Purpose of this presentation was to verify if an open source library (Cantera) could be used to calculate rocket engine performance parameters that are comparable to widely used Gordon and McBride CEA NASA software. Reason why I have decided to write my own library was caused by difficulty in integrating CEA with other software. Ultimate objective of my project is to provide a software that automate optimization of rocket engine conceptual design (e.g. selection of propellants, chamber pressure, etc.).
I wrote my own library called ROPSI (ROcket Propulsion SImulation) that re-use Cantera thermodynamic routines to calculate basic thermodynamic properties of combustion gas mixture. ROPSI calculates on its own shifting equilibrium thermodynamic gas mixture properities and all rocket engine performance characteristics (list of these properties is same as for NASA CEA).
Comparison was done in two primary areas:
- Thermodynamic data (e.g. adiabatic combustion temperature, density, enthalpy, velocity of sound, mass fractions, etc.)
- Rocket engine data (e.g. ideal specific impulse, characteristic velocity, coefficient of thurst)
Thermodynamic data was compared assuming HP setting (constant enthalpy and pressure). Rocket engine performance characteristics were compared assuming frozen equilibrium and shifting equilibrium (infinite combustion area).
Three combinations of propellant mixtures have been subject of comparison to NASA software:
- Liquid hydrogen and liquid oxygen
- Liquid methanol and liquid oxygen
- Gaseous methane and gaseous oxygen
Detailed composition, fractions and combustion pressures of the propellant mixtures is provided in below table:
Following error ranking was used (assumed arbitrarily) to determine how relative (absolute) errors spread across pre-defined accuracy levels:
- Excellent (accuracy below 0.005%)
- Good (accuracy range: 0.005% to 0.05%)
- Acceptable (accuracy range: 0.05% to 0.5%)
- Not acceptable (accuracy below 0.5%)
Assuming all samples are within first three criteria, Cantera and ROPSI could be used for rocket propulsion engine simulation/calculations.
Results of the CEA and ROPSI calculations are summarized on below graph, that presents how all samples (882) are spread across various acceptance ranking/criteria:
These results for given propellants combinations, clearly presents that all samples are within pre-defined acceptance criteria. On base of that I will continue to develop ROPSI library that will perform further calculations of more advanced rocket engine performance characteristics.
Hopefully I will be able to publish more details of this comparison – stay tuned on details (e.g. via my Twitter account).