Predicting mechanical properties and structural stability of metallic systems using artificial intelligence and fractal analysis

This paper presents an integrated methodology for predicting mechanical properties and assessing structural stability of metallic systems through the synergistic application of fractal analysis and artificial intelligence. Leveraging an extensive experimental dataset comprising microstructural images, morphological characteristics, and mechanical properties of carbon steels — including grades 20, U8, U10A, and others — acquired via optical and electron microscopy, the authors developed specialized software for automated computation of the fractal dimension Df from digital micrographs. The results demonstrate that Df exhibits strong correlation with key material parameters such as carbon content, heat treatment condition, average grain size, and degree of structural heterogeneity. Consequently, Df serves as a sensitive quantitative indicator of phase transformations and degradation processes. A neural network was trained on this dataset to predict, from a given microstructure image, the fractal dimension, average grain size, and shear modulus. Notably, structurally stable states — such as those achieved after tempering — consistently correspond to lower Df values compared to metastable conditions like as-quenched martensite. The proposed approach holds significant promise for applications in digital materials science, including automated quality control, detection of decarburized layers, and the design of advanced metallic materials with tailored properties.
EDN: VDUDBT, https://elibrary/vdudbt

References

[1] Gulyaev A.P. Metallovedenie [Metal science]. Moscow, Metallurgiya Publ., 1986. 544 p. (In Russ.).

[2] Potapov A.A. Noveyshie metody obrabotki izobrazheniy [Newest methods of image processing]. Moscow, Fizmatlit Publ., 2008. 496 p. (In Russ.).

[3] Welstead S. Fractal and wavelet image compression techniques. SPIE Press, 1999. 232 p. (Russ. ed.: Fraktaly i veyvlety dlya szhatiya izobrazheniy v deystvii. Moscow, Izd-vo Triumf, 2003. 320 p.)

[4] Mandelbrot B. The fractal geometry of nature. ?Times Books, 1982.? 468 p. (Russ. ed.: Fraktalnaya geometriya prirody. Moscow, IKI Publ., 2002. 656 p.)

[5] Bondar A.S., Streltsova O.V., Chmilenko F.V. [A study on predicting steel properties using artificial neural networks]. Nauka i innovatsii v tekhnicheskikh universitetakh. Mat. 13 Vseros. foruma studentov, aspirantov i molodykh uchenykh [Science and Innovation in Technical Universities. Proc. 13th Russ. Forum of Students, Graduate Students, and Young Scientists]. Sankt-Peterburg, SPbPU Publ., 2019, pp. 5–7. (In Russ.).

[6] Emelyanov V.A., Feklin G.V., Emelyanova N.Yu. Application of artificial intelligence technologies to assess the quality of metals in a metallurgical production. Neyrokompyutery: razrabotka, primenenie [Neurocomputers], 2022, vol. 24, no. 5, pp. 28–35, doi: https://doi.org/10.18127/j19998554-202205-03 (In Russ.).

[7] Kovun V.A., Kashirina I.L. Usage of u-net and w-net neural network architectures for steel samples metallographic analysis. Vestnik VGU. Ser. Sistemnyy analiz i informatsionnye tekhnologii [Proceedings of Voronezh State University. Ser. Systems Analysis and Information Technologies], 2022, no. 1, pp. 101?110, doi: https://doi.org/10.17308/sait.2022.1/9205 (in Russ.).

[8] Soboleva N.N. Application of neural networks for metal microstructure analysis. Perspektivnye materialy, 2022, no. 5, pp. 34–41. (In Russ.).

[9] Gorbenko O.M., Felshtyn M.L., Usachev D.Yu. et al. Automatic recognition of the structures of ferritic-pearlitic steels on the images, obtained by the method of atomic-force microscopy. Nauchnoe priborostroenie, 2012, vol. 22, no. 4, pp. 62–71. (In Russ.).

[10] Ivanova V.S., Balankin A.S., Bunin I.Zh. et al. Sinergetika i fraktaly v materialovedenii [Synergetics and fractals in materials science]. Moscow, Nauka Publ., 1994. 383 p. (In Russ.).

[11] Kabaldin Yu.G. Vvedenie v tsifrovoe materialovedenie [Introduction to digital materials science]. Nizhniy Novgorod, NGTU im. R.E. Alekseeva Publ., 2025. 226 p. (In Russ.).

[12] Dai Q., Liu X., Ma X. et al. Research status and prospects of the fractal analy-sis of metal material surfaces and interfaces. Int J Miner Metall Mater., 2025, vol. 32, pp. 20–38, https://doi.org/10.1007/s12613-024-2961-1

[13] GOST 34473–2018. Stali uglerodistye i legirovannye. Metody opredeleniya mikrostruktury [Pipeline valves. Steel all-welded ball valves for water heat networks. General specifications]. Moscow, Standartinform Publ., 2018. 24 p. (In Russ.).

[14] GOST 8233–56. Stal. Etalony mikrostruktury [Steel. Microstructure standards]. Moscow, Izd-vo standartov, 2004. 13 p. (In Russ.).

[15] Vadavadagi B.H., Bhujle H.V., Khatirkar R.K., Shekhawat S.K. Microstructural correlation with formability aspects of low carbon steels during deep drawing operations. Materials Characterization, 2021, vol. 178, art. 111267, doi: 10.1016/j.matchar.2021.111267