Multichannel optical sensor to quantify particle stability under seepage flow

Abstract

International audienceAn optical sensor was designed to measure the fine solid particles concentration contained in a fluid flow. Thissensor is composed of four light-emitting diodes and four associated light-dependent resistors, allowing the measurement offluid transparency. Given the small device dimensions, it can be placed close to the particles exit from the specimen. Theoptical sensor is able to instantaneously measure a large range of clay or silt concentrations in the effluent, without a significantinfluence of flow rate. The presence of sand grains in fluid flow can be detected. The use of this sensor with a specifictriaxial device allows precise characterization of the initiation and development of the suffusion process on clayey sandspecimens. It is shown that suffusion of clay particles induces a decrease of hydraulic conductivity, which is due to a diffuseprocess of eroded particles filtration. Clay suffusion and filtration processes are influenced by grain-size distribution andalso by grain shape of the coarse fraction. Under a high hydraulic gradient, clay suffusion can be accompanied by sand erosion.The critical value of mean pore velocity to initiate clay suffusion was determined for the types of soils

Influence of angularity of coarse fraction grains on internal erosion process

“…Thus the angularity also has an effect on hydraulic conductivity decrease. In conformity with previous conclusions [Reddi, 2000;Bendahmane, 2008;Marot, 2009Marot, , 2011 this decrease in hydraulic conductivity is attributed to particle filtration which may induce a clogging process.…”

“…The small dimensions of the device allow placing it vertically close to the base of the sample and this position avoids any deposition of the eroded particles between sample and optical sensor. A complete description of the device is given by [Marot, 2011]. Thanks to a previous calibration, the optical sensor allows measuring the clay concentration, C, within the fluid which is expressed by the ratio of clay mass particles and water mass within the fluid.…”

“…As the sensor accuracy allows a fairly precise detection of clay particle erosion (for a clay concentration up to 1.6 mg/g, the maximum relative error is 3%; for a concentration between 1.6 and 12 mg/g the maximum relative error is 5% and a concentration higher than 12 mg/g and up to 90 mg/g, the maximum relative error is 3%, [Marot, 2011]), several suffusion tests are performed on different mixtures of 10% of kaolin and 90% sand. As shown in Figure 3, the grain size distributions of tested mixtures are gap graded.…”

A possible characterization of suffusion susceptibility independent of the hydraulic loading history?

“…By carrying out tests on mixtures of sand and gravel with an upward flow, Skempton & Brogan (1994) concluded that the value of the critical gradient of the suffusion initiation, named i cr , is far lower of the value of the critical heave gradient i c (with i c = γ'/γ w , γ': submerged unit weight of the soil specimen and γ w : unit weight of water). The filtration of some detached particles can induce a clogging process within the soil accompanied with the decrease of the hydraulic conductivity (Reddi et al 2000;Bendahmane et al 2008;Marot et al 2009;2011a;Nguyen et al 2012;Luo et al, 2013), which in turn leads to a decrease of the seepage velocity. Therefore, variations of both seepage velocity and pressure gradient have to be taken into account to evaluate the hydraulic loading.…”

Suffusion susceptibility characterization by triaxial erodimeter and statistical analysis

“…The critical values of these three quantities can then be used to characterize the suffusion initiation (Skempton & Brogan 1994, Moffat & Fannin 2006, Perzlmaier 2007. However a fraction of the detached particles can re-settle or be filtered at the bulk of the porous network (Reddi et al 2000, Bendahmane et al 2008, Marot et al 2009, 2011a, Luo et al 2013. This process can eventually induce local clogging.…”