Prediction of Equivalent Thermal Conduction Resistance of Printed Circuit Heat Exchangers

Shi Haoning,Chang Hongliang,Ma Ting,Wang Qiuwang

DOI: https://doi.org/10.1007/s11630-022-1675-9

IF: 2.013

2022-01-01

Journal of Thermal Science

Abstract:Printed circuit heat exchangers(PCHEs) have great potential to be employed in the advanced nuclear reactor systems. In this work, the equivalent thermal conduction resistance of PCHE is studied. The influences of thermal convection resistance are analyzed. The results indicate that the equivalent thermal conduction resistance of PCHEs with unequal numbers of hot plates and cold plates are sensitive to the thermal convection resistance of hot side and cold side. Specifically, for case C which has unequal number of hot and cold channels, the maximum value of equivalent thermal conduction resistance can be 1.7-2.4 times the minimum value. The equivalent thermal conduction resistance is underestimated under the isothermal boundary. In addition, the non-uniformity of the lengths of all the heat flux lines determines the influence degree of thermal convection resistance on the equivalent thermal conduction resistance. For further investigation, Latin hypercube sampling method is adopted to generate a large number of design points for each PCHE configuration. Based on the sample data, mathematical correlations and artificial neural network(ANN) for prediction of equivalent thermal conduction resistance for each case are developed. The proposed correlations of equivalent thermal conduction resistance for each case have acceptable accuracy of prediction with a wide range covering general engineering applications. The ANN model can achieve much better prediction accuracy than the proposed correlations thus it is recommended in the cases that the prediction accuracy is considered as the priority need.