Effect of aggregation on thermal conduction in ternary molten salt-based nanofluids: insights from a multiscale coupled MD–LBM method

Molten salts serve as primary heat transfer and storage media in thermal energy storage systems. Adding nanoparticles to molten salt to create nanofluids is known to significantly improve the thermal conductivity of the molten. However, nanoparticle agglomeration is inevitable and substantially affects the thermal conductivity of molten salts. Moreover, the mechanisms whereby agglomeration influences thermal conductivity remain unclear. This paper presents an innovative multiscale coupling model that combines molecular dynamics (MD) simulations with the lattice Boltzmann method (LBM) to investigate the thermal conductivity of CuO nanoparticles in ternary NaCl–KCl–LiCl molten salt-based nanofluids. Both nonaggregated and aggregated states were considered. After conducting MD simulations at the microscale to examine the thermal contact resistance at the interface between nanoparticles, we employed the LBM to determine the effective thermal conductivity of the nanofluids at the mesoscale. The findings reveal the formation of significant heat flow channels in nanofluids containing nanoparticles. However, an increase in the thermal contact resistance reduces these channels in agglomerated particles, potentially reducing the thermal conductivity compared with that in the nonaggregated nanofluids. In cluster-like structures, fewer nanoparticles are positioned within heat flow channels, in contrast to chain-like arrangements. This reduction limits the enhancement in the thermal conductivity and minimizes variations in the thermal conductivity due to differences in the aggregate particle number and orientation. Furthermore, the thermal conductivity exhibited notable variations with varying agglomerated nanoparticle diameters at identical mass fractions. Both smaller and larger particles can increase the level of contact thermal resistance, ultimately reducing the thermal conductivity.