Article
Materials Science, Multidisciplinary
Huimin Chen, Zhongwen Shang, Wencong Lu, Minjie Li, Fuping Tan
Summary: A stepwise design strategy based on machine learning was developed to predict the melting points of low-melting alloys, allowing for the design of potential alloys with smaller estimation errors. Candidates with melting points suitable for solder applications and low-cost alternatives were identified, demonstrating the method's potential for exploring other high-performance functional materials.
ADVANCED ENGINEERING MATERIALS
(2021)
Article
Materials Science, Multidisciplinary
Xin Li, Guangcun Shan, Shujie Pang, Chan-Hung Shek
Summary: A multi-stage optimization strategy based on machine learning (ML) was proposed to accelerate the rational design of magnetic high-entropy metallic glasses (HE-MGs), and new HE-MGs with balanced magnetic properties and entropy were successfully developed.
APPLIED MATERIALS TODAY
(2023)
Article
Materials Science, Multidisciplinary
Jie Yin, Qian Lei, Xiang Li, Xiaoyan Zhang, Xiangpeng Meng, Yanbin Jiang, Liang Tian, Shuang Zhou, Zhou Li
Summary: Machine learning-aided alloy design has attracted attention in the materials science community. The prediction accuracy of general machine learning models is limited, but the artificial neural network (ANN) model can improve it. However, ANN suffers from the curse of dimensionality issue. In this study, a novel alloy design strategy using a GRU deep learning model, orthogonal experimental design, and data augmentation was developed to reduce the amount of training data and improve prediction accuracy.
Article
Nanoscience & Nanotechnology
Xiaochen Li, Mingjie Zheng, Xinyi Yang, Pinghan Chen, Wenyi Ding
Summary: In this study, forward and reverse models were established to design the compositions and heat treatment parameters for RAFM steels with the targeted tensile properties. The validity of the intelligent design model was verified using experimental data, and a new type of RAFM steel with higher tensile strength compared to conventional RAFM steels was successfully designed.
MATERIALS SCIENCE AND ENGINEERING A-STRUCTURAL MATERIALS PROPERTIES MICROSTRUCTURE AND PROCESSING
(2022)
Article
Materials Science, Multidisciplinary
Lei Jiang, Changsheng Wang, Huadong Fu, Jie Shen, Zhihao Zhang, Jianxin Xie
Summary: Aluminum alloys with ultra-strength and high-toughness are essential materials in the aerospace industry, but balancing these properties is challenging. This study used a machine learning design system to discover novel aluminum alloys with comparable properties to current state-of-the-art AA7136 alloy. The new alloys have high Mg and Zn content for hardening and refined grain size due to dispersoids.
JOURNAL OF MATERIALS SCIENCE & TECHNOLOGY
(2022)
Article
Materials Science, Ceramics
Qian Zhou, Feng Xu, Chengzuan Gao, Wenxuan Zhao, Lei Shu, Xianqing Shi, Muk-Fung Yuen, Dunwen Zuo
Summary: In this study, a data augmentation generative adversarial network (DAGAN)-driven machine learning (ML) design strategy was proposed to predict the mechanical properties of High-entropy nitride (HEN) ceramics. Optimal descriptor combination was selected using several feature selection algorithms. A property-conditioned DAGAN was established to overcome the data shortage problem, resulting in up to 14.67% improvement in ML model accuracy. Eight super-hard HEN systems with hardness above 40 GPa were discovered, with seven not yet experimentally synthesized. Ternary property diagrams were used to explore the intrinsic effects of chemical descriptors, providing efficient guidance for the design of novel high-performance HENs.
CERAMICS INTERNATIONAL
(2023)
Article
Materials Science, Multidisciplinary
Hongtao Zhang, Huadong Fu, Shuaicheng Zhu, Wei Yong, Jianxin Xie
Summary: A machine learning strategy was proposed to design alloys with remarkable properties by screening key alloy factors and utilizing Bayesian optimization, successfully designing new copper alloys with improved hardness and electrical conductivity, achieving simultaneous improvement of conflicting properties.
Article
Chemistry, Multidisciplinary
Xiangdong Wang, Tian Lu, Wenyan Zhou, Xiaobo Ji, Wencong Lu, Jiong Yang
Summary: New ternary gold alloys with low resistivities were discovered using a interpretable machine learning strategy, with a strong generalization ability of the model. The outputs of the model were analyzed with critical SHAP values and an online web server was developed to share the model.
CHEMISTRY-AN ASIAN JOURNAL
(2022)
Article
Chemistry, Physical
Yanmiao Wu, Zhongwen Shang, Tian Lu, Wenyan Zhou, Minjie Li, Wencong Lu
Summary: This paper presents a study on LMAs using an inverse design strategy, successfully designing and synthesizing two potential LMAs systems with different melting points. The effectiveness and accuracy of the inverse design strategy were verified, and the desired properties of LMAs were achieved.
JOURNAL OF ALLOYS AND COMPOUNDS
(2024)
Article
Materials Science, Multidisciplinary
Bjoern Wiese, Sven Berger, Jan Bohlen, Maria Nienaber, Daniel Hoeche
Summary: This work demonstrates the use of machine learning-based models to predict the relationship between material properties and process parameters in Mg-Gd alloys. The shallow artificial neural networks accurately predict the alloy's properties and outperform the linear regression approach. This machine learning approach has potential applications in alloy system development and online quality monitoring.
MATERIALS TODAY COMMUNICATIONS
(2023)
Article
Materials Science, Multidisciplinary
Yong-fei Juan, Guo-shuai Niu, Yang Yang, Yong-bing Dai, Jiao Zhang, Yan-feng Han, Bao-de Sun
Summary: To efficiently discover high-strength aviation aluminum alloys, a knowledge-aware design system (KADS) was proposed using machine learning methods. With the establishment of an aviation aluminum alloy database and the construction of a feature pool, the transformation from element content to material knowledge to property was realized. A KADS-designed aluminum alloy (KADS-Sc) with superior mechanical strength was successfully fabricated.
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T
(2023)
Article
Materials Science, Multidisciplinary
Feng Zhao, Chenhui Lei, Qingkun Zhao, Huiya Yang, Guoping Ling, Jiabin Liu, Haofei Zhou, Hongtao Wang
Summary: This study proposes a method that combines Gaussian process regression-based machine learning and phase diagrams to search for better compositions in Cu-Co-Si alloys. By training machine learning models, the electrical conductivity and hardness of the alloys were optimized, and the optimal compositions were discovered through contour maps.
MATERIALS TODAY COMMUNICATIONS
(2022)
Article
Materials Science, Multidisciplinary
Umer Masood Chaudry, Kotiba Hamad, Tamer Abuhmed
Summary: Through the use of various machine learning techniques, this study accelerated the process of designing aluminum alloys and found that the model obtained by gradient boosted tree (GBT) could efficiently predict the hardness of alloys.
MATERIALS TODAY COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Xiaobing Hu, Jiajun Zhao, Junjie Li, Zhijun Wang, Yiming Chen, Jincheng Wang
Summary: An intelligent alloy design strategy that combines machine learning and adaptive sampling is proposed and applied to an example of martensitic stainless steel. The feasibility of this strategy is proven through short iterative experiments, and a credible relationship between composition and hardness is demonstrated through a high-confidence prediction model.
ADVANCED THEORY AND SIMULATIONS
(2022)
Article
Dentistry, Oral Surgery & Medicine
W. T. Fu, Q. K. Zhu, N. Li, Y. Q. Wang, S. L. Deng, H. P. Chen, J. Shen, L. Y. Meng, Z. Bian
Summary: The study proposed a novel 3D deep convolutional neural network algorithm, named PAL-Net, for quickly and accurately detecting and segmenting periapical lesions (PALs) associated with apical periodontitis (AP) on cone beam computed tomography (CBCT) images. The algorithm improved the diagnostic performance and speed of dentists and showed comparable or superior segmentation accuracy to existing state-of-the-art algorithms.
JOURNAL OF DENTAL RESEARCH
(2023)
Article
Chemistry, Physical
Yi Zhang, Huadong Fu, Jialin He, Jianxin Xie
Summary: Understanding the formation and elimination mechanism of TCP phases in superalloys is crucial for their development. In this study, C15-Laves phase was identified in a Hf-containing Co-based superalloy, with its formation mechanism, thermal stability, and rapid elimination process elucidated. The findings can serve as a basis for composition design and heat treatment optimization of novel multicomponent Co-based superalloys.
JOURNAL OF ALLOYS AND COMPOUNDS
(2022)
Article
Metallurgy & Metallurgical Engineering
Fan Zhao, Guo-ning He, Ya-zheng Liu, Zhi-hao Zhang, Jian-xin Xie
Summary: Titanium microalloying improves the mechanical properties of vanadium microalloyed steels for hot forging mainly by refining austenite grains, resulting in decreased yield strength, increased elongation, and increased impact energy. Additionally, titanium microalloying promotes the nucleation of intragranular ferrite idiomorphs, which may benefit steels with coarse microstructure caused by critical deformation.
JOURNAL OF IRON AND STEEL RESEARCH INTERNATIONAL
(2022)
Article
Materials Science, Multidisciplinary
Qiang Zhang, Liang Zheng, Hua Yuan, Zhou Li, Guoqing Zhang, Jianxin Xie
Summary: The chemical state of FGH96 superalloy powders remains unchanged after long-term storage, but the oxygen content and NiO/Ni(OH)(2) layer thickness increase in a short period of time and remain stable afterwards. Powders stored in an oxygen atmosphere have the highest oxygen content, while those stored in vacuum have the lowest. The oxygen content of the stored powders affects the workability of the HIPed parts obtained through isothermal compression.
ADVANCED ENGINEERING MATERIALS
(2022)
Article
Chemistry, Physical
Huan Xu, Yuheng Zhang, Huadong Fu, Fei Xue, Xiaozhou Zhou, Jianxin Xie
Summary: The effects of B and C on the solidification microstructures of Co-Ni-Al-W-based superalloys were studied, with B decreasing solidus temperature and enlarging solidification temperature range, promoting the formation of specific phases, while C reducing segregation and suppressing specific phase formation.
JOURNAL OF ALLOYS AND COMPOUNDS
(2022)
Article
Chemistry, Physical
Qianru Yang, Qi Liu, Xinhua Liu, Yu Lei, Yanbin Jiang, Jianxin Xie, Zhou Li
Summary: C70250 alloy plates with different grain morphologies and precipitate contents were produced by horizontal continuous casting. The mechanical properties of the alloy were influenced by the microstructure, with equiaxed grain plates showing the best performance and columnar grain plates exhibiting better plasticity.
JOURNAL OF ALLOYS AND COMPOUNDS
(2022)
Article
Materials Science, Multidisciplinary
Yaliang Zhu, Fangmiao Duan, Wei Yong, Huadong Fu, Hongtao Zhang, Jianxin Xie
Summary: In this study, a machine learning method based on data fusion was proposed to accurately predict creep rupture life of nickel-based superalloys. By using the properties of existing alloys, the method successfully predicted the properties of new alloys and achieved a high accuracy rate.
COMPUTATIONAL MATERIALS SCIENCE
(2022)
Article
Materials Science, Multidisciplinary
Lei Jiang, Zhihao Zhang, Huadong Fu, Shiyu Huang, Dawei Zhuang, Jianxin Xie
Summary: In this study, the corrosion behavior of Al-Zn-Mg-Cu alloys with different components in NaCl solution was experimental studied. The composition changes of micron and nano secondary phases during corrosion were semi-quantitatively characterized. The effect of the type and quantity of the secondary phases on the corrosion resistance and mechanism was discussed. The results showed that AA7050 and AA7136 alloys exhibited localized self-corrosion, while the E2 alloy designed by authors had significantly better pitting resistance due to the absence of secondary phases.
MATERIALS CHARACTERIZATION
(2022)
Article
Materials Science, Multidisciplinary
Xiaozhou Zhou, Yuheng Zhang, Yi Zhang, Huadong Fu, Jianxin Xie
Summary: This study investigates the effect of nickel content on the solidification behavior and hot-tearing susceptibility of cobalt-based superalloys. The results indicate that increasing nickel content leads to slower solidification and higher susceptibility to hot-tearing.
METALLURGICAL AND MATERIALS TRANSACTIONS A-PHYSICAL METALLURGY AND MATERIALS SCIENCE
(2022)
Article
Materials Science, Multidisciplinary
Xingqun He, Huadong Fu, Jianxin Xie
Summary: This study proposes a method of using in-situ composite fiber-reinforcement to regulate the mechanical properties and electrical conductivity of silver-based alloys. By processing the alloy and testing the strength and electrical conductivity at different deformation stages, it is shown that this method can achieve comprehensive performance control of the alloy.
INTERNATIONAL JOURNAL OF MINERALS METALLURGY AND MATERIALS
(2022)
Article
Metallurgy & Metallurgical Engineering
Guangtao Lin, Fan Zhao, Zhihao Zhang, Jianxin Xie
Summary: This study investigates the hot compression behavior of high-silicon electrical steel and the effect of Nb on its properties. The results show that the addition of Nb can inhibit the growth of dynamic recrystallization grains and the generation of microcracks, as well as reduce the ordered degree of the steel.
STEEL RESEARCH INTERNATIONAL
(2023)
Article
Physics, Applied
Yangyang Xu, Guomang Shao, Yumei Zhou, Yu Wang, Sen Yang, Xiangdong Ding, Jun Sun, E. K. H. Salje, Turab Lookman, Dezhen Xue
Summary: A ferroelectric phase transition in barium titanate single crystals under an external electric field is observed, showing scale invariant nucleation and growth of complex domain structures. The energy exponent varies with the external bias and cooling history, with the exponent being near 1.68±0.022 for a single-domain sample after field cooling and 1.66 for a multi-domain sample after zero field cooling under high fields. The complex domain patterns in the multi-domain sample hinder the movement of the phase boundary and generate more small energy signals, resulting in a high critical exponent. The aftershock time distribution remains the same for all switching conditions with Omori exponent near -1 and switching time correlations of -1±0.05 for short times and -2±0.10 for long times.
APPLIED PHYSICS LETTERS
(2023)
Article
Nanoscience & Nanotechnology
Hongji Lin, Shuai Ren, Pengfei Dang, Chunxi Hao, Xuefei Tao, Dezhen Xue, Yu Wang, Hongxiang Zong, Zhenxuan Zhang, Wenqing Ruan, Xiong Liang, Jiang Ma, Xiangdong Ding, Jun Shen
Summary: In this study, a strain glass was found in Ti50-x-yNi50+yNbx due to the synergistic effect of Ni and Nb atoms. The system was led to the strain glass state with the help of 1 at% excess Ni atoms and 2% Nb atoms. A boson-peak-like glassy feature was also observed.
SCRIPTA MATERIALIA
(2023)
Article
Chemistry, Physical
Lei Jiang, Zhihao Zhang, Hao Hu, Xingqun He, Huadong Fu, Jianxin Xie
Summary: One challenge in material design is to rapidly develop new materials or improve existing materials using existing data and knowledge. This study proposes a rapid and effective method of alloy material design through data transfer learning, using existing data to efficiently design new alloys. By transferring 1053 pieces of process-property relationship data of existing AA7xxx commercial aluminum alloys, an optimal three-stage solution-aging treatment process (T66R) was designed for a new type of aluminum alloy (E2 alloy). This method significantly improves the strength and plasticity of the E2 alloy, which is of great importance for lightweight high-end equipment. Microstructure analysis elucidates the mechanism of alloy performance improvement. The study demonstrates that transferring existing alloy data is an effective method for designing new alloys.
NPJ COMPUTATIONAL MATERIALS
(2023)
Article
Materials Science, Multidisciplinary
Yuanchao Yang, Yangyang Xu, Yumei Zhou, Xiangdong Ding, Jun Sun, Turab Lookman, Dezhen Xue
Summary: We study the behavior of martensitic transformation by considering the coupling of volumetric strain and transformation strain within a Ginzburg-Landau framework. Nonthermoelastic features, such as large residual strain, large thermal hysteresis, and incomplete transformation gradually appear with increasing coupling strength. The volume change associated with the transformation due to the coupling is identified as the essential physics for the nonthermoelastic features observed in an ideal thermoelastic martensitic transformation. A materials descriptor based on insights from the model is proposed for features associated with a martensitic transformation.
Article
Materials Science, Multidisciplinary
Pengfei Dang, Lei Zhang, Yumei Zhou, Cheng Li, Xiangdong Ding, Jun Sun, Dezhen Xue
Summary: By combining point defect doping and nanostructure, low-temperature shape-memory alloys (SMAs) have been developed with good actuation properties. The doping of Co into the Ti-Ni host alloy hinders the martensitic transformation at lower temperatures. By post-deformation annealing, the partially recrystallized state with dense dislocations and nanoprecipitates is achieved, effectively strengthening the matrix to prevent irreversible strain under large external loads. The Ti49.2Ni44.8Co6 alloy exhibits a recoverable actuation strain of 4.5% at temperatures below -50 degrees C, and its work output of 36 MJ m(-3) surpasses most SMAs, making it promising for low-temperature actuator applications.
ADVANCED ENGINEERING MATERIALS
(2023)