Issue with 'Pack Primary Phases' hanging indefinitely (Infinite Loop) during 2D to 3D RVE generation

[정의찬]

I am running into a persistent issue where the pipeline hangs indefinitely at the 'Pack Primary Phases' filter.

Context:

• Goal: Generate a 3D synthetic RVE from a single 2D EBSD scan of AA7075 aluminum alloy.

• Input Data: A .ang file pre-processed from an .osc file using a Python script (Standardized to SqrGrid, Radians, Phase 1 Aluminum).

• Pipeline Overview:

  1. Analysis Phase: Import .ang -> Segment Features (Misorientation 5°) -> Clean up small features (Minimum Size 20) -> Generate Ensemble Statistics.

  2. Synthesis Phase: Initialize Synthetic Volume -> Establish Shape Types (Ellipsoid) ->$ Pack Primary Phases.

• Note: I used the 'Create Ensemble Info' filter manually in the early steps because the phase information was not automatically populating during import, but I have confirmed that the data structure is correct before the packing step.

The Problem:

When the pipeline reaches [Filter 17] Pack Primary Phases, it hangs indefinitely with the status message:

"Packing Features || Placing Features"

• Symptoms: The progress bar does not move even after waiting for 24 to 72 hours.

• System Resource: CPU usage hovers around 5% (indicating single-core activity), and there is no memory leak or crash.

• Volume Size: Even with a very small volume (64x64x64, Resolution 1.0), the packing never completes.

Troubleshooting I tried:

• Data Verification: I exported feature data to CSV at intermediate steps. FeatureIds are generated correctly, and EquivalentDiameters show a reasonable distribution (no 1-pixel noise).

• Volume Check: I verified the empty grid generation by saving the .dream3d file right after Initialize Synthetic Volume. The grid dimensions and resolution are correct.

• Data Cleaning: I applied the 'Minimum Size' filter to remove features smaller than 20 pixels to simplify the statistics.

• Parameter Adjustment: I tried increasing the volume dimensions (to 128^3) and changing the resolution, but the issue persists.

Observation:

It seems that the packing algorithm is entering an infinite loop while trying to place the features defined by the input statistics. Since the input data is from a 2D section, I suspect there might be a conflict between the generated size/shape statistics and the 3D packing volume constraints.

Questions:

1. Is there a specific constraint in the Statistics object (e.g., Aspect Ratios or Size Distribution parameters) that typically causes Pack Primary Phases to hang?

2. Could the manual insertion of Create Ensemble Info be causing a hidden mismatch in the Attribute Matrix that affects the packing algorithm?

3. Are there any recommended settings or filters to "relax" the packing requirements to avoid this deadlock?

I have attached my pipeline (.json) and screenshots of the log. Any advice would be greatly appreciated.

Sincerely,

EUICHAN JEONG

Master's Student

Materials Mechanics Laboratory

Department of Material Science and Engineering

Changwon National University  

[Michael Jackson]

Dear Euichan Jeong, 

      DREAM.3D version 6.5.* cannot create a 3D Microstructure from a single 2D Slice. The information that created in the “Create Ensemble Statistics” will be incorrect and break assumptions made by the packing code.

There are no plans to fix this version of DREAM.3D

There is probably nothing that you can do to make it work short of saving a .dream3d file, hand editing the .dream3d file, then reopening the .dream3d to then generate the 3D microstructure.

--
Mike Jackson                    mike.j...@bluequartz.net
BlueQuartz Software         www.bluequartz.net

President/Owner               Dayton, Ohio

Principal DREAM.3D Developer

--
You received this message because you are subscribed to the Google Groups "dream3d-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dream3d-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dream3d-users/6ad201d8-bc14-4587-b1c5-86c35e1c2e0cn%40googlegroups.com.