XGBoost algorithm for predicting heat transfer coefficient of saturated flow boiling in mini/micro-channels

•A universal consolidated database of 11,470 pre-dryout data points was constructed from 41 sources covering 23 working fluids and wide operating ranges.•The XGBoost algorithm was applied to predict saturated flow boiling heat transfer coefficients in mini/micro-channels.•PFI and SHAP analyses were...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of heat and mass transfer Jg. 256; S. 128095
Hauptverfasser: Noh, Hyeonseok, Kim, Jihyeok, Lee, Seunghyun, Kim, Sung-Min, Mudawar, Issam
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Ltd 01.03.2026
Schlagworte:
Extreme gradient boosting
Heat transfer coefficient
Machine learning
Mini/micro-channels
Saturated flow boiling
XGBoost
Mini/micro-channels
Extreme gradient boosting
Heat transfer coefficient
XGBoost
Saturated flow boiling
Machine learning
ISSN:0017-9310
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•A universal consolidated database of 11,470 pre-dryout data points was constructed from 41 sources covering 23 working fluids and wide operating ranges.•The XGBoost algorithm was applied to predict saturated flow boiling heat transfer coefficients in mini/micro-channels.•PFI and SHAP analyses were integrated to enhance the physical interpretability of the data-driven model and to identify key parameters governing boiling heat transfer while mitigating overfitting.•Hyperparameter tuning with Optuna reduced the model’s MAE to 7.18 %, outperforming existing correlations and other machine learning models, while maintaining strong universal predictive capability on unseen data.•The proposed model provides an efficient framework for thermal system design and optimization, enabling reliable prediction of flow boiling heat transfer under diverse operating conditions. Accurate prediction of the heat transfer coefficient in saturated flow boiling within mini/micro-channels is the most critical factor in designing thermal systems for high-heat-flux devices. This study proposes a machine learning technique to predict the heat transfer coefficient of saturated flow boiling using the XGBoost (eXtreme Gradient Boosting) algorithm. The database used in this study consists of 11,470 pre-dryout data points, obtained by removing 1878 post-dryout data points from a total of 13,348 data points collected from 41 sources, employing an XGBoost incipience dryout predicting model. The dataset includes 23 working fluids, hydraulic diameters ranging from 0.19 mm to 6.50 mm, mass flow rates from 19.45 kg/m²s to 1608 kg/m²s, and saturation temperatures from -40 °C to 201.37 °C. The permutation feature importance (PFI) and SHapley Additive exPlanations (SHAP) values were used for feature selection, while Optuna was used for hyperparameter tuning. A total of seven training features—Prf, xdi, Pred, Frfo, Bo, Prg, and Frtp—were selected and used to develop the model. The model achieved a mean absolute error (MAE) of 7.18 %, demonstrating superior predictive performance compared to existing empirical correlations and other machine learning algorithms. This result confirms that XGBoost is an effective and reliable algorithm for predicting the heat transfer coefficient of saturated flow boiling in mini/micro-channels.
ISSN:0017-9310
DOI:10.1016/j.ijheatmasstransfer.2025.128095