See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282059623

Determination of soil stiffness properties

Article · June 2010

CITATION READS

1 555

2 authors, including:

Philippe Reiffsteck
Université Gustave Eiffel
129 PUBLICATIONS   348 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Evaluation of preconsolidation stress from laboratory and in-situ strength parameters View project

Subgrade improvement wit the technique of Deep Soil Mixing View project

All content following this page was uploaded by Philippe Reiffsteck on 16 November 2020.

The user has requested enhancement of the downloaded file.

Ingeokring Newsletter

Determination of soil stiffness properties

H.C. van de Graaf (Lankelma Ingenieursbureau, The Netherlands)
Ph. Reiffsteck (Laboratoire Central des Ponts et Chaussées, Paris, France)

Abstract

A good knowledge of the soil stiffness properties is very important for geotechnical engineering. Stiffness parameters are required as input for a lot of
geotechnical calculations related to settlements or horizontal deformations. Since the latest decennia the available calculation methods are getting
more and more sophisticated. Finite element methods like PLAXIS are nowadays widely used, as they have become accessible not only for specialized
engineers, but for common practice. As a consequence the need for good methods for the assessment of the soil parameters is growing strongly. Soil
stiffness parameters can be determined by in-situ testing or by laboratory testing. Especially for in-situ testing a wide scope of different tests is avail-
able worldwide. However, the choice of the test type depends strongly on the geotechnical tradition of a country. The use of foreign methods in a
country is now stimulated by the development of international standards the scope of in-situ and laboratory testing. In this paper is dealt with a large
scope of tests. For each test a short description of the test principle is given, an example of a test result is shown and the field of application of the test
results is showed. Furthermore the practical advantages of each test are evaluated.

Introduction Description of different in situ testing methods

When making an assessment of stiffness parameters of a
soil, the first stage should be to make a soil model consisting
of a number of layers with different soil stiffness properties.
This soil model can be generated quickly by using CPT’s
(Cone Penetration Tests) and/or borings with sampling. The
second stage consists of determining the parameters of the
different layers by in situ and/or laboratory testing.
On or more of the following test methods may be used:
In situ testing
• Ménard pressuremeter test
Ménard Pressuremeter Test
Developed in France by Louis Ménard more than 50 years
ago, it is nowadays the most common in situ testing tech-
nique for all purposes in France: foundation design, settle-
ments, retaining walls, etc. More than 1000 rigs are carrying
out 15000 tests annually, with a vertical spacing between
tests of 1.5 m. Outside of France the test is mostly used for
special projects only. The principle of this test is that a meas-
uring cell (height 200 mm, diameter 60 mm; see Figure 1) is
inflated with water (allowing measurement of injected vol-
ume) so that its diameter is increasing while the soil is

• Cone pressuremeter test

• Cone Load Test

• Marchetti dilatometer test

• Seismic CPT

• Plate bearing test

Laboratory testing
• Oedometer test

• Triaxial test

Assessment by correlation from CPT

Fig. 1 Ménard pressuremeter equipment

Special Investigation Techniques edition | Spring 2010 | 41

 Ingeokring Newsletter

pushed away horizontally. Above and below this measuring 0 20 40 0 1 2

cell are guard cells that expand using gas. This results in a 0 0
plain strain measurement condition, which makes the inter- 4 EM (MPa) 0,37 pl (MPa)
pretation of test results easier. Inflation of the three cells is 2 6 2 0,44

performed stepwise in 10 stages until ‘failure’, which is de- 10 0,58

fined as an increase in diameter of the measuring cell until 4 15 4 0,90

twice the initial volume. At each stage the volume of water 20 0,94

in the cell is measured, resulting in a curve (see Figure 2). 6 29 6 1,15

30 1,26

800
8 29 8 1,26

31 1,40
700 Vs=535 cm3

corrected pressure p*LM 10 24 10 1,38

volume (cm3) and 10.(V60-V30) (cm3)

600 V60-V30
32 1,54

500 12 30 12 1,59

28 1,69
400

14 27 14 1,70
300
Ménard pressure modulus E M z (m) z (m)
200
16 16
100
Fig. 3 Ménard pressuremeter logs: Pressuremeter modulus (left) and
0 limit pressure (right)
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8
corrected pressure (MPa) limit pressure p l
slotted tube is rather flexible regarding dilatation. As the soil
1 is pushed away when penetrating with the slotted tube, the
Merville -Ménard pressuremeter
0,9 soil has already been subject to displacement when starting
0,8 the actual test. This inconvenience can be overcome by us-
0,7 ing the so called STAF-method, at which the slotted tube is
cell pressure (MPa)

0,6 emptied by inside drilling during its penetration. This is

2m
0,5 4m therefore a self boring pressuremeter installation technique
0,4 8m
but cheaper and with limited risks.
0,3

0,2

0,1 Cone pressuremeter test

0
About 20 years ago the first commercially used cone pres-
0 100 200 300 400 500 600 700 800
3
volume (cm )
suremeter equipment was introduced on the market by
Fugro and Cambridge In-Situ Ltd. The major aim was to in-
Fig. 2 Ménard pressuremeter curves
clude a load-displacement testing modulus in a CPT test.

From this curve the following parameters can be derived:

• Limit pressure Pl

• Creep pressure Pf

• Pressuremeter modulus EM

The pressuremeter modulus EM is derived in an empirical

way from the slope of the load-displacement curve. These
parameters are usually presented related to depth in a pres-
suremeter log, see Figure 3.
a b c d

The preferred installation method is to lower the pressure-

meter cell in a pre-drilled open hole. Where necessary, the Fig. 4 Using a pressuremeter in a Chinese lantern inserted by self bor-
stability of the hole is assured by using bentonite mud. ing: a) Chinese lantern; b) lowering the asymmetric tool; c) rotary
percussion drilling; d) positioning the 44 mm probe in the casing
When hole stability problems are expected, the cell is in-
and performing the expansion test
stalled in a slotted tube and hammered into the ground. The

Special Investigation Techniques edition | Spring 2010 | 42

Ingeokring Newsletter

Qt
> 10.dc lmembrane
cone resistance
dpm pore pressure sleeve friction
q and u fs
CLT 2
25 < dc < 50 mm 60° qc

cone module pressuremeter module

Fig. 5 Cone pressuremeter probe

z z Δh Δh
Δh
The pressuremeter cell is located just above the friction cone settlement
Fig. 7 Principle of the cone load test
sleeve. At depth intervals of e.g. 1 m, the CPT test is inter-
rupted for carrying out a pressuremeter test. The loading
procedure is not stepwise like the Ménard pressuremeter,
but done by a constant flow of water injected into the cell
which results in a constant rate of dilatation. Another differ-
ence from the Ménard test is that there is only one cell (no
guard cell). Like the normal slotted tube version of the Mé-
nard test (which is also a full displacement pressuremeter test),
the cone pressuremeter test starts after displacement of the
soil by the penetration of the probe, so the first part of the
load-displacement curve is missing. Therefore an unloading
and reloading loop is included in each expansion test. Differ-
ent methods are in use to assess the E-modulus from the
slope of the curve. Fig. 8 Cone load test set-up: 1) device allowing to visualize and record
the parameters during the cone loading tests and penetration; 2)
1,4
pump; 3) connection of the hand pump on the hydraulic system
Merville - CPM
1,2 of the CPT; 4) rigid structure anchored in the ground, independ-
1
ent of the thrust machine; 5) displacement transducer; 6) plate
attached to the push rods for measurement of displacements
0,8
pressure (MPa)

7,89 m
3,89 m
0,6
1,89 m

0,4

0,2
E Modulus (MPa)
0
0 20 40 60 80 100
-0,2 0
0 100 200 300 400 500 600 700 800
volume (cm3)
PM1
Fig. 6 Cone pressuremeter curve PM1 bis
PM2
2 CLT raw
CLT corr
TRIAXIAL

4
Cone load test
depth (m)

The principle of this test is to interrupt a CPT test at a certain

6
depth, in order to carry out a load test on the cone while
measuring the vertical displacement of the cone. Like with
the Ménard pressuremeter test, the load is applied in about
8
10 stages until failure. The development of this test started
40-50 years ago and was given a new impulse in 2005 by
Ponts et Chaussées (Paris), Blaise Pascal University 10
(Clermont-Ferrand), Lankelma and Fondasol, by using an
electrical CPT test and a 20 tons CPT rig.
12

Fig. 9a CLT test curves

Special Investigation Techniques edition | Spring 2010 | 43

 Ingeokring Newsletter

2,5

PD1 1 m
2 PD1 2 m
cone pressure load Pp (MPa)

PD1 3 m
C1 4 m
1,5 C1 7,1 m
C1 8 m
C1 9,8 m
C2 4 m
1
C2 7,1 m
C2 8 m
C3 6,5 m
0,5
C3 7,1 m
C3 9,8 m Fig. 11 Dilatometer results
0
0 5 10 15 20 25 30 35 40 45
displacement h (mm)
Seismic CPT
Fig. 9b CLT test curves
This is a standard electrical CPT cone for measuring cone
From the slope of the different parts of the curve the ECLT resistance and local friction, but which also includes one or
and the E50 CLT are derived. The test method has been used more geophones. Every meter the CPT test is stopped for a
for different projects in France and in The Netherlands, in- seismic measurement. For this measurement a compression
cluding validation of the test results by correlating the ob- wave and/or shear wave is generated at soil surface by a
tained E-modulus to the E-modulus of triaxial tests and Mé- hammer. The time elapsed between the impact of the ham-
nard pressuremeter tests. Further validation will be done at mer and the arrival of the wave at the cone is measured.
other projects. From these measurements a diagram results of wave veloc-
ity versus depth. The small strain G-modulus is derived from
this velocity:
Marchetti Dilatometer test
This test was developed in Italy by Sylvano Marchetti, some G = ρVS2
40 years ago. The test equipment consists of a flat spear
equipped with an inflatable membrane, which is pushed The same method can be used by lowering the geophones
into the ground. Pressure readings are taken before inflation in a borehole. The method is more and more used world-
and at 2 mm membrane displacement intervals. From these wide.
readings the dilatometer modulus and other parameters are
derived. This method is used worldwide on a limited scale in
soft soil.

wire
pneumatic
tubing

95 mm 14 mm

60 mm 1,1 mm

flexible P0 P1
membrane
18°

Fig. 10 Marchetti Dilatometer probe (top) and pressure control and

read-out unit (bottom) Fig. 12 Principle of the seismic CPT

Special Investigation Techniques edition | Spring 2010 | 44

Ingeokring Newsletter

Plate bearing test

This method consists of stepwise static loading of a plate.
The E-modulus is derived from the slope of the load-
settlement curve. As settlement under the plate is unknown
it's not a Young’s Modulus. The method is often used for the
compaction control of natural soil or filled materials for shal-
low foundations or road construction.

Fig. 13 Plate bearing test equipment

Description of different laboratory testing methods

Triaxial test
This is a compression test carried out on a soil specimen,
while the specimen is confined in a cell. The first version of
this testing method was the Dutch Cell Test, developed more
then 80 years ago by Keverling Buisman. Some decennia
later the first triaxial tests were carried out in the United
Kingdom. Nowadays, the cell test has disappeared due to
international standardization. The primary aim of the triaxial
test is to determine the shear strength or the internal angle
of friction ρ and the ‘cohesion’, or undrained shear strength.
Over the latest decennia the test is done more and more to
determine also the E-modulus (Young’s modulus) from the
load-settlement curve. There are different testing proce-
dures available:
• Drained or undrained
Fig. 14 Principles of triaxial testing equipment
In The Netherlands the number of triaxial tests carried out
yearly is increasing strongly since 10 years because of the
need for input parameters for finite element computations.
This test allows reproducing the stress history followed by a
soil element close to a structure.
Simulation of a path starting with a Ko consolidation and
then decrease of deviatoric stress as during excavation and
reloading during construction of structure can be easily
done.

• Unconsolidated or consolidated
900

• Isotropic or anisotropic consolidation 800 t5

700 t6
• With or without unloading stage t7
vertical stress (kPa)

600
t8
• Static or cyclical loading 500

400
• Small strain testing using dynamic loading such as
300
bender elements or resonant column
200

• One stage, single stage (three different specimen, con- 100 Merville CU+u

solidated at different stresses, each loaded until failure) 0

0 0,05 0,1 0,15
or multistage (one specimen, consolidated and loaded vertical strain (-)
at three different stresses, at the first two stages loading
Fig. 15 Deformation of a specimen under triaxial loading
is stopped long before failure at small deformation)

Special Investigation Techniques edition | Spring 2010 | 45

 Ingeokring Newsletter

Oedometer test Correlation between soil stiffness and cone resistance

The first oedometer tests were also carried out 80 years ago In literature, relationships between cone resistance and soil
in The Netherlands by Keverling Buisman. The purpose of stiffness can be found (e.g. Baldi et al., 1989).
the drained test is to determine the time-settlement behav-
iour of soil.

Fig. 16 Oedometer equipment

The test specimen in this test is contained in a stiff ring (see Fig. 18 Example of a relationship between cone resistance and soil
stiffness (Baldi et al., 1989)
Figure 16), therefore the deformation of the specimen is
completely different from triaxial testing. Usually the load-
The following table is included in the international standard
ing is done stepwise, the resulting stiffness modulus Eoed is a
Eurocode 7 for geotechnical design (BSI, 2007):
constrained modulus, which is complete different from a
Young’s modulus. In some countries, like The Netherlands,
the stiffness of the specimen is not expressed as an E-
Table 1 Relation between cone resistance and soil stiffness
modulus (strain vs. stress), but as strain vs. the logarithm of
Effective
stress (C). This is because of the fact that for soil, E strongly Cone resis- angle of Drained Young's
Density
depends on the considered strain level. Since a number of tance (qc) from shearing modulus (E’)
index
years the test is also carried out at a Constant Rate of Strain CPT [MPa] resistance [MPa]b
(φ’) [°]a
(CRS test), often combined with the measurement of radial
Very 0.0-2.5 29-32 <10
stress.
Loose 2.5-5.0 32-35 10-20
Me- 5.0-10.0 35-37 20-30
Dense 10.0-20.0 37-40 30-60

1 10 100 1000 log(s'v) 10000

Very >20.0 40-42 60-90
1,40
a) Values given are valid for sands. For silty soil a reduction of 3 should
1,20
be made, for gravels 2 should be added.
Merville
b) E' is an approximation to the stress and time dependent secant
1,00 modulus. Values given for the drained modulus correspond to settle-
ments for 10 years. They are obtained assuming that the vertical stress
0,80
distribution follows the 2:1 approximation. Furthermore, some investi-
void ratio

gations indicate that these values can be 50 % lower in silty soil and 50
0,60
3,77 - 3,80
% higher in gravelly soil. In over-consolidated coarse soils, the modulus
3,74 - 3,77
0,40 7,97 - 8,00 can be considerably higher. When calculating settlements for ground
7,93 - 7,97 pressures greater than 2/3 of the design bearing pressure in ultimate
9,80 - 9,84
0,20 9,73 - 9,76 limit state, the modulus should be set to half of the values given in this
table.
0,00

Fig. 17 Deformation of a specimen under oedometer loading

Special Investigation Techniques edition | Spring 2010 | 46

Ingeokring Newsletter

Please note that such correlations only give a rough indica-
tion. When relevant, for each site a local correlation should
be established by carrying out specific in situ and/or labora-
tory testing. The CPT results may be used for the establish-
ment of the soil model and for the selection of the depth of
the tests. In this way the test results will be more or less rep-
resentative for the considered soil layer, even when the test
is done on a small specimen.
Evaluation of different methods
When selecting a testing method for obtaining stiffness
parameters and when applying the test results in a compu-
tation, the following factors should be taken into account:
• The strain level during the test related to the strain level
of the calculation of the deformation: the smaller the
strain, the higher the stiffness

International standards • The stress level of a test may be completely different

Since 20 years a number of European committees are work-
ing on international standards, replacing national standards.
The most important standard is Eurocode 7 for geotechnical
design (BSI, 2007). In this standard it is specified which spe-
cific test can be used for a certain type of geotechnical engi-
neering project.
Other standards are dedicated to a testing method. For the
tests mentioned in this article, the following standards have
been prepared by different working groups of TC 341:
from the stress level in calculations
• The E-moduli resulting from the different tests are not
automatically Young’s moduli, so they can not be di-
rectly used as input for all computations
• Drilling prior to in situ testing or sampling as well as the
sampling operation itself may have disturbed the sam-
ple, even when the best available techniques have been
used. This disturbance is even a certainty when sampling
cohesionless sand under the ground water table

Table 2 Standards for different test methods In Tables 3 and 4, information on field- and laboratory test-
Document num- ing as described in this article is summarized.
Test type Status
ber
CPT
CEN-ISO-22476-
(mechanical Published (EN ISO 2009)
12 retaining walls
cone)
Shear modulus, G

fondations
CPT (electrical
CEN-ISO-22476-1 Final draft (2009)
cone) tunnels
embankment
Drilling and on soft soils
CEN-ISO-22475-1 Published (EN ISO 2006)
sampling

Ménard pres- 10-6 10-5 10-4 10-3 10-2 10-1 100

CEN-ISO-22476-4 Draft (2007) bender
suremeter shear strain, ln (es)
Laboratory

resonant column
local measurements
Cone pressure-
CEN-ISO-22476-8 No draft available special triaxial
meter triaxial compression test
Marchetti CEN-ISO-22476- surface wave oedometer
Published (TS 2005) down and cross hole
dilatometer 11
In situ

SBP
Seismic CPT - - PMT, PLT, CPT
Plate Loading CEN-ISO-22476-
Draft (2005)
Test 13
secant shear modulus (kPa)

60000
TRIAXIAL
Published (TS, EN ISO BE
Oedometer CEN-ISO-17892-5
2004) 50000 DH
CH
Published (TS, EN ISO SASW
CEN-ISO-17892-8 2004) SC
Triaxial test
40000
4m PMT
CEN-ISO-17892-9 Published (TS, EN ISO
CPM
2004) PAF
30000
CLT C2

20000

10000

0
0,0001 0,001 0,01 0,1 1

strain (-)

Fig. 19 Typical variation of stiffness with strain for most soils

Special Investigation Techniques edition | Spring 2010 | 47

 Ingeokring Newsletter

Table 3 Summary of field test methods

Stress history/initial state Deformation char-
Type Test Strength parameters
parameter acteristics
σho Ko σp E Gmax c φ Drained
SPT C C
Penetration tests
CPT C C
On surface or embedded in bore-
D T
Plate load test hole
Cone loading test C
Pressuremeter D C C
Expansion tests Self-boring pressuremeter D T C D C C
Cone pressurementer C C C
Other tests Seismic cone T T
D: direct measurement; T: theoretically deduced; C: empirical

Table 4 Summary of laboratory test methods

Stress history/initial state Deformation char-
Strength parameters
parameter acteristics
Type Test σho Ko σp E Gmax C φ Drained
Incremental loading D T
Consolidation
Constant rate of strain D D T
tests
Ko D D D T
Conventional T T T T
Triaxial tests
Local measurement D T T T T T T
Cyclic and dy- Resonant column D T T
namic triaxial
tests Bender elements D T T
D: direct measurement; T: theoretically deduced; C: empirical

References and further reading
Atkinson, J.H. & Sallfors, G. (1991). Experimental determina-
tion of stress-strain-time characteristics in laboratory and in
situ tests. In: Associazione Geotecnica Italiana (Eds.), Defor-
mation of Soils and Displacements of Structures, Proceedings
of the 10th European Conference on Soil Mechanics and
Foundation Engineering, Florence, Italy, May 26-30, 1991.
A.A. Balkema, Rotterdam, The Netherlands. Vol. 3, pp. 915-
956.
British Standards Institution (2007). Eurocode 7. Geotechnical
design - Part 2: Ground Investigation and Testing. BS EN 1997-
2: 2007.
Burland J.B. (1989). Ninth Laurits Bjerrum memorial lecture:
“Small is beautiful” - The stiffness of soils at small strains,
Can. Geotech. J., 26, 499-516.

Baldi, G., Bellotti, R., Ghionna, V.H., Jamiolkowski, M., & Lo

Presti, D.C. (1989). Modulus of sands from CPTs and DMTs. In:
Publications Committee of XII ICSMFE (Eds.), Proceedings of
the 12th International Conference on Soil Mechanics and Foun-
dation Engineering, Rio de Janeiro, Brazil, Aug. 13-18, 1989.
A.A. Balkema, Rotterdam, The Netherlands. Vol. 1, pp. 165-
170.

Special Investigation Techniques edition | Spring 2010 | 48

 Ingeokring Newsletter

Lankelma Ingenieursbureau is werkzaam op het gebied van Geo-, Milieu-, en Funderingstechnie. Zij behoort voor wat betreft de Geotechniek tot
de grotere onderzoeksbedrijven van Nederland, met vestigingen in Zuidoostbeemster, Oirschot en Almelo.

Wij willen graag in contact komen met een kandidaat voor de functie van:

COMMERCIEEL GEOTECHNISCH INGENIEUR (M/V)

Zijn of haar taken zullen omvatten:

- Het genereren van offerteaanvragen van grotere opdrachtgevers

- Het opstellen van onderzoeksvoorstellen en aanbiedingen aangaande geotechnisch terrein- en laboratoriumonderzoek
- Het begeleiden en feitelijk rapporteren van het onderzoek

Gedacht wordt aan iemand met een opleiding op universitair- of HBO-niveau met veel affiniteit voor het vak geotechniek. Hij/zij wil zich
zowel commercieel als onderzoekstechnisch ontwikkelen. Daarnaast heeft hij/zij voldoende kennis van geotechnisch advieswerk om een
waardige gesprekspartner te zijn voor de geotechnische adviseurs van onze opdrachtgevers.

Relevante ervaring wordt op prijs gesteld maar is geen harde eis. De coaching en opleiding zal intern worden verzorgd door een zeer er-
varen geotechnisch ingenieur.

Kennis van de Engelse en/of Franse taal is een pluspunt.

De standplaats is in principe Zuidoostbeemster. Part-time detachering bij andere Lankelma-vestigingen behoort tot de mogelijkheden.

Uw sollicitatie kunt u richten tot:

Lankelma Geotechniek Beemster B.V.

T.a.v. de Directie
Nekkerweg 71
Postbus 712
1440 AS Purmerend

Website: www.lankelma.nl
Special Investigation Techniques edition | Spring 2010 | 49

View publication stats