Geosynthetic-reinforced Walls Research Articles

Pullout testing and Particle Image Velocimetry (PIV) analysis of geogrid reinforcement embedded in granular drainage layers

The paper investigates the feasibility of using fine-grained soil as backfill material of geosynthetic-reinforced walls and slopes, through a laboratory study on pullout behavior of geogrids in granular layers. A series of pullout tests was carried out on an HDPE uniaxial geogrid in thin sand and gravel layers that were embedded in clay specimens.Aside from different soil arrangements, the influences of moisture content and overburden pressure on the geogrid pullout behavior is assessed and discussed. The tests were carried out at four different gravimetric water contents (GWC) on the dry and wet sides of the clay optimum moisture content (OMC), and overburden pressure values within the range σv = 25–100 kPa. Particle Image Velocimetry (PIV) was used to capture digital images during the tests, which were processed to help with the interpretation and improved understanding of the soil-geogrid interactions at different GWC values. Results show that embedding geogrid reinforcement in layers of sand or gravel can significantly increase the pullout resistance in an otherwise moist clay backfill, and this improved pullout efficiency is greater at higher overburden pressures. The improvement in pullout capacity was observed in clay specimens compacted at both the dry and wet sides of the OMC.

Limit analysis of lateral earth pressure on geosynthetic-reinforced retaining structures using finite element and second-order cone programming

In this paper, a thorough numerical study is conducted to rigorously evaluate the lateral earth pressures exerted on the retaining walls backfilled with geosynthetic-reinforced soil strata. With this aim, the lower bound theorem of limit analysis is exploited in conjunction with the finite element discretization method. The computational programming adopting second-order cone optimization is employed to model the Mohr-Coulomb yield criterion in its respective nonlinear form. The reinforcement layers are assumed to bear solely axial tension; but not bending moment, which is the common characteristic of all geosynthetic materials. Results show that the horizontal stress field behind retaining structures is heavily disturbed by placing geosynthetic layers, especially in close proximity to the retaining structure. Accordingly, the failure zone shrinks in size and the lateral earth pressure decreases by either adding more reinforcement layers or increasing their relative lengths. Through a comprehensive parametric survey, the influence of several parameters, including internal friction angle, soil-wall and soil-reinforcement interface friction angles, length and number of reinforcements, soil inherent anisotropy and surface loading on the lateral earth pressure is meticulously examined. The employed formulations are rigorously compared with results published in the literature. A design table is also provided so as to effectively estimate the active earth pressure coefficient of geosynthetic-reinforced walls accounting for all contributing parameters.

Seismic responses of vertical-faced wrap-around reinforced soil walls

A series of shaking table tests on geosynthetic-reinforced walls with a height of H = 0.6 m is performed to investigate the seismic performance of prototype walls with H = 6 m. Ground-wall resonance occurs at small values of horizontal peak ground acceleration (HPGA), namely 0.1–0.5g, with an amplification factor of Am = 1.2–2.2 at the crest of the wall. This factor decreases with increasing HPGA. For walls that attain states of maximum horizontal displacement (Dmax)/H > 5%, the failure mechanism consists of a vertical failure surface at a distance from the facing of 0.3H for the upper half of the wall and a Rankine triangular wedge over the lower half of the wall. Mononobe-Okabe theory gives dynamic earth pressure coefficients that agree well with experimental results based on kh = η·HPGA/g, where khand η are the horizontal seismic coefficient and an empirical constant, respectively, with values of η ranging from 1/4 to 1/3. Furthermore, the maximum tensile forces induced by shaking increase as the depth of reinforcement increases, generating a trapezoidal shape rather than the inverted trapezoidal shape proposed in the literature.

Seismic stability of geosynthetic-reinforced walls with variable excitation and soil properties: A discretization-based kinematic analysis

Seismic stability of vertical slopes is investigated using the discretization-based kinematic analysis, aiming at determining the geosynthetic reinforcement force required for preventing slope failure. The earthquake effect is characterized by spatial and time varying seismic inputs in sinusoidal functions. Two modes of seismic inputs are proposed for describing the 2D rather than 1D spatial effect. Further, the change in cyclic amplitude of the acceleration time-history is incorporated into kinematic analysis. The reinforcement force is formulated from the work rate-based balance equation. The pseudo-static/dynamic solutions are sought and compared. Comparisons are highlighted between the numerical results from three different seismic inputs.

Analytical method for calculating natural frequencies of geosynthetic-reinforced wall with full-height concrete facing

Based on the theory of beams on elastic foundations, a new analytical solution for calculating the natural frequencies of reinforced-soil retaining walls is presented. The assumed retaining wall is geosynthetic-reinforced, with a full-height rigid concrete facing (a GRS-FHR wall). The current study takes into account the rigid body motion of these walls for the first time, and two analytical formulae for calculating the natural frequencies of the rigid mode of deformation for reinforced retaining walls are presented through an exact solution. Next, applying the energy method, a new analytical formula for calculating the natural frequency of the flexural deformation mode is developed. Another innovation is the modeling of the foundation soil with both rotational and translational springs, which makes it possible to consider rocking and sliding deformation modes for retaining wall motion. Moreover, the effect of backfill soil interaction is taken into consideration by translational springs along the wall. The results thus obtained are then compared with numerical analysis using Ansys software, and an excellent agreement was observed.

Service-state behavior of reinforced soil walls supporting spread footings: a parametric study using finite-element analysis

Reinforced soil has become a common means of retaining soil and supporting spread footings due to ease of construction and cost-efficiency. However, the behavior of these structures under service state conditions is inherently complex and dependent on the selection of design factors. Factors that govern design include earth pressures, settlements and reinforcement strains, but are not easily attained from closed form solutions. This study aims to provide insight into the complex behavior of reinforced soil walls surcharged by a discrete footing in context of these design factors, facilitating the development of a numerical model. A nonlinear elastic–plastic constitutive soil material model was calibrated to experimental plane strain test data, while the performance of a wall model was compared against full-scale laboratory testing of an identical geosynthetic-reinforced wall from prior literature, demonstrating satisfactory agreement. This agreement enabled a parametric study of the same wall model with a discrete spread footing evaluating the effects of reinforcement type and vertical spacing, footing location, footing dimensions, and toe restraint on lateral earth pressure distributions, wall deformations and reinforcement strains. The results of the study demonstrated that simplified consideration of only reinforcement type on service-state performance of reinforced walls neglects the effects of closely spaced reinforcements on behavior. Use of closely spaced reinforcements created increased earth pressures, but also reduced lateral displacement, footing settlements and reinforcement strains. Furthermore, it enabled achievement of equivalent service state criteria for footings located closer to the wall facing, potentially enabling increased right-of-way or reduced deck length when supporting a bridge superstructure.

Optimization of Design Parameters and Cost of Geosynthetic-Reinforced Earth Walls Using Harmony Search Algorithm

This paper proposes a new approach to optimize the design of geosynthetic-reinforced retaining walls. Minimizing the cost of construction was considered as the optimization criterion. A metaheuristic technique, named Harmony Search Algorithm (HSA), is applied in optimizing the design of geosynthetic-reinforced earth walls. The involved optimization procedures are discussed in a step-wise approach and their applicability is demonstrated on geosynthetic-reinforced walls of height 5, 7 and 9 m. The effects of static and dynamic loads are considered. Results are compared between this study and studies that used Sequential Unconstrained Minimization Technique (SUMT). It is found that the construction cost, for a geosynthetic-reinforced walls optimized by HSA, showed as high as 9.2 % reduction from that of SUMT.

Framework for Limit State Design of Geosynthetic-Reinforced Walls and Slopes

Introduced is a unified limit state design framework for geosynthetic-reinforced slopes and walls. It is demonstrated that limit state design is an essential step in design even though the usual perception is that the performance of such structures is “better than expected.” A brief critical overview of commonly available analysis methods is presented. The typical design methodology which is an extension of conventional limit equilibrium (LE) analysis is discussed. It is shown that with the addition of the safety map tool, an optimized design can be achieved with relative ease. However, while the safety map can be used in design, the basic solution is still incomplete. Subsequently, the LE approach is generalized to include the complete solution, i.e., the required tensile force distribution along each layer considering reinforcement layout and anchorage properties. The modified approach produces a solution needed for stability including the required connection strength. It is based on free body equilibrium ensuring that at each point within the reinforced mass, the factor of safety is the same. That is, unlike the safety map which shows the corresponding safety factor at each location, the presented framework adjusts the required strength of the reinforcement so that the safety factor is the same constant everywhere. Limited parametric studies demonstrate the performance of the new framework as well as that of the safety map approach. These studies show the impact of facing blocks, seismicity, backslope, quality of backfill, length of reinforcement, and effects of secondary short reinforcement. The results are compared with relevant experimental data. The agreement is reasonably good. Finally, a general link between the framework and actual design is briefly discussed.

Behaviour of reinforced structures under simulated toe scouring

ABSTRACT: Four geosynthetic-reinforced model walls backfilled with an idealised two-dimensional uniform medium were built and brought to failure using surcharging and toe cutting to simulate the behaviour of waterfront structures subjected to potential slope toe scouring. External stability analyses – base sliding, overturning, and bearing capacity failure – were performed in order to examine their relevance to the displacement, and ultimate failure state, of the walls. It was shown that a well-designed geosynthetic-reinforced wall can sustain a surcharge and a uniform toe cut up to an equivalent wall height (He) equal to 1.4 times the total wall height (Ht) with negligible horizontal and vertical wall displacement. This can be attributed to the effective mobilisation of tensile force in the reinforcement layers. Safety factors for base sliding and overturning vary within small and acceptable ranges during this stage. However, a drastic increase in wall displacement occurred during ground cutting adjacent to the toe of the tested wall, owing to insufficient bearing capacity beneath the wall. This increase was associated with a rapid decrease in the safety factor against bearing capacity failure beneath the wall. Detailed internal and external stability analyses for the wall behaviour observed here are reported in a companion paper.

Behavior of cantilever and geosynthetic-reinforced walls on deformable foundations

A series of plane-strain model tests was performed to investigate the behavior of cantilever soil retaining walls (CWs) and geosynthetic-reinforced soil retaining walls with a rigid facing (GRS-RWs) placed on non-deformable and deformable foundations with various subgrade reaction moduli ( k v). The walls were designed to have configurations similar to those used in practice, with similar controlling safety factors against sliding. Screw jacks and springs were used to simulate undeformable and deformable grounds, respectively, with various maximum foundation settlements of S max ≒ 0, 2.5, 5, and 10% of the backfill height ( H). Test results show that the GRS-RW has better settlement-tolerating performances, in terms of the tilting angle ( θ), the horizontal displacement ( D h), and the settlement of the crest of the backfill ( D v1), than those of the CW. For both CWs and GRS-BWs, the worst scenario of the wall performance, in terms of D h, θ and D v1, occurred at a moderate foundation settlement of S max/ H ≒ 5% (or k v = 1.8 kPa/mm), rather than at a greater foundation settlement of S max/ H ≒ 10%, which facilitates a tilting-backward displacement mode. Experimental results also indicate that local lateral pressure coefficients against facing ( K f,z) for CWs may reach the at-rest (or K o) state at the central portion; values of K f,z may reach the passive (or K p) state at the lower portion of the wall. In the case of CWs, the measured values of local and global lateral pressure coefficients ( K f,z and K f) tend to increase with increasing maximum foundation settlement. This is not the case for GRS-RWs, which exhibited a relatively settlement-independent response, in terms of K f and K f,z, against facing. To develop relevant limit-equilibrium-based design methods for CWs and GRS-RWs placed on deformable foundations, knowledge of lateral pressure coefficients associated with various displacement and tilting induced by the foundation settlement are required.

Simplified approach for assessing seismic displacements of soil-retaining walls. Part II: Geosynthetic-reinforced walls with rigid panel facing

Critical seismic coefficient ([Formula: see text]) and seismic displacement of some idealized geosynthetic-reinforced walls with full-height rigid panel facing (GRS-FHR walls) were calculated using a pseudo-static-based multi-wedge method in association with Newmark's sliding block theory. The proposed method uses experimentally verified lateral earth pressure coefficients to evaluate connection force at the facing/reinforcement interface subjected to various input seismic loadings. It was found that the values of critical seismic coefficient, [Formula: see text], can be expressed as linear functions of some dominant factors, such as the internal friction angle of backfill (ϕs) and the ratio between reinforcement length (L) and wall height (Ht). Empirical equations for the values of [Formula: see text] as functions of ϕs, L/Ht and other factors were established, and these equations were used in conjunction with seismic displacement evaluation methods proposed in this study. Based on the result of seismic displacement evaluations on a well-documented GRS-FHR wall that experienced the 1995 Hyogoken-Nambu earthquake, the effectiveness of the simplified approach for assessing seismic displacement of GRS-FHR walls was verified. These results were also compared with those calculated for some geosynthetic-reinforced walls with modular block facing that have been reported in a companion paper.

Seismic Displacement of a Geosynthetic-Reinforced Wall in the 1995 Hyogo-Ken Nambu Earthquake

ABSTRACT Seismic displacement of a geosynthetic-reinforced wall with full-height rigid panel facing (Tanata wall) in the 1995 Hyogo-ken Nambu earthquake is calculated using a pseudo-static method based on a 'multi-wedge' failure mechanism. The calculated value of horizontal displacement of the wall is comparable to the measured one. Based on an investigation into the effect of vertical ground acceleration to the seismic displacement of the wall using vertical and horizontal input ground accelerations in recorded seismograms, it is found that the contribution of vertical ground acceleration to the seismic displacement of the Tanata wall is small, because the peak horizontal and vertical ground accelerations are out of phase. Therefore, the use of peak vertical-to-horizontal ground acceleration ratio obtained in an earthquake event for pseudo-static multi-wedge analysis may overestimate the seismic displacement of a geosynthetic-reinforced wall to some extent. The effect of the embedment of the facing on the seismic displacement of the Tanata is also investigated. It is found that the effect of facing embedment to the seismic stability and/or displacement in the case of the Tanata wall is insignificant.

Stress and Strain Monitoring of Reinforced Soil Test Wall

Stresses and strains were measured in a full-scale reinforced test wall constructed to evaluate the design procedure and performance of geosynthetic-reinforced walls constructed with silty-clay backfill over a soft foundation. The measurements consisted of monitoring wall deformation, setlement, strains along the reinforcement, vertical and horizontal stresses, and pore water pressure under the wall. The instrumentations and measurements of wall deformation, soil stresses, and reinforcement strains during and after construction are presented. The wall consisted of three sections reinforced with various geogrid types and spacing. The first section was built with closely spaced, low-strength geogrid. The second section was constructed by using higher strength geogrid at a maximum vertical spacing. The reinforcement and spacing of the third section were based on standard design procedures. Deformations along the reinforcement were monitored with strain gauges. The magnitudes and distributions of the strains varied in the three wall sections. The strains were used to predict the reinforcement loads and stresses in the reinforced soil system. The concept of normalizing reinforcement stress was useful in defining the relative "rigidity" of the wall and in determining the horizontal stress coefficient of the wall. The locus of maximum stresses in the reinforcement did not correspond to the (45 + ϕ/2) line and varied according to the reinforcement rigidity. A comparison was made between the measured and predicted maximum tensile loads in the reinforcements. The results indicated that the loads predicted by the Ko-stiffness method were closer to the measured loads than those predicted by the AASHTO simplified method.

Strain Distribution within Geosynthetic-Reinforced Slopes

Geosynthetic-reinforced slopes are conventionally designed using methods based on limit equilibrium. In order to estimate the factor of safety against internal stability using these methods, the distribution of the reinforcement peak tensile forces with height must be assumed. A linear distribution of reinforcement peak tension with height, with zero tension at the crest and maximum peak tension at the toe of the structure, has often been assumed. Although this assumption may be appropriate for the design of vertical geosynthetic-reinforced walls, little evidence has been collected so far justifying this distribution for the design of geosynthetic-reinforced slopes. A combination of centrifuge testing and digital image analysis is undertaken in order to obtain the strain distribution within geosynthetic-reinforced slopes under prefailure conditions. Specifically, digital image analysis techniques are used to determine the displacement distribution along reinforcement layers in reduced-scale models subjected to increasing g levels. A sigmoid function was useful to fit raw displacement data and estimate the strain distribution along reinforcement layers. Analysis of reinforcement strain results shows that the location of the reinforcement maximum peak strain does not occur near the toe of the structure, but was located approximately at midheight of the reinforced slopes, at the point along the critical failure surface directly below the crest of the slope. The pattern of reinforcement peak strain with height obtained for prefailure conditions is similar to that obtained for failure conditions. The estimated factor of safety is found to be a good indicator of the magnitude of the reinforcement maximum peak strain for geosynthetic-reinforced slopes built with different configurations.

Observed Long-Term Performance of Geosynthetic Walls and Implications for Design

Geosynthetic-reinforced walls have been viewed by the civil engineering profession as a new technology whose acceptable long-term performance is yet to be established. Nevertheless, geosynthetic walls have been in use for almost 25 years. Much of the uncertainty associated with acceptance of geosynthetic-reinforced wall technologies is related to time-dependent deformation. This paper summarizes the creep rates measured in full-scale walls and compares them to creep rates measured in-isolation in the laboratory where the applied load level matches values estimated for the same structures in the field. In the majority of cases, the laboratory in-isolation creep rates were the same as or greater than the measured reinforcement creep rates in full-scale walls, corroborating that reinforcement load levels can be estimated from measured strain data. At the end of wall construction, it appears that the reinforcement is primarily exhibiting creep, with only minor stress relaxation. However, in the long-term, there is a trend toward reinforcement stress relaxation. Furthermore, the long-term behavior observed in the full-scale walls indicates that the reinforcement loads are well below values required to cause creep rupture over the design life of the structures and, in some cases, creep appears to have stopped completely. Finally, the paper offers quantitative guidelines to delineate anticipated poor and good long-term wall performance based largely on level of reinforcement strains and magnitude of post-construction wall deformations.

Design of geosynthetic-reinforced walls : Leshchinsky, D; Perry, E B Geotext GeomemV8, N4, 1989, P311–323

Design of geosynthetic-reinforced walls : Leshchinsky, D; Perry, E B Geotext GeomemV8, N4, 1989, P311–323

On the design of geosynthetic-reinforced walls

An important design aspect of a geosynthetic-reinforced wall is specifying the reinforcement length. A simple limit equilibrium analysis, which considers two possible extreme inclinations of a reinforcement's resistance force, shows that predictions based on the K a-analysis may not be safe. Although rules of thumb to determine embedment length are sometimes superimposed on the K a-analysis prediction in design, it is demonstrated that such ‘rules’ may not always be safe. Design charts for granular backfill illustrate the effects of the reinforcement's force inclination showing quantitatively that the horizontal inclination required slightly larger tensile resistance force, while underestimating the embedment length.