In modern compression ignition engines, the dense liquid fuel is directly injected into high pressure and temperature atmosphere, so the spray transitions from subcritical to supercritical conditions. To gain better control of the spray-combustion heat release process, it is important to have a physically accurate description of the spray development process. This work explored the effect of real-fluid thermodynamics in the computational prediction of multiphase flow for two non-ideal situations: the cryogenic nitrogen and non-cryogenic n-dodecane and ammonia sprays. Three real-fluid equations of state (EoS) such as the Soave-Redlich-Kwong (SRK), Peng-Robinson (PR), and Redlich-Kwong-Peng-Robinson (RKPR) coupled with the real-fluid Chung transport model were implemented in OpenFoam to predict the real-fluid thermodynamic properties. Validations against the CoolProp database were conducted. The RKPR EoS demonstrated an overall better predictive performance compared to the SRK and PR EoS. Due to miscalculations of the thermodynamic properties under supercritical conditions, the cases using the ideal-gas EoS predicted the significantly distinct spray features from the cases using real-fluid EoS. For the cryogenic nitrogen spray simulations, cases using various real-fluid EoS yielded similar spray features because of the low injection rate and thus the weak ambient entrainment process. The reduction of ambient pressure promoted the turbulent mixing process for the n-dodecane spray due to the smaller ambient density and resistance. Compared to ammonia, n-dodecane had higher density and viscosity under supercritical conditions, which led to its higher jet velocity and more concentrated spray feature.
