Fix minor conservation issue + new diagnostic summary for conservation #1338

albeanth · 2023-07-05T21:49:49Z

What is the change?

Downstream testing illustrated some minor conservation issues with non-axial expansion target components (e.g., clad, wire, or duct within a fuel block) during thermal expansion. This PR addresses this and also introduces two things:

minor cleanup of simplifying some repeated and unnecessary code.
diagnostics that allows users and devs to identify mass conservation issues (these were used in place of the existing runLogs so their utility has been illustrated, at least in my opinion!).

Some additional details on item 2. There are two tables that get printed.

massConservationReport: prints the pre- and post-axial expansion assembly mass (as well as the difference) by material and by component if the difference if greater than 1e-9 grams. This 1e-9 g threshold was chosen semi-arbitrarily as the differences between pre- and post-axial expansion were found in downstream testing to range from 0.0 to on the order of 1e-10 g.
detailedMassConservationReport: prints the following component-level information for an assembly: grow fraction, bottom, top, height, and mass for pre- and post-axial expansion. The post-pre component mass is also printed. This summary table gets printed only for debug runs.

Closes #1048.

Why is the change being made?

The axial expansion changer should not lose mass post thermal expansion. This PR fixes that. There are some cases where this is inevitable and these should be documented in the ARMI docs (in a subsequent PR).

Checklist

This PR has only one purpose or idea.
Tests have been added/updated to verify that the new/changed code works.

The code style follows good practices.
The commit message(s) follow good practices.

The release notes (location doc/release/0.X.rst) are up-to-date with any important changes.
The documentation is still up-to-date in the doc folder.
The dependencies are still up-to-date in setup.py.

- added actual aclp duct to illustrate and test for non-conservation

albeanth · 2023-07-06T15:22:34Z

The mass conservation tables look like the following:

massConservationReport

Is assembly-wide and shows the pre- and post-axial expansion mass values for the unique materials and components.

[xtra] <igniter fuel Assembly A0000 at LoadQueue>           pre-exp        post-exp    round(post - pre, 9)
       --------------------------------------------  --------------  --------------  ----------------------
       HT9                                           9.10087881e+04  9.10089401e+04          1.52037573e-01
       UZr                                           1.40720557e+05  1.40720557e+05          0.00000000e+00

       grid                                          6.97388334e+03  6.97388334e+03         -0.00000000e+00
       shield                                        2.12942247e+04  2.12942247e+04          0.00000000e+00
       clad                                          3.39860068e+04  3.39860068e+04          0.00000000e+00
       wire                                          1.84110362e+03  1.84110362e+03         -0.00000000e+00
       duct                                          2.69135695e+04  2.69137216e+04          1.52037573e-01
       fuel                                          1.40720557e+05  1.40720557e+05          0.00000000e+00

detailedMassConservationReport

Breaks downs the massConservationReport by component and can be useful to see how/if mass is being conserved in the assembly post-axial expansion.

Block                                              Component               GrowFrac    (prev) c.bottom    (prev) c.top    (prev) c.height    (post) c.bottom    (post) c.top    (post) c.height     (prev) mass     (post) mass    (post - pre) mass
-------------------------------------------------  ----------------  --------------  -----------------  --------------  -----------------  -----------------  --------------  -----------------  --------------  --------------  -------------------
<grid plate B0000-000 at ExCore XS: A BU GP: A>    <Hexagon: grid>   1.00493359e+00     0.00000000e+00  2.50000000e+01     2.50000000e+01     0.00000000e+00  2.51233397e+01     2.51233397e+01  6.97388334e+03  6.97388334e+03      -1.81898940e-12
<axial shield B0000-001 at ExCore XS: A BU GP: A>  <Circle: shield>  1.00700345e+00     2.51233397e+01  5.25000000e+01     2.75000000e+01     2.51233397e+01  5.28159346e+01     2.76925949e+01  2.12942247e+04  2.12942247e+04       0.00000000e+00
                                                   <Circle: clad>    1.00520603e+00     2.51233397e+01  5.28159346e+01     2.75000000e+01     2.51233397e+01  5.27665056e+01     2.76431659e+01  5.34065821e+03  5.35020789e+03       9.54967709e+00
                                                   <Helix: wire>     1.00520603e+00     2.51233397e+01  5.28159346e+01     2.75000000e+01     2.51233397e+01  5.27665056e+01     2.76431659e+01  2.89316283e+02  2.89833612e+02       5.17328946e-01
                                                   <Hexagon: duct>   1.00493359e+00     2.51233397e+01  5.28159346e+01     2.75000000e+01     2.51233397e+01  5.27590134e+01     2.76356737e+01  3.62350518e+03  3.63096851e+03       7.46333195e+00
<fuel B0000-002 at ExCore XS: B BU GP: A>          <Circle: fuel>    1.01341384e+00     5.28159346e+01  8.00000000e+01     2.75000000e+01     5.28159346e+01  8.06848152e+01     2.78688806e+01  4.69068524e+04  4.69068524e+04       3.63797881e-11
                                                   <Circle: clad>    1.00520603e+00     5.28159346e+01  8.06848152e+01     2.75000000e+01     5.27665056e+01  8.04096715e+01     2.76431659e+01  5.34065821e+03  5.38426628e+03       4.36080648e+01
                                                   <Helix: wire>     1.00520603e+00     5.28159346e+01  8.06848152e+01     2.75000000e+01     5.27665056e+01  8.04096715e+01     2.76431659e+01  2.89316283e+02  2.91678637e+02       2.36235361e+00
                                                   <Hexagon: duct>   1.00493359e+00     5.28159346e+01  8.06848152e+01     2.75000000e+01     5.27590134e+01  8.03946871e+01     2.76356737e+01  3.62350518e+03  3.65408255e+03       3.05773746e+01
<fuel B0000-003 at ExCore XS: C BU GP: A>          <Circle: fuel>    1.01341384e+00     8.06848152e+01  1.07500000e+02     2.75000000e+01     8.06848152e+01  1.08553696e+02     2.78688806e+01  4.69068524e+04  4.69068524e+04       3.63797881e-11
                                                   <Circle: clad>    1.00520603e+00     8.06848152e+01  1.08553696e+02     2.75000000e+01     8.04096715e+01  1.08052837e+02     2.76431659e+01  5.34065821e+03  5.38426628e+03       4.36080648e+01
                                                   <Helix: wire>     1.00520603e+00     8.06848152e+01  1.08553696e+02     2.75000000e+01     8.04096715e+01  1.08052837e+02     2.76431659e+01  2.89316283e+02  2.91678637e+02       2.36235361e+00
                                                   <Hexagon: duct>   1.00493359e+00     8.06848152e+01  1.08553696e+02     2.75000000e+01     8.03946871e+01  1.08030361e+02     2.76356737e+01  3.62350518e+03  3.65408255e+03       3.05773746e+01
<fuel B0000-004 at ExCore XS: C BU GP: A>          <Circle: fuel>    1.01341384e+00     1.08553696e+02  1.35000000e+02     2.75000000e+01     1.08553696e+02  1.36422576e+02     2.78688806e+01  4.69068524e+04  4.69068524e+04       3.63797881e-11
                                                   <Circle: clad>    1.00520603e+00     1.08553696e+02  1.36422576e+02     2.75000000e+01     1.08052837e+02  1.35696003e+02     2.76431659e+01  5.34065821e+03  5.38426628e+03       4.36080648e+01
                                                   <Helix: wire>     1.00520603e+00     1.08553696e+02  1.36422576e+02     2.75000000e+01     1.08052837e+02  1.35696003e+02     2.76431659e+01  2.89316283e+02  2.91678637e+02       2.36235361e+00
                                                   <Hexagon: duct>   1.00493359e+00     1.08553696e+02  1.36422576e+02     2.75000000e+01     1.08030361e+02  1.35666035e+02     2.76356737e+01  3.62350518e+03  3.65408255e+03       3.05773746e+01
<plenum B0000-005 at ExCore XS: D BU GP: A>        <Circle: clad>    1.00520603e+00     1.36422576e+02  1.50000000e+02     1.50000000e+01     1.35696003e+02  1.50774094e+02     1.50780905e+01  2.91308630e+03  2.77271243e+03      -1.40373871e+02
                                                   <Helix: wire>     1.00520603e+00     1.36422576e+02  1.50774094e+02     1.50000000e+01     1.35696003e+02  1.50774094e+02     1.50780905e+01  1.57808882e+02  1.50204492e+02      -7.60438979e+00
                                                   <Hexagon: duct>   1.00493359e+00     1.36422576e+02  1.50774094e+02     1.50000000e+01     1.35666035e+02  1.50740038e+02     1.50740038e+01  1.97645737e+03  1.88172716e+03      -9.47302080e+01
<aclp plenum B0000-006 at ExCore XS: A BU GP: A>   <Circle: clad>    1.00520603e+00     1.50774094e+02  1.75000000e+02     2.50000000e+01     1.50774094e+02  1.75904245e+02     2.51301508e+01  4.85514383e+03  4.85514383e+03       2.72848411e-12
                                                   <Helix: wire>     1.00520603e+00     1.50774094e+02  1.75904245e+02     2.50000000e+01     1.50774094e+02  1.75904245e+02     2.51301508e+01  2.63014803e+02  2.63014803e+02       5.68434189e-14
                                                   <Hexagon: duct>   1.00493359e+00     1.50774094e+02  1.75904245e+02     2.50000000e+01     1.50740038e+02  1.75863378e+02     2.51233397e+01  3.85490024e+03  3.85594532e+03       1.04508710e+00
<plenum B0000-007 at ExCore XS: A BU GP: A>        <Circle: clad>    1.00520603e+00     1.75904245e+02  2.00000000e+02     2.50000000e+01     1.75904245e+02  2.01034395e+02     2.51301508e+01  4.85514383e+03  4.85514383e+03       2.72848411e-12
                                                   <Helix: wire>     1.00520603e+00     1.75904245e+02  2.01034395e+02     2.50000000e+01     1.75904245e+02  2.01034395e+02     2.51301508e+01  2.63014803e+02  2.63014803e+02       5.68434189e-14
                                                   <Hexagon: duct>   1.00493359e+00     1.75904245e+02  2.01034395e+02     2.50000000e+01     1.75863378e+02  2.00986718e+02     2.51233397e+01  3.29409561e+03  3.29498866e+03       8.93049530e-01
<duct B0000-008 at ExCore XS: A BU GP: A>          <Hexagon: duct>   1.00493359e+00     2.01034395e+02  2.25000000e+02     2.50000000e+01     2.00986718e+02  2.26110058e+02     2.51233397e+01  3.29409561e+03  3.28784427e+03      -6.25134671e+00
<SodiumBlock B0000-009 at ExCore XS: A BU GP: A>

keckler

I left some comments throughout the code, and have one more here:

In the PR description you say

Downstream testing illustrated some minor conservation issues with non-axial expansion target components (e.g., clad, wire, or duct within a fuel block) during thermal expansion. This PR addresses this

It is unclear to me exactly what has changed to fix the problem that you were seeing (partly because I don't know the exact problem you were seeing).

Is the fix related to the change of c.height = growFrac * blockHeight to c.height = growFrac * prevCompHeight?

keckler · 2023-07-06T22:31:22Z

armi/reactor/converters/axialExpansionChanger.py

+        self.massConservationReport = None
+        self.detailedMassConservationReport = None


It doesn't appear that there is any need to store these items on self, since they're only really used in one method of this class.

I think that trashing these dicts after each call to axiallyExpandAssembly() would be both safe and more memory efficient.

I put them on self in 29c345e simply to accommodate testing. I could use the mock logs to rebuild them if you'd prefer... But that seemed clunky to me and the easier solution was just to make them class variables.

That said, if we kept them as class instances, we could empty them after they are populated and used. Maybe a compromise of sorts?

Ok, so this should be resolved with 027db32. All of the logging now lives in its own class and only gets populated if the runLog verbosity is either extra or debug. The dicts that get populated will go out of scope with the rest of the expansion changer and should follow suit with the expansion data, etc.

armi/reactor/converters/axialExpansionChanger.py

albeanth · 2023-07-07T14:57:36Z

Is the fix related to the change of c.height = growFrac * blockHeight to c.height = growFrac * prevCompHeight?

@keckler, yes, exactly.

albeanth · 2023-07-07T17:03:13Z

The mass conservation tables look like the following:

massConservationReport

Is assembly-wide and shows the pre- and post-axial expansion mass values for the unique materials and components.
...

Just a note, I did confirm that the summary tables are identical post 027db32.

keckler

Thanks for that change, I think it streamlines this a lot!

armi/reactor/converters/axialExpansionChanger.py

armi/tests/detailedAxialExpansion/refSmallReactorBase.yaml

john-science · 2023-07-10T13:47:58Z

armi/reactor/converters/axialExpansionChanger.py

+                self.massConservationReport["(prev) mass"],
+            )
+        ]
+        self.massConservationReport["round(post - prev, 9)"] = diff


Why is this "self.massConservationReport"? Isn't this just a local variable? Is it used outside this method?

It appears to be a new variable that is local, but is initialized in the constructor.

I put it on self to accommodate some unit testing.

prefer returning it if its not actually needing to be an attribute.

we would have WAY more attributes if we did this every time we wanted to test.

albeanth · 2023-07-13T21:51:17Z

Is the fix related to the change of c.height = growFrac * blockHeight to c.height = growFrac * prevCompHeight?

@keckler, yes, exactly.

This is breaking downstream testing... Converting to draft while I debug.

albeanth added 5 commits July 5, 2023 12:14

simplify block height and c.nDens updates

3c3d49b

fix for using the wrong height (comp vs block)

fcb3efa

add new diagnostics to checks mass conservation

5221f8b

adjust unit tests to use mass conservation report

29c345e

- added actual aclp duct to illustrate and test for non-conservation

release notes

6a6a4d9

albeanth requested review from keckler and john-science July 6, 2023 15:11

albeanth marked this pull request as ready for review July 6, 2023 15:12

albeanth mentioned this pull request Jul 6, 2023

Formal Documentation of Axial Expansion in ARMI #935

Open

keckler requested changes Jul 6, 2023

View reviewed changes

albeanth added 2 commits July 6, 2023 16:07

rm special treatment for ACLP with axial expansion target component

c7b0ac9

reviewer feedback

80444f8

move logging to its own class

027db32

make logging threshhold a class variable

7c70f35

keckler approved these changes Jul 7, 2023

View reviewed changes

armi/reactor/converters/axialExpansionChanger.py Outdated Show resolved Hide resolved

albeanth added 2 commits July 7, 2023 15:51

fix broken unit test

d89ad69

put back verbosity on unit test

434b7b7

jakehader reviewed Jul 8, 2023

View reviewed changes

armi/tests/detailedAxialExpansion/refSmallReactorBase.yaml Show resolved Hide resolved

john-science added the enhancement New feature or request label Jul 10, 2023

john-science reviewed Jul 10, 2023

View reviewed changes

john-science approved these changes Jul 10, 2023

View reviewed changes

albeanth added 2 commits July 10, 2023 07:53

reviewer feedback

9d8362b

Merge branch 'main' into fixMinorConservationIssues_FINAL

ea05e6c

albeanth marked this pull request as draft July 13, 2023 21:50

albeanth added 2 commits July 13, 2023 15:08

reset c.height when axialExpChanger is reset

073f647

Merge branch 'main' into fixMinorConservationIssues_FINAL

2a9cd6f

keckler mentioned this pull request Oct 9, 2023

Be smarter about debug messages in the axial expansion changer #1427

Closed

7 tasks

john-science mentioned this pull request Oct 18, 2023

Misc cleanup #1439

Merged

8 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix minor conservation issue + new diagnostic summary for conservation #1338

Fix minor conservation issue + new diagnostic summary for conservation #1338

albeanth commented Jul 5, 2023 •

edited

albeanth commented Jul 6, 2023 •

edited

keckler left a comment

keckler Jul 6, 2023

albeanth Jul 6, 2023

albeanth Jul 6, 2023 •

edited

albeanth Jul 7, 2023

albeanth commented Jul 7, 2023

albeanth commented Jul 7, 2023

massConservationReport

keckler left a comment

john-science Jul 10, 2023

albeanth Jul 10, 2023

onufer Jul 11, 2023

albeanth commented Jul 13, 2023

		self.massConservationReport = None
		self.detailedMassConservationReport = None

Fix minor conservation issue + new diagnostic summary for conservation #1338

Are you sure you want to change the base?

Fix minor conservation issue + new diagnostic summary for conservation #1338

Conversation

albeanth commented Jul 5, 2023 • edited

What is the change?

Why is the change being made?

Checklist

albeanth commented Jul 6, 2023 • edited

massConservationReport

detailedMassConservationReport

keckler left a comment

Choose a reason for hiding this comment

keckler Jul 6, 2023

Choose a reason for hiding this comment

albeanth Jul 6, 2023

Choose a reason for hiding this comment

albeanth Jul 6, 2023 • edited

Choose a reason for hiding this comment

albeanth Jul 7, 2023

Choose a reason for hiding this comment

albeanth commented Jul 7, 2023

albeanth commented Jul 7, 2023

massConservationReport

keckler left a comment

Choose a reason for hiding this comment

john-science Jul 10, 2023

Choose a reason for hiding this comment

albeanth Jul 10, 2023

Choose a reason for hiding this comment

onufer Jul 11, 2023

Choose a reason for hiding this comment

albeanth commented Jul 13, 2023

albeanth commented Jul 5, 2023 •

edited

albeanth commented Jul 6, 2023 •

edited

albeanth Jul 6, 2023 •

edited