A computational fluid dynamics numerical simulation was performed on Al₂O₃-synthetic oil nanofluids to study heat transfer in a trough solar receiver. Different nanofluid thermal conductivity models were tried. Calculated Nusselt numbers were compared with results of semiempirical model models. It was found that the thermal conductivity model based on Brownian motion predicted heat transfer characteristics well. The relative motion between nanoparticles and base fluid was found important in enhancing heat transfer. Further volume fraction analysis shows that the nanoparticles improved significantly average heat transfer coefficient, which indicates that nanofluids have great potential in solar collector applications.

Sodium sulfate is a common inorganic salt that causes blockage in supercritical water (SCW) reactors. An investigation of its crystallization kinetics is important for the design of deposition-preventing reactors. Microscopic crystallization process of sodium sulfate in SCW was studied with LAMMPS molecular dynamics simulation. Water molecule was calculated with SPC/E model, and ion-ion and ion-water interactions were calculated with Coulumb and Lennard-Jones combined potential energy functions. It shows that the impact of water on electrostatic shielding effect of ion charges decreases at higher temperatures and lower densities. Increasing temperature and density is advantageous to diffusion of ions, which helps ions to collide and form nuclei. Magnitude of nucleation rate is generally in the order of 10²⁹ cm^-3s^-1 in simulated parameters range of SCW.

A hybrid method is proposed to study microcosmic characteristics of nanofluid mulfiphase flows in a lattice Boltzmann approach. Multicomponent lattice Boltzmann model on fine mesh is used in regions where physical parameters, such as density and velocity change tempestuously, otherwise a single component one on coarse mesh is used. In order to keep continuity of physical information (physical parameters) in overlapping regions, principle of mass and momentum conservation is used. It shows that in the model microcosmic characteristics of nanofluid flow can be obtained. Computational efficiency is improved observably compared with a multicomponent method.