Experimental and neural network prediction of the cyclic stability and light absorption characteristics of supercritical CO2 based CNTs nanofluids

Abstract

Traditional working fluids face challenges including constrained heat transfer efficiency and inadequate thermal stability, which are detrimental to the thermal efficiency of energy conversion systems. To overcome these limitations and achieve a profound breakthrough in the field of energy conversion, it is imperative to develop innovative working fluids. Consequently, this study builds a nanofluid circulation flow experimental platform and prepares a supercritical carbon dioxide/hydroxylated multi-walled carbon nanotubes nanofluid. The study conducts a series of meaningful experimental tests and performance predictions, including stability analysis, deposition characterization research, light absorption potential investigation, and artificial neural network prediction. Remarkably, the artificial neural network model exhibits excellent learning capacity in training, validation, and testing, with root mean square error of only 9.069 W/m2, 8.9996 W/m2, and 13.0143 W/m2, respectively. Test results indicate that the nanofluid reaches a steady state after 4 min, with a fluctuation amplitude of only 2.13 %. Furthermore, the deposition density of the nanofluid on nickel sheet saturates after 4 h, confirming its considerable potential for promotion and application. The light absorption capacity and power decay rate of the nanofluid are enhanced by 606–881 W/m2 and 32.3–46.99 %, compared to other typical substances. The transmittance is reduced by 8.51–22.96 % compared to other nanofluids. The nanofluid proves to be an effective alternative to conventional working fluids, providing a practical reference solution for researchers aiming to improve working fluids in the field of energy conversion.