Use of Ménard pressuremeter modulus in finite element

models for retaining walls design

François PHILIP
EGIS Géotechnique, Lyon, France, [email protected]
Ziyad El Balqui
Fugro, Nantes, France, [email protected]
Pierre SCHMITT
Soletanche Bachy, Rueil-Malmaison, France, [email protected]
Jean-Pierre BAUD
Eurogéo, Avrainville, France, [email protected]

ABSTRACT: For the design of the permanent diaphragm walls of the underground stations of line 3 phase 3 of the Cairo
metro project, pressuremeter tests were carried out with an unloading / reloading cycle. The goal was to help to define
the soil behavior laws parameters in finite element calculations with Plaxis software. This work has been further
developed within the framework of the ARSCOP national research project and is divided into three parts:

• Link between the deformation modulus and the pressuremeter modulus, history and recent interpretation for
application to retaining walls calculations using the finite element method.
• Modeling pressuremeter tests with unloading-reloading cycle with Plaxis, sensitivity analysis on soil parameters.
Complexity of the non-linear behavior of soils with regard to the use in the FEM codes of a modulus of
elasticity E.
• Calibration of Hardening Soil Model (HSM) parameters of Plaxis, from the pressuremeter tests results, in order
to best fit with the inclinometer measurements results of the Nasser station.

Keywords: Ménard pressuremeter tests, cyclic test, soil deformation modulus, finite elements calculation, retaining walls
and diaphragm walls.

measurement range of the pressuremeter test (distortion

1. Introduction between 1 to 10%) is higher than the rate due to retaining
wall deformation (distorsion about 0.01%), the modulus
This article presents a work carried out within the is purely deviatoric and the deformation measured with
framework of the national ARSCOP project based on the the probe is horizontal.
retaining walls design and construction of the stations of
the line 3 phase 3 of the Cairo metro, a consortium of 4 This article, after a reminder of the relations between
companies Vinci, Bouygues, Arabco and Orascom. the pressuremeter modulus and the deformation modulus,
will deal with pressuremeter tests with unloading /
It appears that retaining walls design with Finite reloading loop carried out on the Cairo metro
element calculations are increasingly used in addition to construction site. Then a calibration result of the
traditional methods. The case of the Line 3 phase 3 parameters of an HSM law will be presented from the
project of the Cairo metro is a typical example, the finite inclinometer measurements of the diaphragm walls of the
element calculations were required by the developer Nasser station.
N.A.T. (National Authority for Tunnels). In this context
the geotechnical engineer has to face the difficulty of 2. Link between soil moduli and Ménard
choosing the parameters of soil behavior laws to provide
a sufficiently reliable prediction of the forces and the pressuremeter modulus
deformations of the retaining walls especially regarding The link between soil moduli and pressuremeter
the neighbouring structures in a very dense urban area. modulus was initially based on Louis Ménard’s analysis
Another difficulty, because of the lack of international and calculation of shallow foundations settlement.
papers on this topic, is to agree with the other consultants As explained in his articles “Ménard et al [1]&[2]“,
and checkers on the parameters chosen to obtain the Ménard, considering that settlements are nothing but a
approval. combination of isotropic compression and pure shear
strains, relied on two linear elastic analytical
For the engineer, the pressuremeter test allowing an in formulas :
situ measurement of a deformation modulus appears
• Isotropic component : shrinkage of a sphere
ideal. But the use of this modulus to feed computation
submitted to an isotropic loading
models is not direct. The deformation rate of the
 1
𝑤𝑠 = 𝜎. 𝐷 In the case of a normally consolidated soil, for which
9.𝐸𝑀
initial stress conditions may be described by the at rest
• Deviatoric component : dilation of a earth pressure coefficient K0 = 1-sin, and for which
spherical cavity, inflated within an elastic plastic behavior may be described by Coulomb criterion,
1+𝜈
layer 𝑤𝑑 = 𝜎. 𝐷 associated with a passive earth pressure coefficient Kp =
3.𝐸𝑀
(1+sin)/(1-sin), it may easily be shown that in situ
Then Ménard empirically adjusts these theoretical initial conditions theoretically correspond to a deviatoric
formulas in order to match measured settlements. stress 1-3 that is half the ultimate plastic deviator (1-
Having demonstrated that the pressuremeter modulus 3)l.
is purely deviatoric, and observing that some categories It is then possible to make a link with the conventional
of soil are less resistant than others when submitted to secant modulus E50 that serves as a reference in the HSM
specific loading such as pure extension, which tends to model : as a matter of fact it may be shown that, in the
generate tensile stress, he suggests that the pressuremeter case of a hyperbolic behavior, the tangent modulus is in
modulus is finally corrected by a « structural » each location equal to half the secant modulus, so that it
coefficient, , in order to obtain a soil modulus EM/ ≈ can be concluded that E50 = 2.Ei = 2.EM/.
Eoed more representative of usual spheric loading The latter relationship is nowadays more especially
conditions, which include a significant part of isotropic used as it reasonably matches a significant number of
compression, while the same rheological coefficient was back-analysis calculations, based on measured
used to adjust deviatoric part of settlement to the size of displacements of retaining structures in normally
the loading, and completed by classical shape coefficents consolidated soils (higher values being generally
l. Ménard ([3] fixed in 1975 his equation for settlement encountered in overconsolidated soils).
of an isolated footing, keep in French codes from this
moment and always used in Eurocode : When a linear perfectly plastic model is used, such as
𝛼 1+𝜈 𝐷 𝛼 the so-called Mohr-Coulomb model, that does not make
𝑤= 𝜎. 𝜆𝑠 . 𝐷 + 𝜎. 𝐷0 (𝜆𝑑 ) it possible to distinguish between primary loading and
9. 𝐸𝑠 3. 𝐸𝑑 𝐷0
unloading-reloading stress paths, the representative soil
In the case of normally consolidated soils, EM/ modulus that needs to be introduced in the calculation
cannot be considered as a properly-so-called elastic cannot be the same in all circumstances :
modulus, since it clearly corresponds to a primary -when the soil is essentially submitted to a primary
loading, exerted outside of the overconsolidated area ; in loading (foundation built on a normally consolidated
such a case, a link with soil moduli that are nowadays soil), the representative modulus should rather be chosen
often introduced in numerical models could be proposed, close to Ei=EM/ ;
considering for instance the hyperbolic behavior -this is no longer true when the soil is essentially
characteristic of the HSM model Fig. 1, by simply subjected to an unloading-reloading stress path (such as
assimilating EM/ and the initial modulus, Ei, in front of a retaining structure, for which the soil is
representative of the soil behavior in situ (that is in “K0” unloaded due to the excavation and reloaded due to the
conditions). horizontal displacement, or behind the same retaining
structure, where the soil is unloaded due to the horizontal
displacement and possibly reloaded due to prestressing of
anchors).
In the last case, back-analysis calculations generally
show that the representative modulus must be chosen
intermediate between the primary loading modulus Ei =
EM/ and the unloading-reloading modulus Eur,
conventionally equal to 3 . E50 = 6 . EM/.
In practice, a relevant order of magnitude has been
shown to be E = 4.EM/, considering either continuum
numerical models such as finite elements, or traditional
calculations using the subgrade reaction coefficient k,
associated with the classical relationship k = 3.6 /
Figure 1. Graph of hyperbolic behavior and soil moduli (EM/(.a)) [3], where a is the interaction length along
which the retaining structure mobilizes passive earth
This primary loading modulus obviously needs to be reaction.
distinguished, and is substantially lower than the elastic It must be emphasized that both approaches generally
unloading-reloading modulus Eur, when considering a lead to similar results, provided that they rely on the same
triaxial stress path, or than the alternate modulus E a, values of initial primary soil moduli EM/.
when considering the pressuremeter purely deviatoric
stress path, for which Baud et al.[5] propose to retain the In practice, it seems that the only reason why
initial relationship of Menard to define his  coefficient calculations of retaining structure based on pressuremeter
: Ea = EM/2. measurements are rarely used for international projects is
the empirical definition of the  coefficient.
 This is the reason why pressuremeter tests including On Pressiorama diagram [p*LM /p0, EM/p*LM, in Fig.3,
unloading-reloading cycles have been undertaken for the lower sands appears as very homogenoeux level, with a
Cairo metro project. mean E/PL ratio around 7, and extrem values ranging
Test results have been systematically analyzed as part between 4 and 15 ; these rather low values had already
of the national French research project ARSCOP : it may been observed in Cairo sands in previous surveys for
be anticipated that systematizing such tests should enable former Metro lines.
a rational definition of the coefficient  based on the Ea
= EM/2 relationship, that has already been validated by
a significant number of tests.

3. PMT survey for Cairo Metro CML3

Several lines of the Cairo metro have been built in

recent years by the Egyptian JV, joint venture - Vinci,
Bouygues, Arabco and Orascom, on behalf of the NAT.
Geotechnical drilling included core drilling with
identification and geomechanical testing of samples, SPT
profiles, CPT profiles and pressuremeter soundings.
During the successive studies, an increasingly efficient
collaboration was set up between the drilling teams from
Ardaman and the French operators from Eurogéo
specializing in PMT, in order to develop the best pre
boring method for pressuremeter. in the alluvium of the
Nile. mainly sandy-silty, sometimes with more or less
thick clay levels. The risk of drill hole collapse in a
predominantly granular material under water table has
led to the preference of a destructive pre-drilling of 63 to
66 mm in diameter, with the use of a bentonite slurry with
the addition of barite powder, and to perform the tests
using the 44 mm three-cell Ménard probe inside a 63 mm
OD / 49 mm OD outer diameter tube, for each successive
drilling stage for 3 tests. During the placement of the
probe in its slotted tube and the carrying out of tests, a
low flow of bentonite is maintained overflowing from the
Figure 2. PMT results p*LM, EM and EM/p*LM ratio from 3 boreholes
drilling head, so that the mud column and the slotted tube in Cairo Metro Line 3
hold the borehole wall stress conditions before the
beginning of the test as close as possible to the earth
pressure at rest, and ensure less remodeling of the In the sand of Cairo CML3, the average ratio Ea/E M
borehole walls. As drilling progresses by stages, a measured is around 5. This ratio is lower than those found
temporary lining casing (diameter 95-104 mm) is in sands on experimental sites in France and whose
lowered to minimize losses of drilling fluid in a relatively results are presented in “Combarieu et Canepa [9]”. We
permeable medium. believe that this difference is coming from the drilling
The pressuremeter tests used in this article are those of method used in Cairo, which required the use of bentonite
the pressuremeter campaign for the CML3 line, in mud heavier than usual in order to stabilized the borehole
particular for the Sudan, Kit-Kat and Nasser stations. The in sand material under water. This explanation is found
tests were carried out according to the Ménard EN-ISO consistent with the observations of Combarieu and
22476-4 procedure, and all make it possible to determine Canepa [9] who have shown that the drilling method has
an EM module in the pseudoelastic phase and a limit an influence on the value of EM more than on the value
pressure PLM; a part of the tests gave rise to an of Ea..Menard's idea in his article [1] of linking 
unloading-reloading loop in the pseudoelastic phase coefficient to the Ea/EM ratio would therefore integrate
furnishing a cyclic module Ea. the influence of the drilling method in a certain way. We
Figure 2 gives an overview of the results of the 3 can propose (Baud [6] in this ISC6 symposium) to correct
stations cited between 5m and 50m deep, which represent the expression by an index of borehole decompression d
mainly sandy layers. The profile can be distributes in 3 based on curve initial curvature. For Cairo tests, a mean
levels : value d=0.15 has been retained to draw alpha values in
1- from ground surfce to 11 m depth, rather soft soils figure 3. Nevertheless, resulting a value are from 1/4 to
comprises more recent alluvium and urban fills 1/2 and centered on 1/3, current value given for sandy
2- between 11 m and around 22m depth, upper sands soils.
3- from 22m to 50m depth, lower sands.
Figure 3. Classification of PMT results from 3 boreholes in Cairo
Metro Line 3

4. Modeling pressuremeter tests including Figure 4. Pressuremeter test view of Plaxis model

an unloading / reloading cycle The correlations found to obtain the best calibration of
the pressuremeter tests are:
In this part it is presented the calibration results of 2D E50 = 3*EM
Plaxis models trying to reproduce the pressuremeter tests Eur =Ea
with unloading/reloading cycle. As part of the Cairo Line Where: Ea is the alternate modulus of a cyclic
3 Phase 3 metro project, Eurogéo has carried out several pressuremeter test.
pressuremeter tests with an unloading/reloading cycle at Based on the above, it should be noted that only tests
different depths. Only tests that took place in sands are with a ratio Ea/EM larger or equal than 6 can be modeled
treated in this part and have been used fot FEM models with the Plaxis HSM law. In fact, this law requires that
by Z. El Balqui et al. [7]. the Eur modulus to be greater than or equal to 2*E50:
Ea = Eur with Eur ≥ 2*E50 ➔ Eur ≥2*3EM ➔ Ea/EM ≥ 6.
The model is axisymmetric, the borehole is modelled
with a radius of 30cm whose depth is 1m below test’s The table below summarized the tests modeled under
depth. A hydrostatic pressure, calculated with a density Plaxis2D:
of 11 kN/m3, acting as bentonite is applied to the walls of Table 1. Summary of pressure-
the borehole. meter tests modelled with
Plaxis
Test Ratio Ea/EM Model Quality of
calibration
SUDAN z=9m 6,8 HSM Very good

The soil is governed by the Hardening Soil Model SUDAN z=12m 4,8 Good

(HSM): non-linear elastic combined with the Mohr- SUDAN z=24m 4,8 Good
Coulomb plasticity criterion. MASPERO z=15m 4,4 HSM Acceptable
The Hardening Soil Model on Plaxis is characterized
MASPERO z=18m 4,5 HSM Acceptable
by several reference parameters E50,ref, Eoed,ref et Eur,ref, pref
MASPERO z=22,5m 5,9 HSM Very good
and m and allows to take into account for an increase in
the modules with the minor main stress. For the sake of MASPERO z=34,5m 4,7 HSM Acceptable

simplification, this effect of increasing the modules has MASPERO z=37,5m 4,8 HSM Good

been canceled by taking m = 0, then we have : NASSER z=18m 3.0 Poor

E50 = E50,ref KIT-KAT z=12m 6,2 HSM Very good

Eoed = Eoed,ref KIT-KAT z=24m 5.0 Good

Eur = Eur,ref KIT-KAT z=30m 3,8 Poor

KIT-KAT z=33m 3,1 Poor

KIT-KAT z=39m 6,7 HSM Very good

KIT-KAT z=46,5m 4,1 Acceptable

BH z=30m 6.0 HSM Very good

 The calibration quality of the tests having a ratio E a/EM EM Ea (MPa)
around 6 is very good, while those whose ratio is much
lower than 6 are impossible to reproduce on Plaxis with (MPa)
a Hardening Soil Model. Test 3,5 21,8

Below are presented more in details the calibration HSM 1 3,2 27,5
results of two pressuremeter tests: HSM 2 3,2 28,7

- Station Sudan, depth of test z=9m

Table 2. Sudan Plaxis calibration
parameters
It should be noted that to match with the second part
E50 (MPa) Eur (MPa) φ ψ c (kPa)
of the curve beyond the loop it has been necessary to
choose high shear resistance parameters for sand. Lower
HSM 1 20 40 45 10 1 shear resistance values led to a fast plasticity compared
HSM 2 18,5 37 45 12 5 to the curve of the test.
It must also be remembered that we have been limited
by the law HSM on Plaxis which imposes Eur ≥ 2E50,
without this limitation the calibration of the tests whose
ratio Ea/EM is less than 6 could have been possible.

To conclude, this method of calibrating a test with an

unloading / reloading cycle using a Hardening Soil
Model is relatively simple to implement and for the ratio
Ea/EM ≥ 6 the modulus are well reproduced. The
calibrations give Eur = Ea, which shows that the
calculation model is relevent to reproduce the
pressuremeter test in the cyclic part, where the behavior
of the soil is of linear elastic type. Nevertheless some
questions raised. As expected, things are more complex
Figure 5. Sudan pressuremeter test depth z=9m for the simulation of the first loading part of the curve,
Table 3. Sudan Pressuremeter and for which we can only make two extreme
modulus assumptions:
EM Ea (MPa)
(MPa) 1. the tangent modulus of first loading is equal to
Test 5,2 35,2 EM/α, which should correspond to a secant
HSM 1 5,2 36,5
HSM 2 5,2 35,3
modulus E50 of the order of 2EM/α = 6.EM in
accordance with §2; however, this assumption
assumes that the model is able to simulate a
- Kit-Kat station, depth of test z=12m triaxial test and a pressuremeter test with a
Table 4. Kit Kat Plaxis calibration single set of parameters, which is equivalent to
parameters admit that the empirical "coefficient of
E50 (MPa) Eur (MPa) φ ψ c (kPa)
structure" α introduced by Ménard to take into
HSM 1 11,5 23 40 10 5
account precisely the different response from
HSM 2 12 24 38 8 5
the soils according to their nature in these two
fields of extremely different constraints
(vertical compression in one case, horizontal
extension in the other), can be found by a
simple numerical calculation, which is
unlikely.
2. the tangent modulus of first loading is equal to
EM, so a secant modulus of the order of 2.E M,
which seems a priori more logical for the
simulation of a pressuremeter test which is not
triaxial, but which is may be not well adapted
to take into account for the fact that, as
reported by Flavigny and / or Cambou, the
pressuremeter test in extension induces from
the beginning a significant plastification,
which makes that the modulus measured is
perhaps an apparent modulus, lower than the
Figure 6. Kit Kat pressuremeter test depth z=9m
intrinsec modulus that must be introduced into
the law of behavior?
Table 5. Kit Kat Pressuremeter
In any case, the calibrations give an E50 modulus of the
modulus order of 3.5EM, which is well within the predictable range
of 2 to 6. In addition, the fact that it is necessary to The thickness of external DWall of box 1 to 4 has been
introduce unrealistic shear parameters to calibrate the set to 1.5m and the panels size on site have been limited
model well show the difficulty of a model essentially to the width of the hydrofraise tool equal to 2.8m. The
based on triaxial tests to reproduce a pressuremeter test, purpose was to limit as most as possible the displacement
as reported from the beginning by Ménard, a difficulty to minimize the impact on the surrounding buildings.
which is logical to find when looking for the modulus
value. Sequence of works:
Nasser station is the first station crossed by the TBM
5. Presentation of Nasser station which starts from Attaba station. In order to allow the

Figure 7. Plan view of Nasser station TBM to go through Nasser station as early as possible the
Nasser station is one of the major underground station 6 boxes have not been excavated in the same time. Boxes
of the line 3 phase 3 extension. The whole station is 1 to 4 have been first excavated and Boxes 5 and 6 have
divided in 6 boxes where the boxes 1 to 4 are separated started once the raft of Boxes 1 to 4 has been completed
together by three transversal bentonite cement non and the TBM is inside the station.
reinforced diaphragm walls with purpose, in case of
leakage of the plug, to limit the reinjection works to one The sequence of works for box 1 to 4 is as follow:
box (injected plug made of soft gel, bottom at 70m depth 1. Dwalls construction and injected plug
and thickness about 12m). Whereas the boxes 5 and 6 are 2. Pumping tests inside each boxe and draw
separated by two structural reinforced diaphragm walls down of ground water inside boxes 1 to 4 at
required for the service stage. Below the plan view of the -13m
station with the main dimensions and showing the 3. First excavation to 18.3m
buildings arround at a very close distance from the DWall 4. Casting Roof
(sometimes not more than 2m). 5. Excavation to 9.8m
6. Casting Struts 1 axis at 10.5m
The geotechnical context is relatively homogeneous 7. Excavation to 4.3m
along the station, it is encountered from the top to bottom: 8. Casting Struts 2 axis at 5.0m
• A layer of backfill of about 3m 9. Excavation to -0.5m
10. Casting Intermediate slab axis at 0.05m
• A layer of clay of about 9m
11. Excavation to -4.3m
• A layer of upper sand moderately
12. Casting Struts 3 axis at -3.6m
dense to dense of about 10m
13. Excavation to -8.3m
• A layer of lower sand very dense 14. Casting Struts 4 axis at -7.6m
15. Bottom excavation at -12.4m
The ground level is at +20.5m, the deepest excavation 16. Casting Raft
level is at -12.5m and the Ground water level at 18.4m.
17. Removing struts 3 and 4
Top level γ cu c' ϕ' E' α EM
(m asl) (kN/m3) (kPa) (kPa) (°) (Mpa) (-) (Mpa)
Fill 20,5 17 - 0 27 17 0,50 4,2
Clay 17,0 19 80 0 27 18 0,67 8,0
Clayey silt North 12,7 19 70 0 27 16 0,50 4,0
Clayey silt South 12,7 18 50 0 27 11 0,50 2,9
Upper sand 8,0 19 - 0 40 150 0,33 16,0
Lower sand -2,0 20 - 0 40 350 0,33 38,0

Table 6. Nasser design geotech-

nical parameters
 Figure 8. Section of Nasser station
To be noted that Nasser station is still under
construction at the present time of writing this article, and
6. Plaxis HSM calibration results from the available inclinometers results stop just after casting
inclinometers the raft and removing the struts above of boxes 1 to 4 and
before to start the works of boxes 5 and 6.
The calibration calculations were carried out from the
same Plaxis sections as the ones used for the design of In following figures 9 to 13 are presented the calibratio
the diaphragm walls. It should be noted that the design results of section CS 15 North (INC4) during progrssof
has been done with a linear elastic law perfectly plastic building :
using the Mohr Coulomb criteria where the elastic 1- Punping test
modulus have been evaluated by means of the correlation 2- Excavation to Strut 2 at -4.3m
E = EM/α2 [4], with EM the pressuremeter modulus and α 3- Excavation to Strut 3 at -4.3m
the Ménard's reological coefficient. It can be noted that 4- Excavation to Strut 4 at -8.3m
this correlation lead to slightly lower values than the one 5- Excavation to Raft at -12.4m
found for retaining wall E = 4EM/α [8] but it has been
already difficult to convince and get the approval in the
particular context of having in Egypt a consultant not
very familiar with the pressuremeter tests and whose
reference books in english reported the correlation
E = EM/α.
For the calibration calculations, the Plaxis section
corresponding to the inclinometers position have been
used in which the only modification was to use the HSM
constitutive law instead of the linear elastic perfectely
plastic model. Several set of parameters have been tested
and the best fit has been found with the following HSM
parameters:
• Pref = 100 kPa
• m=0
• E50 = 2*EM/α
• Eoed = E50
• Eur = 3*E50
6.1. Calibration results, section CS15
North: Pumping test

INC4 North HSM2 - Pumping test

Inc4 - 04-07-2018 (mm) Pumping test

Plaxis-Ph2 Pumping test

25

20

15

10

Elevation (m) -5

-10

-15
Figure 9. Calibration results – Phase pumping test
-20

-25

-30

-35
-50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20
Ux (mm)
6.2. Calibration results, section CS15 6.3. Calibration results, section CS15
North: Excavation to Strut 2 at -4.3m North: Excavation to Strut 3 at -4.3m

INC4 North HSM2 - Exc ST2

Inc4 - 25-11-2018 (mm) Exc ST2
INC4 North HSM2 - Exc ST3
Inc4 - 17-02-2019 (mm) Exc ST3
Plaxis-Ph5 Strut 1 box 3
25 Plaxis-Ph7 Inter 2 box 3
25

20
20

15
15

10
10

5
5

0
Elevation (m)

0
Elevation (m)

-5
-5

-10
-10

-15
-15

-20
-20

-25
-25

-30
-30

-35
-50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 -35
Ux (mm) -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20
Ux (mm)
Figure 10. Calibration results – Phase Exc. Strut 2
Figure 11. Calibration results – Phase Exc. Strut 3
6.4. Calibration results, section CS15
6.5. Calibration results, section CS15
North: Excavation to Strut 4 at -8.3m
North: Excavation to Raft at -12.4m

INC4 North HSM2 - Exc ST4

Inc4 - 04-03-2019 (mm) Exc ST4 INC4 North HSM2 - Exc RAFT
Plaxis-Ph8 Strut 3 box 3 Inc4 - 15-04-2019 (mm) Exc RAFT
25 Plaxis-Ph9 Strut 4 box 3
25
20
20
15
15

10
10

5
5

0
Elevation (m)

0
Elevation (m)

-5
-5

-10
-10

-15
-15

-20
-20

-25 -25

-30 -30

-35 -35
-50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20
Ux (mm) Ux (mm)
Figure 12. Calibration results – Phase Exc. Strut 4 Figure 13. Calibration results – Phase Exc. To raft
7. Conclusion 8. References
The exercise of reproducing pressuremeter tests of the [1] Louis Ménard et Jean Rousseau « L'évaluation des tassements,
Tendances nouvelles », Sols-Soils-n°01-1962
Cairo metro using a Plaxis finite element calculation
using HSM law allowed to obtain very good calibrations [2] L.Ménard et J.-P. Dauvisis « Etude expérimentale du tassement et
but only when the ratio Ea/EM is at least equal to 6. This de la force portante de fondations superficielles » (Experimental
comes from a limitation of the HSM law in Plaxis which study of settlement and bearing capacity of superficial founda-
tions), Sols-Soils n°10-1964 pp 11-23 (in French, synopsis in En-
imposes Eur,ref ≥ 2*E50,ref. However, there is a need to glish, German, Spanish)
introduce abnormally high shear parameters and the
correlations between HSM and pressuremeter parameters [3] L.Ménard « The interpretation of pressuremeter test results »,
differ from those found for retaining walls calculations. Sols-Soils n°26-1975, pp 5-44

The link between a deformation modulus and the [4] P. Schmitt « De l'importance du suivi pour maîtriser le dimension-
nement des ouvrages géotechniques « (The importance of moni-
pressuremeter modulus should not be limited to the toring to control the design of geotechnical retaining structures )
correlation E = EM/α which leads to particularly wrong Revue française de Géotechnique n°126-127 1er et 2e trimestres
results such as for retaining walls calculations [10]. For 2009 (in French)
retaining walls, it has been shown [8] that it is possible to
[5] J.-P. Baud & M. Gambin “Soil and Rock Classification from High
use a HSM law whose input parameters can be correlated Pressure Borehole Expansion Tests” Geotechnical and Geological
with the pressuremeter modulus as follows: Engineering ISSN 0960-3182 Volume 32 Number 6 - Geotech
• Pref = 100 kPa Geol Eng (2014) 32:1397–1403
•m=0
[6] J.-P. Baud “Soil and Rock Classification from Pressuremeter
• E50 = 2*EM/α Data. Recent Developments and Applications” ISC’6, Budapest,
• Eoed = E50 2020.
• Eur = 3*E50
The calibration exercise of a Plaxis HSM calculation [7] ARSCOP “Projet National nouvelles Approches de Reconnais-
sance des Sols et de Conception des Ouvrages géotechniques avec
of Nasser station’s diaphragm wall, using the correlations le Pressiomètre”) Rapport interne Z. El Balqui, F. Philip & J. Ves-
above, with the results of the inclinometer measurements cuere “ Utilisation des essais pressiométriques Ménard pour l’ap-
gave good results. The curvatures of the retaining wall plication aux calculs des soutènements par éléments finis” (Use of
calculated with Plaxis are found, at each phase of Ménard pressuremeter tests for the calculations of the retaining
structures with finite elements method) (2018) (in French)
construction, to be relatively comparable to the
inclinometer measurements. This calibration exercise, in [8] K. Serrai, C. Plumelle, P. Schmitt “Analysis of measured deflec-
a new context, reinforces the results found previously [8] tions of a diaphragm wall in Colombes using finite element calcu-
and then increases our confidence in finding a good lations” 16th International Conference on Soil Mechanics and Ge-
otechnical Engineering – 2055-2006 – doi: 10.3233/978-1-61499-
prediction of retaining wall deformations following this 656-9-1119
approach.
The use of finite element calculations is in certain [9] O. Combarieu, Y. Canepa, “L’essai cyclique au pressiomètre“
configurations the only way to apprehend a satisfactory (The cyclic test with the pressuremeter) Bulletin des laboratoires
des ponts et chaussées - 233 - JUILLET-AOÛT 2001 - RÉF. 4381
model of the problem, hence the importance of having a - PP. 37-65 (in French)
reliable established calculation methodology. But they
must not unnecessarily replace the traditional method of [10] O.Combarieu, “L'usage des modules de deformation en
subgrade reaction modulus, which has the enormous geotechnique“ (The use of deformation modules in geotechnical
design), Revue française de Géotechnique n°114 1er trimestres
advantage for the engineer to be simple and fast to 2006. (in French)
implement.
We will conclude by recalling the importance of the
instrumentation and the follow-up of the deformations of
structures in interaction with the soil, allowing at the
same time to control the predictions made by the
calculations, to anticipate a possible wrong prediction
allowing for modifying the construction with a
reasonable economic impact, and to continue to improve
models through calibration exercises.