Electronic devices release heat to the atmosphere from the walls, and system miniaturisation results in exponential increase in the heat associated with these high-heat-generating/electronic devices. To analyse this numerically, heat-flux and convective types of boundary condition are considered at the walls, which are the most realistic boundary condition type. In present work, the physical model consists of nanofluid inside an enclosure with one side subjected to constant heat-flux and the other sides exposed to convective boundaries, is solved. The nanofluid is completely confined within the enclosure and flows due to natural convection. Considering the heat-flux varying from 100 W/m2 to 10 kW/m2, three (Ra) Rayleigh numbers are calculated. Results are validated with experimental results also. Results show that increasing Ra and copper nanoparticle concentration results in strengthening the heat transfer and average Nusselt number. Results also show that the thermal boundary layer thickness increases with aspect ratio (AR). Streamline contours show that natural convection strength is higher for low ARs compared to high ARs. For a low Ra (2 × 104), viscosity model is more sensitive than the thermal conductivity model, and for high Ra (2 × 106), the thermal conductivity model is more sensitive than the viscosity model.

Modern automobiles require materials with high thermal properties and less weight for extraordinary performance. Metal matrix composites (MMCs) possess the mechanical properties required for automobile applications. The present work focuses on developing a mathematical model using the finite-difference technique to examine the cumulative and individual influence of silicon carbide (SiC)/aluminum oxide (Al2O3) nanoparticle weight percentage on the MMCs’ time–temperature history curve. The thermophysical properties of MMCs are considered to be isotropic and homogeneous and are calculated using the mixture rule and the Maxwell model. The results reveal that increment in the weight percentage of the nanoparticles (silicon carbide/aluminum oxide) makes the time–temperature history curve steeper. The cumulative effect of the nanoparticles results in a less steep time–temperature history curve compared to the individual nanoparticle, indicating a decreased solidification rate due to low effective thermal conductivity. The two thermal conductivity models show the same time–temperature history curve due to less deviation in the thermal conductivity of MMCs. The mathematical model developed in the present work predicts the time–temperature history curve accurately which matches well with the experimental results.