Surrogate model for energy release rate and structure optimization of double-ceramic-layers thermal barrier coatings system

ABSTR A C T The finite element (FE) model is established based on the microstructure of the double-ceramic-layer thermal barrier coating system (DCL-TBCs). Different structural parameters are sampled by Latin hypercube Sampling method to construct corresponding models. The energy release rates of the interfacial crack, which is between lanthanum zirconate (LZ) layer and yttria-partially stabilized zirconia (YSZ) layer of DCL-TBCs, are calculated by numerical simulation during cooling and a dataset is created according to the results. On the basis of the dataset, a surrogate model is constructed to predict the maximum energy release rate during cooling by the extreme random forest algorithm, and the prediction accuracy of the surrogate model is verified by testing dataset. The thicknesses and porosities of the LZ and YSZ layers of DCL-TBCs, and the length of the interfacial crack are set as input variables, and the maximum energy release rate during cooling process is set as output variables of the surrogate model. Using the surrogate model and the particle swarm optimization (PSO) method, the LZ and YSZ layers with different combination of thicknesses is optimized to minimize the energy release rate of the inter-facial crack of DCL-TBCs.