English
Język Polski
A new ANN rheological model of a comply polymer in temperature spectrum
DOI:
https://doi.org/10.24425/ace.2023.144170Abstract
The article presents modelling using artificial neural networks (ANN) of the phenomenon of creep of comply polymer SIKA PS which can be used in various applications in civil engineering. Data for modelling was gathered in compressive experiments conveyed under a set of fixed conditions of compressive stress and temperature. Part of the data was pre-processed by smoothing and rediscretisation and served as inputs and targets for network training and part of the data was left raw as control set for verification of prognosing capability. Assumed neural network architectures were one- and two-layer feedforward networks with Bayesian regularisation as a learning method. Altogether 55 networks with 8 to 12 neurons in varying structural configurations were trained. Fitting and prognosing verification was performed using mean absolute relative error as a measure; also, results were plotted and assessed visually. In result, the research allowed for formulation of a new rheological model for comply polymer SIKA PS in time, stress and temperature field domain with fitting quality of mean absolute relative error 1.3% and prognosis quality of mean absolute relative error 8.73%. The model was formulated with the use of a two-layer network with 5 + 5 neurons.References
[1] J. Pečnik, et al., “Mechanical performance of timber connections made of thick flexible polyurethane adhesives”, Engineering Structures, vol. 247, no. 13, 2021, DOI: 10.1016/j.engstruct.2021.113125.
[2] A. Kwiecień, P. Krajewski, Ł. Hojdys, M. Tekieli, and M. Słoński, “Flexible Adhesive in Composite-to-Brick Strengthening - Experimental and Numerical Study”, Polymers, vol. 10, no. 4, art. no. 356, 2018, DOI: 10.3390/polym10040356.
[3] A. Kwiecień, K. Rodacki, B. Zając, and T. Kozik, “Mechanical Behavior of Polyurethane Adhesives Applied to Timber Joints in Repair of Historical Timber Structures: An Interdisciplinary Approach” in RILEM Bookseries, vol. 18: Structural Analysis of Historical Constructions, R. Aguilar, et al., Eds. Springer, 2019, pp. 1603-1612, DOI: 10.1007/978-3-319-99441-3_172.
[4] A. M. Stręk, et al., “Highly Dissipative Materials for Damage Protection against Earthquake-Induced Structural Pounding”, Materials, 2021, vol. 14, no. 12, art. no. 3231, 2021, DOI: 10.3390/ma14123231.
[5] B. Zając, ścinane połączenia klejone sztywne i podatne pracujące w podwyższonej temperaturze. Kraków: Wydawnictwo PK, 2018.
[6] Z. Bogusław and K. Arkadiusz, “Thermal Compatibility of Rigid and Flexible Adhesives to Substrates of Historical Structures”, in RILEM Bookseries, vol. 18: Structural Analysis of Historical Constructions, R. Aguilar, et al., Eds. Springer, 2019, pp. 1868-1877, DOI: 10.1007/978-3-319-99441-3_200.
[7] Z. Bogusław, K. Arkadiusz, M. Gams, and T. Tatara, “Strengthening of masonry and concrete structures working in elevated temperatures and mining tremors area”, E3S Web of Conferences, 2019, vol. 106, pp. 1-8, DOI: 10.1051/e3sconf/201910601021.
[8] Z. Waszczyszyn and L. Ziemianski, “Neural networks in the identification analysis of structural mechanics problems”, in Parameter identification of materials and structures, Z. Mróz, G. E. Stavroulakis, Eds. Springer, 2005, pp. 265-340, DOI: 10.1007/3-211-38134-1_7.
[9] I. Floodm and N. Kartam, “Neural network in civil engineering I: Principles and understandings”, ASCE Journal of Computing in Civil Engineering, vol. 8, no. 2, pp. 131-148, 1994, DOI: 10.1061/(ASCE)0887-3801(1994)8:2(131).
[10] I. Flood and N. Kartam, “Neural network in civil engineering II: Systems and application”, ASCE Journal of Computing in Civil Engineering, vol. 8, pp. 149-162, 1994, DOI: 10.1061/(ASCE)0887-3801(1994)8:2(149).
[11] M. Dudzik, A. M. Stręk, “ANN model of stress-strain relationship for aluminium sponge in uniaxial compression”, Journal of Theoretical and Applied Mechanics, vol. 58, no. 2, pp. 385-390, 2020, DOI: 10.15632/jtampl/116804.
[12] A. M. Stręk, M. Dudzik, and T. Machniewicz, “Specifications for modelling of the phenomenon of compression of closed-cell aluminium foams with neural networks”, Materials, 2022, vol. 15, no. 3, 2022, DOI: 10.3390/ma15031262.
[13] M. Dudzik and A. M. Stręk, “ANN architecture specifications for modelling of open-cell aluminum under compression”, Mathematical Problems in Engineering, vol. 2020, art. no. 834317, pp. 1-26, 2020, DOI: 10.1155/2020/2834317.
[14] C. H. Reinsch, “Smoothing by spline functions”, Numerische Mathematik, vol. 10, pp. 177-184, 1967, DOI: 10.1007/BF02162161.
[15] R. Champion, C. T. Lenard, and T. M. Mills, “An introduction to abstract splines”, Mathematical Sciences, vol. 21,pp. 8-26, 1996.
[16] Mathworks documentation, “Csaps”. [Online]. Available: www.mathworks.com/help/curvefit/csaps.html. [Accessed: 21 Nov. 2021].
[17] Mathworks documentation: “Mapminmax”. [Online]. Available: www.mathworks.com/help/deeplearning/ref/mapminmax.html. [Accessed: 21 Feb 2019].
[18] H. Demuth, M. Beale, and M. T. Hagan, Neural Network Toolbox 6 User’s Guide. The MathWorks Inc., 2009.
[19] F. D. Foresee and M. T. Hagan, “Gauss-Newton approximation to Bayesian learning”, Proceedings of the International Joint Conference on Neural Networks (ICNN’97), vol. 3, pp. 1930-1935, 1997, DOI: 10.1109/ICNN.1997.614194.
[20] D. J. C. MacKay, “Bayesian interpolation”, Neural Computation, vol. 4, no. 3, pp. 415-447, 1992, DOI: 10.1162/neco.1992.4.3.415.
[21] Mathworks MATLAB Answers, “Why does the “trainbr” function not require a validation dataset?”. [Online]. Available: https://www.mathworks.com/matlabcentral/answers/405727-why-does-the-trainbr-function-not-require-a-validation-dataset. [Accessed: 28 Apr. 2022].
Downloads
Published
How to Cite
Issue
Section
