Cantera Skill (K-Dense)

Overview

Cantera Skill (K-Dense) is an instruction-based resource with a bundled calculation helper in the scientific-agent-skills collection. It guides closed, adiabatic, homogeneous ideal-gas ignition calculations using Cantera, rather than serving as an autonomous agent or replacing the upstream simulation package. Its scope covers constant-volume and constant-pressure reactors, temperature histories, and mechanism-specific ignition-delay comparisons.

Inputs include a kinetic mechanism, reactor constraint, initial temperature and pressure, mole composition, simulation horizon, output spacing, and numerical controls. Custom mechanisms must be available as Cantera YAML files. The workflow asks users to preserve mechanism provenance and verify the phase and species before running calculations. The helper produces a JSON report, four CSV histories, and a serialized mechanism snapshot; these capture conditions, hashes, normalized composition, delay estimates, conservation diagnostics, and thermodynamic temperature bounds.

Ignition delay is defined as the time of the global maximum temperature derivative on a uniform output grid. A delay is withheld when heating is insufficient or the maximum is too close to a time boundary. Independent runs vary output spacing, solver settings, and horizon to assess numerical resolution. Conservation checks address mass, elements, and the energy quantity appropriate to the reactor constraint.

These diagnostics establish numerical consistency, not experimental agreement or kinetic-mechanism validity. The helper is not a flame solver or general reactor-network builder, and it does not validate heat loss, real-gas effects, surfaces, flow devices, or multistage experimental ignition definitions. Mechanism selection and comparison with a measured ignition observable remain scientific evaluation tasks.

Key Features

• Bundled Python helper for closed, adiabatic ideal-gas ignition under constant-volume or constant-pressure constraints.
• Temperature-derivative ignition-delay definition with minimum-heating and boundary checks; unresolved cases receive a null delay and explanatory status.
• Four independent reactor runs assess sensitivity to output spacing, solver tolerances, internal time-step limits, and simulation horizon.
• Conservation diagnostics cover mass, elemental composition, species mass-fraction consistency, and total internal energy or enthalpy.
• JSON reporting and CSV histories preserve calculation settings, normalized composition, requested species, numerical comparisons, and thermodynamic-range checks.
• Mechanism provenance includes original-file and saved-snapshot hashes, with documented limitations for imported dependencies and snapshot reproducibility.

Use Cases

• Suggested evaluation: compare ignition delays from candidate mechanisms at matched conditions while checking each mechanism's validated range and using the same delay definition.
• Suggested evaluation: investigate whether a computed temperature-rise peak remains stable when output spacing, solver controls, and simulation horizon change.
• Suggested evaluation: prepare auditable homogeneous-reactor calculation records containing conditions, histories, mechanism artifacts, and conservation diagnostics.
• Suggested evaluation: examine whether a temperature-based delay is suitable for comparison with a particular experiment, especially where species-based or multistage definitions differ.

How to Use

1. Read the Cantera Skill instructions and confirm that a closed, adiabatic, homogeneous ideal-gas model fits the intended study.
2. Select a mechanism with an appropriate validated condition range. Preserve its source, citation, modifications, and dependencies; verify the ideal-gas phase and all required species.
3. Follow the linked configuration example and execution procedure in the Skill file. Set the reactor constraint, temperature in K, pressure in Pa, time in seconds, and mole amounts—not mass fractions. The example is a numerical starting point, not universal mechanism guidance.
4. Run the bundled helper in the documented Python environment, using a fresh output directory. Inspect the JSON report and all four CSV histories; consult the upstream reactor equations when interpreting the constraint.
5. Evaluate delay status, refinement changes, conservation, and thermodynamic-range checks separately. Extend or refine the calculation if the peak is unresolved.
6. Report the mechanism provenance, conditions, delay definition, output spacing, and numerical limitations alongside results. Retain original mechanism dependencies in addition to the saved snapshot.