Guidance on using two-phase models for predicting the performance of a nanofluid minichannel heat sink

This paper exposes the root cause by which two-phase models overestimate the heat transfer rate compared to single-phase models in a nanofluid minichannel. It solves this problem by representing nanofluids as Newtonian nanofluid bubbles (second phase) dispersed in water (first phase). The key aspect is the nanoparticle volume fraction Φ in the second phase less than 100 %. This representation differs to the use of water as the first phase and solid nanoparticles as the second phase, at Φ = 100 %. This paper applies the Eulerian, mixture, and the VoF models with 1 % volume of Al₂O₃ nanoparticles in water. At the second-phase nanoparticle volume fraction of Φ = 100 %, all three models fail to provide accurate predictions. Improved results from all three models are obtained at second phase nanoparticle volume fractions lower than Φ = 10 % with good data collapse among the three models for Nusselt number and friction factor.