Search results

Kriging surrogate model has demonstrated a powerful ability to be applied to a variety of engineering challenges by emulating time-consuming simulations. However, when it comes to problems with high-dimensional input variables, it may be difficult to obtain a model with high accuracy and efficiency due to the curse of dimensionality. To meet this challenge, an improved high-dimensional Kriging modeling method based on maximal information coefficient (MIC) is developed in this work.

The hyperparameter domain is first derived and the dataset of hyperparameter and likelihood function is collected by Latin Hypercube Sampling. MIC values are innovatively calculated from the dataset and used as prior knowledge for optimizing hyperparameters. Then, an auxiliary parameter is introduced to establish the relationship between MIC values and hyperparameters. Next, the hyperparameters are obtained by transforming the optimized auxiliary parameter. Finally, to further improve the modeling accuracy, a novel local optimization step is performed to discover more suitable hyperparameters.

The proposed method is then applied to five representative mathematical functions with dimensions ranging from 20 to 100 and an engineering case with 30 design variables.

The results show that the proposed high-dimensional Kriging modeling method can obtain more accurate results than the other three methods, and it has an acceptable modeling efficiency. Moreover, the proposed method is also suitable for high-dimensional problems with limited sample points.

The effects of infiltrant-related factors during post-processing on mechanical performance are fully considered for three-dimensional printing (3DP) technology. The factors contain infiltrant type, infiltrating means, infiltrating frequency and time interval of infiltrating.

A series of printing experiments are conducted and the parts are processed with different conditions by considering the above mentioned four parameters. Then the mechanical performances of the parts are tested from both macroscopic and microscopic papers. In the macroscopic view, the compressive strength of each printed part is measured by the materials testing machine – Instron 3367. In the microscopic view, scanning electron microscope and energy dispersion spectrum are used to obtain microstructure images and element content results. The pore size distributions of the parts are measured further to illustrate that if the particles are bound tightly by infiltrant. Then, partial least square (PLS) is used to conduct the analysis of the influencing factors, which can solve the small-sample problem well. The regression analysis and the influencing degree of each factor are explored further.

The experimental results show that commercial infiltrant has an outstanding performance than other super glues. The infiltrating action will own higher compressive strength than the brushing action. The higher infiltrating frequency and inconsistent infiltrating time interval will contribute to better mechanical performance. The PLS analysis shows that the most important factor is the infiltrating method. When compare the fitted value with the actual value, it is clear that when the compressive strength is higher, the fitting error will be smaller.

The research will have extensive applicability and practical significance for powder-based additive manufacturing.

The impact of the infiltrating-related post-processing on the performance of 3DP technology is easy to be ignored, which is fully taken into consideration in this paper. Both macroscopic and microscopic methods are conducted to explore, which can better explain the mechanical performance of the parts. Furthermore, as a small-sample method, PLS is used for influencing factors analysis. The variable importance in the projection index can explain the influencing degree of each parameter.