NUMERICAL ANALYSIS FOR PREDICTING OF SELF COMPACTING CONCRETE MIXTURES USING ARTIFICIAL NEURAL NETWORKS | ||||
| JES. Journal of Engineering Sciences | ||||
| Article 1, Volume 40, No 6, November and December 2012, Page 1575-1597 PDF (326.36 K) | ||||
| Document Type: Research Paper | ||||
| DOI: 10.21608/jesaun.2012.114530 | ||||
 View on SCiNiTO | ||||
| Author | ||||
| Ashraf M. Heniegal | ||||
| Assistant prof., Civil Strut. Dept., Faculty of Industrial Education, Suez Canal University, Suez, Egypt. | ||||
| Abstract | ||||
| Application of artificial neural networks (ANNs) model to design the mix component of self compacting concrete (SCC) with desirable properties, compressive strength and slump flow, is described in this research Artificial Neural Networks (ANNs) have recently been introduced as an efficient artificial intelligence modeling technique for applications involving a large number of variables, especially with highly nonlinear and complex interactions among input/output variables in a system without any prior knowledge about the nature of these interactions. Various types of ANN models are developed and used for different problems. In this paper, an artificial neural network of the feed-forward back-propagation type has been applied for the prediction of self compacting concrete mixtures. The main targets are SCC components and the inputs interred are compressive strength and slump flow. Due to the complex non-linear effect of compressive strength and slump flow properties on the SCC components, the ANN model is used to predict the components of SCC parameters (mix components). SCC component parameters were outputted according to a multi mixes taken from 34 papers [1-34] related with self compacting concrete which contains the compressive strength and slump flow test results. Mix component values are considered as the aim of the prediction. A total of 225 specimens were selected from the laboratory results of about 34 researches. The system was trained and validated using 150 training mixes chosen randomly from the data set and tested using the remaining 75 mixes. About 20 mixes of experimental SCC not found in the entered data were performed experimental in order to simulate the program and compare between experimental and predicted mix design. Results indicate that SCC components can be predicted with reliable values to the experimental results using the ANN method. | ||||
| Keywords | ||||
| Self compacting concrete; Compressive strength; Slump flow; Neural Network | ||||
|
Statistics Article View: 101 PDF Download: 192 |
||||
