Experiments on earth dams breaching. Monitoring instrumentation and methods

Abstract

A research project on laboratorial characterization of the breach processes involved in the failure by overtopping of earth dams is currently ongoing at LNEC and IST. The breach evolution processes is a complex phenomenon that requires the understanding of both hydraulic and geotechnical components involved as well as the interaction between the hydrodynamic erosion and the geotechnical discrete failure episodes. A campaign of laboratorial tests allowed relating the dynamics of the breach with the hydraulic and geotechnical properties of the embankments and understanding how these last influence the failure mechanisms of the embankment. This campaign is being held in a medium scale facility specifically built for the purpose in the fluvial hydraulics pavilion of Hydraulic and Environment Department of LNEC (Fig. 1). A strict control of the experimental work is the base for reliable characterizations of breach temporal evolution and the following effluent hydrograph (herein designated as breach hydrograph). As a starting point, soils for earth embankments laboratorial tests were carefully selected and the laboratorial construction process was realistic and accurate. Beyond these aspects, consistent characterization of breach temporal evolution and therefore, the breach hydrographs estimates were achieved by monitoring the phenomena main variables. All the equipment and measuring/operating methods adopted in the experiments were compatible with the phenomena to monitor and it was assured that all these resources were coordinated with the breach temporal evolution. In this paper three different approaches for estimating breach hydrographs are presented and discussed. Two are indirect estimates: i) from stage-discharge relations; and ii) by watervolume balances within the reservoirs. One is a direct estimate – a novel method that represents an attempt to estimate breach effluent hydrographs based on LSPIV and breach area detection. This last estimate is a step forward to study the dynamics of the breach since it contains phenomenological information on the breaching process. In particular, the velocity field maps in vicinity of the breach required for this estimate allow to for a better understanding of the hydraulic and geotechnical phenomena involved in the breaching process as well as the interaction between the hydrodynamic erosion and the geotechnical discrete failure episodes. The way the instrumentation and measuring/operating methods were combined in order to allow obtaining these three different approaches for breach hydrographs estimation is herein scrutinized. A great attention is given to the way as breach instantaneous velocity was monitored as well as breach geometric evolution was recorded. The following instrumentation and methods were used to monitor the variables that were considered determinants for the characterization of the initiation and breach evolution process: i) 1 high power green laser coupled with a cylindrical lens to create a laser sheet whose intersection with the media interfaces allows the delineation of the breach geometry under water and the free surface above it; ii) 2 CCD video cameras: Photonfocus (200 fps) - to monitor the breach geometric evolution; Mikrotron (up to 1000 fps) - to acquire surface flow images near the breach to estimate velocity maps with an LSPIV algorithm; iii) 2 HD video cameras (25 fps) to monitor the downstream face of the dam and the crest during the overtopping failure; iv) 5 resistive and 7 acoustic probes distributed inside the reservoir and the settling basin (resistive) and around the breach (acoustic) for water level control (Fig. 1); v) geotechnical samples collection during the construction for control of compaction energies and water content of the embankments. The breach hydrographs estimates discussed in this paper resulted from accurate data gathered with recent and suitable instrumentation and measurement methods. The instrumentation and measuring methods adopted in these experiments were chosen among several others that were also tested and either did not serve the purpose or not reached the desired accuracy degree. Therefore the final instrumentation set-up endorsed the achievement of accurate data allowing to discuss the applicability extent of each breach hydrograph estimates herein presented.