Page - 856 - in Book of Full Papers - Symposium Hydro Engineering

Constructing a mathematical model to perform a probabilistic analysis, we will obtain the output in the form of the density distribution of the stress safety factors. To gain the optimum f’c, we should constantly change f’c as an input to find a distribution of the stress safety factors with the allowable failure percentage (for example 20% is acceptable for most gravity dams). Performing the analysis on Dasht-e-Palang dam, the application of the method mentioned above will result in the graphs shown in Fig. 9. Trial and error yields that assuming a normal distribution for f’c of the samples with a mean of 7 MPa (known as fcr) and a variation coefficient of 30% (i.e. a standard deviation of 2.1 MPa), we will have a f’c equal to 7-0.824*2.1=5.2 MPa. Performing tests to determine E in the lab will make a more accurate estimation of f’c possible. Fig. 9 Dasht-e-Palang dam probabilistic analysis results 6. CONCLUSIONS CMDs simplify the method of construction at the cost of causing uncertainties in the analysis and design of the dam. Applying an approach based on probabilistic methods with the concept of performance-based design will enable us to overcome these unwanted variations. In this paper, using a probabilistic approach in the designing process of three CMDs (currently under construction) has been described. The main benefits are listed as follows: 1- Minimizing the variations of aggregates in the Hardfill using an optimal method for screening and fractioning without any crushing. 2- More authentic estimation of the effect of weak rock parameters on the sliding safety factors of the dam body. 3- Better estimation of the layered properties of the CMD body. 4- Using less cementitious material by optimizing f’c. ACKNOWLEDGEMENTS The authors would like to give special thanks to A.Mohammadi and R.Erami from Bushehr Regional Water Authority (BRWA) for their kind contributions. 856