Add compositional dependence on viscosity in the Steinberger material model #5260

poulamiiroy · 2023-07-11T01:02:04Z

This pull request supersedes the PR #5248.

Added compositional dependence on viscosity in the Steinberger material model. The original Steinberger material model uses one viscosity profile for all of the compositions. So, I have added prefectors which are multiplied by original viscosity based on volume fraction.

I have read the guidelines in our CONTRIBUTING.md document.

For all pull requests:

I have followed the instructions for indenting my code.
I have tested my new feature locally to ensure it is correct.
I have created a testcase for the new feature/benchmark in the tests/ directory.
I have added a changelog entry in the doc/modules/changes directory that will inform other users of my change.

naliboff

@poulamiiroy - Thank you for the contribution and breaking the original PR into smaller contributions! I think with a few moderate adjustments the PR will be in good shape. Please let me know if you have any questions about my comments, and thank you again for the contribution!

naliboff · 2023-07-11T21:37:50Z

include/aspect/material_model/steinberger.h

@@ -274,6 +274,7 @@ namespace aspect
        std::vector<double> conductivity_reference_temperatures;
        std::vector<double> conductivity_exponents;
        std::vector<double> saturation_scaling;
+        std::vector<double> prefactors;


Suggested change

std::vector<double> prefactors;

std::vector<double> viscosity_prefactors;

Can you also move this variable down a few lines and add documentation that described what it represents. The variable is currently lumped with other variables related to thermal conductivity.

naliboff · 2023-07-11T21:38:23Z

source/material_model/steinberger.cc

               const SymmetricTensor<2,dim> &,
               const Point<dim> &position) const
    {
      const double depth = this->get_geometry_model().depth(position);
      const double adiabatic_temperature = this->get_adiabatic_conditions().temperature(position);
+      // Calculate volume fractions from mass fractions


Can the following four lines be removed?

naliboff · 2023-07-11T21:41:13Z

source/material_model/steinberger.cc

@@ -189,11 +194,14 @@ namespace aspect
      // For an explanation on this formula see the Steinberger & Calderwood 2006 paper
      const double vis_lateral_exp = -1.0*lateral_viscosity_lookup->lateral_viscosity(depth)*delta_temperature/(temperature*adiabatic_temperature);
      // Limit the lateral viscosity variation to a reasonable interval
+


I would remove this extra line, and instead add it between lines 195 and 196.

naliboff · 2023-07-11T21:45:31Z

source/material_model/steinberger.cc

      const double vis_lateral = std::max(std::min(std::exp(vis_lateral_exp),max_lateral_eta_variation),1/max_lateral_eta_variation);

      const double vis_radial = radial_viscosity_lookup->radial_viscosity(depth);
+      const double vis_compositional = MaterialUtilities::average_value (volume_fractions, prefactors, viscosity_averaging);


Suggested change

const double vis_compositional = MaterialUtilities::average_value (volume_fractions, prefactors, viscosity_averaging);

// Volume fraction averaged value of the composition-dependent viscosity prefactors

const double vis_prefactor = MaterialUtilities::average_value (volume_fractions, prefactors, viscosity_averaging);

I wanted to keep it as vis_compositional since I am already renaming "prefactors" as "viscosity_prefactors" as you have suggested. What do you think?

naliboff · 2023-07-11T21:46:47Z

source/material_model/steinberger.cc

@@ -268,11 +276,16 @@ namespace aspect

      // Evaluate the equation of state properties over all evaluation points
      equation_of_state.evaluate(eos_in, eos_outputs);
+      // std::vector<Vector<double>> viscosities(in.n_evaluation_points(), this->n_compositional_fields());


Can you remove this and additional commented lines below if they are no longer needed?

naliboff · 2023-07-11T21:50:12Z

source/material_model/steinberger.cc


-      return std::max(std::min(vis_lateral * vis_radial,max_eta),min_eta);
+      return std::max(std::min(vis_lateral * vis_radial * vis_compositional,max_eta),min_eta);


I confess I don't understand this equation, as it looks like viscosity values are being multiplied together. Would it be possible to add the equation here (or further above) in a comment, along with a brief description?

Agreed, as written it looks like multiplication of different viscosities. Would you mind renaming

vis_lateral to temperature_prefactor

vis_compositional to compositional_prefactor

vis_radial to eta_ref?

I second @bobmyhill renaming scheme for the values

naliboff · 2023-07-11T21:51:32Z

source/material_model/steinberger.cc

+                                                "geometric, or maximum composition.");
+          prm.declare_entry ("Prefactors", "1.e21",
+                             Patterns::Anything(),
+                             "List of viscosities for background mantle and compositional fields,"


Just to be sure, are the "Prefactors" in fact viscosity values, or are they dimensionless composition-dependent factors that the viscosity is multiplied by?

Follow up - depending on the answer, I would adjust the parameter name and description to be a bit more specific.

Prefactors are the dimensionless composition-dependent factors, for example, if one composition needs to have 10 times higher viscosity than ambient, then prefactor for this composition would be 10.

naliboff · 2023-07-11T21:52:27Z

source/material_model/steinberger.cc

-
+          viscosity_averaging = MaterialUtilities::parse_compositional_averaging_operation ("Viscosity averaging scheme",
+                                prm);
+          // viscosities = Utilities::parse_map_to_double_array (prm.get("Viscosities"),


Remove commented lines?

Thanks a lot John for reviewing my pull request. I will follow what you have suggested. Please let me know for further discussion.

jdannberg

This looks goof to me, thank you for contributing this! Can you address the few small comments I left, and then I think this is good to be merged!

jdannberg · 2023-07-13T18:28:19Z

source/material_model/steinberger.cc

+          // if (in.requests_property(MaterialProperties::viscosity))
+          // out.viscosities[i] = viscosity(in.temperature[i], in.pressure[i], in.composition[i], in.strain_rate[i], in.position[i]);


If these lines are not needed anymore, simply remove them.

jdannberg · 2023-07-13T18:28:47Z

source/material_model/steinberger.cc

+//  out.viscosities[i] = MaterialUtilities::average_value (volume_fractions[i], viscosities, viscosity_averaging);
+// out.viscosities[i] = viscosities[i];


Same here, these lines can be removed.

jdannberg · 2023-07-13T18:30:15Z

source/material_model/steinberger.cc

-                             "model. Larger values will be cut off.");
+                             "model. Larger values will be cut off."),


I do not think you want to change this. This actually needs to be a semicolon.

jdannberg · 2023-07-13T18:30:44Z

source/material_model/steinberger.cc

+                                                Patterns::Selection("arithmetic|harmonic|geometric|maximum composition"),
+                                                "When more than one compositional field is present at a point "
+                                                "with different viscosities, we need to come up with an average "
+                                                "viscosity at that point.  Select a weighted harmonic, arithmetic, "


Suggested change

"viscosity at that point. Select a weighted harmonic, arithmetic, "

"viscosity at that point. Select a weighted harmonic, arithmetic, "

jdannberg · 2023-07-13T18:34:45Z

source/material_model/steinberger.cc

+          // viscosities = Utilities::parse_map_to_double_array (prm.get("Viscosities"),
+          // list_of_composition_names,
+          // has_background_field,
+          // "Viscosities");


Suggested change

// viscosities = Utilities::parse_map_to_double_array (prm.get("Viscosities"),

// list_of_composition_names,

// has_background_field,

// "Viscosities");

I think John is just asking to remove these lines.

jdannberg · 2023-07-13T18:38:44Z

tests/multicomponent_steinberger.prm

+subsection Steinberger model
+    set Use lateral average temperature for viscosity = false
+    set Data directory = $ASPECT_SOURCE_DIR/data/material-model/steinberger/
+ #  set Derivatives file names = /home/poulami/software/aspect_30_09_2022/aspect/data/material-model/steinberger/test-steinberger-compressible/constant_material_small_hefesto_derivatives.txt


If you're not using that line, simply remove it.

jdannberg · 2023-07-13T18:42:09Z

tests/multicomponent_steinberger.prm

+    set Use lateral average temperature for viscosity = false
+    set Data directory = $ASPECT_SOURCE_DIR/data/material-model/steinberger/
+ #  set Derivatives file names = /home/poulami/software/aspect_30_09_2022/aspect/data/material-model/steinberger/test-steinberger-compressible/constant_material_small_hefesto_derivatives.txt
+ set Radial viscosity file name = radial-visc.txt #, radial-visc.txt, radial-visc.txt     ## lithosphere at top


Suggested change

set Radial viscosity file name = radial-visc.txt #, radial-visc.txt, radial-visc.txt ## lithosphere at top

set Radial viscosity file name = radial-visc.txt # lithosphere at top

Same here, remove everything that is not used.

jdannberg · 2023-07-13T18:42:26Z

tests/multicomponent_steinberger.prm

+    set Data directory = $ASPECT_SOURCE_DIR/data/material-model/steinberger/
+ #  set Derivatives file names = /home/poulami/software/aspect_30_09_2022/aspect/data/material-model/steinberger/test-steinberger-compressible/constant_material_small_hefesto_derivatives.txt
+ set Radial viscosity file name = radial-visc.txt #, radial-visc.txt, radial-visc.txt     ## lithosphere at top
+ set Lateral viscosity file name = temp-viscosity-prefactor.txt  ##, temp-viscosity-prefactor.txt, temp-viscosity-prefactor.txt


Suggested change

set Lateral viscosity file name = temp-viscosity-prefactor.txt ##, temp-viscosity-prefactor.txt, temp-viscosity-prefactor.txt

set Lateral viscosity file name = temp-viscosity-prefactor.txt

jdannberg · 2023-07-13T18:44:20Z

data/material-model/steinberger/llsvp_vs_vp.txt

It does not look like you are actually using this file. Is that intentional? If you do not want to use it in your test case below, simply remove it from your pull request.

jdannberg · 2023-07-13T18:51:03Z

source/material_model/steinberger.cc

+      //The lateral variation of viscosity due to lateral temperature (Visc_lT)
+      //Visc_lT = exp [(-1)*(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT)], Eq. 6 of the paper


I think what @naliboff was asking for was a more detailed explanation of why these things are multiplied together. I have made a suggestion below:

Suggested change

//The lateral variation of viscosity due to lateral temperature (Visc_lT)

//Visc_lT = exp [(-1)*(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT)], Eq. 6 of the paper

// We here compute the lateral variation of viscosity due to temperature (vis_lateral) as

// V_lT = exp [-(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT))] as in Eq. 6 of the paper.

// We get H/nR from the lateral_viscosity_lookup->lateral_viscosity function.

@jdannberg Thanks a lot for your review! I made the changes.

jdannberg

It looks like many of the tests that use the steinberger material model are failing now. They all have the error message

An error occurred in line <943> of file </__w/aspect/aspect/source/material_model/utilities.cc> in function
    double aspect::MaterialModel::MaterialUtilities::average_value(const std::vector<double, std::allocator<double> >&, const std::vector<double, std::allocator<double> >&, const aspect::MaterialModel::MaterialUtilities::CompositionalAveragingOperation&)
The violated condition was: 
    volume_fractions.size() == parameter_values.size()
Additional information: 
    The volume fractions and parameter values vectors used for averaging
    have to have the same length!

To figure out what might be the problem, I would suggest to run these tests locally on your machine. For example, you could run them in the debugger to see what the two values are (volume fractions and viscosity prefactors) and which of these actually has the wrong length.

jdannberg · 2023-07-14T16:12:09Z

tests/multicomponent_steinberger/screen-output

You will need to update this test output with the one from the official tester (and same for the statistics file). The file that you added here still, for example, includes the output path on your own computer (Writing graphical output: /home/poulami/software/Aspect_globalscale/aspect/tests/multicomponent_steinberger/solution/solution-00001)

jdannberg · 2023-07-14T17:23:54Z

/rebuild

bobmyhill

A couple of suggested changes to variable names that will make code comprehension easier.

bobmyhill · 2023-07-14T23:49:53Z

source/material_model/steinberger.cc


-      return std::max(std::min(vis_lateral * vis_radial,max_eta),min_eta);
+      return std::max(std::min(vis_lateral * vis_radial * vis_compositional,max_eta),min_eta);


Agreed, as written it looks like multiplication of different viscosities. Would you mind renaming

vis_lateral to temperature_prefactor

vis_compositional to compositional_prefactor

vis_radial to eta_ref?

bobmyhill · 2023-07-15T00:28:07Z

source/material_model/steinberger.cc

@@ -187,13 +189,22 @@ namespace aspect
        delta_temperature = temperature-adiabatic_temperature;

      // For an explanation on this formula see the Steinberger & Calderwood 2006 paper
+      // We here compute the lateral variation of viscosity due to temperature (vis_lateral) as
+      // V_lT = exp [-(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT))] as in Eq. 6 of the paper.
+      // We get H/nR from the lateral_viscosity_lookup->lateral_viscosity function.
      const double vis_lateral_exp = -1.0*lateral_viscosity_lookup->lateral_viscosity(depth)*delta_temperature/(temperature*adiabatic_temperature);


Please rename vis_lateral_exp to log_temperature_prefactor

A follow-up question on these lines.

The equation in the comments is written as V_lT = exp [-(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT))], while the actual equation on line 195 has delta_temperature/(temperature*adiabatic_temperature).

Is the equation in the comments referring to the same equation as the one on line 195? If yes, can you confirm which one is correct or expand the documentation a bit further to explain the differences between the two equations?

@naliboff I suggested to write the equation like this, this is how it is written in the Steinberger and Calderwood paper.
temperature is the same as T_adiabatic + dT

naliboff

@poulamiiroy - A few additional comments on top of those from @bobmyhill.

naliboff · 2023-07-15T03:04:15Z

source/material_model/steinberger.cc


-      return std::max(std::min(vis_lateral * vis_radial,max_eta),min_eta);
+      return std::max(std::min(vis_lateral * vis_radial * vis_compositional,max_eta),min_eta);


I second @bobmyhill renaming scheme for the values

naliboff · 2023-07-15T03:10:28Z

source/material_model/steinberger.cc

@@ -187,13 +189,22 @@ namespace aspect
        delta_temperature = temperature-adiabatic_temperature;

      // For an explanation on this formula see the Steinberger & Calderwood 2006 paper
+      // We here compute the lateral variation of viscosity due to temperature (vis_lateral) as
+      // V_lT = exp [-(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT))] as in Eq. 6 of the paper.
+      // We get H/nR from the lateral_viscosity_lookup->lateral_viscosity function.
      const double vis_lateral_exp = -1.0*lateral_viscosity_lookup->lateral_viscosity(depth)*delta_temperature/(temperature*adiabatic_temperature);


A follow-up question on these lines.

The equation in the comments is written as V_lT = exp [-(H/nR)*dT/(T_adiabatic*(T_adiabatic + dT))], while the actual equation on line 195 has delta_temperature/(temperature*adiabatic_temperature).

Is the equation in the comments referring to the same equation as the one on line 195? If yes, can you confirm which one is correct or expand the documentation a bit further to explain the differences between the two equations?

jdannberg

Okay, the code looks much better now and most of the tests are succeeding as well! 👍

But there is still one test that is failing (steinberger_phase_fractions_background). The test finishes now, but the results are different, so you'll need to figure out why the output is different (i.e. which column in the graphical output is different? Is it the viscosity? What has changed compared to before?).

In addition, you have accidentally added some lines in your last commit that are just printing statements you commented out.

jdannberg · 2023-07-15T04:37:09Z

source/material_model/steinberger.cc

+// std::cout<<"size vf"<<volume_fractions.size()<<std::endl;
+// std::cout<<"size vis_pre"<<viscosity_prefactors.size()<<std::endl;


You added some lines again here that are code you commented out and that is no longer needed. Please remove them.

jdannberg · 2023-07-15T04:38:26Z

source/material_model/steinberger.cc

+// std::cout<<"size vf"<<volume_fractions.size()<<std::endl;
+// std::cout<<"size vis_pre"<<viscosity_prefactors.size()<<std::endl;
+      const double compositional_prefactor = MaterialUtilities::average_value (volume_fractions, viscosity_prefactors, viscosity_averaging);
+      // std::cout<<"size vf"<<volume_fractions.size()<<std::endl;


Same here, please remove this line.

jdannberg · 2023-07-15T04:38:52Z

source/material_model/steinberger.cc

@@ -303,7 +305,7 @@ namespace aspect
                  chemical_compositions.push_back(in.composition[i][c]);

              mass_fractions = MaterialUtilities::compute_composition_fractions(chemical_compositions, *composition_mask);
-
+// std::cout<<"mf"<<mass_fractions<<std::endl;


Please remove this line as well.

bobmyhill · 2023-07-21T02:53:49Z

@poulamiiroy Ignore my last; thank you for making the variable name changes we suggested.
Do you need any help with figuring out why the tests are failing?

bobmyhill · 2023-07-21T03:03:20Z

source/material_model/steinberger.cc

@@ -551,7 +580,15 @@ namespace aspect
                                  + " fields of type chemical composition."));

          has_background_field = (equation_of_state.number_of_lookups() == n_chemical_fields + 1);
-
+          std::vector<std::string> list_of_composition_names = this->introspection().get_composition_names();


I think you want fields that correspond to chemical compositions, rather than all compositional fields:

Suggested change

std::vector<std::string> list_of_composition_names = this->introspection().get_composition_names();

std::vector<std::string> list_of_composition_names = this->introspection().chemical_composition_field_names();

This might fix a couple of the tests.

bobmyhill · 2023-07-21T03:21:54Z

source/material_model/steinberger.cc

+          if (equation_of_state.number_of_lookups() == 1)
+            list_of_composition_names.resize(1, "background");
+          if ((equation_of_state.number_of_lookups() == 1) && (has_background_field))
+            list_of_composition_names.resize(0, "background");


Are you sure that these lines cover all the potential use cases of the material model?

The way you have implemented this, if the number of lookups is equal to 1, the vector of composition names is replaced by {"background"}. If additionally a background field is implied, the vector is empty {}.

What happens when the number of lookups is not equal to 1? You should test that parse_map_to_double_array does the correct thing in these cases.

You might be also be interested in mirroring the changes I made here, using the new map parser:
https://github.com/geodynamics/aspect/pull/5039/files#diff-cfe301f322bebee87432e53137677e6372743e5e142733d8fb63b5599c9322cb

poulamiiroy · 2023-07-28T12:20:59Z

Thank you for your comments! I am working on it. I'll ping you when the PR needs a new review. :)

bobmyhill · 2023-07-28T13:07:19Z

source/material_model/steinberger.cc

+          // if ((equation_of_state.number_of_lookups() > 1) && (has_background_field))
+          // list_of_composition_names.resize(equation_of_state.number_of_lookups(), "background");
+          if (equation_of_state.number_of_lookups() > 1)
+            list_of_composition_names.resize(n_chemical_fields, "background");


The line that you've added here appends to the list_of_composition_names vector with entries called "background" (e.g. {"a", "b", "background", "background", "background", "background", ...}.

I doubt that is what you want to do. The background field is the first compositional field, not the last. And there is only ever one of them.

Have another look at the link above. I suspect "insert" will be more useful to you than "resize".

If you're ever unsure of what a C++ function does, you can always look it up: https://cplusplus.com/reference/vector/vector/resize/
https://cplusplus.com/reference/vector/vector/insert/

bobmyhill · 2023-08-14T17:12:27Z

Closed in favour of #5362

poulamiiroy added 8 commits July 10, 2023 11:49

add_steinberger_multicomponent

ef336d8

add llsvp material file and one test file

a9d4542

add steinberger multicomponent header file

1210975

add steinberger multicomponent source file

1454948

add steinberger multicomponent parameter file

c48b29d

add test input file

e3d1a2b

add test output

70fa91e

add indentations

6f88ba5

naliboff requested changes Jul 11, 2023

View reviewed changes

poulamiiroy added 6 commits July 12, 2023 23:16

add header file

0bf41c8

add source file

847aa75

add parameter file

7c5e5cd

screen output

16c2e3f

add header

62d958b

add source

30d8106

jdannberg reviewed Jul 13, 2023

View reviewed changes

address comments

2e3bdbd

jdannberg reviewed Jul 14, 2023

View reviewed changes

jdannberg added the ready to test label Jul 14, 2023

bobmyhill requested changes Jul 15, 2023

View reviewed changes

naliboff reviewed Jul 15, 2023

View reviewed changes

poulamiiroy added 2 commits July 14, 2023 20:48

address test issue

96f5301

address test issues with indentation

0668153

jdannberg reviewed Jul 15, 2023

View reviewed changes

bobmyhill requested changes Jul 21, 2023

View reviewed changes

poulamiiroy added 2 commits July 28, 2023 14:14

Resize vector of composition names

adfe2f0

Fix Test

13dd4bf

bobmyhill reviewed Jul 28, 2023

View reviewed changes

poulamiiroy added 2 commits August 11, 2023 17:30

Fix background field

83dcfbb

add indent

3b5d825

bobmyhill mentioned this pull request Aug 12, 2023

Add compositional dependence on viscosity in the Steinberger material model #5362

Merged

bobmyhill closed this Aug 14, 2023

	std::vector<double> prefactors;
	std::vector<double> viscosity_prefactors;

-      const double vis_compositional = MaterialUtilities::average_value (volume_fractions, prefactors, viscosity_averaging);
+      // Volume fraction averaged value of the composition-dependent viscosity prefactors
+      const double vis_prefactor = MaterialUtilities::average_value (volume_fractions, prefactors, viscosity_averaging);


		return std::max(std::min(vis_lateral * vis_radial,max_eta),min_eta);
		return std::max(std::min(vis_lateral * vis_radial * vis_compositional,max_eta),min_eta);

		// if (in.requests_property(MaterialProperties::viscosity))
		// out.viscosities[i] = viscosity(in.temperature[i], in.pressure[i], in.composition[i], in.strain_rate[i], in.position[i]);

		// out.viscosities[i] = MaterialUtilities::average_value (volume_fractions[i], viscosities, viscosity_averaging);
		// out.viscosities[i] = viscosities[i];

		"model. Larger values will be cut off.");
		"model. Larger values will be cut off."),

	"viscosity at that point. Select a weighted harmonic, arithmetic, "
	"viscosity at that point. Select a weighted harmonic, arithmetic, "

	set Radial viscosity file name = radial-visc.txt #, radial-visc.txt, radial-visc.txt ## lithosphere at top
	set Radial viscosity file name = radial-visc.txt # lithosphere at top

	set Lateral viscosity file name = temp-viscosity-prefactor.txt ##, temp-viscosity-prefactor.txt, temp-viscosity-prefactor.txt
	set Lateral viscosity file name = temp-viscosity-prefactor.txt

		//The lateral variation of viscosity due to lateral temperature (Visc_lT)
		//Visc_lT = exp [(-1)(H/nR)dT/(T_adiabatic*(T_adiabatic + dT)], Eq. 6 of the paper

		// std::cout<<"size vf"<<volume_fractions.size()<<std::endl;
		// std::cout<<"size vis_pre"<<viscosity_prefactors.size()<<std::endl;

	std::vector<std::string> list_of_composition_names = this->introspection().get_composition_names();
	std::vector<std::string> list_of_composition_names = this->introspection().chemical_composition_field_names();

Add compositional dependence on viscosity in the Steinberger material model #5260

Add compositional dependence on viscosity in the Steinberger material model #5260

Conversation

poulamiiroy commented Jul 11, 2023

For all pull requests:

naliboff left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jdannberg left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jdannberg left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jdannberg commented Jul 14, 2023

bobmyhill left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

naliboff left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jdannberg left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bobmyhill commented Jul 21, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

poulamiiroy commented Jul 28, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bobmyhill commented Aug 14, 2023