Constructing porosity database for Al−Si alloy castings through 3D cellular automata model and machine learning

A coupled three-dimensional cellular automata (CA) model has been used to predict the hydrogen porosity in an Al−Si alloy as a function of thermal boundary conditions. By quantifying the porosity distribution from simulations, a porosity defect database was established, representing a cooling rate ranging from 0.25 to 50 °C/s at an initial hydrogen content of 3.0×10⁻³ mL/g. Based on the database, four machine learning algorithms including support vector machine (SVM), random forest (RF), K-nearest neighbors (KNN), and gradient boosting machine (GBM) were trained and compared for each porosity characteristic to identify the optimal model. For the prediction of porosity percentage, the determination coefficient (R²) and the root mean square error (RMSE) on the test set reached 0.95 and 0.042, respectively. The predicted porosity distribution agreed well with experiments, indicating that the model can be used to map the porosity size in large casting components.