Citation: TANO, F. ...et al., Numerical modelling of the nonlinear mechanical behavior of multilayer geosynthetic system for piggyback landfill expansions.Geotextiles and Geomembranes, 44 (6), pp. 782-798.Additional Information:• This paper was accepted for publication in the journal Geotextiles of the GSYs and of the interfaces between GSYs must be considered. Designers, however, often 31 use simplistic assumptions without further analyzing the implications of these assumptions on the 32 results. Such simplifications mainly concern the nonlinear axial stiffness of GSYs, the strain 33 softening at interfaces between GSYs, and the difference between the compressive and tensile 34 behavior of GSYs. By, considering these key aspects, the present study aims to understand the 35 extent to which the results of numerical calculations can be influenced both by the differing 36 compressive and tensile behavior of GSYs and by the assumption of strain softening at interfaces 37 between GSYs. For this purpose, several numerical models are implemented by using the finite-38 difference code FLAC 2D on a typical PBLE that involves four GSYs and six interfaces. The 39 present work also applies comprehensive, state-of-the-art numerical modelling to study the 40 interactions between multiple layers of GSYs. This study also investigates the nonlinear axial 41 stiffness of GSYs through a series of uniaxial tensile tests. The numerical results show that, if the 42 GSY axial compressive and tensile characteristics are the same, then tensile force is minimized, 43 which induces significant compressive force in the GSYs. The results also indicate that 44 3 neglecting strain softening at the interface between GSYs affects interface shear stresses, 45 displacements of GSYs at the interface, and the GSY force distribution, potentially rendering the 46 model unrealistic. Including strain softening, however, allows the assessment (location) of 47 unstable areas along the interface where large displacements occur. 48 49

This report presents a new large-scale test apparatus (LSTA) developed to assess the efficiency of a geosynthetic reinforcement for the limitation of deformations of a geosynthetic lining system (GLS) over a 0.5 m wide cavity. Two experiments were conducted. The first one involved a geosynthetic clay liner, a nonwoven needle-punched geotextile and a high-density polyethylene geomembrane. For the second experiment, a 50 kN/m polyvinyl alcohol geogrid was imbedded within the sand layer below the geosynthetic clay liner to provide reinforcement above the cavity. An overburden pressure varying from 10 to 100 kPa was applied to the top of the GLS. Strain gauges were used to measure the strain within the geogrid and the geomembrane. The results proved that the 50 kN/m geogrid reinforcement beneath the geomembrane reduced the maximum strain within the geomembrane, compared to the case where the geomembrane was unreinforced, by 25% on average. The results showed that the overall strain within the geomembrane was 31% to 42% less than that of the geogrid above the cavity. Finally, the results showed that the spatial distribution of the strain within the geomembrane and that of the geogrid differed because of a conical shape of the collapsed zone.

Geosynthetics have been extensively used in landfills as a lining system to prevent leachate infiltration into groundwater. In piggy-back landfill expansion (PBLE), consisting of building a new landfill over an existing one, a lining system is implemented between the old and new waste. In this context, interface failure (stability) and deformation (integrity) of the lining system should be considered for the design. Such stability and integrity mainly depend on the PBLE geometry and the mechanical properties of the geosynthetics. Comprehensive numerical modelling simulations were performed to show how these factors influence the shear stresses, shear displacements, translational stability and the axial strains/forces within the various geosynthetics. The numerical modelling was conducted using the finite difference code FLAC 2D, focusing on a typical PBLE and considering geosynthetic interface strain softening, the nonlinear stiffness of geosynthetics, and the differentiation between the compressive and tensile behaviours of geosynthetics. The simulations showed that the lateral length of the PBLE, the type of geomembrane (textured or smooth) and the level of the leachate table in new waste are the factors that most influenced the mechanical behaviour of the lining system and its stability. Finally, a parameter called the stability ratio was proposed as a complement to the traditional factor of safety, to analyse the progressive slippage along the geosynthetic interfaces. The numerical results indicated that interface failure concurrently begins at the rightmost part of the lower flat area of the PBLE and near the corner of the inner slope before spreading out to the left as backfilling progresses.

Numerical modelling approaches can aid in designing geotechnical constructions involving geosynthetics. However, the reliability of numerical results depends on how the model is developed, the constitutive model, and the set of parameters used. By comparing the numerical results with experiment, the present work verifies a numerical modelling technique developed to model multilayered geosynthetic lining systems for landfills. The numerical modelling technique involves strain softening at interfaces and allows the axial stiffness of the geosynthetics to evolve as a function of strain. This work focuses on a two-dimensional finite-difference model, which is used to simulate three types of experimental tests: conventional uniaxial tensile tests, direct shear tests, and a large-scale test that was used to assess the overall mechanical behaviour of a reinforced geosynthetic system that spanned over a cavity. This reinforced geosynthetic system consisted of a 50 kN/m polyvinyl alcohol geogrid reinforcement embedded in a layer of sand, a geosynthetic clay liner, a high-density polyethylene geomembrane, and a non-woven needlepunched geotextile. The uniaxial tensile tests, direct shear tests, and the large-scale test were numerically modelled and the numerical results were compared with experimental results. The results of the numerical modelling technique presented very closely match the results of the three experimental tests, which indicates that the numerical model correctly predicted the measured data.