Publicou 6 edições por ano
ISSN Imprimir: 1940-2503
ISSN On-line: 1940-2554
Indexed in
APPLICATION OF GENE EXPRESSION PROGRAMMING FOR ESTIMATING TOTAL EMISSIVITY OF H2O−CO2 MIXTURES IN AIR−FUEL COMBUSTION WITHOUT SOOT FORMATION
RESUMO
In the present study, a gene expression programming algorithm has been applied to propose new and accurate correlations for estimating total emissivity of CO2−H2O homogeneous mixtures in air−fuel combustion environment without soot formation at atmospheric condition. The main parameters of the correlations include temperature (T: 300−2500 K), partial pressure of water vapor (pw: 2.0265−20.265 kPa), partial pressure of carbon dioxide (pc: 4.053−20.265 kPa), and mean beam length (L: 0.01−25 m). The RADCAL statistical narrow-band model was used in order to generate 78,000 values of total emissivity to be used as the benchmark data for the correlations. 34,620 total emissivity data points were selected for developing the correlations and 43,380 data points were selected for the optimization and testing the capability of the correlations. All the benchmark data were split into two sub-data sets based on temperature (i.e., the first data set: 300 K ≤ T <1200 K, the second data set: 1200 K ≤ T ≤ 2500 K). For each sub-data set, different correlations have been developed. The average absolute relative deviations of the estimated results from the benchmark data are 3.6% of the low temperature, and 3.9% of the high temperature data sets.