With the rapid development of supercritical fluid, it has been widely applied to various industrial technology fields such as supercritical fluid extraction, thermal and nuclear power generation systems, air conditioning and refrigeration, and solar power generation because of its unique physical and transport properties. At present, some studies have found that the speed of sound sharply changes near the supercritical point and the supercritical region can be divided into “liquid-like” and “gas-like” regions, and the concept of “pseudo-boiling” was proposed. However, the structural characteristics of supercritical fluids in different regions and their evolution laws versus time are rarely available in the literature. Therefore, in this paper, the density fluctuations in the “liquid-like” and “gas-like” regions of a supercritical fluid at different isotherms and densities have been investigated and analyzed by molecular dynamic simulation. The results show that near the critical temperature, the density time series curve has a large fluctuation at the “pseudo-critical” point. As the system temperature and the deviation from the “pseudo-critical” point increase, the density fluctuation gradually decreases. In addition, with the density increase or decrease, the difference of density fluctuation between different isotherms decreases. This can indicate that the degree of density deviation from the “pseudo-critical” point is the main reason for density fluctuations. Moreover, supercritical fluid is a type of a heterogeneous medium with considerable density fluctuation and can be regarded as a mixture of high density (“liquid-like”) and low density (“gas-like”) fluids; therefore, the heterogeneous character declines with temperature, and the deviation from the “pseudo-critical” point increases. In this paper, nonlinear analysis has been based on the time series of density and the autocorrelation function, attractor phase diagram, and recurrence plots have been mainly used to conduct it. The studies confirm that the density fluctuations in the different region are deterministic chaotic or stochastic systems. Based on the time delay obtained from the autocorrelation function, the chaotic and stochastic characteristics of the calculated conditions can be preliminarily interpreted. It is demonstrated that the lower system temperature and closer to the pseudo-critical point, the time series curve has strong predictability. All of the time series only showed two different non-linear characteristics according to the phase diagram of the attractor and the recurrence plots. The result reveals that when the isotherm is near the critical point, a large density range represent chaotic characteristics and as the temperature increases, this range gradually decreases. Moreover, when T =1.075 T c, all the calculation conditions exhibit stochastic systems and chaotic systems no longer exist. Finally, a study on the sample entropy of different calculation conditions was conducted indicating a change trend in which the sample entropy of the chaotic system is less than that of the stochastic system. Additionally, as the deviation from the pseudo-critical point increases, the entropy converges in a small region and the difference between the different isotherms is reduced. At the same time, the effect of the system temperature on the sample entropy gradually decreases, and the system density plays a dominant role. This illustrates that a certain ordered structure in the chaotic system and a stronger disordered structure in the stochastic system exists and proves that this special structure of supercritical fluid exists in a certain range of temperature and pressure. These finding may facilitate to understand the characteristics of supercritical fluids and provide theoretical support for further exploration.
Read full abstract