MeshMaterialFilter bug with material ids

[egor1abs]

Bug Description

Initially, when reading the documentation, the implementation of bins placement in MeshMaterialFilter is not very clear.

If you look at the example made by @shimwell on the topic of R2S dose rate calculation, you can see mat_iron.id = 1. However, if you change it to another number, for example, more than 6, the bins parameter will be a strange set of values that have nothing to do with the original IDs of the materials.
Because of this, when trying to calculate neutron transport with get_microxs_and_flux method and using the MeshMaterialFilter as domain, the error "Could not find material X specified on tally filter." occurs. The material number indicated in the message has nothing to do with the actual number of materials existing in the model.

It seems to me that the problem of interpreting the number of material id at the C++ code.

Environment

Version | 0.15.3-dev86
Commit Hash | a64cc96
Date/Time | 2025-08-04 12:55:20
MPI Processes | 1
OpenMP Threads | 32

Nuclear library: ENDFB8.0-hdf5

[GuySten]

I've looked at shimwell's example and it does not include MeshMaterialFilter.

Can you include your exact failing script for easier debugging?

[egor1abs]

@GuySten, sorry, I've attached a wrong link 😅
Here is the correct link: https://github.com/fusion-energy/neutronics-workshop/blob/testing-MeshMaterialFilter/tasks/task_11_CSG_shut_down_dose_tallies/3_R2S_regularmesh_shutdown_dose_rate.ipynb

However, you can create any material and try to change its mat_id as I described above, and you will encounter the same problem if you try to create a openmc.MeshMaterialFilter object and look at the bins parameter after calculating material_volumes.

p.s. changed hyperlink in the original text

[GuySten]

When I change the material id of mat_iron.id to 10 I reproduce your crash.
However when I run the example unmodified it also do not work and I get the error message:
ValueError: The length of fluxes (1012) should be equal to the length of micros (1012) and the length of materials (1000).

I am not sure when this example worked but it is clearly does not work anymore!

[egor1abs]

That's interesting. Everything works for me with mat_id=1.
Are you using the latest version of OpenMC?

[GuySten]

I am running latest commit of the develop branch

[egor1abs]

Perhaps this code will work: https://github.com/fusion-energy/neutronics-workshop/blob/testing-MeshMaterialFilter/tasks/task_11_CSG_shut_down_dose_tallies/simple.py
It's almost the same, and the same problem occurs when changing the id of the material.

[shimwell]

The example linked is on a PR that has not been merged. I was waiting for the more complete r2s methods to be added to openmc before finishing it.

For r2s workflows i think this pr is the basis of any new examples.
#3508

I don't have any complete r2s examples in the neutronics workshop for now.

[egor1abs]

Okay, but I'm not referring to the entire script, but rather to a specific part of it with material definition and the logic behind MeshMaterialFilter when considered it separately.

Initially, I encountered this problem when working with a large model, and R2Smanager does not seem to solve the MeshMaterialFilter issue.

[GuySten]

Hopefully #3520 will fix the problem.

[egor1abs]

Thank you @GuySten!
I'll check it out tomorrow and let you know the results!

[GuySten]

@egor1abs, This issue should be open until my PR is merged. Then it will be automatically closed.
I do not have permission to reopen so you should do it.