Andreev S.M., Kolokoltsev M.V., Savinov D.A., Dubrovin V.K., Petrochenko E.V. The Usage of Artificial Neural Networks for Predicting Strength Properties of Sand-Clay Mixtures for Casting Molds

DOI: 10.18503/1995-2732-2023-21-4-23-33

Abstract

This research is mainly aimed at studying the usage of artificial neural network (ANN) models for testing and predicting strength properties of dry and wet sandclay mixtures used for molding in manufacturing the cast products. Deformation resistance of the mold was used as the main parameter expressing strength properties of sandclay mixtures. A perceptron with multiple hidden layers was used as the chosen ANN structure, and a supervised method using experimental data was applied as the training method. Synaptic weights were determined using a basic error backpropagation algorithm. Training time and the resulting accuracy of ANN were studied to predict the strength parameter of sandclay mixtures depending on the mixture composition, moisture and temperature. The result of the study is a formed structure of the perceptron of the neural network model and determined values of synaptic weights. A computer program, using the developed neural network model, predicted the quantitative effect of individual components of wet and dry sandclay mixtures on their ultimate compressive strength expressed with deformation resistance. The obtained parameter prediction results were compared with the experimental results received by other researchers. It has been established that the calculated predicted data regarding the effect of a mixture composition on deformation resistance are not in conflict with the experimental data. The studies conducted with the application of the developed software contributed to the assessment of strength of molding sands, having compositions that have not been studied yet. This will expand ranges of assessment of potential hot cracking in castings.

Keywords

sandclay mold, artificial neural network, molding sand, component composition, mold compliance, stress state modeling, neural network model

For citation

Andreev S.M., Kolokoltsev M.V., Savinov D.A., Dubrovin V.K., Petrochenko E.V. The Usage of Artificial Neural Networks for Predicting Strength Properties of Sand-Clay Mixtures for Casting Molds. Vestnik Magnitogorskogo Gosu-darstvennogo Tekhnicheskogo Universiteta im. G.I. Nosova [Vestnik of Nosov Magnitogorsk State Technical University]. 2023, vol. 21, no. 4, pp. 23-33. https://doi.org/10.18503/1995-2732-2023-21-4-23-33

Sergey M. Andreev – DrSc (Eng.), Head of the Department of Automated Control Systems, Nosov Magnitogorsk State Technical University, Magnitogorsk, Russia. Email: This email address is being protected from spambots. You need JavaScript enabled to view it.. ORCID 0000-0003-0735-6723

Mikhail V. Kolokoltsev – student of NMSTU Project School, Nosov Magnitogorsk State Technical University, Magnitogorsk, Russia. Email: This email address is being protected from spambots. You need JavaScript enabled to view it..

Dmitry A. Savinov – student of Secondary School No. 5, Magnitogorsk, Russia. Email: This email address is being protected from spambots. You need JavaScript enabled to view it..

Vitaly K. Dubrovin – DrSc (Eng.), Associate Professor of the Department of Pyrometallurgy and Foundry Technologies, South Ural State University, Chelyabinsk, Russia. Email: This email address is being protected from spambots. You need JavaScript enabled to view it..

Elena V. Petrochenko – DrSc (Eng.), Professor of the Department of Foundry Processes and Materials Science, Nosov Magnitogorsk State Technical University, Magnitogorsk, Russia.

1. Bekhalov V.N. Mechanical properties of the casting mold during compaction by pressing. Liteinoe proizvodstvo [Foundry Production]. 1965;(4):18-21. (In Russ.)

2. Smelyakov N.N., Kosarikov N.F. Ispravlenie porokov otlivok [Correction of defects in castings]. Moscow: Mashgiz, 1950, 120 p. (In Russ.)

3. Savinov A.S., Tuboltseva A.S. Calculation of the movement of the moisture front edge in a wet sand-clay mold. Vestnik Magnitogorskogo gosudarstven-nogo universiteta im. G.I. Nosova [Vestnik of Nosov Magnitogorsk State Technical University]. 2010;(2):30-33. (In Russ.)

4. Savinov A.S. Metodologicheskie osnovy otsenki i prognozirovaniya napryazhenno-deformirovannogo sostoyaniya sistemy «otlivka – forma» dlya preduprezhdeniya obrazovaniya goryachikh treshchin v fasonnykh otlivkakh: dis. … dokt. tekhn. nauk [A methodological framework for assessing and predicting the stress and strain state of the “casting – mold” system to prevent hot cracking in shaped castings. Doctoral dissertation]. Magnitogorsk, 2016. 382 p.

5. Evstigneev A.I., Odinokov V.I., Petrov V.V., Dmitriev E.A. Matematicheskoe modelirovanie napryazhenno-deformiruemogo sostoyaniya obolochkovykh form pri lite po vyplavlyaemym modelyam: monografiya [Mathematical modeling of the stress and strain state of shell molds during investment casting: monograph]. Vladivostok: Dalnauka, 2009, 130 p. (In Russ.)

6. Velikanov G.F., Primak I.N., Brechko A.A. Strength of molding sands. Liteinoe proizvodstvo [Foundry Production]. 1986;(3):10-13. (In Russ.)

7. Mutilov V.P. Elastic properties of molding sands. Liteinoe proizvodstvo [Foundry Production]. 1974;(1):30-31. (In Russ.)

8. Matveenko I.V., Ivanov E.I., Rezchikov E.A. Rheo-logical tests of molding sands. Formovochnye smesi [Molding Sands]. 1977;(12):11-14. (In Russ.)

9. Matveenko I.V., Ivanov E.I., Rezchikov E.A. Defor-mation properties of molding sands under complex stress state. Liteinoe proizvodstvo [Foundry Produc-tion]. 1977;(10):17-18. (In Russ.)

10. Karpenko V.M., Filipenko E.V. Development of an experimental device for testing the deformation prop-erties of the molding sand. Vestnik Natsionalnogo tekhnicheskogo universiteta Ukrainy "Kievsky politekhnichesky institut. Seriya: Mashinostroenie [Bulletin of National Technical University of Ukraine Kyiv Polytechnic Institute. Series: Mechanical Engi-neering]. 2010;(58):129-133. (In Russ.)

11. Avdokushin V.P., Doroshenko S.P., Surguchev E.A. et al. Modeling of rheological properties of molding sands. Liteinoe proizvodstvo [Foundry Production]. 1979;(8):13-14. (In Russ.)

12. Kolokoltsev V.M., Savinov A.S., Feoktistov N.A., Postnikova A.S. Change in ductility of a wet sand-clay mold under influence of geometric and thermophysical properties of the casting. Izvestiya vysshikh uchebnykh zavedeniy. Chernaya metallurgija [Izvestiya. Ferrous Metallurgy]. 2020;63(10):829-835. (In Russ.) DOI: 10.17073/0368-0797-2020-10-829-835

13. Balandin G.F., Stepanov Yu.A. On the force interaction of a hardening casting and a mold. Liteinoe proizvodstvo [Foundry Production]. 1962;(4):37-41. (In Russ.)

14. Aksenov P.N., Trukhov A.P. A device for testing wet molding sands for hot compliance. Liteinoe proizvod-stvo [Foundry Production]. 1979;(3):20-21. (In Russ.)

15. Lin C.-J., Wu N.-J. An ANN model for predicting the compressive strength of concrete. Applied Sciences. 2021;11(9):3798.

16. Bilim C., Atiş C.D., Tanyildizi H., Karahan O. Predicting the compressive strength of ground granulated blast furnace slag concrete using artificial neural net-work. Advances in Engineering Software. 2009;40(5):334-340.

17. Seyhan A., Tayfur G., Karakurt M., Tanog M. Artificial neural network (ANN) prediction of compressive strength of VARTM processed polymer composites. Computational Materials Science. 2005;1(34):99-105.

18. Savinov A.S., Tuboltseva A.S., Sinitsky E.V. Study on the deformation capacity of sand-clay mixtures. Vestnik Magnitogorskogo gosudarstvennogo universiteta im. G.I. Nosova [Vestnik of Nosov Magnitogorsk State Technical University]. 2011;(1):29-32. (In Russ.)

19. Savinov A.S., Tuboltseva A.S. Calculation of the compliance of a wet sand-clay mold. Innovatsionnye tekhnologii v obespechenii kachestva, energoeffek-tivnosti i ekologicheskoy bezopasnosti. Povyshenie konkurentosposobnosti metallurgicheskikh i mashi-nostroitelnykh predpriyatii v sovremennykh usloviyakh [Innovative Technologies in Ensuring Quality, Energy Efficiency and Environmental Safety. Increasing Competitiveness of Metallurgical and Machine-Building Enterprises in Current Conditions]. Magnitogorsk: Publishing House of Nosov Magnitogorsk State Technical University, 2010, pp. 176-179. (In Russ.)

20. Savinov A.S., Tuboltseva A.S. Simulation of deformation ability of dry sandclay molds. Progressivnye tekhnologii v sovremennom mashinostroenii [Progres-sive Technologies in Modern Mechanical Engineering: Collected Papers of the 6th International Scientific and Technical Conference]. Penza: Privolzhsky dom znaniy, 2010, pp. 66-70. (In Russ.)

21. Savinov A.S., Dziuba V.P., Frolushkina K.A., Baibatyrov R.M. Modelirovanie napryazhennogo sostoyaniya syrykh peschano-glinistykh liteynykh form [Modeling of the stress state of wet sand-clay casting molds]. Magnitogorsk: Nosov Magnitogorsk State Technical University, 2012, 10 p. (In Russ.)

22. Savinov A.S., Tuboltseva A.S., Dziuba V.P., Frolushkina K.A. Issledovanie fizicheskikh svoystv sukhikh peschano-glinistykh liteinykh form [Study on the physical properties of dry sand-clay casting molds]. Magnitogorsk: Nosov Magnitogorsk State Technical University, 2012, 41 p. (In Russ.)

23. Parsunkin B.N., Andreev S.M., Obukhova T.G. et al. Adaptive statistical models synthesized on the basis of artificial neural networks. Vestnik Magnitogorskogo gosudarstvennogo universiteta im. G.I. Nosova [Vestnik of Nosov Magnitogorsk State Technical University]. 2012;4(40):68-71. (In Russ.)

24. Haykin S. Neironnye seti: polny kurs [Neural net-works: A comprehensive foundation]. Moscow: Wil-liams Publishing House, 2006, 1104 p. (In Russ.)