Sensitivity analysis for uncertainty quantification in earth dams modeling (Case study: Maku dam)

Due to the existence of various uncertainties in the materials and modeling of these structures, the quantification of these uncertainties requires sensitivity analysis and reducing the number of variable parameters to reduce cost and time in modeling. In this research, numerical analysis of Maku dam located in West Azerbaijan province which is considered as a case study is performed using FLAC2D software and the static response is obtained. The effect of construction layers and the amount and shape of contours of stress and displacement were compared and validated with previous researches. The 49 random variables (RVs) (of materials properties in the zones of the core, shell, filter, drainage, alluvium and bedrock) was considered. By performing sensitivity analysis and using the Tornado diagram, the number of variable parameters is reduced and 18 important RVs were identified for the modeling and analysis of the earth dam. The results showed that altering the dry density (γd) of the shell causes a change of about 8% and altering the Poisson's ratio (ν) and also the modulus of elasticity (E) of the core leads to a change of about 7% in earth dams’ response. Finally, parameters of dry density (γd), Poisson's ratio (ν), modulus of elasticity (E) and internal friction angle (ϕ) were identified as more sensitive parameters that have the greatest impact on the response of earth dams.