4. Conclusion
The paper presents an analytical model named EFR (EFfet Retard - Delayed effect) for the prediction of pore water pressure inside embankment dams. This model is able to take into account delayed effects between the changes of reservoir water level and the monitoring device located inside homogeneous dams, due to the hydraulic diffusivity. Applying the model to dam monitoring data permits to obtain the corrected measurements which represent measurements under identical loading conditions over time. These corrected measurements can therefore highlight and quantify the irreversible evolution trends occurring under constant loads. A good estimation of irreversible effect enables us to understand better the evolution response of dam over time and to detect early the abnormal events. At the same time, the delayed hydrostatic effect can be isolated and quantified.
An application of the model to a French embankment dam is provided in the paper. The corrected measurements of pore water pressure are obtained using the equivalent reservoir water level computed with the optimized characteristic time T0. The dispersion of raw measurements has been strongly reduced by applying this model. Particularly, the diffusion models, the numerical methods of computing the equivalent reservoir water level and the different versions of irreversible term are discussed. The results show that both the diffusion models and the numerical methods can perform a good linear regression for the study case, and a linear term is sufficient to model the irreversible effect. Besides, a comparison between the HST and EFR models is carried out and confirmed the good performance of the EFR model.
