Neural network approximation of deformation curves under uniaxial tension of steel and silumin specimens

The purpose of the study is developing and testing of the new computational technique for approximation of deformation curves of steel and silumin specimens under uniaxial tension. A scheme of testing steel and silumin specimens for uniaxial tensile is presented. The experiment was carried out in the mechanical testing laboratory of the Department of Applied Mathematics and Mechanics of the Voronezh S_tate Technical University. The experimental deformation curve of a steel specimen was approximated by P. Ludwig’s equation. Prediction of the true stress from the logarithmic strain using a pretrained artificial neural network with a multilayer perceptron architecture is discussed. The neural network model was trained using the RProp (resilient backpropagation) method. The software implementation of the neural network approximation was carried out in a framework of the open source for data analysis — Knime Analytics Platform. A scheme for the implementation of a multilayer perceptron that solves the approximation problem is considered. The simulation results are compared by the values of the mean squared error (MSE) of the approximation. The neural network approximation is turned out to be an order of magnitude more accurate for the steel specimen than the approximation by the P. Ludwig equation. The neural network approximation provided even a smaller MSE value for a silumin specimen than that or a steel specimen. It is revealed that changing the architecture of an artificial neural network affects the quality of modeling. With an increase in the number of hidden layers, the accuracy of the approximation increases. Neural network approximation is an effective approach to solving the problem of the analytical description of experimental deformation curves and leaves the possibility of using a universal technique for a variety of materials and different types of tests.

1. Smirnov A. S., Konovalov A. V., Kanakin V. S. Neural network modeling of the rheology of the AlMg6 alloy under the dispersoid barrier effect and the inhibition of dynamic relaxation processes / Diagnostics, Resource and Mechanics of materials and structures. 2020. Issue 6. P. 10 – 26 [in Russian]. DOI: 10.17804/2410-9908.2020.6.010-026

2. Vasilyev A. N., Kuznetsov E. B., Leonov S. S. Neural network method of identification and analysis of the model of deformation of metal structures under creep conditions / Sovr. Inf. Tekhnol. IT-obrazov. 2015. N 11. Vol. 2. P. 360 – 369 [in Russian].

3. Chatzidakis S., Alamaniotis M., Tsoukalas L. H. Creep Rupture Forecasting: A Machine Learning Approach to Useful Life Estimation / International Journal of Monitoring and Surveillance Technologies Research. 2014. N 2. P. 1 – 25. DOI: 10.4018/ijmstr.2014040101

4. Wang H., Zhang W., Sun F. and Zhang W. A comparison study of machine learning based algorithms for fatigue crack growth calculation / Materials. 2017. N 10. P. 543. DOI: 10.3390/ma10050543

5. Rovinelli A., Sangid M. D., Proudhon H. and Ludwig W. Using machine learning and a data-driven approach to identify the small fatigue crack driving force in polycrystalline materials / Computational Materials. 2018. N 4. Article number 35. DOI: 10.1038/s41524-018-0094-7

6. Liu H., Liu S., Liu Z., et al. Prognostics of damage growth in composite materials using machine learning techniques / IEEE International Conference on Industrial Technology. 2017. P. 1042 – 1047. DOI: 10.1109/ICIT.2017.7915505

7. Hanief M., Wani M. F. Artificial neural network and regression-based models for prediction of surface roughness during turning of red brass (C23000) / Journal of Mechanical Engineering and Sciences (JMES). 2016. Vol. 10. Issue 1. P. 1835 – 1845. DOI: 10.15282/jmes.10.1.2016.8.0176

8. Shevchenko E. C., Teraud W. V. Modelling of mechanical characteristics of deformable flat specimens under creep by machine learning methods / Journal of Mechanical Engineering and Sciences (JMES). 2021. Vol. 14. Issue 1. P. 6393 – 6405.

9. Khlivnenko L. V., Pyatakovich F. A. Practice of neural network modeling: a textbook for universities. 2nd edition. — St. Petersburg: Lan’, 2021. — 200 p. [in Russian]

10. Nielsen E. Practical analysis of time series: forecasting with statistics and machine learning. — St. Petersburg: Dialektika, 2021 — 544 p. [in Russian].