ANALYSIS OF NEURAL NETWORK MODELS WITH A FULLY CONNECTED ARCHITECTURE USED IN CALCULATIONS OF PUSHING RESISTANCE IN REINFORCED CONCRETE SLABS
DOI:
https://doi.org/10.36773/1818-1112-2026-139-1-64-78Keywords:
reinforced concrete slabs, punching resistance, neural network modeling, complex stress-strain state, assessment of the accuracy of neural network models, digital modelingAbstract
In reinforced concrete slabs, punching shear occurs in a small area surrounding a column, resulting in a complex stress-strain state involving bending, shear, and torsion deformations, leading to brittle failure in the ultimate limit state. Numerous experimental and theoretical studies conducted to date have resulted in the development of a number of analytical computational models for punching shear resistance, the most accurate of which have been incorporated into current regulatory and technical documents. However, these models lack sufficient accuracy and require improvement. The rapid development of neural network technologies in recent years has enabled the use of machine learning in many engineering calculations, including predicting punching shear resistance. This paper proposes a hypothesis for digital modeling of punching resistance using neural network technologies, which will reduce the number of experimental tests of real slab samples. Neural network models with a fully connected architecture have been developed, and the influence of the number of hidden layers and the number of neurons in a hidden layer on model accuracy has been studied. The influence of the number of samples in the training set on the accuracy of the neural network model was tested. Commonly accepted and widely used mathematical statistics in modern practice were used to evaluate the accuracy of the neural network models: Pearson correlation coefficient (r), determination coefficient (R2), mean-normalized root mean square error CV(RMSE), mean-normalized mean absolute error CV(MAE), as well as the average model error (d) obtained from the error vector (δ), and the coefficient of variation (Vd) of the error vector δ [41]. As a result of the conducted research, despite the relatively high accuracy of the developed neural network models, the digital modeling hypothesis was rejected. Nevertheless, the authors believe that it is possible to create a neural network model with sufficient accuracy for use in digital modeling.
References
Shen, Y. Interpretable Machine Learning Models for Punching Shear Strength Estimation of FRP Reinforced Concrete Slabs / Y. Shen, J. Sun, S. Liang // Crystals. – 2022. – № 259.
The Machine-Learning-Based Prediction of the Punching Shear Capacity of Reinforced Concrete Flat Slabs: An Advanced M5P Model Tree Approach / M. H. Abdallah, Z. A. Thoeny, S. N. Henedy [et al.] // Appl. Sci. – 2023. – Vol. 13. – P. 8325. – DOI: 10.3390/app13148325.
Silva Júnior, F. E. S. Machine learning models to predict the punching shear strength of reinforced concrete flat slabs / F. E. S. Silva Júnior, W. J. S. Gomes // Rev. IBRACON Estrut. Mater. – 2023. – Vol. 16, № 4, Art. 16405. – DOI: 10.1590/S1983-41952023000400005.
Молош, В. В. Сопротивление срезу при продавливании самонапряженных плоских железобетонных элементов без поперечного армирования : дис. ... канд. тех. наук : 01.07.14 / Виктор Викторович Молош. – Брест, 2014. – 226 л.
Kinnunen, S. Punching of concrete slabs without shear reinforcement / S. Kinnunen, H. Nylander. – Stockholm, Sweden : KTH Royal Institute of Technology, 1960. – 112 p.
Broms, C. E. Elimination of flat plate punching failure mode / C. E. Broms // ACI Structural Journal. – 2000. – Vol. 97. – P. 94–101.
Broms, C. E. Concrete flat slabs and footings-design method for punching and detailing for ductility. Ph. D. Thesis / C. E. Broms. – Stockholm, Sweden : KTH Royal Institute of Technology, 2005. – 114 p.
Tian, Y. Strength Evaluation of Interior Slab-Column Connections / Y. Tian, J. O. Jirsa, O. Bayrak // ACI Structural Journal. – 2008. – Vol. 105. – P. 692–700.
Stasio, D. Transfer of Bending Moment between Flat Plate Floor and Column / D. Stasio, M. R. V. Buren // ACI Structural Journal. – 1960. – Vol. 57. – P. 299–314.
Moe, J. Shearing Strength of Reinforced Concrete Slabs and Footings under Concentrated Loads / J. Moe. – Skokie, Illinois : Portland Cement Association, 1961. – 130 p.
GB 50010–2010. Code for design of concrete structures : national standard of the People's Republic of China / MOHURD (Ministry of Housing and Urban-Rural Development). – Beijing : China Architecture & Building Press, 2015. — P. 232–236.
ACI 318-19. Building code requirements for structural concrete and commentary : American Concrete Institute standard / American Concrete Institute. – Farmington Hills, MI : ACI, 2019. – 624 p.
Muttoni, A. Punching shear strength of reinforced concrete slabs without transverse reinforcement / A. Muttoni // ACI Structural Journal. – 2008. – Vol. 105. – P. 440–450.
A Modified Compression Field Theory Based Analytical Model of RC Slab-Column Joint without Punching Shear Reinforcement / L. F. Wu, T. C. Huang, Y. L. Tong, S. X. Liang // Buildings. – 2022. – Vol. 12. – 226 p.
Punching Shear Strengths of RC Slab-Column Connections: Prediction and Reliability / P. Chetchotisak, P. Ruengpim, D. Chetchotsak, S. Yindeesuk // KSCE Journal of Civil Engineering. – 2018. – Vol. 22. – P. 3066–3076.
Mohammed, T. A. Post-Earthquake Fire Punching Shear Behavior of GFRP-Reinforced Slab-Column Connections / T. A. Mohammed, S. Shirtaga // Hindawi Advances in Civil Engineering. – Vol. 23 (February 2023). – Article ID 7596032. – 25 p. – DOI: 10.1155/2023/7596032.
Geetha, N. K. Overview of machine learning and its adaptability in mechanical engineering / N. K. Geetha, P. Bridjesh // Materials Today. – 2020. – Vol. 4. – P. 395–399.
Machine-learning interpretability techniques for seismic performance assessment of infrastructure systems / S. Mangalathu, K. Karthikeyan, D. C. Feng, J. S. Jeon // Engineering Structures. – 2022. – Vol. 250. – Art. 112883.
Development of data‐driven prediction model for CFRP‐steel bond strength by implementing ensemble learning algorithms / S. Z. Chen, D. C. Feng, W. S. Han, G. Wu // Construction and Building Materials. – 2021. – Vol. 303. – Art. 124470.
Implementing ensemble learning methods to predict the shear strength of RC deep beams with/without web reinforcements / D. C. Feng, W. J. Wang, S. Mangalathu [et al.] // Engineering Structures. – Vol. 2021. – Vol. 235. – Art. 111979.
A probabilistic bond strength model for corroded reinforced concrete based on weighted averaging of non‐fine‐tuned machine learning models / B. Fu, S. Z. Chen, X. R. Liu, D. C. Feng // Construction and Building Materials. – 2022 – Vol. 318. – Art. 125767.
Nasiri, S. Machine learning in predicting mechanical behavior of additively manufactured parts / S. Nasiri, M. R. Khosravani // Journal of Materials Research and Technology. – 2021. – Vol. 14. – P. 1137–1153.
Hoang, A. N. D. Estimating punching shear capacity of steel fibre reinforced concrete slabs using sequential piecewise multiple linear regression and artificial neural network / A. N. D. Hoang // Measurement. – 2019. – Vol. 137. – P. 58–70.
Nguyen, H. D. Development of extreme gradient boosting model for prediction of punching shear resistance of r/c interior slabs / H. D. Nguyen, G. T. Truong, M. Shin // Engineering Structures. – 2021. – Vol. 235. – Art. 112067.
Explainable machine learning models for punching shear strength estimation of flat slabs without transverse reinforcement / S. Mangalathu, H. Shin, E. Choi, J. S. Jeon // Journal of Building Engineering. – 2021. – Vol. 39. – Art. 102300.
Truong, G. T. Assessment of punching shear strength of FRP-RC slab-column connections using machine learning algorithms / G. T. Truong, H. J. Hwang, C. S. Kim // Engineering Structures. –2022. – Vol. 255. – Art. 113898.
A machine learning framework for predicting the shear strength of carbon nanotube‐polymer interfaces based on molecular dynamics simulation data / A. Rahman, P. Deshpande, M. S. Radue [et al.] // Composites Science and Technology. – 2021. – Vol. 207. – Art. 108627.
Ilawe, N. V. Breaking Badly: DFT‐D2 Gives Sizeable Errors for Tensile Strengths in Palladium‐Hydride Solids / N. V. Ilawe, J. A. Zimmerman, B. M. J. Wong // Journal of Chemical Theory and Computation. – 2015. – Vol. 11. – P. 5426–5435.
Compressive Strength Prediction of High-Strength Concrete Using Long Short-Term Memory and Machine Learning Algorithms / H. G. Chen, X. Li, Y. Q. Wu [et al.] // Buildings. – 2022. – Vol. 12. – Art. 302.
Data-Driven Compressive Strength Prediction of Fly Ash Concrete Using Ensemble Learner Algorithms / M. S. Barkhordari, D. J. Armaghani, A. S. Mohammed, D. V. Ulrikh // Buildings. – 2022. – Vol. 12. – Art. 132.
Shear Strength Prediction of Slender Steel Fiber Reinforced Concrete Beams Using a Gradient Boosting Regression Tree Method / A. Shatnawi, H. M. Alkassar, N. M. Al-Abdaly [et al.] // Buildings. – 2022. – Vol. 12. – Art. 550.
Jiang, Y. M. Compressive Strength Prediction of Fly Ash Concrete Using Machine Learning Techniques / Y. M. Jiang, H. Y. Li, Y. S. Zhou // Buildings. – 2022. – Vol. 12. – Art. 690.
Shen, Y. X. Explainable machine learning-based model for failure mode identification of RC flat slabs without transverse reinforcement / Y. X. Shen, L. F. Wu, S. X. Liang // Engineering Failure Analysis. – 2022. – Vol. 141. – Art. 106647.
Fu, B. A machine learning-based time-dependent shear strength model for corroded reinforced concrete beams / B. Fu, D. C. Feng // Journal of Building Engineering. – 2021. – Vol. 36. – Art. 102118.
Головко, В.А. Нейросетевые технологии обработки данных / В. А. Головко, В. В. Краснопрошин. – Минск : БГУ, 2017. – 264 с.
Hou, R. Prediction of the shear capacity of ultrahigh-performance concrete beams using neural network and genetic algorithm / R. Hou, Q. Hou // Scientific Reports. – 2023. – https://www.nature.com/ articles/s41598-023-29342-0 (дата обращения: 17.01.2026).
Shen, L. Reliability Analysis of RC Slab-Column Joints under Punching Shear Load Using a Machine Learning-Based Surrogate Model / L. Shen, Y. Shen, Sh. Liang // Buildings. – 2022. – Vol. 12. – Art. 1750. – DOI: 10.3390/buildings12101750.
Xu, H. H. Dependent Evidence Combination Based on Shearman Coefficient and Pearson Coefficient / H. H. Xu, Y. Deng // IEEE Access. – 2018. – Vol. 6. – P. 11634–11640.
Abellan-Garcia, J. Properties prediction of environmentally friendly ultra-high-performance concrete using artificial neural networks / J. Abellan-Garcia, F. Gomez, N. Torres Castellanos // European Journal of Environmental and Civil Engineering. – 2022. – Vol. 26, № 6. – P. 2319–2343.
Du, G. Prediction of the compressive strength of high-performance self-compacting concrete by an ultrasonicrebound method based on a GA-BP neural network / G. Du, L. Bu, Q. Hou // PLoS ONE. – 2021. – Vol. 16, № 5. – Art. e0250795. – DOI: 10.1371/journal.pone.0250795.
Основы проектирования строительных конструкций = Асновы праектавання будауничых канструкцый : СН 2.01.01. – Введ. 16.19.2019. – Минск : РУП «Стройтехнорм» ; Мин. арх. и стр.-ва Респ. Беларусь, 2020. – 89 с.
Hinkle, D. E. Applied statistics for the behavioral sciences / D. E. Hinkle, W. Wiersma, S. G. Jurs. – 5th ed. – Boston ; New York : Houghton Mifflin Company, 2003. – 756 p.
Статистика : учеб. для студ. учреждений сред. проф. образования / В.С. Михтарян, В. Г. Минашкин, Р. А. Шмойлова, Н. А. Садовникова ; под ред. В. С. Михтаряна. – 12-е изд. перераб. и доп. – М : Академия, 2013. – 304 с.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The work is provided under the terms of Creative Commons public license Attribution-NonCommercial 4.0 International (CC BY-NC 4.0). This license allows an unlimited number of persons to reproduce and share the Licensed Material in all media and formats. Any use of the Licensed Material shall contain an identification of its Creator(s) and must be for non-commercial purposes only. Users may not prevent other individuals from taking any actions allowed by the license.
