to compare with experimental results in the updated figures. Because only recombination flux ($K_r C^2$) really permeates out of the sample.

lin-yang-ly · 2024-10-22T15:27:23Z

Morning @simopier , I just removed the data and model from TMAP7, now it is ready for review!

simopier

Great stuff! I left quite a few comments, but you're getting very close. Let's discuss this.

Thank you for your hard work, @lin-yang-ly!

doc/content/verification_and_validation/index.md

test/tests/val-2a/gold/experiment_data_paper.csv

test/tests/val-2a/gold/val-2a_TMAP4_out.csv

doc/content/verification_and_validation/val-2a.md

test/tests/val-2a/val-2a_base.i

test/tests/val-2a/tests

(Ref. idaholab#188)

Co-authored-by: Pierre-Clement Simon <[email protected]>

lin-yang-ly · 2024-10-25T20:57:20Z

@simopier I finally solved the issue about different number on time column of csvdiff test. I used the larger tolerance and a constant time step in the test case.

Besides, I have several questions for the PR:

Do we need to add exodus test for heavy case? The reason why I ask is I found the exodus file is around 500 MB in heavy case with finer mesh and time step.

I try the adaptivity mesh method from the input file you sent to me, and I found the adaptivity mesh is not as good as the previous one. Because the previous mesh is the hand-crafted refined mesh. However, the adaptivity mesh may not find the reasonable mesh as soon as possible. This makes the running time longer than previous. The simulation may need at least two days to calculate the final results. Do you think we should keep current adaptivity mesh or turn back to the hand-crafted refined mesh?

Except for these questions, the PR is ready for review.

test/tests/val-2a/val-2a.i

use constant time step update tolerance update tolerance 2 high tolerance with low time step loose tolerance lower tolerance mum to nm for better converge update output parameters alter time step update time step update time step maybe very high tolerance A test for better performance roll back Apply suggestions from code review Co-authored-by: Pierre-Clement Simon <[email protected]>

simopier

It seems like you override a previous commit that contained my suggestions. I unresolved these few conversations. Please make sure not to do it in the future.

I added some more suggestions and requests, mostly for the documentation.

The last thing we need to figure out if how to reduce the tolerance (and potentially increase the time step) of this case. I'll work on it and let you know if I figure something out.

doc/content/verification_and_validation/val-2a.md

test/tests/val-2a/comparison_val-2a.py

test/tests/val-2a/val-2a.i

Co-authored-by: Pierre-Clement Simon <[email protected]>

lin-yang-ly · 2024-10-28T16:23:13Z

Morning @simopier, I just updated the PR with your comments. Just let me know if you have more suggestions!

simopier

A couple more comments. Getting close!

doc/content/verification_and_validation/val-2a.md

test/tests/val-2a/comparison_val-2a.py

doc/content/verification_and_validation/val-2a.md

test/tests/val-2a/val-2a.i

Co-authored-by: Pierre-Clement Simon <[email protected]>

lin-yang-ly · 2024-10-28T18:10:52Z

Thank you, PC! Just updated with your comments. Ready for review again!

- use AD to improve convergence - tighten tolerances - increase max time step to reduce computational time - remove heavy tests since the test now runs in less than 3 seconds - update documentation - regold

cticenhour

@simopier asked me to leave a few small comments on top of his last commit, and I found a few more items that should be addressed. This looks great overall!

doc/content/verification_and_validation/val-2a.md

test/tests/val-2a/tests

doc/content/verification_and_validation/val-2a.md

Co-authored-by: Casey Icenhour <[email protected]>

lin-yang-ly · 2024-10-29T23:34:42Z

@simopier asked me to leave a few small comments on top of his last commit, and I found a few more items that should be addressed. This looks great overall!

Thank you for the comments! @cticenhour

simopier

Looks good to me! Thank you for this contribution, @lin-yang-ly!!

simopier · 2024-10-30T11:25:39Z

@cticenhour, it's ready for your final review.

simopier self-assigned this Oct 15, 2024

simopier added the V&V Relevant to V&V label Oct 15, 2024

simopier requested changes Oct 15, 2024

View reviewed changes

lin-yang-ly marked this pull request as ready for review October 15, 2024 17:15

lin-yang-ly marked this pull request as draft October 15, 2024 17:15

lin-yang-ly force-pushed the add_val2a branch from 8f74f38 to 8d9a306 Compare October 17, 2024 23:39

simopier requested changes Oct 24, 2024

View reviewed changes

lin-yang-ly force-pushed the add_val2a branch from 33b856a to 006cfb1 Compare October 25, 2024 04:32

lin-yang-ly and others added 13 commits October 25, 2024 08:52

Draft version

8755bcc

(Ref. idaholab#188)

Update documentation

a8e92db

Apply suggestions from code review

c7425d4

Co-authored-by: Pierre-Clement Simon <[email protected]>

Simplify input files

bf35b71

Update figures in documentation

4489c32

Fix typos

e196afd

Update model

6f31341

update_model

74a12cc

Update documentation

86b6c3e

Remove the data from TMAP7

4d862ca

fix typo

96369ce

Apply suggestions from code review

27fefd3

Co-authored-by: Pierre-Clement Simon <[email protected]>

Update documentation

ee540cd

lin-yang-ly force-pushed the add_val2a branch from 006cfb1 to 52c1ffe Compare October 25, 2024 14:52

simopier reviewed Oct 25, 2024

View reviewed changes

test/tests/val-2a/val-2a.i Show resolved Hide resolved

lin-yang-ly force-pushed the add_val2a branch from bd761cc to 5e0dce1 Compare October 25, 2024 22:21

update mesh and add exodus

165bc2b

lin-yang-ly force-pushed the add_val2a branch from 5e0dce1 to 165bc2b Compare October 25, 2024 22:24

lin-yang-ly added 2 commits October 25, 2024 17:09

change nm to mum

ab0dbf7

update interval with time_step_interval

f0f2174

simopier requested changes Oct 27, 2024

View reviewed changes

lin-yang-ly and others added 3 commits October 28, 2024 08:25

Apply suggestions from code review

7ee8e31

Co-authored-by: Pierre-Clement Simon <[email protected]>

plot figure from python script

967ffe4

update gold files with latest input

6c043ea

simopier marked this pull request as ready for review October 28, 2024 17:20

simopier requested changes Oct 28, 2024

View reviewed changes

lin-yang-ly and others added 3 commits October 28, 2024 11:39

Apply suggestions from code review

94ecc89

Co-authored-by: Pierre-Clement Simon <[email protected]>

add factor explaination

fa1ad2b

update limited gold files

f61dc23

improve convergence and clean up:

fe118d5

- use AD to improve convergence - tighten tolerances - increase max time step to reduce computational time - remove heavy tests since the test now runs in less than 3 seconds - update documentation - regold

simopier force-pushed the add_val2a branch from 15e3450 to fe118d5 Compare October 28, 2024 22:23

cticenhour requested changes Oct 29, 2024

View reviewed changes

Apply suggestions from code review

4cd5dad

Co-authored-by: Casey Icenhour <[email protected]>

update typo

4755cf1

simopier approved these changes Oct 30, 2024

View reviewed changes

cticenhour approved these changes Oct 30, 2024

View reviewed changes

cticenhour merged commit aee97ea into idaholab:devel Oct 30, 2024
9 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add validation case val2a #189

Add validation case val2a #189

lin-yang-ly commented Oct 14, 2024 •

edited by simopier

Loading

lin-yang-ly commented Oct 14, 2024

moosebuild commented Oct 15, 2024 •

edited

Loading

moosebuild commented Oct 15, 2024 •

edited

Loading

simopier left a comment

lin-yang-ly commented Oct 15, 2024

lin-yang-ly commented Oct 17, 2024

lin-yang-ly commented Oct 21, 2024

chaibhave commented Oct 21, 2024

lin-yang-ly commented Oct 21, 2024

lin-yang-ly commented Oct 22, 2024

simopier left a comment

lin-yang-ly commented Oct 25, 2024 •

edited

Loading

simopier left a comment

lin-yang-ly commented Oct 28, 2024 •

edited

Loading

simopier left a comment

lin-yang-ly commented Oct 28, 2024

cticenhour left a comment •

edited

Loading

lin-yang-ly commented Oct 29, 2024

simopier left a comment

simopier commented Oct 30, 2024

Add validation case val2a #189

Add validation case val2a #189

Conversation

lin-yang-ly commented Oct 14, 2024 • edited by simopier Loading

lin-yang-ly commented Oct 14, 2024

moosebuild commented Oct 15, 2024 • edited Loading

moosebuild commented Oct 15, 2024 • edited Loading

Coverage

simopier left a comment

Choose a reason for hiding this comment

lin-yang-ly commented Oct 15, 2024

lin-yang-ly commented Oct 17, 2024

lin-yang-ly commented Oct 21, 2024

chaibhave commented Oct 21, 2024

lin-yang-ly commented Oct 21, 2024

lin-yang-ly commented Oct 22, 2024

simopier left a comment

Choose a reason for hiding this comment

lin-yang-ly commented Oct 25, 2024 • edited Loading

simopier left a comment

Choose a reason for hiding this comment

lin-yang-ly commented Oct 28, 2024 • edited Loading

simopier left a comment

Choose a reason for hiding this comment

lin-yang-ly commented Oct 28, 2024

cticenhour left a comment • edited Loading

Choose a reason for hiding this comment

lin-yang-ly commented Oct 29, 2024

simopier left a comment

Choose a reason for hiding this comment

simopier commented Oct 30, 2024

lin-yang-ly commented Oct 14, 2024 •

edited by simopier

Loading

moosebuild commented Oct 15, 2024 •

edited

Loading

moosebuild commented Oct 15, 2024 •

edited

Loading

lin-yang-ly commented Oct 25, 2024 •

edited

Loading

lin-yang-ly commented Oct 28, 2024 •

edited

Loading

cticenhour left a comment •

edited

Loading