2Applications inT/H ProblemsPARTFlowBehaviorsPercentagesofANNsusedinFlow RegimeIdentificationdifferentT/HproblemsPressureDrop5.1%Void Fraction26.53%30.61%Heat transfer problemsflowregimepressuredropCritical HeatFlux?voidfractionCHF16.33%Onsetof Nucleate BoilingLONBHTCHeattransfercoefficient2.04%2.04%boilingcurve17.35%Boiling Curve11西步交通大学
11 Applications in T/H Problems PART 2 Percentages of ANNs used in ➢ different T/H problems Flow Regime Identification ➢ Pressure Drop ➢ Void Fraction Flow Behaviors ➢ Critical Heat Flux ➢ Onset of Nucleate Boiling ➢ Heat transfer coefficient ➢ Boiling Curve Heat transfer problems
2Applications in T/H ProblemsPARTFlow regimeidentificationInputANN ModelTrainingOutputMean and SDbubbly, slug andvalues,FeatherBPN, SOM....Flow regimeschurnflowsdataA,B,C,..PressuredropTrainingInputANN ModelOutputPressure dropG,q,Tsat, VgMostly BPNPressure dropdataVoid FractionInputANN ModelTrainingOutputVoid fractionP, D, x, Pg/pI, μg/μlBPN,RBFNVoid fractiondataRe, We, Fr....12西步交通大学?
12 PART 2 Applications in T/H Problems ➢ Flow regime identification Input ANN Model Training Output Mean and SD values, Feather A,B,C，. BPN，SOM. bubbly, slug and churn flows data Flow regimes ➢ Pressure drop Input ANN Model Training Output G,q,Tsat,Vg,. Mostly BPN Pressure drop data Pressure drop ➢ Void Fraction Input ANN Model Training Output P, D, x, ρg/ρl , μg/μl , Re, We, Fr. BPN，RBFN Void fraction data Void fraction
2Applications inT/HProblemsPARTCritical heatfluxInputANN ModelTrainingOutputCHF data bankCHFCHF loop-upP G, D, L, D ,X....BPN, GNN....TableHeattransfercoefficientInputTrainingANNModelOutputG, Xm,Xi, X。rMostly BPNHTC dataNu or HTCTsat , D, Tsat ....BoilingCurveInputANN ModelTrainingOutputP, G, Dhi , DTsat steady/boiling regionsBPNHeat fluxtransient statemapindex13天通大学
13 PART 2 Applications in T/H Problems ➢ Critical heat flux Input ANN Model Training Output P, G, D, L, D ,X. BPN，GNN. CHF data bank CHF loop-up Table CHF ➢ Heat transfer coefficient Input ANN Model Training Output G, Xm , Xi , Xo , Tsat , D, Tsat . Mostly BPN HTC data Nu or HTC ➢ Boiling Curve Input ANN Model Training Output P, G, Dhi , DTsat , steady/ transient state index BPN boiling regions map Heat flux
西安交通大学XIANJIAOTONG UNIVERSITY03RelevantStudiesinNuTHel1896
Relevant Studies in NuTHeL