Impact of non-ideal transport modeling on supercritical flow simulation

The simulation of a supercritical fluid flow requires sophisticated models for real gas thermodynamic and non-ideal phenomena. They both are presently addressed through the simulation of a non-reacting and reacting high pressure H₂/O₂ splitter-plate configuration. In particular, the diffusion velocity of species is evaluated through the gradient of chemical potential (${\mathbf{d}}_l^{\text{NI}}=X_l{\left(?{\mu }_l\right)}_T$) expressed with the Peng–Robinson equation of state, or with the classical low-pressure approach ${\mathbf{d}}_l^{\mathrm{I}}=?X_l,$ which only uses the gradient of the lth species molar fraction, X_l. In addition, the high pressure binary diffusion coefficients are estimated by the correction of Kurochkin et al. or with the Takahashi approach. The results for the non-reaction case are consistent with the literature for mean and rms values using ${\mathbf{d}}_l^{\mathrm{I}}$. The use of ${\mathbf{d}}_l^{\text{NI}}$ has a limited impact but the temperature profiles become steeper. In the reactive case, the two approaches lead to a difference of 50 K on the average temperature just downstream of the injector and about 100 K further downstream. A non-ideal transport is then required for the modeling of supercritical flow simulation.