Advanced Deep Learning‐Based 3D Microstructural Characterization of Multiphase Metal Matrix Composites

Abstract

Cast near eutectic Al-Si alloys with addition of transition elements such as Cu, Fe, and Ni are commonly used materials in the aerospace and automotive industries. [1,2] The microstructure of these alloys is characterized by a 3D interconnected network formed by eutectic and primary Si and several Ni-, Fe-, and Cu-rich aluminides embedded in the Al matrix. [3-7] Under prolonged service time at high temperature (up to around 300-350 C), the aluminum matrix is overaged, what deteriorates its strength and creep properties. To improve the strength and creep resistance of these Al-Si alloys, additional ceramic reinforcements such as short fibers and particles can be used. [8-10] It has been shown that the micromechanical behavior of such composites strongly depends on the orientation of the fibers, the spatial distribution of the particles,

Virtual microstructure design for steels using generative adversarial networks

“…The DCGAN, Cycle GAN, and Pix2Pix algorithms created realistic microstructures for engineering and functional materials, which are expected to play a pivotal role in either theoretical or ML modeling for microstructure prediction. The status of the generated virtual micrograph was considerably enhanced in comparison to the existing synthetic GAN‐generated micrographs, 23‐27 and has reached a level wherein generated micrographs are qualitatively indistinguishable from the ground truth. The completeness is, of course, dependent upon the size of the training dataset, the types of microstructure, the deep net architectures, and the choice of appropriate hyper‐parameters, and so on.…”

“…Despite major issues with GAN, which are mode collapse, non-convergence, and training instability, 22 GAN has been one of the most interesting ideas in machine learning (ML) of the past 10 years. 13 Although conventional ML approaches based on supervised learning are well established in the materials research community, [23][24][25][26][27][28][29][30][31][32] GAN algorithms have just begun to be used for the materials research.…”

“…One neural network, called the generator, generates new data instances from randomly distributed sample data, and the other, the discriminator, judges them for authenticity; in other words, the discriminator determines whether each instance of data it examines belongs to the genuine training dataset or not. In contrast to the recent boom for deep learning-based artificial intelligence for use in both the chemistry and materials science research fields, [23][24][25][26][27][28][29][30][31][32] the unsupervised learning-based GAN is yet to be spotlighted in both the fields. Nevertheless, we have seen a certain degree of progress in the GAN utilization for the design of molecules, [34][35][36][37][38][39] and the drug discovery, [40][41][42] the inverse design of materials, 43 the design of crystal structure of inorganic materials.…”

Machine Learning-Assisted Identification of Copolymer Microstructures Based on Microscopic Images

“…Recently, enabled by modern algorithms and high-performance computing, artificial intelligence (AI) has been progressively applied in engineering and science domains . Several works have reported AI technology’s application in recognizing materials’ micrographs. − These studies have achieved various goals, including classification (such as recognition of known materials), generation (such as 3D image reconstruction of materials using generated models or super-resolution image generation), and segmentation (semantic segmentation of micrographs to achieve goals like material-defect detection, reading the number of particles or the percentage of different compositions in a composite material), which significantly increase the speed of materials development.…”

Phase Prediction of High Carbon Pearlitic Steel: An Improved Model Combining Mind Evolutionary Algorithm and Neural Networks

“…As each algorithm has its own scope of application, the selection of an appropriate ML algorithm has become the crucial operation in the construction of the ML system, [ 25 ] which significantly affects the predictive accuracy and generalization ability. So far, there have been several studies [ 28–30 ] aimed at improving the accuracy of model in predicting the targeted features to exhibit promising potential in guiding new steel design. As an analytical model technique, artificial neural network (ANN) models exhibit excellent performance in dealing with various types of nonlinear or complex problems, including prediction and optimization problems, or regression and classification ones.…”