A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Compressive Strength Prediction of Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) Using Artificial Neural Networks
Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) is a cementitious composite that contains fibers, leading to superior mechanical properties. This research employs Artificial Neural Network (ANN) techniques to predict the compressive strength of UHPFRC. A comprehensive literature review was undertaken to collect mixed design information from previous experimental studies, resulting in the formation of a database consisting of 200 data points. This study integrated the Multi-Layer Perceptron Neural Network (MLPNN) method, a commonly employed technique in Artificial Neural Networks, for comparable prediction tasks. The developed model consisted of 10 input parameters such as the amount of cement, fine aggregate, superplasticizer, supplementary cementitious materials, fillers, water-to-binder ratio, fiber type, volume of fibers, aspect ratio, and tensile strength of fiber. Only one hidden layer was employed in the model and the optimum number of neurons for this layer was determined by a systematic performance analysis. The database was divided into 80% training and 20% testing data and complimented by k-fold cross-validation to prevent overfitting. The final model exhibited the coefficient of determination (R2) and mean squared error (MSE) values of 0.883 and 85.57 respectively for test data. The predicted and actual compressive strength values showed reasonable agreement with the exception of a few points where minor deviations were observed. Furthermore, a sensitivity analysis of parameters was undertaken to study the influence of each material parameter on the compressive strength of UHPFRC based on the weights assigned by the ANN model. This analysis showed that the water-to-binder ratio has the highest influence followed by fine aggregates and supplementary cementitious materials. Also, it was revealed that the amount of filler materials and the fiber type only had a minor effect on the compressive strength of UHPFRC.
Compressive Strength Prediction of Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) Using Artificial Neural Networks
Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) is a cementitious composite that contains fibers, leading to superior mechanical properties. This research employs Artificial Neural Network (ANN) techniques to predict the compressive strength of UHPFRC. A comprehensive literature review was undertaken to collect mixed design information from previous experimental studies, resulting in the formation of a database consisting of 200 data points. This study integrated the Multi-Layer Perceptron Neural Network (MLPNN) method, a commonly employed technique in Artificial Neural Networks, for comparable prediction tasks. The developed model consisted of 10 input parameters such as the amount of cement, fine aggregate, superplasticizer, supplementary cementitious materials, fillers, water-to-binder ratio, fiber type, volume of fibers, aspect ratio, and tensile strength of fiber. Only one hidden layer was employed in the model and the optimum number of neurons for this layer was determined by a systematic performance analysis. The database was divided into 80% training and 20% testing data and complimented by k-fold cross-validation to prevent overfitting. The final model exhibited the coefficient of determination (R2) and mean squared error (MSE) values of 0.883 and 85.57 respectively for test data. The predicted and actual compressive strength values showed reasonable agreement with the exception of a few points where minor deviations were observed. Furthermore, a sensitivity analysis of parameters was undertaken to study the influence of each material parameter on the compressive strength of UHPFRC based on the weights assigned by the ANN model. This analysis showed that the water-to-binder ratio has the highest influence followed by fine aggregates and supplementary cementitious materials. Also, it was revealed that the amount of filler materials and the fiber type only had a minor effect on the compressive strength of UHPFRC.
Compressive Strength Prediction of Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) Using Artificial Neural Networks
Lecture Notes in Civil Engineering
Dissanayake, Ranjith (editor) / Mendis, Priyan (editor) / De Silva, Sudhira (editor) / Fernando, Shiromal (editor) / Konthesingha, Chaminda (editor) / Attanayake, Upul (editor) / Gajanayake, Pradeep (editor) / Wijesundara, R. S. S. A. (author) / Wijesundara, K. K. (author) / Bandara, N. M. S. H. (author)
International Conference on Sustainable Built Environment ; 2023 ; Kandy, Sri Lanka
Proceedings of the 14th International Conference on Sustainable Built Environment ; Chapter: 13 ; 167-178
2024-08-28
12 pages
Article/Chapter (Book)
Electronic Resource
English
Compressive behavior of ultra-high performance fiber-reinforced concrete (UHPFRC) confined with FRP
| British Library Online Contents | 2018
Compressive behavior of ultra-high performance fiber-reinforced concrete (UHPFRC) confined with FRP
| British Library Online Contents | 2018
French Standards for Ultra-High Performance Fiber-Reinforced Concrete (UHPFRC)
| Springer Verlag | 2017
Shear Strength of Ultra High Performance Fiber Reinforced Concrete (UHPFRC) Precast Bridge Joint
| Springer Verlag | 2012