Abstract：The hydrogen embrittlement issue in ferrite steels has always been a hot topic of concern for researchers, particularly the unclear influence of different carbon contents on the hydrogen resistance of iron and steel materials. In this study, a high-precision machine learning force field (MLFF) for the iron-carbon-hydrogen system was constructed by combining first-principles calculations with machine learning algorithms. Molecular dynamics simulations were performed to investigate the diffusion behavior of hydrogen atoms in steels with different carbon contents. The high-precision MLFF was trained using a neural network (NN) model based on first-principles molecular dynamics (AIMD) results of multiple configurations. Various tests were conducted to ensure that the machine learning force field could accurately describe the statistical and dynamic properties of the iron-carbon-hydrogen system. Using this MLFF, molecular dynamics simulations were performed on ferrite steels with different carbon contents, and the hydrogen diffusion coefficients were calculated. It was found that the hydrogen diffusion coefficient generally decreased with increasing carbon content, in good agreement with experimental results. The algorithm model established in this study can analyze the influence of carbon content on the hydrogen resistance of iron and steel materials, which is of significant importance for studying hydrogen-induced damage in steel materials and composition design.
[1]VENEZUELA J, LIU Q, ZHANG M, et al.A review of hydrogen embrittlement of martensitic advanced high-strength steels[J].Corros Rev, 2016, 34(3):153-86[2]DWIVEDI S K, VISHWAKARMA M.Hydrogen embrittlement in different materials: A review[J].Int J Hydrogen Energ, 2018, 43(46):21603-16[3]ZHAO H, CHAKRABORTY P, PONGE D, et al.Hydrogen trapping and embrittlement in high-strength Al alloys[J].Nature, 2022, 602(7897):437-41[4]ROBERTSON I M, SOFRONIS P, NAGAO A, et al.Hydrogen embrittlement understood[J].Metall Mat Trans A, 2015, 46(6):2323-41[5]BIRNBAUM H K, SOFRONIS P.Hydrogen-enhanced localized plasticity—a mechanism for hydrogen-related fracture[J].Mat Sci Eng A, 1994, 176(1-2):191-202[6]BINHAN SUN D W, XU LU, DI WAN, DIRK PONGE, XIANCHENG ZHANG.Current Challenges and Opportunities Toward Understanding Hydrogen Embrittlement Mechanisms in Advanced High-Strength Steels: A Review[J].Acta Metall Sin, 2021, 34(6):741-54[7]SKYLARIS C-K.A benchmark for materials simulation[J].Science, 2016, 351(6280):1394-5[8]HAMMES-SCHIFFER S.A conundrum for density functional theory[J].Science, 2017, 355(6320):28-9[9]QIU X, ZHANG K, KANG Q, et al.Investigation of hydrogen diffusion behavior in 12Cr2Mo1R (H) steel by electrochemical tests and first-principles calculation[J].Anti-Corrosion Methods and Materials, 2022, 69(1):17-28[10]HE Y, LI Y, CHEN C, et al.Diffusion coefficient of hydrogen interstitial atom in α-Fe,γ-Fe and ε-Fe crystals by first-principle calculations[J].Int J Hydrogen Energ, 2017, 42(44):27438-45[11]JIANG D, CARTER E A.Diffusion of interstitial hydrogen into and through bcc Fe from first principles[J].Physical Review B, 2004, 70(6):064102-064102[12]SANCHEZ J, FULLEA J, ANDRADE M, et al.Ab initio molecular dynamics simulation of hydrogen diffusion in α-iron[J].Physical Review B, 2010, 81(13):132102-132102[13]FAN X, MI Z, YANG L, et al.Application of DFT Simulation to the Investigation of Hydrogen Embrittlement Mechanism and Design of High Strength Low Alloy Steel[J].Materials, 2022, 16(1):152-152[14]ZHANG B, ASTA M, WANG L-W.Machine learning force field for Fe-H system and investigation on role of hydrogen on the crack propagation in α-Fe [J].[J].Computational Materials Science, 2022, 214(111709):1-11[15]ZHANG W, WENG M, ZHANG M, et al.Revealing Morphology Evolution of Lithium Dendrites by Large‐Scale Simulation Based on Machine Learning Force Field[J].Advanced Energy Materials, 2023, 13(4):2202892-2202892[16]CHMIELA S, VASSILEV-GALINDO V, UNKE O T, et al.Accurate global machine learning force fields for molecules with hundreds of atoms[J].Science Advances, 2023, 9(2):e-a[17]COUNTS W, WOLVERTON C, GIBALA R.First-principles energetics of hydrogen traps in α-Fe: Point defects[J].Acta Materialia, 2010, 58(14):4730-41[18]JIA W, CAO Z, WANG L, et al.The analysis of a plane wave pseudopotential density functional theory code on a GPU machine[J].Computer Physics Communications, 2013, 184(1):9-18[19]JIA W, FU J, CAO Z, et al.Fast plane wave density functional theory molecular dynamics calculations on multi-GPU machines [J].[J].Journal of Computational Physics, 2013, 251(1):102-115[20]LIN L.Adaptively compressed exchange operator[J].Journal of chemical theory and computation, 2016, 12(5):2242-9[21]KANG J, WANG L-W.First-principles Green-Kubo method for thermal conductivity calculations[J].Physical Review B, 2017, 96(2):020302-020302[22]KOWSARI M, ALAVI S, ASHRAFIZAADEH M, et al.Molecular dynamics simulation of imidazolium-based ionic liquids. I. Dynamics and diffusion coefficient [J].Journal of Chemical Physics, 2008, 129(22):2038-2045[23]JACK D, JACK K.Invited review: carbides and nitrides in steel[J].Materials Science and Engineering, 1973, 11(1):1-27[24]HUANG Y, KANG J, GODDARD III W A, et al.Density functional theory based neural network force fields from energy decompositions[J].Physical Review B, 2019, 99(6):064103-064103[25]ORIANI R A.The diffusion and trapping of hydrogen in steel[J].Acta Metall, 1970, 18(1):147-57[26]KIUCHI K, MCLELLAN R.The solubility and diffusivity of hydrogen in well-annealed and deformed iron [M]. Perspectives in Hydrogen in Metals. Elsevier. 1986: 29-52.[27]GADGEEL V L, JOHNSON D L.Gas-phase hydrogen permeation and diffusion in carbon steels as a function of carbon content from 500 to 900 K[J].Journal of Materials for Energy Systems, 1979, 1(2):32-40
2024, Vol. 31 