Prediction model of flow boiling heat transfer for R407C inside horizontal smooth tubes based on RBF neural network

Abstract The flow boiling heat transfer inside horizontal smooth tubes is studied for refrigerant mixture R407C and a prediction model is proposed based on RBF neural network. The factors strongly affecting the flow boiling have been assumed as the inputs, such as mass flux ( G ), heat flux ( q ), quality ( x ), saturation temperature ( T sat) and tube inner diameter ( D ). At the same time, the flow boiling heat transfer coefficient ( h ) as the output. The K-means clustering algorithm is applied to design RBF network. In addition, the prediction results is significantly improved compared with the four frequently used conventional correlations. For the network model of heat transfer, the average deviation, absolute average and root-mean-square deviations are -0.9%, 5.5% and 10.9%, respectively. Hence, the simulation results prove that the model based on RBF neural network is feasible to forecast the flow boiling heat transfer coefficient of R407C, and optimization design evaporators used R407C.