Smart Geotechnics for Smart Societies – Zhussupbekov, Sarsembayeva & Kaliakin (Eds)

© 2023 The Author(s), ISBN 978-1-003-29912-7

Open Access: www.taylorfrancis.com, CC BY-NC-ND 4.0 license

Modeling of geogrids in the analysis of the stability of reinforced

soil structures with the finite element method

A.V. Kuznetsova
Tensar Innovative Solutions, Russia

ABSTRACT: Finite elements models (FEM) programs have become an integral part of geo­
technical analysis. However, the inclusion of geosynthetics in the model often gives results
that differ from the actual behavior of the structure. The classic way is the modeling of
a geogrid in the form of a membrane with specified strength parameters. The list of these
parameters and methods for their determination are not spelled out clearly enough in the
design and test standards. The author proposes to investigate both the finite element and ana­
lytical models and estimate their cohesion with the monitoring results and ability to fully con­
sider the effect of geogrids.

1 INTRODUCTION

Reinforced soil was used for retaining systems since ancient times. Ziggurat in Mesopotamia,
Roman marinas, the Great Wall of China and a lot of other structures were built using this
technology. But today’s methodology of design and construction has started its development
in 1960s by Henri Vidal. Until the two last decades of technology development mainly analyt­
ical design methods have been used by engineers. But today finite elements models (FEM)
serve to check the validity of the design for complex geometry, geology or loading. The classic
way is the modeling of a geogrid in the form of a membrane with specified strength param­
eters. The list of these parameters and methods for their determination are not spelled out
clearly enough in the design and test standards. In this study author defines time and load
dependent parameters for both analytical and FEM methods, provides comparison of results
of both approaches and makes propositions for the further development of the model.

2 REINFORCED SOIL MODELING

2.1 Analytical methods

A number of analytical design methods have been developed since 1960s. All of them analyze
the balance of forces and mainly differ in terms or partial factors and a degree of conservatism.
Reinforcing elements are described by their design strength calculated as short-term tensile
strength reduced by a row of durability parameters including creep, installation damage and
environmental influence. The purpose of the analysis is to check the balance between acting
and resisting forces as shown in Figure 1.
Analytical methods have been widely used during the last three decades and are showing
good correspondence with the actual structure behavior.

DOI: 10.1201/9781003299127-357

2318
Figure 1. Scheme for calculating the internal stability of a reinforced soil retaining wall.
H – total height; Нi – distance from the top to the wedge base; L – length of geosynthetic material; φ’ – fill angle
of friction; θ – angle from horizontal to the failure surface; Q2 and q1 – surcharge; Еah and Eav – horizontal and
vertical components of active force and surcharge; Wi – failing wedge weight; Zi - required resistance force; Ri –
base resistance force.

2.2 Finite element analysis (FEM)

Firstly used for homogeneous soil structures, FEM now covers all types of reinforced soil
applications, including retaining walls. Geogrids are implemented as membrane elements with
given stiffness or modulus. Soil models are well calibrated with actual soil behavior, but geogrid
models are far from that point. For that purpose, a number of real laboratory tests have been
modeled with the use of FEM, like tensile strength test or pull-out test, shown in Figure 2 and
others. But the behavior of a full structure still cannot be predicted accurately.

Figure 2. Geogrid model used for correlation.

FEM highly overestimates facing deformation due to the inability to model the actual grid
influence on the fill material.

3 DESIGN STRENGTH AND STRAIN OF A GEOSINTETIC ELEMENT

3.1 Design strength

Design strength is described in a number of national standards and manuals for reinforcement
soil design (BS8006, EBGEO and etc.) as a strength, guaranteed on the last year of service

2319
after material was affected by creep, installation, aggressive environment, microbiology and
UV. For polymer geogrids and geotextiles creep is the main reduction parameter and it could
drop the strength to 30 – 60 % of quality control short-term strength. Figure 3 shows the dif­
ference in creep behavior of HDPE geogrid depending on the load applied. For this particular
material creep test the long-term creep strength for 120 years is set as 24kN/m.

Figure 3. A series of creep test.

This number is used for ultimate limit state (ULS) analysis, that models a case where geo­
grid is ruptured. But for the serviceability limit state (SLS) that investigates geomaterial strain
during the post-construction period, allowable load for geogrid should be limited even stricter.
For reinforced soil retaining wall geogrid strain during the service life is limited by 1% only.
Due to this design strength for the material above is set as 10kN/m for SLS analysis.

3.2 Creep related stiffness

FEA uses the stiffness of material that is generally obtained as load (strength) divided by
strain. From geometry point of view stiffness could be compared looking at the degree of lines
in Figure 3. First, it differs dramatically depending on the load applied, but Figure 4 indicates
that even for the single load it changes during the test.

Figure 4. Stiffness deviation during the test.

2320
 In Russian design manual for geosynthetic materials FEM ODM 218.3.120-2020 short-term
stiffness at 2 or 5% load is recommended as a performance related parameter. During the ser­
vice life HDPE geogrid is subject to permanent load equal to 5 – 45% of its short-term
strength. It could hardly be correlated with short-term 2% strain load recommended in ODM.
For example, 50kN/m strong HDPE grid shows a short-term stiffness at 2% equals to 500kN/
m but the 120 years creep test isochrone provides 262 kN/m for 2% strain.
Apart from loading speed and type there is one more significant factor for polymer geo­
grids – design temperature in the soil massive. ISO/TR 20432:2007 proposes to define it as the
temperature which is halfway between the average yearly air temperature and the normal
daily air temperature for the hottest month at the site. Creep behavior related to test speed
and temperature is shown in Figure 5.

Figure 5. Stiffness deviation due to loading speed and temperature.

Summarizing the above, the list of input data is the following:

– design life;
– design temperature;
– durability parameters;
– actual load in reinforcing element as partial from short-term strength.
To provide this data, geogrid should have been tested for creep at all range of loads from 5%
to maximum creep related strength that differs from 40 to 70% depending on polymer type.
During a finite element analysis operator or software defines the actual load in reinforced
element and uses the corresponded stiffness. For example, Plaxis includes Elastoplastic (N-e)
geogrid model that requires stiffnesses at all load proportions and chooses the proper one
automatically.

4 FINITE ELEMENT MODEL

4.1 Soil slope model

To highlight the difference between soil and reinforced soil models two structures are shown.
First one is a Mohr-Coulomb simple soil 6m high embankment with no geosynthetic material.
Main design properties are shown in Table 1.

2321
 Table 1. Design properties for soil slope.
Parameter Fill Foundation

Cohesion, kPa 1 -
Angle of friction, ° 36 -
Su, kPa - 125
Unit weight, kN/m3 19 18

Stability analysis shown in Figure 6 provides a factor of safety equal to 1,22.

Figure 6. Simple slope FEM analysis.

The same slope modeled by analytical method shows a non-circular slip surface with quiet
close nature and safety factor of 1,206 as per Figure 7.

Figure 7. Simple slope analytical analysis.

Based on this, we can assert a high convergence of finite element and analytical modeling
the results.

4.2 Reinforced soil model

Second model represents a 6m high reinforced soil structure with facing wrapped around
sandbags. Main design properties are shown in Table 2, section is shown in Figure 8.

2322
Table 2. Design properties for soil slope.
Parameter Reinforced soil Retained fill Foundation Soilbags

Material model H-S (Drained) H-S (Drained) M-C M-C

E50ref 25.000 20.000 - -
Eoedref 25.000 20.000 - -
Eurref 75.000 60.000 - -
m 0.5 0.5 - -
E, kN/m2 - - 37.500 30.000
v - - 0.35 0.30
λ* - - - -
k* - - - -
c’, kPa 1 0.1 - 10
su,kPa - - 125 -
Angle of friction, ° 36 30 - 30
ψ, ° 0 0 - -
Unit weight sat, kN/m3 19 19 18 19
Unit weight sat, kN/m3 19 19 18 19

Figure 8. Reinforced soil structure FEM model.

For geogrid an elastoplastic model was obtained with secant stiffness EA1 = 262kN/m and
design strength Np,1=23.5 kN/m that came from a set of creep tests under loads from 5 to 65%
of short-term strength. No surcharge was used. Stability analysis shows factor of safety 2,11
and 10,5cm surface deformation as per Figure 9.

Figure 9. Reinforced soil structure FEM stability and deformation analysis.

2323
 The same structure modeled in analytical software shows a different way of failure with
much lower factor of safety (1,657) as per Figure 10. Grid strain is lower than 0,25% or 1,5cm.

Figure 10. Reinforced soil structure stability and deformation analysis in analytical software.

Modeling a real reinforced soil structure on the automobile road “Perm - Berezniki” in
Perm region of Russia under in-service load resulted in 25cm movement shown in Figure 11.

Figure 11. Reinforced soil wall at “Perm - Berezniki”.

Construction is going to start in the beginning of 2023 with accompanying monitoring. Fur­
ther scientific work is being carried out to compare the results of modeling and monitoring.

5 ALTERNATIVE APPROACH

Lees & Dobie (2021) propose to include stabilisation effect of geogrid into the analysis. This
effect was measured during biaxial testing in a unit cell as an ability to reduce the horizontal

2324
stress and strain once loaded vertically. The biaxial test was then simulated by FEA using
Plaxis 3D software followed by a back-calculation of existing trial wall behavior. It was
recommended to add stabilisation effect as an additional 9° to an angle of friction for
reinforced soil areas with σ3 < 20kPa or 14kPa cohesion for reinforced soil areas with σ3 >
20kPa.
This approach was proved to follow the actual reinforced wall behavior way more accurate
than classic FEM modeling of a membrane elements. But stabilisation effect must be defined
for each individual grid + fill combination during the specific and not standardized test.

6 DISCUSSIONS

This is an important finding in the understanding of the long-term creep related stiffness being
the key design parameter together with classic service life design strength. The present study
confirmed the findings about insufficiency of current FEM models. Previously this fact was
not so evident because of wide usage of short-term stiffness.

7 CONCLUSION

At this moment, analytical modeling methods better correspond to the real life behavior of
reinforced soil structures. Numerical modeling is used for complex geometry cases and is able
to include reinforcement effect. To do this, a full range of creep tests is required for specifying
long-term stiffness at design tension level in reinforcement. Secant stiffness EA1 is obtained
from a creep test isochrone for required service life and not from a short-term tensile test.
Long-term stiffness is 50-100% lower than short-term one and this reveals a problem of FEM
overestimating structure deformations when compared to either analytical designs or actual
soil structure behavior. Discovering of stabilisation effect that geosynthetics provides to a fill
material moves FEM models closer to actual monitored structure deformation but requires
a series of specific and not standardized tests.

ACKNOWLEDGEMENT

The author would like to acknowledge Lau Joe Jiunn of Tensar International - Malaysia for
providing a sensitivity FEM analysis and Mikhail Strapchuk of the JSC “Stroyproekt Insti­
tute” for project support.

REFERENCES

British Standards Institution. 2016. BS8006-1:2010 +A1:2016 Code of practice for strengthened/reinforced
soils and other fills. London: BSI Standards Limited
German Geotechnical Society. 2012. Recommendations for Design and Analysis of Earth Structures using
Geosynthetic Reinforcements – EBGEO, Second Edition. Germany: Wilhelm Ernst & Sohn
NTC GEOPROECT. 2012. ODM 218.3.120-2020 Guidelines for the design of road embankments on soft
foundation soils using geosynthetic materials. Moscow: Rosavtodor
ISO/TR 20432:2007 Guidelines for the determination of the long-term strength of geosynthetics for soil
reinforcement.
Lees A. & Dobie M. 2021 Finite Element Modeling of a Mechanically Stabilized Earth Trial Wall, Trans­
portation Research Record Journal of the Transportation Research Board.

2325