CARM Computational Workbench Environment

As part of a National Science Foundation SBIR project, REI has developed a computational workbench environment which allows for the evaluation and optimization of reduced chemical kinetic mechanisms. The workbench is tightly integrated with the CARM software to provide seamless generation and testing of reduced chemical kinetic mechanisms.

Workbench features:

  • Provides capabilities for rapid setup, execution, storage and retrieval of results
  • Combines state-of-the-art techniques in automated chemical kinetic mechanism reduction for combustion systems with recently developed, advanced tools for creating problem solving environments
  • Allows the range of validity of reduced mechanisms to be thoroughly and rigorously characterized
  • Automates the comparison of reduced and detailed chemistry over a wide range of conditions using advanced visualization tools
  • Automatically optimizes reduced mechanisms to the users' specification
  • Produces reduced mechanisms that seamlessly integrate into a variety of CFD codes
  • Provides for a significant reduction in human effort

Genetic algorithm results for methane/air combustion. This figure shows decrease in the CO mole fraction error compared to detailed chemistry The genetic algorithm mimics biological evolution by selecting the “fittest” members of a “population” of designs to pass on traits to successive generations.

 

Pseudo-gradient results for auto-ignition time of a stoichiometric heptane-air mixture at 1 atm. In the example above, the error was reduced nearly a factor of 100 from the starting mechanism in less than 10 iterations.