A new multicomponent diffusion formulation for the finite-volume method

Daniel N. Pope, George Gogos

Research output: Contribution to journalConference articlepeer-review

Abstract

A new multicomponent formulation, which is appropriate for use with the finite-volume method, has been developed to accurately describe the diffusion velocity. The new formulation is presented and applied to the numerical simulation of n-heptane fuel droplet combustion in a zero-gravity, forced convection environment at 1 atm. Combustion is modeled using finite-rate chemical kinetics and a one-step overall reaction. Results obtained using the complete formulation are compared to the results obtained while assuming (1) thermal diffusion (Soret effect) is negligible and (2) thermal diffusion is negligible and all binary diffusion coefficients are the same. The effect these assumptions have on the results at a fixed Reynolds number (Re =10) is investigated for a low (300 K) and a high (1200 K) ambient temperature. The use of a single binary diffusion coefficient produces results that are significantly different from the results obtained using the complete formulation. These differences include a much lower maximum temperature (700 K lower), a "longer" flame and lower (8-20%) values for the evaporation constant and drag coefficient. Thermal diffusion caused only minor changes (∼1%) in the numerical predictions for the maximum temperature, evaporation constant and drag coefficient.

Original languageEnglish (US)
Article numberIMECE2004-59551
Pages (from-to)87-93
Number of pages7
JournalAmerican Society of Mechanical Engineers, Heat Transfer Division, (Publication) HTD
Volume375
Issue number1
DOIs
StatePublished - 2004
Event2004 ASME International Mechanical Engineering Congress and Exposition, IMECE - Anaheim, CA, United States
Duration: Nov 13 2004Nov 19 2004

ASJC Scopus subject areas

  • Mechanical Engineering
  • Fluid Flow and Transfer Processes

Fingerprint

Dive into the research topics of 'A new multicomponent diffusion formulation for the finite-volume method'. Together they form a unique fingerprint.

Cite this