Molecular Dynamics Simulation on Evaporation of a Suspending Difluoromethane Nanodroplet
Xinghui Wu,Zhen Yang,Yuanyuan Duan
DOI: https://doi.org/10.1016/j.ijheatmasstransfer.2020.120024
IF: 5.2
2020-01-01
International Journal of Heat and Mass Transfer
Abstract:Nanodroplet evaporation is a basic process widely existing in nature and industrial applications. Difluoromethane (CH2F2, also called "R32") has attracted more and more attention in refrigeration, heat pump, Organic Rankine Cycle and other fields because of its excellent thermophysical characteristics. In this work, the evaporation process of R32 nanodroplet in large space is studied by means of molecular dynamics simulation. This work focuses on the dynamic evaporation characteristics of droplets under different conditions (droplet size, initial temperature, and ambient temperature) in the evaporation process, and quantitatively analyzes the effects of different parameters on the evaporation rate. The simulation results were compared with the results calculated from the diffusion-based model and the kinetic model. Based on the dynamic law of evaporation process, such as density, temperature field and velocity distribution, the calculation accuracy of each model is evaluated, the reasons for the deviation between model prediction and simulation results are also discussed. The results show that the increase of ambient temperature and initial saturation temperature can lead to a significant increase of droplet evaporation rate, and shorten the pre-heating time needed for droplet to reach quasi-steady evaporation stage. The increase of droplet size also increases pre-heating time, but has no significant effect on evaporation rate. For the evaporation scenario studied in this paper (Knudsen number (Kn) ~ 1), the existing modification methods of the kinetic model can enhance the prediction accuracy of the evaporation rate of R32 droplets in the quasi-steady evaporation stage. The prediction deviation of diffusion-based model applied directly is large, the accuracy of prediction can be significantly improved by correcting the thermal conductivity considering the scale effect.