Numerical analysis of performance uncertainty of heat exchangers operated with nanofluids

Abstract

In this paper, we analyse the performance of two types of heat exchangers with nanofluid as the working fluid in turbulent flow regime (Re≈4,000–180,000). Based on the experimental uncertainty of the thermophysical properties of the nanofluids, we use the Stochastic Collocation Method in combination with a deterministic simulation programme to estimate the expected value and variance of the targeted engineering results. We find that the uncertainty in the thermal conductivity of the nanofluid has the largest impact on the uncertainty in the heat exchanger performance, while the uncertainty in the density can be neglected. The uncertainties in the Nusselt number, friction factor and several figures of merit are smaller than the change in these performance estimators due to a change in nanoparticle concentration. Predictions for heat exchanger performance agree much better with experimental data when used with empirical heat transfer correlations developed specifically for nanofluids than with the general Gnielinski correlation developed for pure fluids.We also perform a correlation analysis of the relationships between heat exchanger performance enhancement and pressure drop to show that they are strongly correlated. We find that the relationship between the concentration of nanoparticles and the Nusselt number is statistically insignificant. The Re−Nurelationship is significant, indicating the importance of flow conditions. The correlation between nanoparticle concentration and friction factor is significant and strong. This result suggests that the optimisation of the thermal-hydrodynamic behaviour should be sought in a parameter other than the nanoparticle volume fraction.

Keywords

• Nanofluid
• Heat exchanger
• Figure of merit
• Stochastic collocation method
• Statistical analysis
• Correlation analysis