Al-Si alloys are widely used in automotive casting components while microporosity has always been a detrimental defect that leads to property degradation. In this study, a coupled three-dimensional cellular automata (CA) model has been used to predict the hydrogen porosity as functions of cooling rate and initial hydrogen concentration. By quantifying the pore characteristics, it has been found that the average equivalent pore diameter decreases from 40.43 to 23.98 µm and the pore number density increases from 10.3 to 26.6 mm^–3 as the cooling rate changes from 2.6 to 19.4°C/s at the initial hydrogen concentration of 0.25 mL/100 g. It is also notable that the pore size increases as the initial hydrogen concentration changes from 0.15 to 0.25 mL/100 g while the pore number remains stable. In addition, the linear regression between secondary dendrite arm spacing and the equivalent pore diameter has been studied for the first time, matching well with experiments. This work exhibits the application of CA model in future process optimization and robust condition design for advanced automotive parts made of Al-Si alloys.

An active area of MGI (Materials Genome Initiative)/MGE (Materials Genome Engineering) is to accelerate the development of new materials by means of active learning and “digital trial-error” using a prediction model of material property. Machine learning methods have widely been employed for predicting crystalline materials properties with crystal graph neural networks (CGNN). The prediction accuracy of the state-of-the-art (SOTA) CGNN models based on big models and big data is generally higher. However, for the development of some classes of materials, the datasets obtained by experiments are usually lacking due to costly experiments and measurement costs. The lack of datasets will impact the accuracy of CGNN models and may result in overfitting during training models. This paper proposes a simplified crystal graph convolutional neural network (S-CGCNN) which possesses higher prediction accuracy while reducing the vast amount of train datasets and computation costs. The S-CGCNN model has successfully predicted properties of crystalline materials, such as piezoelectric materials and dielectric materials, and increased the prediction accuracy up to 12%–20% than existing SOTA CGNN models. Furthermore, the distribution map between properties and compositions of materials has been built to screen the latent space of candidate materials efficiently by principal component analysis.

The Scheil–Gulliver equation is essential for assessing solid fractions during alloy solidification in materials science. Despite the prevalent use of the Calculation of Phase Diagrams (CALPHAD) method, its computational intensity and time are limiting the simulation efficiency. Recently, Artificial Intelligence has emerged as a potent tool in materials science, offering robust and reliable predictive modeling capabilities. This study introduces an ensemble-based method that has the potential to enhance the prediction of the partitioning coefficient (k) in the Scheil equation by inputting various alloy compositions. The findings demonstrate that this approach can predict the temperature and solid fraction at the eutectic temperature with an accuracy exceeding 90%, while the accuracy for k prediction surpasses 70%. Additionally, a case study on a commercial alloy revealed that the model’s predictions are within a 5°C deviation from experimental results, and the predicted solid fraction at the eutectic temperature is within a 15% difference of the values obtained from the CALPHAD model.

Conventional trial-and-error method is usually time-consuming and expensive for multi-objective optimization of Mg alloys. Although machine learning exhibits great potential to accelerate related research studies, machine learning prediction of properties of Mg alloys is often a prediction of a single target at a time. To address this, this paper integrates non-dominated sorting genetic algorithm III multi-objective optimization algorithm with light gradient boosting machine algorithm to simultaneously optimize yield strength, ultimate tensile strength, and elongation of Mg alloys. This is the first time that simultaneous machine learning optimization of these three objectives has been achieved for Mg alloys.

Additive Manufacturing (AM) is revolutionizing aerospace, transportation, and biomedical sectors with its potential to create complex geometries. However, the metallic materials currently used in AM are not intended for high-energy beam processes, suggesting performance improvement. The development of materials for AM still faces challenge because of the inefficient trial-and-error conventional methods. This review examines the challenges and current state of materials including aluminum alloys, titanium alloys, superalloys, and high-entropy alloys (HEA) in AM, and summarizes the high-throughput methods in alloy development for AM. In addition, the advantages of high-throughput preparation technology in improving the properties and optimizing the microstructure mechanism of major additive manufacturing alloys are described. This article concludes by emphasizing the importance of high-throughput techniques in pushing the boundaries of AM materials development, pointing toward a future of more effective and innovative material solutions.

Future-oriented Science & Technology (S&T) Strategies trigger the innovative developments of advanced materials, providing an envision to the significant progress of leading-/cutting-edge science, engineering, and technologies for the next few decades. Motivated by Made in China 2025 and New Material Power Strategy by 2035, several key viewpoints about automated research workflows for accelerated discovery and smart manufacturing of advanced materials in terms of AI for Science and main respective of big data, database, standards, and ecosystems are discussed. Referring to classical toolkits at various spatial and temporal scales, AI-based toolkits and AI-enabled computations for material design are compared, highlighting the dominant role of the AI agent paradigm. Our recent developed ProME platform together with its functions is introduced briefly. A case study of AI agent assistant welding is presented, which is consisted of the large language model, auto-coding via AI agent, image processing, image mosaic, and machine learning for welding defect detection. Finally, more duties are called to educate the next generation workforce with creative minds and skills. It is believed that the transformation of knowledge-enabled data-driven integrated computational material engineering era to AI⁺ era promotes the transformation of smart design and manufacturing paradigm from “designing the materials” to “designing with materials.”

Bipolar electrochemistry allows testing and analysing the crevice corrosion, pitting corrosion, passivation, general corrosion, and cathodic deposition reactions on one sample after a single experiment. A novel two-dimensional bipolar electrochemistry setup is designed using two orthogonal feeder electrode arrangements, allowing corrosion screening tests across a far wider potential range with a smooth potential gradient to be assessed. This two-dimensional bipolar electrochemistry setup was applied here to simultaneously measure for the simultaneous measurement of the nucleation and propagation of pitting and crevice corrosion under a broad range of applied potential on type 420 stainless steel, which has a very short localised corrosion induction time. It reduces the error from corrosion induction to corrosion competition, and all pits and crevice corrosion have no lacy cover. Results show crevice corrosion can gain current density and easier to support its nucleation and propagation at different potential regions more easily than pitting corrosion.

Machine-learning interatomic potentials have revolutionized materials modeling at the atomic scale. Thanks to these, it is now indeed possible to perform simulations of ab initio quality over very large time and length scales. More recently, various universal machine-learning models have been proposed as an out-of-box approach avoiding the need to train and validate specific potentials for each particular material of interest. In this paper, we review and evaluate four different universal machine-learning interatomic potentials (uMLIPs), all based on graph neural network architectures which have demonstrated transferability from one chemical system to another. The evaluation procedure relies on data both from a recent verification study of density-functional-theory implementations and from the Materials Project. Through this comprehensive evaluation, we aim to provide guidance to materials scientists in selecting suitable models for their specific research problems, offer recommendations for model selection and optimization, and stimulate discussion on potential areas for improvement in current machine-learning methodologies in materials science.

Cu_xO with flower-like hierarchical structures has attracted significant research interest due to its intriguing morphologies and unique properties. The conventional methods for synthesizing such complex structures are costly and require rigorous experimental conditions. Recently, the X-ray irradiation has emerged as a promising method for the rapid fabrication of precisely controlled Cu_xO shapes in large areas under environmentally friendly conditions. Nevertheless, the morphological regulation of the X-ray-induced synthesis of the Cu_xO is a multi-parameter optimization task. Therefore, it is essential to quantitatively reveal the interplay between these parameters and the resulting morphology. In this work, we employed a high-throughput experimental data-driven approach to investigate the kinetics of X-ray-induced reactions and the impact of key factors, including sputtering power, film thickness, and annealing of precursor Cu thin films on the morphologies of Cu_xO. For the first time, the flower-like Cu_xO nanostructures were synthesized using X-ray radiation at ambient condition. This research proposes an eco-friendly and cost-effective strategy for producing Cu_xO with customizable morphologies. Furthermore, it enhances comprehension of the underlying mechanisms of X-ray-induced morphological modification, which is essential for optimizing the synthesis process and expanding the potential applications of flower-like structures.

Metal-organic frameworks (MOFs), renowned for structural diversity and design flexibility , exhibit potential in catalysis. However, the pursuit of higher catalytic activity through defects often compromises stability, requiring a delicate balance. Traditional trial-and-error method for optimizing synthesis parameters within the complex chemical space is inefficient. Herein, taking the typical MOF UiO-66(Ce) as an illustrative example, a closed loop workflow is built, which integrates machine learning (ML)-assissted prediction, multi-objective optimization (MOO) and experimental preparation to synergistically optimize the defect content and thermal stability of UiO-66(Ce) for efficient hydrogenation of dicyclopentadiene (DCPD). An automatic data extraction program ensures data accuracy, establishing a highquality database. ML is employed to explore the intricate synthesis-structureproperty correlations, enabling precise delineation of pure-phase subspace and accurate predictions of properties. After two iterations, MOO model identifies optimal protocols for high defect content (>40%) and thermal stability (>300°C). The optimized UiO-66(Ce) exhibits superior catalytic performance in hydrogenation of DCPD, validating the precision and reliability of our methodology. This ML-assisted approach offers a valuable paradigm for solving the trade-off riddle in materials field.