Журналы →  CIS Iron and Steel Review →  2023 →  №1 →  Назад

Metal Forming and Tubemaking
Название Use of machine learning methods for determinatuon of the boundary conditions coefficients in a FEM task for the case of accelerated cooling of hot-rolled sheet metal
DOI 10.17580/cisisr.2023.01.10
Автор A. G. Zinyagin
Информация об авторе

Bauman Moscow State Technical University (Moscow, Russia):

A. G. Zinyagin, Cand. Eng., Associate Prof., e-mail: ziniagin_ag@bmstu.ru


The article considers an approach to combine the FEM method with the machine learning method in modeling the process of sheet metal cooling in a laminar cooling unit. A model of rolled products cooling based on the FEM has been developed; it considers the variable properties of the material and phase transformations. The high importance of taking into account physical processes, which occur on the surface of rolled products during cooling, is shown, namely influence of the surface temperature of rolled surface on the heat transfer coefficient. In first iteration data from literature was used for this dependenсe, afterwards it was adapted for the concrete case using iteration method. Especial importance of this phenomenon for calculation of cooling processes for rolled heavy plates (with thickness more than 30 mm) is shown. Two ways to calculate heat dissipation from phase transformations based on the Avrami formula and using the curve of relationship between heat capacity and temperature are given; they are used in a model depending on availability of data for the examined steel grade. Heat transfer coefficient was determined using machine learning methods in order to increase accuracy of calculations. The training set was built on the basis of industrial data, cleared from serial production factor and errors in sensors data signals. Several machine learning models were examined, the model based on gradient boosting of the catboost library displayed the best results. The optimal model parameters were selected using the GridSearchCV method of the Sklearn library or other builtin methods. The most important factors (feature importance) were those that provide especial influence on the heat transfer coefficient - water flow, thickness of rolled products, temperature range of cooling, chemical composition of steel.

The research was carried out within the program of strategic academic leadership of Russian Federation “Prioritet-2030”, which is directed on support of the programs for development of high school educational organizations, as well as within the scientific project “PRIOR/SN/NU/22/SP5/26” “Creation of innovative digital tools for application of artificial intellect and advanced statistical analysis of big data in technological processes of metal production” and also within the framework of scientific collaboration between Vyksa Steel Works and Bauman Moscow State Technical University.

Ключевые слова Rolled plates, cooling, machine learning, heat transfer coefficient, finite element method
Библиографический список

1. Efron L. I. Metal science in the “large” metallurgy. Pipe steels. Moscow: Metallurgizdat. 2012. 696 p.
2. Pogorzjelskiy V. I., Litvinenko D. A., Matrosov Yu. I., Ivanitskiy A. V. Controlled rolling. Moscow: Metallurgiya. 1979. 184 p.
3. Zinyagin A. G. Improvement of rolling and cooling processes for sheets from pipe steel grades at the 5000 rolling mill. Dissertation of Candidate of Technical Sciences. Specialty 05.02.09 “Technologies and machines of metal forming:. Moscow. 2014. 159 p.
4. Zinyagin A. G., Muntin A. V., Ilyinskiy V. I., Nikitin G. S. Mathematical simulation of the process of sheet accelerated colling at the 5000 rolling mill. Problemy chernoy metallurgii i materialovedeniya. 2013. No. 1. pp. 9-15.
5. Ryabchikov M. Yu., Ryabchikova E. S., Shmanev D. E., Kokorin I. D. Management on steel strip cooling in flexible production of galvanized rolled sheet. Izvestiya vysshikh uchebnykh zavedeniy. Chernaya metallurgiya. 2021. Vol. 64. No. 7. pp. 519-529.
6. Poletskov P. P., Alekseev D. Yu., Kuznetsova A. S., Gulin A. E., Emaleeva D. G., Adishchev P. G. Analysis of cooling routes for rolled sheets on the base of computer modeling. Vestnik MGTU im. G. I. Nosova. 2022. No. 4. pp. 102-109.
7. Poletskov P. P., Kuznetsova A. S., Gulin A. E., Emaleeva D. G., Alekseev D. Yu. Study of the heat treatment process of sheet metal in the software package “Deforv-3D”. Norwegian Journal of Development of the International Science. 2022. No. 97. pp. 46-49.
8. Sedykh I. A., Istomin V. A. Neural net modeling of strip cooling process in a hot rolling mill. Vestnik VGU. Series: Sistemnyi analiz i informatsionnye tekhnologii. 2018. No. 2. pp. 116-125. DOI: 10.17308/sait.2019.2/1296
9. Ixin Li, Xianliang Hu, Artificial neural network approximations of Cauchy inverse problem for linear PDEs. Applied Mathematics and Computation. 2022. Vol. 414. 126678. DOI: 10.1016/j.amc.2021.126678
10. Jun Takeuchi, Yukio Kosugi. Neural network representation of finite element method. Neural networks. 1994. Vol. 7. Iss. 2. pp. 389-395. DOI: 10.1016/0893-6080(94)90031-0
11. Pothina H., Nagaraja K. V. Finite Element Method-Based Artificial Neural Network. In: ICT Analysis and Applications. Lecture Notes in Networks and Systems. Springer, Singapore. 2022. Vol. 314. DOI: 10.1007/978-981-16-5655-2_73
12. Zenkevich O., Morgan K. Finite elements and approximations. Moscow: Mir. 1986. 318 p.
13. Balandin M. Yu., Shurina E. P. Methods of solving the SLAE with large dimensionality. Novosibirsk: Izdatelstvo NGTU. 2000. 70 p.
14. George A., Lyu J. Numerical solving of large sparse system of equations. Moscow: Mir. 1984. 333 p.
15. Tolubinskiy V. I. Heat exchange during boiling. Kiev: Naukova dumka. 1980. 316 p.
16. Saeychev V. D., Nevskiy S. A., Ilyashenko A. V. On the mechanism of accelerated cooling in thermal strengthening of rolled product. Izvestiya vysshikh uchebnykh zavedeniy. Chernaya metallurgiya. 2017. Vol. 60. No. 12. pp. 1005-1007. DOI: 10.17073/0368-0797-2017-12-1005-1007
17. Physical constants of some commercial steels at elevated temperatures (based on measurements made at the National Physical Laboratory, Teddington). Butterworths Scienctific Publations. 1953.
18. Sun C. G., Han H. N., Lee J. K. A Finite Element Model for the Prediction of Thermal and Metallurgical Behavior of Strip on Run-out-table in Hot Rolling. ISIJ International. 2002. Vol. 42. No. 4. pp. 392–400.
19. Saha S. K., Kumar A. Coupled Temperature-Microstructure Model for Predicting Temperature Distribution and Phase Transformation in Steel for Arbitrary Cooling Curves. ASME. J. Thermal Sci. Eng. Appl. 2021. June. DOI: 10.1115/1.4048212
20. Zhao H., Hu X., Cui J., Xing Z. Kinetic Model for the Phase Transformation of High-Strength Steel Under Arbitrary Cooling Conditions. Met. Mater. Int. 2019. Vol. 25. No. 2. pp. 381–395. DOI: 10.1007/s12540-018-0196-2
21. Suehiro M., Oda T., Senuma T. Development of mathematical model for predicting transformation of high-carbon steel during cooling on runout table and its application to on-line temperature control of hot strip mill. Nippon Steel Technical Report. 1995. Vol. 67. pp. 49-56.
22. Bamberg M., Prinz B. Determination of heat transfer coefficients during water cooling of metals. Materials Science and Technology. 1986. Vol. 2. No. 4. pp. 410–415.
23. Gan Y. F., Chien C. C., Jang J. Y. et al. Experimental and 3-D Numerical Heat Transfer Analyses of Dual Spray Water Nozzle Impingement on a Flat Plate. Heat Mass Transfer. 2020. Vol. 56. pp. 2397–2412. DOI: 10.1007/s00231-020-02870-5
24. Mukhamediev R., Symagulov A., Kuchin Y., Yakunin K., Yelis M. From Classical Machine Learning to Deep Neural Networks: A Simplified Scientometric Review. Applied Sciences. 2021. Vol. 11. No. 12.

Полный текст статьи Use of machine learning methods for determinatuon of the boundary conditions coefficients in a FEM task for the case of accelerated cooling of hot-rolled sheet metal