Integrating 3D CNN and phase-field simulation for TiAlN coating property prediction via 3D microstructure

Abstract

Microstructure evolution in service significantly influences the properties of advanced materials. Numerical simulation can effectively capture microstructure development and provide abundant high-fidelity data. However, effective 3D microstructure-informed property prediction methods are lacking due to the complexity and richness of 3D microstructural data. In this work, a novel approach combining phase-field simulation and a 3D convolutional neural network is proposed to explore the composition-process-structure-property relationship in Ti_1-xAl_xN coatings. A large dataset of 4,962 simulated 3D microstructures under various heat treatment conditions was first generated using phase-field simulations. Then, a reconstructible feature extraction model was trained to compress each 48 × 48 × 48-grid microstructure into a 128-dimensional latent vector with a reconstruction accuracy of up to 99%. Using the extracted features, a microstructure-based hardness prediction model was constructed, achieving a low prediction error of 1.6 GPa (ca. 5.3% error for an average hardness of 30.8 GPa). The results demonstrate the effectiveness of 3D microstructure-informed deep learning for accurate property prediction, providing a promising tool for the data-driven design of high-performance materials.

Keywords

3D convolutional neural network / phase-field simulation / property prediction / TiAlN coating

References

[1] Bostanabad R,Li X.Computational microstructure characterization and reconstruction: review of the state-of-the-art techniques.Prog Mater Sci2018;95:1-41

[2] Cai J,Yan X.A process-synergistic active learning framework for high-strength Al-Si alloys design.NPJ Comput Mater2025;11:1721

[3] Jiang L,Hu H,Fu H.A rapid and effective method for alloy materials design via sample data transfer machine learning.NPJ Comput Mater2023;9:979

[4] Zhang H,Zhu S,Xie J.Machine learning assisted composition effective design for precipitation strengthened copper alloys.Acta Mater2021;215:117118

[5] Gao T,Yang S.Data-driven design of novel lightweight refractory high-entropy alloys with superb hardness and corrosion resistance.NPJ Comput Mater2024;10:1457

[6] Giles SA,Broderick SR.Machine-learning-based intelligent framework for discovering refractory high-entropy alloys with improved high-temperature yield strength.NPJ Comput Mater2022;8:926

[7] Jia B,Yan L.Tribological properties and machine learning prediction of FeCoCrNiAlN high entropy coatings.Surf Coat Technol2024;477:130341

[8] Zou H,Zhang L.Integrating machine learning and CALPHAD method for exploring low-modulus near-β-Ti alloys.Rare Met2024;43:309-23

[9] Liu Z,Jin L.Towards high stiffness and ductility - The Mg-Al-Y alloy design through machine learning.J Mater Sci Technol2025;221:194-203

[10] Yang W,Huang X,Wang P.Effects of Cu content and Sintering temperature on microstructure and mechanical properties of SiCp/Al-Cu-Mg composites through experimental study, CALPHAD-type simulation and machine learning.J Mater Res Technol2024;33:2216-25

[11] Qiu R,Bäcke O,Andrén H.Atom probe tomography investigation of 3D nanoscale compositional variations in CVD TiAlN nanolamella coatings.Surf Coat Technol2021;426:127741

[12] Knutsson A,Rogström L.Microstructure evolution during the isostructural decomposition of TiAlN - A combined in-situ small angle X-ray scattering and phase field study.J Appl Phys2013;113:213518

[13] Tai P,Shen T.Microstructure evolution and nanohardness of nanostructured TiAlN coating under N+ ion irradiation.Surf Coat Technol2022;441:128494

[14] Salama H,Steinbach I.The interplay between the martensitic transformation rate and the rate of plastic relaxation during martensitic transformation in low-carbon steel, a phase-field study.NPJ Comput Mater2025;11:1499

[15] Gong P,Wynne BP,Rainforth WM.The effect of thermomechanical controlled processing on recrystallisation and subsequent deformation-induced ferrite transformation textures in microalloyed steels.J Mater Sci2018;53:6922-38

[16] Zou Y,Hu Z.Austenite stability and its effect on the toughness of a high strength ultra-low carbon medium manganese steel plate.Mater Sci Eng A2016;675:153-63

[17] Li Z,Chu D,Cui Y.A coupled crystal-plasticity and phase-field model for understanding fracture behaviors of single crystal tungsten.Int J Plast2022;157:103375

[18] Zhao P,Wang Y.An integrated full-field model of concurrent plastic deformation and microstructure evolution: application to 3D simulation of dynamic recrystallization in polycrystalline copper.Int J Plast2016;80:38-55

[19] Roy AM,Acar P,Sundararaghavan V.Combining crystal plasticity and phase field model for predicting texture evolution and the influence of nuclei clustering on recrystallization path kinetics in Ti-alloys.Acta Mater2024;266:119645

[20] Cai T,Zhang Z,Huang HH.The application of phase-field method in magnetoelastic functional materials.Microstructures2025;5:2025059

[21] Xu Y,Hu C.Magnetic-field driven domain wall evolution in rhombohedral magnetostrictive single crystals: a phase-field simulation.Microstructures2025;5:2024052

[22] Zhao Y.Understanding and design of metallic alloys guided by phase-field simulations.NPJ Comput Mater2023;9:1038

[23] Wang Y,Liu X.Simulation of microstructure evolution in Mg alloys by phase-field methods: a review.Crystals2022;12:1305

[24] Yang C,Liu B.Phase-field-lattice Boltzmann simulation of dendrite growth under natural convection in multicomponent superalloy solidification.Rare Met2020;39:147-55

[25] Yang C,Su X.Multiphase-field and experimental study of solidification behavior in a nickel-based single crystal superalloy.Acta Mater2019;175:286-96

[26] Oommen V,Goswami S,Karniadakis GE.Learning two-phase microstructure evolution using neural operators and autoencoder architectures.NPJ Comput Mater2022;8:876

[27] Ma B,Huang H,Liu W.Deep learning-based image segmentation for Al-La alloy microscopic images.Symmetry2018;10:107

[28] Kim Y,Jung J.Exploration of optimal microstructure and mechanical properties in continuous microstructure space using a variational autoencoder.Mater Des2021;202:109544

[29] Chowdhury A,Yener B.Image driven machine learning methods for microstructure recognition.Comput Mater Sci2016;123:176-87

[30] Cang R,Yao H,Ren Y.Improving direct physical properties prediction of heterogeneous materials from imaging data via convolutional neural network and a morphology-aware generative model.Comput Mater Sci2018;150:212-21

[31] Long X,Shen Z.Identification of Mechanical Properties of Thin-Film Elastoplastic Materials by Machine Learning.Acta Mech Solida Sin2023;36:13-21

[32] Zhu L,Chen Q.Prediction of ultimate tensile strength of Al-Si alloys based on multimodal fusion learning.Mater Genome Eng Adv2024;2:e26

[33] Kondo R,Masuoka Y,Asahi R.Microstructure recognition using convolutional neural networks for prediction of ionic conductivity in ceramics.Acta Mater2017;141:29-38

[34] Chen L,Du Y.Thermal stability and oxidation resistance of Ti-Al-N coatings.Surf Coat Technol2012;206:2954-60 PMCID:PMC3587463

[35] Biava G,Vaz RF.Evaluation of high temperature corrosion resistance of CrN, AlCrN, and TiAlN arc evaporation PVD coatings deposited on Waspaloy.Surf Coat Technol2022;438:128398

[36] Rachbauer R,Stergar E.Decomposition pathways in age hardening of Ti-Al-N films.J Appl Phys2011;110:023515

[37] Liu Z,Suehiro M.Thermodynamics of interfacial segregation in solute drag.Mater Sci Eng A1998;247:222-8

[38] Zhang T,Zhang L.Pareto optimal driven automation framework for quantitative microstructure simulation towards spinodal decomposition.MRS Commun2023;13:877-84

[39] Chavee L,da Silva Pires M,Haye E.A mechanistic approach of oxidation resistance, structural and mechanical behaviour of TiAlN coatings.Appl Surf Sci2022;586:152851

[40] Chen L,Wang AJ,Zhou SZ.Effect of Al content on microstructure and mechanical properties of Ti-Al-Si-N nanocomposite coatings.Int J Reract Met Hard Mater2009;27:718-21

[41] Sun X,Zhou J.Structure, mechanical and thermal properties of TiAlBN/TiAlN multilayers.Surf Coat Technol2022;441:128512

[42] Peng YH,Xu YX,Du Y.Interface enhanced mechanical and thermal properties of TiSiN/TiAlN multilayers.J Alloys Compd2021;861:158571

[43] Chen L,Wang SQ,Xu H.Mechanical properties and microstructural evolution of TiN coatings alloyed with Al and Si.Mater Sci Eng A2009;502:139-43

[44] Chen Y,Yu C.Enhanced thermal stability and fracture toughness of TiAlN coatings by Cr, Nb and V-alloying.Surf Coat Technol2018;342:85-93

[45] Zou HK,Chang KK,Du Y.Enhanced hardness and age-hardening of TiAlN coatings through Ru-addition.Scr Mater2019;162:382-6

[46] Zhang J,Du JW,Kong Y.Evolution of microstructure, mechanical and thermal properties with varied oxygen contents in TiAlON coatings.Int J Refract Met Hard Mater2023;111:106074

[47] Mendez A,Santiago J.Effect of Al content on the hardness and thermal stability study of AlTiN and AlTiBN coatings deposited by HiPIMS.Surf Coat Technol2021;422:127513

[48] Chaar A,Johansson-jöesaar M.Microstructural influence of the thermal behavior of arc deposited TiAlN coatings with high aluminum content.J Alloys Compd2021;854:157205

[49] Forsén R,Odén M.Decomposition and phase transformation in TiCrAlN thin coatings.J Vac Sci Technol Vac Surf Films2012;30:061506

[50] Chen L,Mayrhofer PH,Li J.The influence of age-hardening on turning and milling performance of Ti-Al-N coated inserts.Surf Coat Technol2008;202:5158-61