Académique Documents
Professionnel Documents
Culture Documents
ISP7 Baud-Gambin-Heintz AboutMnardmodulus Englishversion
ISP7 Baud-Gambin-Heintz AboutMnardmodulus Englishversion
net/publication/344943256
CITATIONS READS
0 1,348
3 authors, including:
Jean-Pierre Baud
EUROGEO
21 PUBLICATIONS 38 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Jean-Pierre Baud on 28 October 2020.
Translated from French version for ISP7 (S. Varaksin, Apageo, and the authors)
RÉSUMÉ – Dès la définition du module pressiométrique dans une plage expérimentale assez
large pour aller de la pression des terres au repos à la pression de fluage, Louis Ménard a
qualifié ce module de « pseudo-élastique » pour rappeler que son module était à la fois
analogue à un module d’Young, mais non pas élastique, et caractéristique d’une phase de
déformation faible à modérée mais non linéaire et non réversible. L’existence d’une
proportionnalité constante entre le module EM et un supposé « module d’Young » du sol est
démentie par la pratique des essais pressiométriques en autoforage sans déplacement ni
décompression du sol.
ABSTRACT – With his PMT Louis Ménard introduced a pressuremeter E-modulus within a
rather large pressure range bounded by the earth pressure at rest and the creep pressure. He
called it a “pseudo-elastic” modulus, thus claiming it is like a Young’s modulus, but not elastic,
characterizing small to moderate, non-linear and non-reversible strains,. No proportionality is
observed between the EM modulus and an assumed “Young’s modulus” of the soil when
realizing self-bored PMT by the RotoSTAF selfboring method, preventing borehole wall
displacement and decompression.
Civil engineers specialized in geotechnique are used to identify the behavior of actual soils with
“continuum mechanics”, unfortunately reduced most of the time to elasticity and elastoplasticity,
which exactly apply only to a homogeneous, isotropic, that is an ideal medium. However the
history of soil mechanics warns us about some fundamental risks linked to this practice: indeed
the many failed projects remind us that the upper layer of the Earth's crust is neither continuous
nor isotropic. By creating measuring devices, first for the oedometric modulus on samples,
secondly in situ for the pressuremeter modulus, and making soil compressibility measurements
possible, Terzaghi (1920) and Menard (1955) pioneered and were in the same time the first to
encourage the review of the real soil behavior on the survey site, avoiding the adoption of
models based on inappropriate simplifications. Thus it doesn't seem adequate still to consider
systematically that the forecast of the behavior of a structure in a soil is reduced to a site
modelling of soil elements characterized by an elastic Young’s modulus and a friction angle
coupled with a cohesion, whereas:
- soils behavior do not resume in a single deformation modulus, but in an infinite number of
moduli depending on the considered stress level and the deformation under increasing
or decreasing stresses always occurs in an elastoplastic mode (Ménard 1961, Ménard et
Rousseau, 1961);
1
Symposium International ISP7/PRESSIO 2015
- the compression modulus is not equal to the traction modulus (Briaud 2013), thus soils
have no Young’s modulus (Schlosser 2014);
- granular soils and overconsolidated clays exhibit a phenomenon of dilatancy in certain
stress conditions;
- soils get different behaviors depending the applied stress level; they can be perfectly
elastic, pseudo-elastic, plastic, quasi liquid, and the transition between those different
phases is progressive, variable in space and time during the application of stresses
imposed by the construction and earthworks.
Due to automatization and computer improvements, the measuring equipment as well as
the interpretation of pressuremeter tests have greatly evolved. Thus a new type of pressure
volume regulator, the GeoPac® coupled with a new recording rugged computer and
GeoVision software, has been presented during the 6th International Symposium on
Pressuremeter ISP6 included in the 18th ICSMG in Paris (Arsonnet and al. 2013a). Mainly
four prominent breakthroughs are currently possible with the help of this new equipment:
- a precise automatic compensation of the membrane resistance of the probe taking into
account the gap of time between the pressure generated in the measuring cell and in the
guard cells due to the head losses;
- a complete automatization of the test procedure; the automatic calibration of the initial
volume of the probe, so to bring the pressure for the first pressure hold of the test into
equilibrium with the earth pressure at rest, this being equivalent to the “lift-off” of the
Cambridge Insitu Ltd probes;
- a (p, V) diagram showing no inflection point any more.
As a result the precision of the average radial deformation is of the order of 10-5.
All those developments, together with the self-boring procedures already described
elsewhere (Whittle 1999, Dalton 2005, Arsonnet and al. 2005, 2013a) and the
automatization of the running of the test, must be considered as a capital gain, for good test
practice that renders the behaviour of the soil more representative.
The foundation design of a simple construction can be conducted following the "direct"
method proposed by Ménard (Ménard 1963, 1965), by merely taking in account the
pressuremeter modulus EM and the conventional limit pressure pLM, as a result of the
interpretation of the pressuremeter curve. However, for major civil engineering projects, it is
generally necessary to verify soil displacements not only during loading conditions but also
during unloading that occurs in deep excavations, tunnels, during the life of a structure,
vibrations or dynamic impacts generated under the foundation of wind mills for example
(Gambin, 2009). These complex constructions require more detailed study of every
pressuremeter curve.
First of all, we want to submit a typical test performed using the equipment that we
described above. We have chosen a test that has already been published in order to
illustrate a simple hyperbolic model (Baud and al. 2013b) with the details of the measured
data that allows us to determine the corresponding soil properties. It has to be noted that
this test was performed according to more rigorous regulations than those usually applied:
- for example the contact pressure of the probe in the slotted, tube is regulated in order to
make sure that the initial point of the test curve corresponds to an experimental value p0
quite close to the conventionally calculated value of the earth pressure at rest;
2
Symposium International ISP7/PRESSIO 2015
- moreover the initial part of the test is performed with very small pressure increments in
order to be able to measure the corresponding very small strains*.
Figure 1: Example of the results of a Ménard type test, performed by using the self-boring RotoStaf drill
technique measuring p0 by very small pressure steps close to p0, then continued by standard pressure
increments up to the phase of large deformations, close to pLM. Stiff sandy clay, belonging to the
Cenomanian stratigraphic stage, located at the west border of the Parisian Basin (Le Mans). The test is
performed at 18 meters depth.
In the above figure 1 where the classical V=f(p) curve is replaced by dR/R0= f(p), it already
appeared that the hyperbolic model fits well with to the pressure volume curve, this for a visual
precision, when plotting the points of test in a rectangle of approximately. 100 cm², according to
the format of printing. The computation of the secant modulus and, the tangent modulus by
derivation of this hyperbolic model is also quite close to the results obtained by computation
considering the test points except for small initial deformations for which the moduli are
substantially higher the closer you get to the initial measurement. The diagram of figure 2
represents the decreasing of moduli as a function of the deformation. It clearly shows the
difference between the measurements of small strains and deformations, and the usual
measuring range of the modulus EM.
It has been verified that this is not an artifact linked to this type of test, but that the results are it
is repeatable in various types of soils on condition that the probe has been installed
successfully using the self-drilling method of the a slotted tube and thus avoiding any
*This test procedure was not a one-off, in one borehole or during a single site investigation. It has been
verified in different types of soils during other soil investigations surveys.
3
Symposium International ISP7/PRESSIO 2015
deformation of the initial state of the soil, neither through compression nor through relaxation. It
also means you have to start the test by very small pressure increments at least until the middle
of the pseudo-elastic phase is reached. Further on the way to increasing plasticity the number
of test points has no influence any more on the results of the modelisation based on these
points.
From the same set of experimental data, the detailed study of the test can be continued to
model the decrease of the modulus. The inverted S-shaped "pressure, modulus" curves (Fig.1,
right-hand ordinate), plotted both by the secant modulus and the tangent modulus, are fitted to
the double-hyperbola model (Baud et al. 1992) in accordance with EN ISO 22476-4. This model
is commonly used for the V = f (p) curve of any PMT and also fits very well on the point data
series: V = f (p). The two asymptotes of this adjustment do take place as following: the right on
the right merges with the asymptote of the direct hyperbolic curve of the test, which is the true
limit pressure pL. The other on the left is simply limited to the p = 0 value being the asymptotic
curve portion between 0 and p0 being obviously virtual.
In figure 2 the results of these fittings on the same test are represented in a semi-logarithmic
diagram. The strains on the x-axis in a logarithmic scale, the secant and tangent moduli and the
4
Symposium International ISP7/PRESSIO 2015
stress values on the y-axis in an arithmetic scale. It clearly shows that the module decay curve,
as a function of the strain, is not as regular as its representation by Atkinson and al. (1991),
which has become a classic, would suggest: indeed it reveals a distinct phase of lesser
decrease, which corresponds to the “pseudo elastic” phase, as defined by Ménard, to
determine the pressuremeter modulus EM.
The slope change between the phase of micro deformations and Ménard's classic “pseudo-
elastic” phase is well indicated by the "zoom" on the initial points of small strain in figure 3.
Figure 3: Shape of the pressuremeter curve at the beginning of the test, generated by small pressure
increments and very small deformations.
The representation of the module decay curve as a function of the applied pressure, based on
the experimental data, can relatively simply be transposed, in graphic form, in a
“deformation/relative modulus” cartesian diagram. The algebraic expression that is also
possible, would be long to be detailed. It is sufficient to note that, for extremely small strains,
when the deformation tends towards zero, the modulus has a tendency to reach a finite Gmax
value. This allows us to express the reduced moduli values G/Gmax in a range going from 1
and 0. The complete "S" shaped curve presented in figure 4, is obtained by means of a single
pressuremeter test performed from an in situ measure of p0 up to an expansion of the probe
close to pLM It could be represented in a diagram with three axis: G/Gmax, ε, p. To simplify
here, the characteristic points of the test are indexed through their p value. Several remarkable
properties can be observed.
First of all, for deformations of the order of 10 -2, the module decay curve is not quick and
monotonous, as it could be supposed by looking at the curves proposed by Atkinson and al.
(1991) and later by Tani (1994) and Tatsuoka and al. (1997). On the contrary the curve exhibits
a flat section exactly like Ménard had foreseen and declared as the section in which his
pseudo-elastic modulus is measured.
Next, the tendency for the values of ε close to zero and G close to Gmax allows to extrapolate
the value of p0 so that G/Gmax is exactly 1: in our case, this value is approximately 4% less
than the in-situ measurement. It is compatible with an assumed value K 0 = 0,5 and an average
density of 1,94 Mg/m3 of the saturated, sandy clay in which our test was performed at 18 meters
depth. This Cenomanian soil layer in the region of Le Mans, is considered to be only slightly
5
Symposium International ISP7/PRESSIO 2015
overconsolidated after having received during Quaternary the stress of the ice cap, and then
post-glacial rebound and superimposition of the river network. The quite precise knowledge of a
p0 value, obtained by extrapolation, allows us to indicate as well the decreasing evolution of
p0/p during the test, according to the same abscissa ε. Thus, this “S” shaped curve has a rapid
and regular decrease between 10-3 and 10-1. When comparing it with the module decay curve,
three phases become apparent. First, in the phase of a small deformations, as much as 10 -2,
the overlapping of the two curves indeed corresponds to the linear elastic behavior at the origin.
Then, the curves distinctly diverge during the Ménard pseudo elastic phase, until ε = 10-1.
Finally during the large deformations in the plastic phase, the two curves join and cross each
other slightly above the conventional value of the doubling of the volume of the cavity.
Figure 4: The first complete E/Emax “S" shaped curve” as a function of ε, measured and interpreted,
based on a single pressuremeter test.
6
Symposium International ISP7/PRESSIO 2015
In the general manual D60 of the interpretation of PMTs, initially edited in 1967, Ménard already
recommended to use the results of cyclic pressuremeter tests for settlement calculation of
structures founded in deep excavations and for the definition of the subgrade reaction
coefficients under vibrating machines (Ménard 1975). Only 34 years later a first attempt was
made to normalize a stress controlled cyclic Ménard pressuremeter test. Nevertheless it did not
surpass the experimental stage of the normalization procedure (NFP XP 94-110-2 2001)
despite the intentions to introduce cyclic tests in the current investigation practice (Canepa et
Combarieu 2001) and despite the conviction expressed by analysts of the pressuremeter curve
that the reload curve would be is most useful for the determination of the intrinsic parameters of
the soil (Monnet 1990).
The precision of measurements of 1/ 50th cm3 reached by a G type manual-controlled
pressuremeter, operating in high sensibility mode (Ménard 1961), has been improved
nowadays by an automatized test performance conducted by GeoPAC type and HyperPAC
type pressuremeters, able to run cyclic tests according to various predefined loading and
unloading operation modes.
The curves of the second to the nth loading phases, cycled between two chosen pressures in
the pseudo-elastic part of the test, always tend to become more or less linear, thus inciting
theoreticians to recognize a nearly linear elastic soil reaction and the possibility to measure
“the” Young’s modulus of the soil (J. Monnet 1999, 2013). However real soil behaviour does not
generate perfectly linear moduli. All the cycles remain 'almond-shaped', which means that they
have a hyperbolic unload - reload curvature, as shown by the examples here below.
When after one or several cycles the volume of the probe is held constant for some hours or
tens of hours, the soil will relax till the stabilization of the pressure decrease is reached, thus
indicating the value of earth pressure at rest. When up from this point a second test is carried
out the first load moduli EM obtained before and after the relaxation are equal not only in their
value but also in the shape of the curve, point by point. This ascertainment, that was verified in
the Fontainebleau sand and in the smectite clays of the "Rupélian" stage in the region of
Hurepoix (C. Saade, 1981), seems to be attributable to most kinds of soils.
The measurement of the first virgin loading module, under conditions of self-drilling without
disturbance neither decompression at the probe-borehole interface, remains an essential soil
characteristic. It is more relevant than the multiple loading measurements. The contribution of
cyclic loadings is certainly interesting and perhaps fundamental (Ménard and Rousseau 1961)
but one should not base on the unload-reload the hope of eliminating the imperfections of the
hole from the decompression and the absence of perfect support of the wall before starting the
test.
7
Symposium International ISP7/PRESSIO 2015
Figure 5. Single cycle test. Fontainebleau sand, self-drilling STAF depth 5 meters, Bruyère-le-Châtel
(91).
For this purpose, the precision of volume measurement is compulsory in order to obtain a
representative measurement of very high moduli. This from the first loading, either when the
second loading induces deformation three to four time smaller than the first one. The two
exemples here below exhibit test with cycles:
- Figure 7 in a calcareous rock with very few fissures and relatively soft. The core
calcareous deposits being the historical construction material of Paris. Some creep in
this material is observed as soon as 7 MPa.
- Figure 8 in a sedimentary rock of much higher resistance, the Urgonien of Alpilles,
dolomitic calcareous deposit that does not exhibit any creep at 25 MPa.
Figure 7. Test with a cycle between one and five MPa in soft rock (cores calcareous deposits
Sartrouville, 78) and zoom on the loading unloading cycle.
In these “nears elastic” loops, the difference in volume between two points of measurement is
between 0.1 and 1 cm², or a slope dV/dP of 0.2 and 0.3 cm²/MPa, or a deformation of the
borehole of 10-5 to 10-4.
8
Symposium International ISP7/PRESSIO 2015
Figure 8. With three loading cycles between 0 and 25 MPa, without creep neither failure of the
Urgonien calcareous deposit.
Those examples do illustrate how one switches, from the domain of soils with a (p, V) or (p,ε)
exhibiting a hyperbolic curve from the origin, to the domain of rock with cementation and
fracturatio. The (p, V) relation becomes with an initial inverse curvature, the concavity towards
axis of pressure indicating a precise pressuremeter measurement of the closing under radial
stress of the fractures and micro fissures and the inter granular porosity.
Under a stress of 25 MPa (calcareous deposit of figure 8) or until 50 MPa (sand stone of figure
1, Baud and al. 2013c), the closing of the fissures is not terminated and the rock did not exhibit
the beginning of the expansion behaviour. It is in this type of rock that the test should proceed
as far as possible to obtain an initial shearing of the rock matrix, after complete closing of the
fissures; this will be tried with the Hyperpac prototype to 100 MPa.
5. Conclusion
The behavior of cyclic test in rock without reaching failure may help to compare and understand
the nature of cyclic moduli in soils and rocks:
- the strain reached during the first loading in rock is the same at maximum pressure after the
second and third cycle and most probably in case of next cycles, identical with of gap of only
10-6.This is within the range of the experimental error.
- In rock, the increase of secant modulus according to cycles (here between 2 and 25 MPa) is
only due to a partial and incomplete reopening of fissures which have been closed by the first
loading.
- In soils, the increase of the modulus partially originates from the reorganization of grains
under the influence of micro shears, after which there is no return to initial conditions further to
the successive unloading phases, this when they are relatively quickly performed after the first
loading.
- The important difference as compared to firm rocks is that, for soils and weak rocks, there is
a reinitiation of shear immediately at the beginning of successive loadings. This is observed by
the hyperbolic shape of the curve.
9
Symposium International ISP7/PRESSIO 2015
- This irreversibility however is limited in time, since a long phase of rest after loading cycles
allows soil to return to the initial structure of soil grains or rock elements. At the end, recovery
will be complete
6. Expression of thanks
Many thanks to the clients which leave available their sites for the performance of tests: Eiffage
for the figures 1 to 4, CEA-DAM for the figures 5 and 6, Colas-DTP for the figure 7, Botte
Fondations for the figure 8.
7. References
Arsonnet G., Baud J.-P. & Gambin M. (2005) Réalisation du forage pour essais
pressiométriques par un système de tube fendu autoforé (STAF), ISP5 – PRESSIO 2005,
Actes Symp. Intern. Paris, Gambin, M.P., Magnan, J.-P., Mestat, P. (eds.) 22-24 août 2005,
Paris : Presses des Ponts. vol.1 pp 31-45.
Arsonnet G., Baud J.-P., Gambin M. & Youssef W. (2013a) Le GéoPAC®, un contrôleur
pression-volume automatisé pour les essais pressiométriques de qualité. Actes d’ISP6,
CFMS, XVIIIème CIMSG, Paris, consultable sur http://www.geotech-fr.org/ressources-
documentaires/congres/symposium-isp6.
Arsonnet G., Baud J.-P. & Gambin M. (2013b) RotoSTAF®, une amélioration déterminante de
l’autoforage du pressiomètre Ménard. Actes d’ISP6, CFMS, XVIIIème CIMSG, Paris,
consultable sur http://www.geotech-fr.org/ressources-documentaires/congres/symposium-
isp6.
Atkinson J.H. & Sällfors G. (1991) Experimental determination of soil properties. 10e ECSMFE,
Firenze, 3, pp 915-956.
Baud J.-P., Gambin M. & Uprichard S. (1992) Modeling and Automatic Analysis of a Ménard
Pressuremeter Test, Géotechnique et Informatique, Actes du Colloque ENPC, 29 septembre
– 1er octobre 1992, Paris, Presses des Ponts, pp 25-32
Baud J.-P. & Gambin M. (2013a) Détermination du coefficient rhéologique α de Ménard dans le
diagramme Pressiorama®. Actes du 18ème CISMG, Paris 2013.
Baud J.-P., Gambin M. & Schlosser F. (2013b) Courbes hyperboliques contrainte–déformation
au pressiomètre Ménard autoforé, diagramme Pressiorama®. Actes du 18ème CISMG, Paris
2013.
Baud J.-P., Gambin M. & Heintz R. (2013c) 50 MPa Ménard PMTs Help Linking Soil and Rock
Classifications. Actes d’ISP6, pendant le 18ème CIMSG, Paris, disponible sur
http://www.cfms-sols.org/actes-du-colloque. Ce document est également disponible sur le
site d’ISP6 http://www.geotech-fr.org/ressources-documentaires/congres/symposium-isp6.
Bisgambiglia J.-A. (1976) Essais pressiométriques cycliques. Le Pressiomètre dynamique.
Essais alternés. Mémoire de fin d’étude. Ecole Centrale des Arts et Manufactures, Sèvres.
Bourgeois E., Coquillay S. & Mestat P., (2005) Exemples d’utilisation d’un modèle élasto-
plastique avec élasticité non-linéaire pour la modélisation d’ouvrages géotechniques, Bull.
Liaison P&C n°256-257, Paris.
Briaud J.-L. (2013) The Pressuremeter Test: Expanding its Use, para. 7, Proc. 18th ICSMGE,
Paris, consultable sur http://www.cfms-sols.org/actes-du-colloque.
Combarieu O. & Canepa Y (2001) L’essai cyclique au pressiomètre, Bull. Liaison des P & C, n°
233, pp 37-65.
Dalton C. (2005), United Kingdom Experience with Pressuremeter 1982-2005, Symposium
ISP5-Pressio2005, Presses des Ponts, Paris, vol.2, pp 201-208.
Gambin M. (2009, 2012) Réflexions sur les fondations des éoliennes, document transmis au
CFMS, non publié. Disponible par
http://www.apageo.com/upload/medias/documents/8_michel-gambin---communication-
recentes-fruk---2010_321.pdf
10
Symposium International ISP7/PRESSIO 2015
11
View publication stats