Assessment of different full-spectrum correlated K-distribution methods in radiative heat transfer of NH₃-H₂ flames

Because of the strong radiative property of NH₃, the emission and absorption parts in radiative source term of ammonia-containing fuel flames are equally important, which is quite different from traditional hydrocarbon-air flames. Therefore, the conclusions of radiative property models applicable to traditional hydrocarbon-air flames cannot be extended to ammonia-containing fuel flames. In this work, the accuracy of five Full-Spectrum Correlated K-distribution (FSCK) methods, i.e., FSCK-1 (Modest et al., 2002), FSCK-2 (Cai et al., 2014), FSCK-3 (Liu et al., 2020), RC-FSK (Solovjov et al., 2018) and ω-ALDF RC-FSK (André F et al., 2022), is evaluated in 24 one-dimensional NH₃-H₂ flames in 6 different pressures and 4 two-dimensional NH₃-H₂ flames in 2 different pressures prior to engineering applications. The results show that owing to the partial invalidation of the correlated spectral assumption, all FSCK methods demonstrate distinct errors in the low-temperature ammonia-rich regions. FSCK-1 requires an appropriate reference temperature to ensure accuracy in high-temperature regions, whereas other FSCKs perform excellently here due to the conservation of emission part of radiative source term. ω-ALDF RC-FSK and FSCK-2 display errors in H₂O-containing low-temperature oxidizer regions, especially for high-pressure flames. To guarantee accuracy, no fewer than 32 Gauss nodes are required for ω-ALDF RC-FSK, whereas 16 nodes suffice for the other models.