# Pnm Flow Eval

> Evaluates a topology-based pore network model's ability to predict flow-permeable surface area and hydraulic conductance in granular materials from micro-CT images. Use when the user wants to benchmark on Sphere Packing & High-Explosive Micro-CT Samples, or asks about evaluating this task. Reports conductance_ratio.

- Skill: `qhjqhj00/pnm-flow-eval` (Agent Skill)
- Install (CLI): `npx skillmds add qhjqhj00/pnm-flow-eval`
- Raw SKILL.md: https://api.skillmd.com/api/skills/qhjqhj00/pnm-flow-eval/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: qhjqhj00 (https://skillmd.com/u/qhjqhj00)
- Updated: 2026-09-08
- Page: https://skillmd.com/skills/qhjqhj00/pnm-flow-eval

---


# pnm-flow-eval

> Towards replacing physical testing of granular materials with a Topology-based Model — Venkat et al. (2021) (arXiv:2109.08777, 2021)

## What this evaluates

Evaluates a topology-based pore network model's ability to predict flow-permeable surface area and hydraulic conductance in granular materials from micro-CT images.

## Datasets

- **Sphere Packing & High-Explosive Micro-CT Samples** — total 6; splits: test (6); repo https://github.com/sci-visus/MSCEER

## Metrics

- `conductance_ratio` **(primary)** — range: other
  - Ratio of experimentally measured Fisher conductance ($C_f$) to computed PNM conductance ($C_{pnm}$). A value near 1 indicates accurate prediction; the model typically underestimates, yielding ratios of 1.4–3.0.

## Input / output format

**Input**: Micro-CT volumetric image of packed granular material (spheres or explosive crystals) at specified resolution.

**Output**: Computed flow-permeable surface area ($S$) and PNM conductance ($C_{pnm}$) derived from the Morse-Smale complex pore network.

## Scoring recipe

```python
ratio = experimental_Cf / computed_Cpnm
surface_area_ratio = experimental_Sf / computed_S
ranking_match = (rank(experimental_Cf) == rank(computed_Cpnm))
return ratio, surface_area_ratio, ranking_match
```

## Common pitfalls

- Model conservatively underestimates conductance due to resistive network assumptions, so ratios > 1 are expected.
- Micro-CT resolution limits smooth solid/void interfaces, affecting surface area accuracy.
- Dead-end pores may be ignored in the topological decomposition, impacting flow predictions for high-aspect-ratio crystals.

## Evidence (verbatim from paper)

> We evaluate the effectiveness of our PNM by comparison with experimentally measured surface area and volume flow rate for three different sphere packing distributions using the Fisher apparatus. Our computed conductance $C_{pnm}$ underestimates the Fisher measured conductance, $C_f$ by a factor of 2.43-2.66 and $C_{iso}$, conductance computed using the isosurface area is approximately the same as the Fisher measured conductance.

## Citation

```bibtex
@misc{venkat2021topology,
  title={Towards replacing physical testing of granular materials with a Topology-based Model},
  author={Venkat et al. (2021)},
  year={2021},
  note={arXiv:2109.08777}
}
```

- arXiv: 2109.08777