The role of physical metallurgical relationships in enhancing alloy properties prediction and design: A case study on Q&P steel

Abstract

Traditional alloy design typically relies on a trial-and-error approach, which is both time-consuming and expensive. Whilst physical metallurgical (PM) models offer some predictive capabilities, their reliability is limited by errors accumulating across space scales. To address this, this study proposes a novel framework that combines PM knowledge graphs (PMKGs) with graph neural networks (GNNs) to predict the tensile properties of quenching and partitioning steels, using genetic algorithms for dual-objective optimization. Compared to traditional artificial intelligence (AI) models, this framework shows significant advantages in predicting ultimate tensile strength (UTS) and total elongation (TEL) with higher accuracy and stability. Notably, the R² for TEL prediction improved by approximately 15%. Furthermore, this framework successfully balances UTS and TEL, resulting in the design of alloys with superior overall properties. The designed alloys, with a composition of approximately 0.3 wt.% C, 3 wt.% Mn, 1.2 wt.% Si, and minor amounts of Cr and Al, achieve a UTS exceeding 1500 MPa and TEL near 20%, aligning with PM principles and validating the rationality and feasibility of this method. This study offers new insights into applying AI in complex multi-objective alloy design, highlighting the potential of integrating expert knowledge with GNNs.

Keywords

alloy design / dual-objective optimisation / graph neural networks / knowledge graphs / physical metallurgy

References

[1] Yu H, Fu J, Wang C, et al. Robust additive manufacturable Ni superalloys designed by the integrated optimisation of local elemental segregation and cracking susceptibility criteria. Acta Mater. 2024; 266:119658.

[2] Li H, Zhao X, Zhang H, Li D, Li J. Effect of the oxide layer structure on the decarburization behavior of 60Si2Mn spring steel in dry air: experimental and first-principle study. Mater Char. 2023; 196:112651.

[3] Xu XY, Huang CP, Wang HY, Li YZ, Huang MX. Rate-dependent transition of dislocation mechanisms in a magnesium alloy. Acta Mater. 2024; 263:119474.

[4] Li Y, Martín DS, Wang J, Wang C, Xu W. A review of the thermal stability of metastable austenite in steels: martensite formation. J Mater Sci Technol. 2021; 91: 200-214.

[5] Qi X, Huan P, Chen X, et al. Corrosion resistance and mechanism of X100 pipeline steel laser-metal active gas hybrid welds with Cr containing welding wire in NS4 solution. Corrosion Sci. 2023; 221:111329.

[6] Zeng X, Yi H, Zeng Z, et al. Track-Rex: a universal toolbox for tracking recrystallization nucleation and grain growth behaviors in polycrystalline materials. J Mater Sci Technol. 2024; 197: 149-159.

[7] Li Y, Chen S, Wang C, Martín DS, Xu W. Modeling retained austenite in Q& steels accounting for the bainitic transformation and correction of its mismatch on optimal conditions. Acta Mater. 2020; 188: 528-538.

[8] Zang C, Zhao X, Xia H, et al. Design method of immiscible dissimilar welding (Mg/Fe) based on CALPHAD and thermodynamic modelling. Mater Des. 2024; 243:113050.

[9] Wang C, Zhang C, Yang Z, Su J, Weng Y. Microstructure analysis and yield strength simulation in high Co-Ni secondary hardening steel. Mater Sci Eng, A. 2016; 669: 312-317.

[10] Galindo-Nava EI, Rainforth WM, Rivera-Díaz-del-Castillo PEJ. Predicting microstructure and strength of maraging steels: elemental optimisation. Acta Mater. 2016; 117: 270-285.

[11] Shen C, Wang C, Wei X, Li Y, van der Zwaag S, Xu W. Physical metallurgy-guided machine learning and artificial intelligent design of ultrahigh-strength stainless steel. Acta Mater. 2019; 179: 201-214.

[12] Li Y, Wang L, Zhu K, Wang C, Xu W. An integral transformation model for the combined calculation of key martensitic transformation temperatures and martensite fraction. Mater Des. 2022; 219:110768.

[13] Geng X, Wang F, Wu HH, et al. Data-driven and artificial intelligence accelerated steel material research and intelligent manufacturing technology. Mater Genome Eng Adv. 2023; 1(1): e10.

[14] Jin X, Luo H, Wang X, et al. Data mining accelerated the design strategy of high-entropy alloys with the largest hardness based on genetic algorithm optimisation. Mater Genome Eng Adv. 2024; 2(2): e49.

[15] Rahaman M, Mu W, Odqvist J, Hedström P. Machine learning to predict the martensite start temperature in steels. Metall Mater Trans A. 2019; 50(5): 2081-2091.

[16] Galimzyanov BN, Doronina MA, Mokshin AV. Machine learning-based prediction of elastic properties of amorphous metal alloys. Phys Stat Mech Appl. 2023; 617:128678.

[17] Guo Z, Sha W. Modelling the correlation between processing parameters and properties of maraging steels using artificial neural network. Comput Mater Sci. 2004; 29(1): 12-28.

[18] Li H, Li Y, Huang J, et al. Physical metallurgy guided industrial big data analysis system with data classification and property prediction. Steel Res Int. 2022; 93(8):2100820.

[19] Han S, Wang C, Zhang Y, Xu W, Di H. Employing deep learning in non-parametric inverse visualization of elastic-plastic mechanisms in dual-phase steels. Mater Genome Eng Adv. 2024; 2(1): e29.

[20] Clarke AJ, Speer JG, Miller MK, et al. Carbon partitioning to austenite from martensite or bainite during the quench and partition (Q&) process: a critical assessment. Acta Mater. 2008; 56(1): 16-22.

[21] Wang C, Zhu K, Hedström P, Li Y, Xu W. A generic and extensible model for the martensite start temperature incorporating thermodynamic data mining and deep learning framework. J Mater Sci Technol. 2022; 128: 31-43.

[22] Zhou J, Cui G, Hu S, et al. Graph neural networks: a review of methods and applications. AI Open. 2020; 1: 57-81.

[23] Shi X, Zhou L, Huang Y, Wu Y, Hong Z. A review on the applications of graph neural networks in materials science at the atomic scale. Mater Genome Eng Adv. 2024; 2(2): e50.

[24] Schutt KT, Arbabzadah F, Chmiela S, Muller KR, Tkatchenko A. Quantum-chemical insights from deep tensor neural networks. Nat Commun. 2017; 8(1):13890.

[25] Dai M, Demirel MF, Liang Y, Hu J-M. Graph neural networks for an accurate and interpretable prediction of the properties of polycrystalline materials. npj Comput Mater. 2021; 7(1): 103.

[26] Li Y, Wang C, Zhang Y, et al. Thermodynamically informed graph for interpretable and extensible machine learning: martensite start temperature prediction. Calphad. 2024; 85:102710.

[27] Zhang Z, Liu S, Zhang Y, et al. Optimisation of low-power femtosecond laser trepan drilling by machine learning and a high-throughput multi-objective genetic algorithm. Opt Laser Technol. 2022; 148:107688.

[28] Wang C, Wei X, Ren D, Wang X, Xu W. High-throughput map design of creep life in low-alloy steels by integrating machine learning with a genetic algorithm. Mater Des. 2022; 213:110326.

[29] Chen S, Wang C, Shan L, Li Y, Zhao X, Xu W. Revealing the conditions of bainitic transformation in quenching and partitioning steels. Metall Mater Trans A. 2019; 50(9): 4037-4046.

[30] Chen S. Bainitic Transformation Characteristics and Accurate Tailoring of Complex Microstructure in High Strength Automotive Steels, PhD Thesis. Northeastern University; 2021.

[31] Carpio M, Calvo J, García O, Pedraza JP, Cabrera JM. Heat treatment design for a QP steel: effect of partitioning temperature. Metals. 2021; 11(7):1136.

[32] Yan S, Liu X, Liu WJ, et al. Comparative study on microstructure and mechanical properties of a C-Mn-Si steel treated by quenching and partitioning (Q&) processes after a full and intercritical austenitization. Mater Sci Eng, A. 2017; 684: 261-269.

[33] Kim JH, Kwon M-H, Gu G, Lee JS, Suh D-W. Quenching and partitioning (Q&) processed medium Mn steel starting from heterogeneous microstructure. Materialia. 2020; 12:100757.

[34] Xiong Z, Jacques PJ, Perlade A, Pardoen T. Ductile and intergranular brittle fracture in a two-step quenching and partitioning steel. Scripta Mater. 2018; 157: 6-9.

[35] Arlazarov A, Ollat M, Masse JP, Bouzat M. Influence of partitioning on mechanical behavior of Q& steels. Mater Sci Eng, A. 2016; 661: 79-86.

[36] Seo EJ, Cho L, Estrin Y, De Cooman BC. Microstructure-mechanical properties relationships for quenching and partitioning (Q&) processed steel. Acta Mater. 2016; 113: 124-139.

[37] de Diego-Calderón I, De Knijf D, Monclús MA, et al. Global and local deformation behavior and mechanical properties of individual phases in a quenched and partitioned steel. Mater Sci Eng, A. 2015; 630: 27-35.

[38] Tkachev E, Borisov S, Borisova Y, Kniaziuk T, Belyakov A, Kaibyshev R. Austenite stabilization and precipitation of carbides during quenching and partitioning (Q&) of low-alloyed Si-Mn steels with different carbon content. Mater Sci Eng, A. 2024; 895:146212.

[39] Cheng YY, Zhao G, Xu DM, Mao XP, Bao SQ, Yang GW. Comparative study on microstructures and mechanical properties of Q& steels prepared with hot-rolled and cold-rolled C-Si-Mn sheets. J Mater Res Technol. 2022; 20: 1226-1242.

[40] Gao P, Li F, An K, Zhao Z, Chu X, Cui H. Microstructure and deformation mechanism of Si-strengthened intercritically annealed quenching and partitioning steels. Mater Char. 2022; 191:112145.

[41] Wang L, Liang Y, Zhao F, et al. Achieving high tensile properties and impact toughness in ultrahigh strength lean alloy steel by quenching and partitioning treatment. Arch Civil Mech Eng. 2023; 24(1): 32.

[42] Guan J, Liu M, Tian J, Chen Z, Xu G. Effects of isothermal transformation at the quenching temperature on the microstructure and mechanical properties of a medium-carbon steel. Trans Indian Inst Met. 2021; 74(12): 3265-3272.

[43] Hu P, Su Y, Li J, Zuo Z. Texture and mechanical properties of quenching and partitioning steel. Mater Res Express. 2023; 10(7):076504.

[44] Zhao ZZ, Liang JH, Zhao AM, Liang JT, Tang D, Gao YP. Effects of the austenitizing temperature on the mechanical properties of cold-rolled medium-Mn steel system. J Alloys Compd. 2017; 691: 51-59.

[45] He F. Study on Microstructure and Mechanical Properties Control Technology of Q&P Steel under Continuous Annealing Process PhD thesis. Central Iron and Steel Research Institute; 2021.

[46] Jiao G. Studies on the Q&P Heat Treatment Process of Car Used High-Strength Steel, Master Thesis. Northeastern University; 2012.

[47] Li S. Microstructure and Properties Evolution and Mathematical Model of Cold-Rolled High-Strength Steel during Q&P Process Master thesis. Northeastern University; 2022.

[48] Li W. Study on Microstructure and Mechanical Properties of Low-Carbon Quenching and Partitioning Steels PhD thesis. University of Science and Technology Beijing; 2016.

[49] Lin Z. Study on Q&P Heat Treatment Process and Deformation Mechanism of Super High Strength Medimum Manganese Steel PhD thesis. University of Science and Technology Beijing; 2017.

[50] Lu B. Research on Microstructure Evolution and Mechanical Properties of HighStrength Medium Mn Q&P Steel Master thesis. Northeastern University; 2018.

[51] Zhang J. Research on Heat-Treatment and Deformation Mechanism of Low-Carbon High-Strength Q&P Steels PhD thesis. Northeastern University; 2015.

[52] Zhang J. Research on Microstrcuture and Mechanical Properties of Q&P High-Property Steel Master thesis. Northeastern University; 2011.

[53] Wei X, Zwaag Svd, Jia Z, Wang C, Xu W. On the use of transfer modeling to design new steels with excellent rotating bending fatigue resistance even in the case of very small calibration datasets. Acta Mater. 2022; 235:118103.

[54] Tong Z, Liang Y, Sun C, Rosenblum DS, Lim A. Directed graph convolutional network. arXiv:2004. 2020:13970.

[55] Iskander M, Shrive N. The effect of the shape and size of initial flaws on crack propagation in uniaxially compressed linear brittle materials. Theor Appl Fract Mech. 2020; 109:102742.

[56] Xia P, Vercruysse F, Petrov R, Sabirova I, Castillo-Rodríguez M, Verleysenc P. High strain rate tensile behavior of a quenching and partitioning (Q&) Fe-0.25C-1.5Si-3.0Mn steel. Mater Sci Eng, A. 2019; 745: 53-62.

[57] Mohammadi Zahrani M, Ketabchi M, Ranjbarnodeh E. Microstructure development and mechanical properties of a C-Mn-Si-Al-Cr cold rolled steel subjected to quenching and partitioning treatment. J Mater Res Technol. 2023; 22: 2806-2818.

[58] Jirková H, Kučerová L, Mašek B. The effect of chromium on microstructure development during Q-P process. Mater Today Proc. 2015; 2: S627-S630.

[59] Park T, Jr LH, Hu X, et al. Crystal plasticity modeling of 3rd generation multi-phase AHSS with martensitic transformation. Int J Plast. 2019; 120: 1-46.

[60] Moor ED, Speer JG, Matlock DK, Kwak J, Lee S. Effect of carbon and manganese on the quenching and partitioning response of CMnSi steels. ISIJ Int. 2011; 51(1): 137-144.

[61] Cho L, Seo EJ, Cooman BCD. Near-Ac3 austenitized ultra-fine-grained quenching and partitioning (Q&) steel. Scripta Mater. 2016; 123: 69-72.

[62] Shen C, Wei X, Wang C, Xu W. A deep learning method for extensible microstructural quantification of DP steel enhanced by physical metallurgy-guided data augmentation. Mater Char. 2021; 180:111392.

RIGHTS & PERMISSIONS

2024 The Author(s). Materials Genome Engineering Advances published by Wiley-VCH GmbH on behalf of University of Science and Technology Beijing.