pfc5-ae-energy

Instrument and post-process PFC 5.0 asphalt damage studies using bond-break hits, clustered AE events, macro energy density, source fields, stage summaries, and auditable CSV outputs.

SKILL.md

GitHub

PFC5 AE And Energy

Use this skill when a calibrated bonded/interface PFC5 asphalt model needs damage timing, crack/AE hit statistics, event clustering, or macro energy curves.

Required Inputs

  • PFC2D/PFC3D 5.0, contact/bond failure law and event callback evidence;
  • stress-strain CSV with units/sign;
  • hit CSV with time, position, mode and optional force/tensor fields;
  • clustering thresholds and loading-stage definitions;
  • elastic modulus used by macro energy approximation.

Workflow

  1. Decide Level 1 hits, Level 2 clustered events, or a separately validated tensor route.
  2. Prove PFC5 callback activation with a counter/history.
  3. If native PFC energies are required, enable set energy mechanical on before the measured interval and inventory the partitions actually exposed by every active contact model. Export raw hits and stress-strain data without GUI-only transformations.
  4. Cluster hits with scripts/cluster_ae_events.py when event catalogs are required.
  5. Compute macro input/elastic/dissipated energy with scripts/ae_energy_metrics.py.
  6. Tie events and figures to saved stages and preserve raw IDs.
  7. Validate thresholds and mechanisms against experimental observations where available.

Read ae-contract.md before choosing classifications.

Working Rules

  • One bond break is a hit, not automatically one laboratory AE event.
  • Macro stress-strain energy is not direct contact-energy release.
  • PFC5 mechanical energy tracking is off by default, and the Burger model supplies no own energy partition. Missing native energy cannot be reconstructed from a later activation or silently replaced by the macro approximation.
  • Do not claim moment-tensor mechanisms without the tensor construction and units.
  • Keep scalar moment symbols distinct from tensor component names.
  • Classification denominator and threshold must be defined together and used consistently.

Output Contract

Return raw hit/event schemas, clustering parameters, energy method/units, stage summaries, figures/tables, validation caveats and runtime status.

Local Contents

  • references/ae-contract.md — hit/event/energy definitions.
  • scripts/cluster_ae_events.py — deterministic time-space clustering.
  • scripts/ae_energy_metrics.py — macro energy-density computation.
  • agents/openai.yaml — interface metadata.