Accuracy of narrow-band and global models for radiative transfer in H 2O, CO 2, and H 2OCO 2 mixtures at high temperature

Abstract

The accuracy of several narrow-band (SNB, CK, CKFG) and global (WSGG, SLW, ADF, ADFFG) gas infrared radiative property models applied to radiative transfer in a planar geometry with different types of temperature profiles is studied. The considered gaseous mixtures are H 2ON 2, CO 2N 2 and H 2OCO 2N 2. Reference solutions are provided by line-by-line (LBL) calculations. All model parameters are based on the same spectroscopic data bases so that only the intrinsic accuracy of each model is tested. All narrow-band models lead in most cases to accurate results, but errors induced by the transmissivity-based models SNB and CKFG increase with wall reflectivity if the reflected radiation is assumed spectrally uncorrelated with gaseous transmissivity. Global models are less time consuming than narrow-band models but are generally less accurate and limited to media with gray boundaries and/or participating particles. The WSGG model leads in many cases to very important errors. The relative accuracy of the SLW and ADF models is typically about 10–20% but care must be taken in the choice of the reference temperature. The ADFFG model is the most accurate global model but requires greater computing times than the ADF and SLW models. For long range sensing of hot gases, only the fictitious-gas based models CKFG and ADFFG lead to accurate results. In the case of mixtures containing H 2O and CO 2, the spectral uncorrelation assumption is accurate for narrow-band models and its implementation results only in greater computing times for the CK model. On the contrary, this assumption is not generally accurate for the whole spectra and specific parameters must be generated from the joint distribution function of the absorption coefficients in the case of global models.