During extreme high-water events, the phreatic water level in levees will rise over time due to infiltration of water. This can promote slope instability or internal erosion, and eventually lead to structural failure. A potential solution is the application of an impermeable seal, such as a geotextile, to the levee’s outer slope to locally reduce the inflow of water. In this study, the spatiotemporal effect of a seal on the phreatic surface level is investigated experimentally, both at laboratory scale for a homogeneous sand levee, and at full-scale for a more realistic levee design. On the two-dimensional laboratory scale, it was found that application of a seal does not significantly change the steady-state phreatic level, as expected from a theoretical perspective. However, the time for the phreatic surface level to reach steady state after a sudden external water rise was found to increase 25% to 50% in the cases with a seal. Similar results were found for the full-scale three-dimensional experiments, which showed that details of the soil-structure interface significantly influenced the effectiveness of the impermeable seal, increasing the time to steady state between 12% and 25%. A simple numerical transient groundwater flow model confirms that the quality of the seal governs the response of the phreatic level. This model required the inclusion of an interface layer to properly model the imperfect soil-seal conditions. It is concluded that application of an impermeable seal to a levee before sudden water rise does not influence the new steady-state phreatic level. However, the seal slows down the infiltration process, especially for a case where the outer slope is damaged.

A series of plate loading tests on clay has been conducted in the centrifuge. The aim of the tests is to create a data set, which is freely downloadable, to validate numerical tools that account for geometrical nonlinearities. The tests include two sources of geometrical non-linearities. The first source is the reducing clay layer thickness below the plate, which causes an increase in resistance. The second source is the backflow of the clay around the tip of the plate. The backflow has a reducing effect on the plate resistance. This paper outlines four tests: two involving a wide plate and two with a small plate. Each plate geometry is investigated under both smooth and rough side model boundaries. An material point method (MPM) schematization is used for numerical analysis. The schematization and parameter selection are initially validated by comparing the MPM results against CPTu data in each test. The numerical analysis examines the impact of a finite layer thickness by analyzing various layer thicknesses. Furthermore, the analysis shows the influence of the backflow on the plate resistance by analyzing different ratios of shaft to plate width. In this study, the pore pressures below the plate and vertical and horizontal displacement fields are considered in addition to the load displacement curves. The MPM simulations are in good agreement with the centrifuge data.

This study aimed to assess the suitability of lime treatment for use in dikes in the Netherlands. The effect of this technique on the behavior of a Dutch clay was addressed by comparing the detailed response of lime-treated and natural samples at different lime contents (1.25% and 2.25%) and curing periods. A series of laboratory tests consisting of index classification, constant rate of strain, and triaxial and hole erosion tests were performed. The results demonstrated that lime treatment altered the soil response. Differences were observed in the physical, compressibility, strength, and erodibility properties. It was found that lime improved considerably the resistance to compression and erosion, but the effect on hydraulic conductivity was limited. The triaxial test results showed that lime treatment was particularly effective at low stress (<25 kPa) and low strain levels (<10%). During shearing, lime-treated samples exhibited dilative tendencies and enhanced effective strength properties until a stress-strain state was reached that was believed to be related to the breakage of the bonding structure of the sample. The findings of this study demonstrate that the merits of lime treatment can be of particular benefit in dike applications, particularly when the focus is on improving soil erosion resistance.

A series of undrained cyclic direct simple shear (CDSS) tests on dense Toyoura sand has been performed with the aim to investigate the influence of the stiffness of the DSS device on test results. To this end, springs were installed to reduce deliberately the stiffness of the apparatus. It is shown that the cyclic resistance of the sand depends strongly on the rigidity of the apparatus frame. In particular, as the stiffness of the DSS device increases, the number of loading cycles required to reach liquefaction decreases. This pronounced apparatus-stiffness dependence is of great practical concern in geotechnical engineering because it directly implies that the CDSS response of a soil sample can be predominantly controlled by the stiffness of the apparatus and not by the soil behavior alone. In addition, the test results indicate that the effect of equipment compliance in cyclic undrained DSS testing can be minimized when the ratio of the stiffness of the tested sand sample to the stiffness of the apparatus has a significantly low value.

The dynamic behavior of a peat deposit in the north of the Netherlands is described. The organic content ranges from 70% to 95%, which is high compared to the organic content generally presented in publications on the dynamic behavior of peats. Shear wave velocities vs and correspondingly small-strain shear moduli G0 closely match values stated in the literature. Correlations stated in the literature for predicting G0 proved to be applicable. Resonant column and cyclic direct simple shear tests were performed to establish the shear modulus reduction curves and damping curves. Excess pore pressure development during testing indicates dilatant behavior. The general trend shows nearly flat shear modulus reduction and damping curves at small strains regardless of organic content. Cyclic direct simple shear tests on humified material showed a larger pore pressure buildup than found in tests on non-to-moderately humified material. Differences in degree of humification did not result in significant differences in the shear modulus reduction curve, including G0 values. Large scatter was found in the damping curves. For the humified material, tested at low stress level, a discontinuity in the damping curve is found at shear strain of 3%, which corresponds to a rapid pore pressure buildup in the tests.

Direct observation of gas in peat layers, generated by slow degradation in anoxic conditions, raised concern in the Netherlands about its potential impact on the geotechnical response of dykes founded on peat. To address this issue, an experimental investigation was initiated, aimed at quantifying the main consequences of the presence of gas on the mechanical response of peats. The results of a series of triaxial tests on natural peat samples flushed with carbonated water are presented and discussed. Controlled amounts of gas were exsolved by undrained isotropic unloading, and the samples were sheared under undrained conditions. During gas exsolution, the samples suffered volumetric expansion, at a rate which is ruled by the relative compressibility of the fluid and the soil skeleton. The gas in the pore fluid dominates the stress-strain response upon undrained shearing, causing lower excess pore pressure compared to fully saturated samples. The experimental results suggest that local fabric changes occur during gas exsolution. However, for the amounts of gas investigated, these fabric changes seem to be almost reversible upon compression. Although the ultimate shear strength is hardly affected by gas, the reduction in the mobilised shear strength at given axial strain thresholds is dramatic, compared to fully saturated samples. The study suggests that the presence of gas must be cautiously accounted for at low stresses, when a reference stiffness is chosen for serviceability limit states, and when operative shear strength definitions, based on mobilised strength for given strain thresholds, are chosen in the assessment of geotechnical structures on peats.

Peats are soils containing a significant component of organic matter. Biochemical degradation of this fraction generates gases such as CO₂, H₂S and CH₄, which tend to saturate the pore water eventually resulting in exsolution and expansion. The effects of these gases on the hydro-mechanical behaviour of peats are under investigation at Delft University of Technology. The results of a series of triaxial tests are discussed, in which gas was exsolved under controlled conditions by flushing natural samples with carbonated water, and undrained isotropic unloading and shear were performed. A significant reduction in the effective stress acting on the soil skeleton was observed during undrained unloading due to gas exsolution. However, different stages were observed in time, which appear to be ruled by the very high compressibility of peat. The mechanical response upon shearing is dominated as well by the ratio between the compressibility of the fluid and the soil skeleton. Although the ultimate strength does not differ much between the samples tested, the mobilised shear strength for a given axial strain does, which has to be accounted for cautiously in the choice for an operative shear strength.