Science Journal of Civil Engineering and Architecture.
March, Volume 2014, ISSN: 2276-6332
© Author(s) 2014. This work is distributed under the Creative Commons Attribution 3.0 License.
Using Neural Networks for the prediction of Permeability in Rock for Derik Embankment Earth Dam
Houtan Jebelli1, Melika sharifironizi2, Vandad Mazarei3 and Esmail Aflaki4
1Department of Construction Engineering, University of Nebraska-Lincoln ,NE,US
2Department of Civil & Environmental Engineering & Earth Sciences, University of Notre Dame, Notre Dame, IN 46556, US.
3Department of Civil and Environmental Engineering, Oregon State University, OR, US.
4Department of Civil and Environmental Engineering, Amirkabir University of Technology, Tehran, Iran
Accepted 27th February, 2014; Available Online 30 March, 2014.
In this study, an artificial neural network has been modeled that is able to predict the permeability of the jointed rocks of the Derik embankment earth dam in the north west of Iran. In jointed rocks the nonlinear relationship exists between the absorption rate and pressure so water pressure test frequently used for evaluating the permeability of jointed rock masses. Different pressures were used during the water pressure test that may result changes in the behavior of rock masses. An artificial neural network is a biologically inspired computing method that is capable of predicting the permeability values of rock masses with high accuracy. The neural networks have shown high potential in interpreting the raw data obtained from Leugon test and predicting the final permeability of a testing section in a borehole. The result showed that, the network that contains five hidden neurons has the best results among other networks and the neural networks are capable of predicting permeability values of jointed rock with high accuracy. This method can be used to analyze the permeability of jointed rocks of dams and grouting zones.
Keyword:Permeability, Artificial Neural, Network, jointed rocks, water pressure test