Complex CSP for chemistry reduction and analysis

Research output: Contribution to conferencePaperpeer-review


The method of Computational Singular Perturbation (CSP) for the analysis and reduction of complicated chemical mechanisms has been extended to the complex eigensystem. The characteristic time scale for each species is defined by using the time scales of the independent modes weighted by radical pointers, and the time scale of each species normalized by a characteristic time scale of the system is used as a criterion in determining the quasi-steady state species. Furthermore, for oscillatory modes the radical pointer and the importance index of the previous CSP theory are re-defined. Results show that the time scales of chemical species change dramatically and nonmonotonically, and the oscillatory modes appear frequently in large chemical reaction mechanisms. The present method is then employed to generate a four-step and a ten-step reduced mechanism for H2/air and CHVair oxidation, respectively. The validity of these reduced mechanisms are evaluated based on the responses of the Perfectly Stirred Reactors (PSR) and the one-dimensional planar propagating premixed flames. Comparisons between the reduced and detailed chemistries over a wide range of pressure and equivalence ratio show good agreement on the flame speed, temperature, and structure. A software package based on the present algorithm is compiled to generate reduced mechanisms for complex chemical mechanisms. The validity and efficiency of the present algorithm is demonstrated.

Original languageEnglish (US)
StatePublished - Dec 1 2001
Event39th Aerospace Sciences Meeting and Exhibit 2001 - Reno, NV, United States
Duration: Jan 8 2001Jan 11 2001


Other39th Aerospace Sciences Meeting and Exhibit 2001
Country/TerritoryUnited States
CityReno, NV

All Science Journal Classification (ASJC) codes

  • Space and Planetary Science
  • Aerospace Engineering


Dive into the research topics of 'Complex CSP for chemistry reduction and analysis'. Together they form a unique fingerprint.

Cite this