Abstract




 
   

IJE TRANSACTIONS C: Aspects Vol. 31, No. 6 (June 2018) 1149-1157   

downloaded Downloaded: 0   viewed Viewed: 108

  INVESTIGATION OF MECHANICAL PROPERTIES OF SCC-POLYMERIC CONCRETE WITH BACKPROPAGATION NETWORK
 
Ali Heidari and M. Hashmpour
 
( Received: November 05, 2017 – Accepted: February 07, 2018 )
 
 

Abstract    Acrylic polymer with high stability against chemical materials is a good choice in producing of concrete which is subjected to chemical attack. In this study, self-compacting concrete (SCC) made of acrylic polymer, Nano silica and micro silica has been investigated. The results of the experimental tests showed that the addition of microsilica and polymer content, decreases tensile, compressive, and bending strength. The addition of nanosilica and an increase in polymer content, increases the bending strength of concrete and decreases the tensile and compressive strength. Due to the fact that in the laboratory, number of samples are limited and the amount of variation is small, the proper results cannot be achieved. With the help of neural networks, estimating any amount within the range of input data is possible. In this paper, in addition to the experimental results, the performance of the neural network of backpropagation on the strength of self-compacting polymeric concrete has been dealt with. The results of the neural network showed that using the mean squared normalized error function, resilient backpropagation training function, a tangent-sigmoid and log sigmoid transfer functions, and 5 neurons in each of the hidden layers in a two-layer backpropagation neural network (BNN), have good results in which the regression is 0.95 and the error is 0.17.

 

Keywords    Backpropagation neural network, Polymeric concrete, self-compacting concrete, Acrylic Polymer

 

چکیده    پلیمر اکریلیک به­دلیل مقاومت بالا در برابر مواد شیمیایی گزینه مناسبی در ساخت بتن­های مقاوم در برابر حملات شیمیایی است. در این مطالعه به بررسی بتن پلیمری خودمتراکم ساخته شده با پلیمر اکریلیک، نانوسیلیس و میکروسیلیس پرداخته شده­است. نتایج حاصل از آزمایش نمونه­ها نشان داده که افزودن میکروسیلیس و افزایش میزان پلیمر، مقاومت کششی، فشاری و خمشی را کاهش می­دهد. افزودن نانوسیلیس مقاومت خمشی بتن را افزایش و مقاومت کششی و فشاری با افزایش پلیمر کاهش یافته است. این مقدار کاهش به نسبت سایر طرح­ها کمتر است. به­علت آن­که در آزمایشگاه ساخت نمونه­ها محدود و میزان تغییرات مقادیر اندک است نمی­توان به نتیجه مناسبی دست یافت. به کمک شبکه­های عصبی هر مقداری را که در محدوده داده­های ورودی باشد را می­توان تخمین زد. در این مقاله علاوه بر نتایج آزمایشگاهی، عملکرد شبکه عصبی انتشار برگشتی بر تخمین مقاومت بتن خودمتراکم پلیمری پرداخته شده­است. نتایج شبکه عصبی نشان داده که یک شبکه عصبی انتشار برگشتی دو لایه که در آن از تابع خطای استاندارد میانگین، تابع آموزش انعطاف­پذیر، تابع تحریک لوگ سیگموئید و تانژانت سیگموئید، و 5 نرون در هر یک از لایه­های پنهان استفاده شود، نتایج مناسبی حاصل شده که در آن رگرسیون 95/0 و مقدار خطا 17/0 است.

References    1. Giustozzi, F., “Polymer-modified pervious concrete for durable and sustainable transportation infrastructures”, Construction and Building Materials, 111, (2016), 502-512. 2. Hong, S., Kim, H. and Park, S.-K., Optimal mix and freeze-thaw durability of polysulfide polymer concrete. Construction and Building Materials, 127, (2016), 539-545. 3. Shen, D., Wang, T., Chen, Y., Wang, M. and Jiang, G., Effect of internal curing with super absorbent polymers on the relative humidity of early-age concrete. Construction and Building Materials, 99, (2015), 246-253. 4. Heidari, A. and Zabihi, M., Self-Compacting Concrete Incorporating Micro-SiO2 and Acrylic Polymer. Advances in Civil Engineering, 2014, (2014), 1-6. 5. Heidari, A., Hashempour, M. and Tavakoli, D.. Using of Backpropagation Neural Network in Estimating of Compressive Strength of Waste Concrete. Soft Computing in Civil Engineering, Vol. 1, No. 1, (2017), 54-64. 6. Kalantari, Z. and Razzaghi, M.S., “Predicting the buckling capacity of steel cylindrical shells with rectangular stringers under axial loading by using artificial neural network”, International journal of engineering, Transaction B: Applications, Vol. 28, No. 8, (2010), 1154-1159.7. Heidari, A. and Karimpor, M., Earth pressures and design of narrow MSE walls by neural network. Paper presented at the 3rd International Conference on Bridge, Tehran. (2008).8. Salajegheh, E. and Heidari, A., Optimum design of structures against earthquake by adaptive genetic algorithm using wavelet networks. Structural and Multidisciplinary Optimization, Vol. 28, No. 4, (2004), 277-285.9. Salajegheh, E. and Heidari, A., Optimum design of structures against earthquake by discrete wavelet neural network. Paper presented at the 7th International Conference on Computational Structures Technology, Lisbon, Portugal. (2004).10. Taylor, J. G. and Mannion, C. L. T., Theory and Application of Neural Networks, SpringerVerlag, (1992).11. Freeman, J., Simulating Neural Networks. New York, Addison-Wesley Publishing Company, (1994).12. Salajegheh, E. and Heidari, A., Optimum design of structures against earthquake by wavelet neural network and filter banks, Earthquake Engineering & Structural Dynamics, Vol. 34, No. 1, (2005), 67-8213. Dogan, E., Reference evapotranspiration estimation using adaptive neuro-fuzzy inference system, Journal of Irrigation and Drainage, Vol. 58, No. 5, 2008, 617-628. 14. Heidari, A. and Heidari, M., Seismic Earth Pressure on Retaining Wall by Artificial Neural Network, Mathematics Applied in Science and Technology, 4, (2012), 11-19. 15. Heidari, A., Tavakoli, D. and Fakharian, P. Approximation of special sheet values using artificial neural network. Quarterly Journal of Engineering Modeling, Vol. 11, No. 35, (1392), 46-62. (In Farsi). 16. Kamalloo, A., Ganjkhanlou, Y., Aboutalebi, H and Nourani, H., “Modeling of compressive strength of metakaolin based geopolymers by the use of artificial neural network”, International journal of engineering, Transaction B: Basics, Vol. 23, No. 2, (2010), 145-152.





International Journal of Engineering
E-mail: office@ije.ir
Web Site: http://www.ije.ir