Artificial neural network prediction models for nanofluid properties and their applications with heat exchanger design and rating simulation

Abstract

Energy is arguably driving society toward a prolific future in the current century. Energy demand has been increasing due to the continuous rise in the global population and basic needs. Energy integration is one of the promising ideas for alleviating energy shortages. Energy integration is performed by using the equipment and recirculating the heat back to the system via heat exchange between the heat source and the coolant. Conventional coolants have limited potential due to their low thermal conductivities. Nanoparticles, nanometer-sized solids with relatively high thermal conductivity, have been used to enhance the thermal properties of coolants, and the mixtures are called nanofluids. Nevertheless, the experiment consumes a lot of cost and time; using an Artificial Neural Network (ANN), a non-linear statistical data correlation model builder, is now considered acceptable and appropriate for correlating the complex relationship between inputs and outputs. This study describes a novel combination of ANN and conventional process simulation for heat transfer using nanofluid in a heat exchanger. In addition, the sensitivity of parameters for heat exchanger design and rating were investigated. The ANN development includes 2723 datasets with various nanofluid types and properties. The predictions of the new ANN nanofluid predictive model are closer to reality than other numerical methods, as shown in the simulation. The maximum error from the result was only 4.1%. The work reproduces the conventional simulation for better adaptability and great performance with acceptable error. Finally, the study combines ANN and conventional process simulation methods that effectively examine nanofluid enhancement's performance on heat exchangers. The parameter sensitivity test is discussed based on a plate heat exchanger, including an increase in heat transfer coefficient (around 7%), maintained pressure drop, and performance efficiency coefficient. Show more