Skip to content

WORKAROUNDS.md

Latest commit

 

History

History
100 lines (72 loc) · 4.62 KB

WORKAROUNDS.md

File metadata and controls

100 lines (72 loc) · 4.62 KB

Workarounds and Known Issues

This document tracks workarounds applied in this repository that circumvent limitations in the upstream EMMO ontology or in the external battery simulation software. These should be addressed in the upstream projects when possible.

TTL Ontology Workarounds

1. Upper Voltage Cut-Off (added locally)

Issue: The upstream EMMO battery ontology does not include a mapping for the upper voltage cut-off parameter. Only the lower voltage cut-off (electrochemistry_534dd59c_904c_45d9_8550_ae9d2eb6bbc9) was defined.

Workaround: Added a local class :modellib_upper_cutoff_voltage in assets/battery-model-lithium-ion.ttl with mappings for:

  • BPX: ['Parameterisation','Cell','Upper voltage cut-off [V]']
  • CIDEMOD: ['simulation','triggers','voltage','upper']
  • BattMo: ['Control','upperCutoffVoltage']

Upstream fix needed: Add UpperVoltageCutOff to the EMMO electrochemistry domain ontology with proper class hierarchy.

2. BattMo.jl uses same key as BattMo.m

Issue: BattMo.jl and BattMo.m use the same JSON input format, so they share the same battmoKey annotation property in the TTL. If the formats diverge in the future, a separate battmoJlKey annotation property would need to be added to the TTL.

Current approach: Both battmo.m and battmo.jl map to the same URI (bmli_e5e86474_8623_48ea_a1cf_502bdb10aa14) in OntologyParser.key_map.

3. Missing BattMo mappings for some BPX parameters

Issue: Several BPX parameters have no BattMo equivalent in the TTL:

  • Cell density (Density [kg.m-3])
  • Cell volume (Volume [m3])
  • Cell specific heat capacity (Specific heat capacity [J.K-1.kg-1])
  • Electrode area (Electrode area [m2])
  • External surface area (External surface area [m2])
  • Electrolyte conductivity activation energy
  • Electrolyte diffusivity activation energy
  • Entropic change coefficients (positive/negative electrode)

Explanation: These are cell-level or thermal properties that BattMo handles differently (e.g., through its own thermal model or geometry configuration). Some would require structural changes to BattMo's JSON format to support directly.

Code Workarounds

4. Function representation conversion

Issue: BPX uses Python string expressions for functional parameters (OCP, conductivity, diffusivity), while BattMo uses structured function objects with functionFormat, argumentList, and either expression or functionName.

Workaround: The ParameterMapper class handles bidirectional conversion:

  • BPX → BattMo: String expressions are wrapped in {"functionFormat": "string expression", ...}
  • BattMo → BPX: Function objects are unwrapped to extract the expression string
  • Named BattMo functions (e.g., computeOCP_Graphite_Torchio) cannot be converted to BPX and result in a comment string: # Named function: ... (requires manual conversion)

Upstream fix needed: BattMo.jl/BattMo.m should support string expression evaluation, or a registry of standard functions with their analytical forms should be maintained.

5. Porosity ↔ Volume Fraction

Issue: BPX uses Porosity directly, while BattMo historically used volumeFraction (where porosity = 1 - volumeFraction). Some BattMo input files use porosity directly.

Workaround: The PreprocessInput class computes porosity = 1 - volumeFraction when processing BattMo input that contains volumeFraction. If volumeFraction is not present (i.e., the input already has porosity), no conversion is performed.

6. TTL syntax error (fixed)

Issue: The original TTL file had emmo::hasMeasurementUnit (double colon) for the BatteryCellSurfaceArea entry, and a duplicated 'Parameterisation' in the BPX key path.

Fix: Corrected to emmo:hasMeasurementUnit and removed the duplicate from the path.

Upstream fix needed: Fix should be applied to the upstream EMMO battery model ontology.

Suggestions for External Projects

BattMo.jl / BattMo.m

  • Consider supporting string expression evaluation for OCP and transport property functions, which would simplify interoperability with BPX/PyBaMM
  • Consider adding cell-level thermal properties (density, specific heat, volume) to the JSON input format for better BPX compatibility

PyBaMM / BPX

  • Consider adding support for tabulated data (lookup tables) as an alternative to string expressions for OCP and transport properties, which is the standard in BattMo

EMMO Battery Ontology

  • Add UpperVoltageCutOff class
  • Add transport efficiency / Bruggeman coefficient mappings
  • Add electrolyte activation energy mappings for BattMo format
  • Add entropic change coefficient mappings for BattMo format