Discovery of marageing steels: machine learning vs. physical metallurgical modelling

Shen, Chunguang; Wang, Chenchong; Rivera-Dı́az-del-Castillo, P.E.J.; Xu, Dake; Zhang, Qian; Zhang, Chi; Xu, Wei

doi:10.1016/j.jmst.2021.02.017

Cited by 12 publications

(7 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Coarsening rate constants k (m 3 s −1 ) of Ni-based superalloys at 1123 K reported in the literature [53,75] and obtained in this work. CMSX-2 [53] Hou et al [75] C2 C4 By comparing with the recent alloy designs based on machine learning or other data-driven methods, [80][81][82][83][84] it is reasonable to conclude that the high accuracy and efficiency of our design strategy is attributed to three aspects. First, the multicomponent diffusion couple method, which was used for the highthroughput preparation of multicomponent diffusion couples in a simple sample, possesses higher efficiency than traditional diffusion couple methods.…”

Section: Tablementioning

confidence: 99%

High‐Throughput Method–Accelerated Design of Ni‐Based Superalloys

Liu

Wang

Wang³

et al. 2022

Adv Funct Materials

View full text Add to dashboard Cite

Ever-increasing demands for superior alloys with improved hightemperature service properties require accurate design of their composition. However, conventional approaches to screen the properties of alloys such as creep resistance and microstructural stability cost a lot of time and resources. This work therefore proposes a novel high throughput-based design strategy for high-temperature alloys to accelerate their composition selections, by taking Ni-based superalloys as an example. A numerical inverse method is used to massively calculate the multielement diffusion coefficients based on an accurate atomic mobility database. These coefficients are subsequently employed to refine the physical models for tuning the creep rates and structural stability of alloys, followed by unsupervised machine learning to categorize their composition and determine the range of the composition with optimal performance. By using a strict screening criterion, two sets of composition with comprehensively optimal properties are selected, which is then validated by experiments. Compared with recent data-driven methods for materials design, this strategy exhibits high accuracy and efficiency attributed to the high-throughput multicomponent diffusion couples, self-developed atomic mobility database, and refined physical models. Since this strategy is independent of the alloy composition, it can efficiently accelerate the development of multicomponent high-performance alloys and tackle challenges in discovering novel materials.

show abstract

Section: Tablementioning

confidence: 99%

High‐Throughput Method–Accelerated Design of Ni‐Based Superalloys

Liu

Wang

Wang³

et al. 2022

Adv Funct Materials

View full text Add to dashboard Cite

show abstract

“…With the vigorous development of computational science, it is a new trend to examine the influence of alloying elements on the properties of materials by leveraging algorithms to model rationally. Additionally, common ML algorithms were adopted to investigate the intrinsic relations of materials especially in steels, including random forest (RF), linear regression (LR), support vector regression (SVR), multi-layer perceptron (MLP), convolutional neural network (CNN), and K-nearest neighbor (KNN) [ 179 , 180 , 181 , 182 , 183 , 184 , 185 , 186 ]. Yupeng Diao et al [ 179 ] proposed a ML prediction model for comprehensive properties and successfully employed the efficient global optimization algorithm to optimize multi-objective mechanical properties for carbon steels.…”

Section: Applications Of Multi-scale Computational Simulations and Ml...mentioning

confidence: 99%

“…Physical metallurgical (PM) method has been employed as an efficient strategy to develop distinguished mechanical properties and illustrate the mechanisms of strength increment. Furthermore, Chunguang Shen et al [ 186 ] introduced PM parameters into ML modeling and established a ML model guided by PM, and these physical parameters can be easily obtained, which are assisted by thermodynamic software calculations. The precision of best prediction results of the PM model is apparently lower than the ML model.…”

Section: Applications Of Multi-scale Computational Simulations and Ml...mentioning

confidence: 99%

See 1 more Smart Citation

Recent Advances on Composition-Microstructure-Properties Relationships of Precipitation Hardening Stainless Steel

et al. 2022

View full text Add to dashboard Cite

Precipitation hardening stainless steels have attracted extensive interest due to their distinguished mechanical properties. However, it is necessary to further uncover the internal quantitative relationship from the traditional standpoint based on the statistical perspective. In this review, we summarize the latest research progress on the relationships among the composition, microstructure, and properties of precipitation hardened stainless steels. First, the influence of general chemical composition and its fluctuation on the microstructure and properties of PHSS are elaborated. Then, the microstructure and properties under a typical heat treatment regime are discussed, including the precipitation of B2-NiAl particles, Cu-rich clusters, Ni3Ti precipitates, and other co-existing precipitates in PHSS and the hierarchical microstructural features are presented. Next, the microstructure and properties after the selective laser melting fabricating process which act as an emerging technology compared to conventional manufacturing techniques are also enlightened. Thereafter, the development of multi-scale simulation and machine learning (ML) in material design is illustrated with typical examples and the great concerns in PHSS research are presented, with a focus on the precipitation techniques, effect of composition, and microstructure. Finally, promising directions for future precipitation hardening stainless steel development combined with multi-scale simulation and ML methods are prospected, offering extensive insight into the innovation of novel precipitation hardening stainless steels.

show abstract

“…Machine learning, a data-driven approach, has been employed to predict the properties of HEAs as well as several other alloys. Furthermore, material researchers have found it can overcome the limitations of the above-mentioned approaches [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34] . Zhang et al 18 , Wen et al 27,30 , Zheng et al 29 , Klimenko et al 32 , Guo et al 23 , and Li et al 26 developed machine learning models which predict the mechanical properties of the HEAs.…”

Section: Introductionmentioning

confidence: 99%

A neural network model for high entropy alloy design

et al. 2023

View full text Add to dashboard Cite

A neural network model is developed to search vast compositional space of high entropy alloys (HEAs). The model predicts the mechanical properties of HEAs better than several other models. It’s because the special structure of the model helps the model understand the characteristics of constituent elements of HEAs. In addition, thermodynamics descriptors were utilized as input to the model so that the model predicts better by understanding the thermodynamic properties of HEAs. A conditional random search, which is good at finding local optimal values, was selected as the inverse predictor and designed two HEAs using the model. We experimentally verified that the HEAs have the best combination of strength and ductility and this proves the validity of the model and alloy design method. The strengthening mechanism of the designed HEAs is further discussed based on microstructure and lattice distortion effect. The present alloy design approach, specialized in finding multiple local optima, could help researchers design an infinite number of new alloys with interesting properties.

show abstract

Discovery of marageing steels: machine learning vs. physical metallurgical modelling

Cited by 12 publications

References 42 publications

High‐Throughput Method–Accelerated Design of Ni‐Based Superalloys

High‐Throughput Method–Accelerated Design of Ni‐Based Superalloys

Recent Advances on Composition-Microstructure-Properties Relationships of Precipitation Hardening Stainless Steel

A neural network model for high entropy alloy design

Contact Info

Product

Resources

About