Prediction of Flow Boiling Curves Based on Artificial Neural Network

WU Jun-mei,SU Guang-hui

DOI: https://doi.org/10.3969/j.issn.1000-6931.2007.03.012

IF: 1.279

2007-01-01

Nuclear Instruments and Methods in Physics Research Section B Beam Interactions with Materials and Atoms

Abstract:The effects of the main system parameters on flow boiling curves were analyzed by using an artificial neural network(ANN) based on the database selected from the 1960s.The input parameters of the ANN are system pressure,mass flow rate,inlet subcooling,wall superheat and steady/transition boiling,and the output parameter is heat flux.The results obtained by the ANN show that the heat flux increases with increasing inlet sub cooling for all heat transfer modes.Mass flow rate has no significant effects on nucleate boiling curves.The transition boiling and film boiling heat fluxes will increase with an increase of mass flow rate.The pressure plays a predominant role and improves heat transfer in whole boiling regions except film boiling.There are slight differences between the steady and the transient boiling curves in all boiling regions except the nucleate one.