In this talk, we present nonequilibrium Molecular Dynamics (MD) simulation results on thin-film evaporation in a nano-pump driven by phase change, using liquid argon confined between parallel platinum plates. Our simulation captures a self-regulating and self-sustaining net flow in a statistically stable steady-state, featuring a thin-film evaporation region near the heater and a one-dimensional condensing interface at the condenser side. We first explore the dynamics of evaporating menisci for different channel heights and provide insights into the underlying flow physics using velocity vectors and temperature contours. Notably, our MD simulations reveal evaporation from the adsorbed layer—a region traditionally considered nonevaporating. This discovery offers a potential explanation for the deviations from theoretical maximum evaporation rates observed in recent experiments. We also compare the MD findings with predictions from continuum-based thin-film evaporation models. Additionally, we examine the temperature profiles and the resulting temperature jumps across the interfacial region under varying heat flux conditions, and investigate their relationship with the energy dynamics of atoms crossing the liquid-vapor interface. Finally, using Lagrangian particle tracking techniques, we assess the validity of the Hertz-Knudsen-Schrage relations and extract mass accommodation coefficients for the steady-state condensing surface. This comprehensive analysis provides new insights into nanoscale phase change dynamics and has implications for enhancing evaporation models.

The study of small to micro-scale heat exchangers and related thermal management systems has emerged as a main research and industrial area due to the ever increasing need to dissipate high heat fluxes from modern electronic devices. The use of air cooling and later choices like the flow of a single-phase liquid have now both reached their limits. As a consequence, thermal engineers are now turning their attention to flow boiling using mostly dielectric fluids as one of the most viable solutions to meet the increasing thermal loads. Other applications include small scale refrigerators, fuel cells, batteries, solar photovoltaic panels and radar systems. In these cases, the transfer of the generated heat will enable efficient operation, while maintaining the operating temperature within design levels. Additional advantages of flow boiling include the small variation in the temperature of the electronic device along the flow path, reducing thermal stresses, enabling system longevity and, in certain cases, avoiding catastrophic failures. The prevailing high heat transfer coefficients allow smaller flow rates and hence smaller pumps and lower power consumption by the thermal management system. 

The presentation will cover extensive research in flow boiling in small to micro-scale tubes and heat exchangers with rectangular micro-channels as well as more recent research in two additional geometries, namely: flow boiling in a simple microgap (single wide channel) and in a micro-scale heat exchanger with micro-pin fins. Fundamental and challenging issues that are currently being investigated in order to facilitate adoption of these small to micro-scale evaporators include the definition of the macro to micro scale dimensions, the prevailing flow regimes and the effect of mass flux, heat flux, channel aspect ratio and length plus material and surface characteristics on pressure drop, heat transfer rates and critical heat flux. Flow visualisation experiments were performed with a high-speed high-resolution camera to record the prevailing flow patterns.  Pressure drop measurements and finally heat transfer coefficients were measured. The results with the microgap heat exchanger formed the lower benchmark and demonstrated lower pressure drop but also lower possible maximum heat transfer rates than the micro-channel heat exchanger. On the contrary the micro-pin fin heat exchanger can be used for applications requiring higher heat flux values than those possible with the micro-channel heat exchanger.  Generally accepted correlations predicting flow patterns and subsequently, pressure drop and heat transfer coefficients that could be used in design remain a challenge. In the case of flow boiling in micro tubes and channels, and based on a significant number of data points, we correlated our data, finally proposing relationships predicting the transition boundaries of the reported flow regimes, pressure drop and heat transfer rates. These can be used with confidence in the design of thermal management systems. The critical heat flux and its dependence on the geometry and mass flux was recorded for now as the upper limit for operational use of these heat exchangers, while design correlations will be recommended.

With the widespread application of modern electrical equipment in harsh low-temperature environments, the issue of icing has increasingly become a critical factor that threatens the safe and stable operation of such equipment. The initial icing process plays a decisive role in the evolution of the entire icing process, and research on it will promote the development of anti-icing technology. In general, the initial icing process is typically triggered by droplets impact on the cold surface under the influence of an electric field, which involves complex multiphase flow, heat transfer and solidification phenomena. In this talk, the numerical and experimental studies on the impact dynamics and solidification behavior of droplet impact upon subcooled wall in an electric field will be reported. To understand the fundamental dynamic behaviors, research is first conducted on the charging characteristics of droplet under an electric field. On this basis, visual experiments and numerical simulations arecombined to explore the dynamic behaviors of droplet impacting a wall in an electric field, and the influences of wettability and structural characteristics of the wall are discussed. Thereafter, facing to the problem of droplet impacting subcooled wall in an electric field, the intricate microscopic mechanisms governing the electric field-mediated heterogeneous nucleation of supercooled water on surfaces subject to electric fields are firstly unraveled by molecular dynamics simulations. Furthermore, a numerical simulation model that couples multiphase flow, charge movement, heat and mass transfer as well as phase change was developed to explore the impact dynamics and solidification behavior of droplet impacting subcooled wall in an electric field. Especially, the solidification process of droplet impact on superhydrophobic surfaces with microstructures under an electric field are systematically analyzed, revealing the failure mechanisms of hydrophobic surfaces under low-temperature conditions. Based on the above research, optimization methods were proposed to effectively reduce liquid-solid contact time which greatly enhanced anti-icing performance on superhydrophobic surfaces. This study not only deepens the understanding of droplet impact and phase change behavior under an electric field but also provides valuable theoretical and engineering guidance for the optimization of anti-icing coatings on electrical equipment. With broad application prospects, this research contributes to improving the long-term operational reliability of power systems in harsh climatic conditions.

Materials Informatics (MI) is a novel approach that aims to accelerate materials development by bridging materials science and data science. At the core of this approach is the feature to identify and utilize correlations within data, not just relying on the principles of physics and chemistry alone. This enables us to fill gaps between available data and composition or structure of high performance materials, leading to the identification of "optimal" materials. By adopting a data-driven approach alongside traditional insight-based methods, more efficient materials development becomes achievable. In fact, use of MI for thermal science/engineering started later than other fields but now there are a growing number of reports, and it turned out that Thermal Energy Engineering and MI are quite compatible. We utilize MI to develop materials with optimal thermal properties such as thermal conductivity, thermoelectricity, and thermal radiation. Additionally, we are working on integrating robotics and machine learning to automate materials experiments in the real world. Thus far we have realized the design and fabrication of optimized phononic and photonic nanostructures for thermal transport, which have rise to not only high-performance materials but also insights into new (or complex) physical phenomena. 

This talk will cover an introduction of MI, provide details on current MI approaches in nanoscale thermal processes, and offer views on how to expand into other fields in thermal engineering.