Complex casting parts rely on sand cores that are both high-strength and can be easily decored after casting. Previous works have shown the need to understand the influences on the decoring behavior of inorganically bound sand cores. This work uses black box and explainable machine learning methods to determine the significant influences on the decoring behavior of inorganically bound sand cores based on experimental data. The methods comprise artificial neural networks (ANN), extreme gradient boosting (XGBoost), and SHapley Additive exPlanations (SHAP). The work formulates five hypotheses, for which the available data were split and preprocessed accordingly. The hypotheses were evaluated by comparing the model scores of the various sub-datasets and the overall model performance. One sand-binder system was chosen as a validation system, which was not included in the training. Robust models were successfully trained to predict the decoring behavior for the given sand-binder systems of the test system but only partially for the validation system. Conclusions on which parameters are the main influences on the model behavior were drawn and compared to phenomenological–heuristical models of previous works.

De-coring is an essential process in the casting process chain, determining the quality and cost of production. In this study, a coupled multibody system (MBS) and finite element modeling (FEM) technique is presented to study the mechanical loads during the de-coring process. The removal of cast-in sand cores from the inner regions of the cast part by de-coring or knocking out is a complex process with dynamic loads. Currently, the process relies upon empirical knowledge and tests. Inorganic sand cores pose additional challenges in the success of the de-coring process. Increasing complexity in geometry and stringent environmental regulations compel a predictive process in the earlier stages of design. Predicting the process’ success is challenged by the dynamic non-linearities of the system. The dynamic characteristics and the interaction between hammer and casting were studied here for the first time using an industrial-based test rig, and a novel modeling approach was formulated. The results of the developed model are in good compliance with the experiments. The methodology presented in this study can be used to include a varying number of hammers and loads. The proposed approach presents the possibility to discretize the process and qualitatively assess the process parameters for optimization.