Add K-HPMR Thermomechanics Model by soonklee · Pull Request #831 · idaholab/virtual_test_bed

soonklee · 2026-06-25T19:33:03Z

No description provided.

soonklee · 2026-06-25T19:34:28Z

ref #830

GiudGiud

is the technical report out for this model? I don't see any new reference.

The approach for mechanics seems similar to Kendrick's thesis at MIT, you might want to cite it there

It would be nice to point to mesh convergence studies in that report too.

Have we done homogenized thermo-mechanics at the core scale before @gambka ? Is it an established methodology?

GiudGiud · 2026-06-28T03:10:31Z

+  [null]
+    type = NullKernel
+    variable = temp
+    block = ${hp_blocks}
+  []


Suggested change

[null]

type = NullKernel

variable = temp

block = ${hp_blocks}

[]

you don't need to do that, just block restrict the variable

block restriction made on variable

I looked into removing null block, but wanted to explain why I eneded up keeping it rather than block-restricting temp from hp_blocks. temp on hp_blocks wasn't meant to be physically meaningful as Sockeye owns the heat pipe thermal-fluid physics, and coupling occurs through hp_temp_aux at heat_pipe_ht_surf on the solid side. The reason it needs to be included is that hp_block still needs to be part of the solid mechanics mesh, as heat pipes have to deform coherently with the rest of the core geometry. The mesh boundary/sideset definitions of top, bottom, side_mirror are whole-geometry constructs spanning the full model including hp_blocks. If temp doesn't cover hp_blocks, every BCs and postprocessor referencing those shared sidesets breaks with a "variable not defined on this boundary" error.

I did consider the alternative ways, but those required surgically touching every BCs and postprocessor referencing those boundaries, with more places for a future edit to silently reintroduce the same failure. While NullKernel isn't ideal for a genuine nonlinear variable, since temp is a single field shared across both non_hp_blocks and hp_blocks (bookkeeping only), it can't be split into two variable types within a system. Given that constraint, this seemed like the right tool for the job. Rational for using NullKernel is specified within the input.

Excluding selected blocks from a sideset takes a single ParsedGenerateSideset with an "exclude_subdomains" argument

The mesh boundary/sideset definitions of top, bottom, side_mirror are whole-geometry constructs spanning the full model including hp_blocks. If temp doesn't cover hp_blocks, every BCs and postprocessor referencing those shared sidesets breaks with a "variable not defined on this boundary" error.

this is exactly the argument for adding specific boundaries for temperature. you're currently applying boundary conditions on sidesets facing elements where only a nullkernel exists. The effect of that is unknown.

gambka · 2026-07-02T20:00:49Z

is the technical report out for this model? I don't see any new reference.

The approach for mechanics seems similar to Kendrick's thesis at MIT, you might want to cite it there

It would be nice to point to mesh convergence studies in that report too.

Have we done homogenized thermo-mechanics at the core scale before @gambka ? Is it an established methodology?

We've used homogenization techniques for thermal solutions for other applications. I'm not aware of full core scale thermo-mechanics applications, but I'd have to read up on the approach used here. There are established homogenization methods. Applications of these techniques have been mostly applied to upscaling of lower-length scale models and their information into engineering scale studies. An example is provided here: 10.1016/j.commatsci.2014.12.039

soonklee

Thanks for the thorough review @GiudGiud. Comments are addressed below.

Changes include:

Figures resized to <500kb as requested
Mesh file properly tracked with git lfs, including cleaning the raw binary out of early commit history
Markdowns revised as requested
Input files revised as requested, plus a couple of minor clarifying comments added

The model is updated and ready for further review.
Note that this branch was rebased to fix the lfs history.

…el.md Co-authored-by: Guillaume Giudicelli <guillaume.giudicelli@gmail.com>

soonklee

Minor revisions made to inputs.

GiudGiud · 2026-07-15T18:47:32Z

Needs testing. If you're not adding it I'll do it when I have time

soonklee · 2026-07-15T22:19:06Z

Needs testing. If you're not adding it I'll do it when I have time

I haven't written a tests spec before, but I'll get one set up and push it to PR soon.

soonklee · 2026-07-16T20:10:18Z

I added the initial regression test structure following the existing examples. I would appreciate feedback on whether the current test settings are appropriate.

… tests/hpc_tests stucture

soonklee · 2026-07-22T03:25:08Z

Hi @GiudGiud,
Please see following updates for tests/hpc_tests.
Latest push includes:

Fix and regenerate hp_fine mesh input and gold to correctly designate outer_reflector block for bison
Document NullKernel rationale for temp on hp_blocks (Sockeye owns heat pipe physics)
Regenerate steady_K_SM and load_following_K_SM gold CSVs
Add ss symlink for load_following_K_SM, following the existing Na convention
List standalone tests for Griffin and Sockeye on mrad/steady_K_SM/tests and for Bison on mrad/hpc_tests due to fine mesh
Confirmed steady/load_following coupled_tests on mrad/hpc_tests with loosen tolerance

soonklee

Pushed gold and test matrix. Ready for re-review.

GiudGiud

testing looks good. I'll turn it on right now.

I ll send you an invite to be part of the VTB organization on github so the tests run automatically.
The Direwolf HPC recipe is the one we ll need to watch here

GiudGiud · 2026-08-14T20:09:01Z

@@ -0,0 +1,705 @@
+################################################################################
+## NEAMS Micro-Reactor Application Driver                                     ##
+## Heat Pipe Microreactor Steady State                                        ##


Suggested change

## Heat Pipe Microreactor Steady State ##

## Heat Pipe Microreactor load follow ##

GiudGiud · 2026-08-14T20:22:41Z

+  [null]
+    type = NullKernel
+    variable = temp
+    block = ${hp_blocks}
+  []


Excluding selected blocks from a sideset takes a single ParsedGenerateSideset with an "exclude_subdomains" argument

The mesh boundary/sideset definitions of top, bottom, side_mirror are whole-geometry constructs spanning the full model including hp_blocks. If temp doesn't cover hp_blocks, every BCs and postprocessor referencing those shared sidesets breaks with a "variable not defined on this boundary" error.

this is exactly the argument for adding specific boundaries for temperature. you're currently applying boundary conditions on sidesets facing elements where only a nullkernel exists. The effect of that is unknown.

GiudGiud · 2026-08-14T20:23:49Z

@@ -0,0 +1,318 @@
+################################################################################
+## NEAMS Micro-Reactor Application Driver                                     ##
+## Heat Pipe Microreactor Steady State                                        ##


Suggested change

## Heat Pipe Microreactor Steady State ##

## Heat Pipe Microreactor Load following transient ##

GiudGiud · 2026-08-14T20:24:25Z

+## Heat Pipe Microreactor Steady State                                        ##
+## BISON Child Application input file                                         ##
+## Thermomechanical model (Heat conduction, Thermal Expansion, Thermal Stress)##
+## FY26 Summer Update for Solid Mechanics Implementation                      ##


Suggested change

## FY26 Summer Update for Solid Mechanics Implementation ##

## of partially-homogenized geometry ##

## FY26 Summer Update for Solid Mechanics Implementation ##

GiudGiud · 2026-08-14T20:25:11Z

+## Heat Pipe Microreactor Steady State                                        ##
+## Sockeye Grandchild Application input file                                  ##


Suggested change

## Heat Pipe Microreactor Steady State ##

## Sockeye Grandchild Application input file ##

## Heat Pipe Microreactor Load following transient ##

## Sockeye Grandchild Application input file ##

GiudGiud requested changes Jun 28, 2026

View reviewed changes

GiudGiud reviewed Jun 28, 2026

View reviewed changes

Comment thread microreactors/mrad/load_following_K_SM/K-HPMR_GRIFFIN_tr.i Outdated

soonklee force-pushed the k_hpmr_thermomechanics branch from 56b0ffd to bb62443 Compare June 30, 2026 16:31

soonklee added 2 commits July 1, 2026 19:38

Add K-HPMR Thermomechanics Implementation ref issue idaholab#830

4cf5dbf

Address review comments and track HPMR mesh with Git LFS

946b94c

soonklee force-pushed the k_hpmr_thermomechanics branch from 7912db0 to 946b94c Compare July 2, 2026 00:47

Clarify nonlinear comments for disp_trans

520e28e

soonklee commented Jul 2, 2026

View reviewed changes

GiudGiud reviewed Jul 6, 2026

View reviewed changes

soonklee and others added 4 commits July 6, 2026 15:12

Update doc/content/microreactors/mrad/mrad_k_hpmr_thermomechanics_mod…

68cfb34

…el.md Co-authored-by: Guillaume Giudicelli <guillaume.giudicelli@gmail.com>

Update doc/content/microreactors/mrad/mrad_k_hpmr_thermomechanics_mod…

3830fa7

…el.md Co-authored-by: Guillaume Giudicelli <guillaume.giudicelli@gmail.com>

Update doc/content/microreactors/mrad/mrad_k_hpmr_thermomechanics_mod…

efdfb7a

…el.md Co-authored-by: Guillaume Giudicelli <guillaume.giudicelli@gmail.com>

Address review comments and add refs for load following description

671f068

soonklee commented Jul 7, 2026

View reviewed changes

Comment thread microreactors/mrad/load_following_K_SM/K-HPMR_GRIFFIN_tr.i

joshuahansel requested changes Jul 8, 2026

View reviewed changes

Comment thread microreactors/mrad/load_following_K_SM/K-HPMR_SOCKEYE_tr.i Outdated

Comment thread microreactors/mrad/load_following_K_SM/K-HPMR_SOCKEYE_tr.i Outdated

Comment thread doc/content/microreactors/mrad/mrad_k_hpmr_thermomechanics_results.md

Address comments for Sockeye input

4c58ada

joshuahansel approved these changes Jul 9, 2026

View reviewed changes

soonklee requested a review from GiudGiud July 15, 2026 17:25

Add K-HPMR regression tests

bcf1007

GiudGiud reviewed Jul 16, 2026

View reviewed changes

Comment thread microreactors/mrad/mesh/tests

Comment thread microreactors/mrad/steady_K_SM/tests Outdated

Comment thread microreactors/mrad/steady_K_SM/tests

Made minor edits, update steady/load-following gold data, and confirm…

fde3384

… tests/hpc_tests stucture

soonklee commented Jul 22, 2026

View reviewed changes

soonklee requested a review from GiudGiud August 13, 2026 15:08

GiudGiud added the new_model Adding a new model to the VTB label Aug 13, 2026

GiudGiud requested changes Aug 14, 2026

View reviewed changes

	## Heat Pipe Microreactor Steady State ##
	## Heat Pipe Microreactor load follow ##

	## Heat Pipe Microreactor Steady State ##
	## Heat Pipe Microreactor Load following transient ##

	## FY26 Summer Update for Solid Mechanics Implementation ##
	## of partially-homogenized geometry ##
	## FY26 Summer Update for Solid Mechanics Implementation ##

		## Heat Pipe Microreactor Steady State ##
		## Sockeye Grandchild Application input file ##

Uh oh!

Conversation

soonklee commented Jun 25, 2026

Uh oh!

soonklee commented Jun 25, 2026

Uh oh!

GiudGiud left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gambka commented Jul 2, 2026

Uh oh!

soonklee left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

soonklee left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

GiudGiud commented Jul 15, 2026

Uh oh!

soonklee commented Jul 15, 2026

Uh oh!

soonklee commented Jul 16, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

soonklee commented Jul 22, 2026

Uh oh!

soonklee left a comment

Choose a reason for hiding this comment

Uh oh!

GiudGiud left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!