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

net/publication/308265296

Study on the practices for preconsolidation stress evaluation from oedometer

tests

Conference Paper · May 2016

CITATIONS READS

5 3,246

4 authors:

Priscilla Paniagua Jean-Sebastien L’Heureux

Norwegian Geotechnical Institute Norwegian Geotechnical Institute
47 PUBLICATIONS 205 CITATIONS 130 PUBLICATIONS 1,492 CITATIONS

SEE PROFILE SEE PROFILE

S. L. Yang Tom Lunne

Norwegian Geotechnical Institute Norwegian Geotechnical Institute
37 PUBLICATIONS 576 CITATIONS 137 PUBLICATIONS 4,655 CITATIONS

SEE PROFILE SEE PROFILE

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

Shallow CPT interpretation View project

Green sOil stAbiLisation (GOAL) View project

All content following this page was uploaded by Jean-Sebastien L’Heureux on 24 April 2017.

The user has requested enhancement of the downloaded file.

Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

Study on the practices for preconsolidation stress evaluation from

oedometer tests
P. Paniagua
Norwegian Geotechnical Institute, Norway, priscilla.paniagua.lopez@ngi.no

J.-S. L'Heureux, S.Y. Yang, T.L. Lunne

Norwegian Geotechnical Institute, Norway

ABSTRACT
The preconsolidation stress (pc') is determined from high-quality oedometer test results by a
variety of interpretation methods. Sample quality in the database varies from very excellent to fair
(i.e. class 1-2). Results show that the five methods applied give very similar pc' values when the
samples are of good quality, with pc' varying up to 14% between the five methods.

Keywords: oedometer test, preconsolidation stress, OCR

These methods have in common the

1 INTRODUCTION assumption of a change in soil stiffness from
a stiffer to a softer response near pc';
The standard oedometer tests provide the however, their procedures vary greatly.
basis for investigating one-dimensional (1D) Ideally, interpretation of pc' should not
soil deformation characteristics and soil depend on the chosen interpretation method
stress history. During the tests, soft and nor require subjectivity.
compressible soil generally exhibit bilinear One of the principal objectives of the
stress-displacement behavior when the data ongoing internal Strategic Project "SP8- Soil
are plotted on a semi-logarithmic graph. Parameters in Geotechnical Design"
Below a vertical effective yield stress (σ'vy), (GEODiP) at the Norwegian Geotechnical
the displacements are small; and above that Institute (NGI) is to help achieving a more
threshold value, the soil strains until it consistent interpretation of laboratory and in-
reaches a more compressed structured. The situ results, and to test/develop procedures
σ'vy is known as preconsolidation stress (pc'). for the choice of characteristic design
The ratio between pc' and the in-situ vertical parameters in soft clay. As part of the project
effective stress (σ'v) is known as SP8-GEODiP, the present work compares the
overconsolidation ratio (OCR). results of applying some of the
Most of the clays included in this study are abovementioned interpretation methods [i.e.
not mechanically over consolidated but have Casagrande (1936), Janbu (1969), Pacheco
an apparent over consolidation due to aging: Silva (1970), Becker et al. (1987) and
an apparent pc' considered to be created by Karlsrud (1991)] for determination of pc' and
delayed consolidation (Bjerrum, 1967) and OCR on high quality (i.e. fair to excellent)
various chemical processes, i.e. not by oedometer test results from clay samples
mechanical unloading. The apparent pc' is collected both on- and off-shore. The purpose
considered as a yield stress that can be is to evaluate the applicability of these
estimated from oedometer test results through methods in a wide range of clays; in order to
a variety of methods like those proposed by suggest a more consistent interpretation of
e.g. Casagrande (1936), Janbu (1969), pc'. A short discussion on the evaluation of pc'
Pacheco Silva (1970), Butterfield (1979), on samples of poor quality is presented.
Becker et al. (1987), Oikawa (1987), Burland Finally, the paper evaluates the interpretation
(1990), Karlsrud (1991), Jacobsen (1992), methods in terms of technicalities, source of
Onitsuka et al. (1995) and Boone (2010). errors and practical purposes.

1
Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

2 METHODS FOR EVALUATING PC' 2.3 Janbu (1969)

Janbu (1969, 1989) suggested to base the
2.1 Previous works on this topic determination of pc' on plots of tangent
After Casagrande (1936) proposed a method constrained modulus values versus σv'.
for defining the preconsolidation pressure, However, he did not specify in detail how to
alternative approaches have been suggested, do the pc'-determination. The procedure used
mostly based on empirical observations here is explained in the sketch shown in
regarding the stress and deformations Figure 1b. Janbu's approach uses the
patterns exhibited by the soil during resistance concept. He defined the tangent
oedometer tests. Grozic et al. (2003, 2005), modulus (or constrained modulus), M, as the
Clementino (2005) and Boone (2010) present ratio of the change in effective stress (δσ') to
detailed and graphical summaries of these the change in strain (δε) for a particular load
approaches. increment (i.e. M = δσ'/δε). For a low stress
In particular, Grozic et al. (2003, 2005)
level, around σ'v0, the resistance against
evaluated multiple methods for interpreting
deformation (M0) is large. When σ' increases,
pc' in low plasticity clays and concluded that
this high resistance decreases appreciably
uncertainties affect all methods, and that
owing to partial collapse of the grain
some were more difficult and ambiguous
skeleton. Resistance reaches a minimum (Mn)
than others in application.
As detailed by Boone (2010), scale effects around pc'. Subsequently when σ' is increased
can clearly influence the Casagrande's beyond pc', the resistance increases linearly
method. He details also that the conclusion with increasing σ'. In the overconsolidated
drawn by Grozic et al. (2003) on range M1 (the average between M0 and Mn) is
recommending Becker et al. (1987) and often used in design. Behaviour in the normal
Onitsuka et al. (1995) methods, in addition to consolidation stress range can be
Casagrande (1936) and Janbu (1989) approximated by a linear oedometer modulus
methods, may prove difficult and frustrating M. Hence, for σ' > pc', M = m (σ'-σr') where
in practice, and lead to no better results given m is the modulus number and σr' is the
the level of uncertainty that exist when using intercept on the σ' axis and it called the
these methods. reference stress.

2.2 Casagrande (1936) 2.4 Pacheco Silva (1970)

Casagrande (1936) uses an empirical The method proposed by Pacheco Silva
construction from the void ratio, e, and the (1970) to determine pc' is widely used in
logarithm of vertical effective stress, σv', Brazil. It uses an empirical construction from
curve. To determine pc', a geometrical the e – log σ'v curve.
approach is followed based on Figure 1a: (i) The pc' is determined graphically as shown in
choose by eye the point of minimum radius Figure 1c and explained as follows: (i) draw
(or maximum curvature) on the consolidation a horizontal line (A–B) passing through the
curve (point A in Fig. 1), (ii) draw a initial void ratio (eo) of the specimen, i.e., the
horizontal line from point A, (iii) draw a line specimen void ratio before any load has been
tangent to the curve at point A, iv), bisect the applied; (ii) extend the straight line portion of
angle made by steps (ii) and (iii), (v) extend the virgin compression curve (C–D) until it
the straight-line portion of the virgin intercepts line A–B; (iii) from the point of
compression curve up to where it meets the intersection of lines A–B C–D, draw a
bisector line obtained in step (iv). The point vertical line down until it intercepts the e –
of intersection of these two lines is the most log p′ curve (Point E); (iv) from Point E draw
probable preconsolidation stress (point B of a horizontal line until it intercepts line C–D
Fig. 1). The maximum possible (Point F); and then (v) the stress value in the
preconsolidation stress is shown as point D horizontal axis associated with Point F is pc′.
while E represents the minimum possible Pacheco Silva (1970) method is independent
value of the preconsolidation stress. of the drawing scale (Pinto, 1992).

2
Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

Figure 1. Methods for evaluating pc': (a) Casagrande (Holtz and Kovacs, 1981), (b) Janbu (Lunne et al.
2008), (c) Pacheco Silva (Grozic et al. 2003), (d) Becker (Becker et al. 1987) and (e) Karlsrud (Karlsrud
and Hernandez-Martinez, 2013).

2.5 Becker et al. (1987) As detailed in Karlsrud and Hernandez-

Becker et al. (1987) proposed the work Martinez (2013), the modulus behaves as
method. They defined the incremental work follows (see Figure 1e):
as: a) During loading from zero to the in-situ σv',
the modulus generally increase gradually and
then tends to reach a plateau defined as the
(1) maximum re-loading modulus, M0. The
modulus then drops off more or less linearly
Where and are the effective stresses to a minimum level defined as ML (i.e. Mn),
at the end of i+1 and i loading increments, with corresponding stress defined as σ'ML1.
respectively; and and are the natural After this stress is reached the modulus
strains at the end of i+1 and i loading increases linearly, but for some clays the
increments, respectively. modulus is constant up to a stress level
For overconsolidated specimens (Figure 1d), defined as σ'ML2 before it starts to increase
a linear relationship is fitted in the data below linearly. Janbu’s modulus number, m, defines
in-situ stress and the post-yield (i.e. above the rate of increase beyond this point. This
pc') data is approximate with a linear line defines an M = 0 intercept on the stress
relationship. The intersection of those two axis defined as pr', which is the same
fitted lines represents pc'. definition as used by Janbu (1969). Note that
for very stiff as well as disturbed clays, pr'
2.6 Karlsrud (1991)
may be negative.
Karlsrud (1991) and Karlsrud and b) The definition of pc' is taken as the average
Hernandez-Martinez (2013) proposed a stress at which the tangent modulus starts to
slightly modified version of the Janbu (1969, drop off, until it starts to climb up again
1989) method. along the virgin modulus line.

3
Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

3 APPLICATION OF METHODS TO shows as an example the variation of sample

INTERPRET PC' quality with depth for the data included in the
analysis.
3.1 Soil index properties ∆e/e0
(a) 0
The 169 oedometer tests results (94% as 0,01 0,02 0,03 0,04 0,05 0,06 0,07
0
CRSC and 6% as incremental loading - IL) 2
2
1

used in the present study come from 18 2 2 2 12

2
2 22 2 22
5 11 1 2 1
different sites. The samples were taken with 1
1
2
22 22
22
2
2
2
1
1
2 1 21 2 2 121
2

either 72 mm piston sampler, miniblock

2 1 1
2 1 11
12 2 1 1 2
10
2 2 2 21 1 2 2 22 221 12 1
10,5 2
sampler or block sampler. Table 1 presents a
1
2 2 1
2 2 1 2
2 1 2 22 2

Depth (m)
summary of some index properties of the clay
2
1 1
2 12 2 1
1 2 1 2 221 1 22
1
materials tested. In particular, water content 15 2
2
12
11
22 1
2
2
(w), unit weight (γT), plasticity index (IP),
clay content (clay) and sensitivity (St). 20
2
2 1
1

Table 1. Basic properties of the clays studied

Site
w γT IP Clay St 25 Very good
(%) (kN/m³) (%) (%) (-) to Excellent Good to fair
Eidsvoll 25-35 19-20 13-19 37-48 2-5 1

Emmerstad 40-48 17-18 3-12 27-40 77-225 30

Hvalsdalen 31-39 18-20 9-18 40-49 5-20 ∆e/e0
Kløfta-
Nybakk
32-46 18-19 8-19 33-46 7-135 (b) 0 0,01 0,02 0,03 0,04 0,05
Leirsund 30-39 19 9-18 36-49 5-20 0
Nybakk- 2
3
30-45 17-19 7-28 13-67 1-170 3
Slomarka 4 3 2 3
5
3
3 3 4 33 3 2332 3 33
Lierstranda 32-42 18-20 13-19 31-36 7-15 2 3
3
3
23
4 3 3
2
3
Glava 30-35 18-20 15-30 30-60 7-10 4
3
2 4 2
33 2
Ellingsrud 34-40 18-19 5-8 37 15-61
Klett 25-35 19-20 4-10 30-35 10-240 10 3
4
3
3
3
3 233

Kvenild 30-46 17-19 10-14 31-47 22-63 2

Depth (m)

3
Stjordal 33-43 17-19 7-8 37-38 200
Klett-
26-35 19-20 6-13 30-33 10-160 15
Bårdshaug 3
Nykirke 25-35 20 4-9 20-55 65-80
Onsøy 60-65 15-16 32-42 20-60 4,5-6 20
Ghana 90-145 13-15 70-95 45-65 2-6
Johan
22-38 18-20 20-42 30-45 1-3
castberg
Bothkennar 66-72 15-16 42-53 17-35 8-13
25
Very good Good to fair
to Excellent

Table 2. Evaluation of sample quality, after 30

Lunne et al. (1998). ∆e/e0
OCR ∆e/eo 0 0,01 0,02 0,03
1 to 2 < 0,04 0,04-0,070 0,070-0,14 > 0,14 0
2 to 4 < 0,03 0,03-0,050 0,020-0,10 > 0,10 (c) 5
5

4 to 6 < 0,02 0,02-0,035 0,035-0,07 > 0,07 4 5 5

1: very 5 5
5 4 5 4
Qua- 2: good to 4: very 5
good to 3: poor
lity fair poor
excellent 6

10 5 5 4

3.2 Quality of oedometer tests results

Depth (m)

Very good Good to fair

to Excellent
The quality of the samples tested is evaluated 15
Leirsund Klett
based on the initial void ratio and the axial Bothkennar
Burswood
Klett-Bårdshaug
Kløfta-Nybakk
strain at the in-situ stress according to the 20 Nybakk-Slomarka
Eidsvoll
Kvenild
Laminaria
classification proposed by the Norwegian Ellingsrud Lierstranda
Emmerstad
Geotechnical Society (NGF, 2013). This
Norwegian sea
25 Ghana Nykirke

classification is based on Lunne et al. (1998).

Glava Onsøy
Hvalsdalen OSBS

The quality criteria is presented in Table 2.

Johan castberg Stjordal, parsell Trondheim

30
The data is categorized as high quality (i.e.
Figure 2. Sample quality for oedometer tests, for
class 1 and class 2) which varies from good
OCR between a) 1 and 2, b) 2 and 4 and c) 4 and
to fair and very good to excellent. Figure 2 6. The OCR is shown over the symbols.

4
 Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

700 700 700

(a) (b) (c) ?

600 600 600
pc' (kPa) from Casagrande's method

pc' (kPa) from Karlsrud's method

pc' (kPa) from Karlsrud's method

500 500 500

400 400 400

300 300 300

200 200 200

100 Quality 2
100 Quality 2
100 Quality 2
Quality 1 Quality 1 Quality 1

0 0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
pc' (kPa) from Janbu's method pc' (kPa) from Janbu's method pc' (kPa) from Casagrande's method

Figure 3. Comparison of pc' obtained by: a) Janbu's and Casagrande's methods, b) Janbu's and
Karlsrud's methods and c) Casagrande's and Karlsrud's methods.
8 8 8
(a) (b) (c)
OCR from Casagrande's method

6 6 6
OCR from Karlsrud's method

OCR from Karslrud's method

4 4 4

2 2 2

Quality 2 Quality 2
Quality 2
Quality 1 Quality 1
Quality 1
0 0 0
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
OCR from Janbu's method OCR from Janbu's method OCR from Casagrande's method

Figure 4. Comparison of OCR' obtained by: a) Janbu's and Casagrande's methods, b) Janbu's and
Karlsrud's methods and c) Casagrande's and Karlsrud's methods.

3.3 Results and interpretation In particular, the differences up to 12% are

Figure 3 and Figure 4 show the values of pc' between Janbu's and Casagrande's methods,
and OCR obtained with Casagrande's, Janbu's and Janbu's and Karlsrud's methods. This
and Karlsrud's approaches. These methods result is somehow surprising considering that
give very similar values. Janbu values tend to the Janbu's and Karlsrud's methods are based
be slightly higher than Casagrande (as on the same assumptions. However, in
observed by Grozic et al. 2003) and Karlsrud Karlsrud's method, pc' is calculated as an
values. Karlsrud's and Casagrande's methods average from two single values whereas in
seem to result in similar values. Janbu's method pc' is a single point.
The values show more scatter for high pc'and Since all of them are graphical methods, the
higher OCR that adds uncertainties to the interpreter must have good plot resolution
data. It seems that the data with quality 2 (large enough scale in the range of analysis)
tend to add more deviation to the expected in order to clearly define tangents and choose
trend. Some exceptions are observed for data the values with high accuracy.
with quality 1, that come from Johan Karlsrud and Hernandez-Martinez (2013)
Castberg and Kvenild sites. Data from Johan comment that a comparative study of
Castberg site was difficult to interpret different other methods for defining pc' from
regarding the definition of the tangents in the oedometer tests was applied to 15 of the
graphical methods. Kvenild data is reported oedometer tests in the database used by them.
as quality 1 data, however, this couldn't be Karlsrud's method comes out just about equal
confirmed by ∆e/eo values. The pc' and OCR to the average of the other methods (0,8% on
values vary between 4-12%. These the high side). Our data shows a higher
differences are more visible for high values deviation (i.e. up to 7,5%) when the quality
of pc' (i.e. pc' > 400-500 kPa) and OCR (i.e. of the sample decreases, and the pc' and OCR
OCR > 4). increases.

5
 Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

700 700 700

(a) (b) (c)

600 600 600
pc' (kPa) from Pacheco Silva's method

pc' (kPa) from Pacheco Silva's method

pc' (kPa) from Pacheco Silva's method

500 500 500

400 400 400

300 300 300

200 200 200

Quality 2 Quality 2 Quality 2

100 100 100
Quality 1 Quality 1 Quality 1

0 0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
pc' (kPa) from Janbu's method pc' (kPa) from Casagrandes's method pc' (kPa) from Karlsrud's method

700 700 700

(d) (e) (f)

600 600 600
pc' (kPa) from Becker's method

pc' (kPa) from Becker's method

pc' (kPa) from Becker's method

500 500 500

400 400 400

300 300 300

200 200 200

Quality 2 Quality 2 Quality 2

100 100 100
Quality 1 Quality 1 Quality 1

0 0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
pc' (kPa) from Janbu's method pc' (kPa) from Casagrande's method pc' (kPa) from Karlsrud's method

Figure 5. Comparison of pc' obtained by: a) Pacheco Silva's and Janbu's methods, b) Pacheco Silva's
and Casagrande's methods, c) Pacheco Silva's and Karlsrud's methods, d) Becker's and Janbu's
methods, e) Becker's and Casagrande's and f) Becker's and Karlsrud's methods.

8 8 8

(a) (b) (c)

6 6 6
Pacheco Silva OCR

Pacheco Silva OCR

Pacheco Silva OCR

4 4 4

2 2 2
Quality 2 Quality 2 Quality 2
Quality 1 Quality 1 Quality 1

0 0 0
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
OCR from Janbu's method OCR from Casagrande's method OCR from Karslrud's method
8 8 8

(d) (e) (f)

6 6 6
OCR from Becker's method

OCR from Becker's method

OCR from Becker's method

4 4 4

2 2 2
Quality 2 Quality 2 Quality 2
Quality 1 Quality 1 Quality 1

0 0 0
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
OCR from Janbu's method OCR from Casagrande's method OCR from Karlsrud's method

Figure 6. Comparison of OCR obtained by: a) Pacheco Silva's and Janbu's methods, b) Pacheco Silva's
and Casagrande's methods, c) Pacheco Silva's and Karlsrud's methods, d) Becker's and Janbu's
methods, e) Becker's and Casagrande's and f) Becker's and Karlsrud's methods.

6
Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

Internationally, most oedometer tests are behaviour after sampling and unloading, or
performed as IL tests (with 24 h load steps). recompression, up to the point at which it
The test in the database used in the present reaches first the in-situ condition and then the
study are mostly CRSC tests. It is generally apparent pc' it has experienced in the past,
observed that, pc' from IL tests will be 15 - and then the "virgin" compression beyond pc'.
20% lower than CRSC tests. This unloading-reloading process might
cause more disturbance in the skeleton and
3.4 Additional methods applied to samples therefore quality reduction.
with variable quality When comparing all methods, a very
Twelve tests results of the ones studied with significant difference is that Casagrande's
Casagrande (1936), Janbu (1969) and method and Pacheco-Silva's method as well
Karlsrud (1991) methods (i.e. tests with as most international methods use double
quality 1 and quality 2, six of each one) were logarithmic scales. Whereas Janbu's method
chosen to apply Pacheco Silva (1970) and suggests that a linear scale should be used to
Becker et al. (1987) methods. In addition, six avoid misinterpretations since double
tests results with quality 3 were included to logarithmic scales can hide large scatter in
study the effect of sample quality in the the data.
application and determination of pc'. One of the main difficulties in applying the
Figure 5 and Figure 6 shows the values of pc' graphical methods is the definition of the
and OCR obtained with these approaches for tangents to the data. Usually a stress-
the different specimens evaluated. As deformation curve from a CRSC odometer
observed before, Janbu's method tends to test has more data points than traditional
have higher values of pc' than the other incremental loading methods and the "best-
methods. When compare to Pacheco Silva's fit" lines require some judgement and
method, Janbu's method results up to 14% subjectivity.
difference in OCR and 10% difference in pc'. 700

Pacheco Silva's and Becker's methods fit 600

(a)
pretty well with Casagrande's and Karlsrud's
pc' (kPa) from Pacheco Silva's method

methods. Some differences up to 5% are 500

observed between the estimated values of pc' 400

and OCR. More deviation from the expected
trend is observed for high pc' (i.e. pc' > 500 300

kPa) and OCR (i.e. OCR > 4-5) values, and 200
in particular for samples with quality 2 and 3.
The difference between Pacheco Silva's and 100 Quality 3
Quality 2

Becker's methods reach 6% for pc' and OCR. 0

Quality 1

0 100 200 300 400 500 600 700

pc' (kPa) from Becker's method
4 EVALUATION OF METHODS 8

(b)
After these results, no strong pc' and OCR
variations are observed for high quality
OCR from Pacheco Silva's method

6
samples interpreted using Casagrande's,
Karlsrud's, Pacheco Silva's and Becker's
methods. Janbu's method gives the highest 4

deviations. One reason could give at Janbu's

method does not clearly specify the steps to 2
follow for a graphical interpretation, Quality 3
therefore, it depends on the own user Quality 2
Quality 1
experience and judgement. The observation 0
0 2 4 6 8
that more scatter is observed for high pc' and OCR from Becker's method
OCR values might be due to the fact that a Figure 7. Comparison between a) pc' and b) OCR
soil specimen undergoes higher reloading from Pacheco Silva's and Becker's methods.

7
Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

Table 3. Final evaluation of the pc' interpretation methods applied in the present study

In particular, it was experienced during the 6 REFERENCES

application of the methods, that a small
change in the slope of the tangents produced Becker, D.B., Crooks, J.H.A., Been, K., &
significant changes in pc'. In addition, the Jefferies, M.G. (1988). Work as a criterion for
scale type of the plot might increase the determining in-situ and yield stresses in clays: Reply.
Can. Geotech. J. 25, 848-850.
difficulties regarding reading the stress Becker, D.R., Crooks, J.H.A, Been, K. & Jefferies,
values. M.G. (1987). Work as criterion for determining in-situ
Table 3 presents a final evaluation of the & yield stresses clays. Can. Geotech. J. 24: 549-564.
methods in terms of technicalities (i.e. small Bjerrum, L. (1967). Engineering Geology of
details of the procedure), source of errors and Norwegian Normally-Consolidated Marine Clays as
Related to Settlements of Buildings. Géotechnique 17:
practical purposes (i.e. according to the 83-118.
experience of its application for pc' Boone, S.J. (2010). A critical reappraisal of
determination). "preconsolidation pressure" interpretations using the
oedometer test. Can. Geotech. J. 47, 281-296.
Burland, J.B. (1990). On the compressibility &
5 CONCLUSIONS shear strength of nat. clays. Géotechnique 40: 329-
378.
Interpretation of 129 oedometer test results Butterfield, R. (1979). A natural compression law
show that, in general, the value of pc' and for soils. Géotechnique 24: 469-479.
OCR does not depend on the interpretation Casagrande, A. (1936). The determination of the
preconsolidation load and its practical significance.
method used when evaluating high quality Proc. First Intern. Conf. on Soil Mech. & Found. Eng.,
samples. This is especially true for low Cambridge, 60-64.
values of preconsolidation stress and Clementino, R.V. (2005). Discussion: An
overconsolidation ratio. oedometer test study on the preconsolidation stress of
For samples having high values of pc' and glaciomarine clays. Can. Geotech. J. 42, 972-974.
Grozic, J.L.H., Lunne, T. & Pande, S. (2003). An
OCR, differences in interpretation methods oedometer test study on the preconsolidation stress of
can lead up to 14% in estimates of pc' and/or glaciomarine clays. Can. Geotech. J 40, 857-872.
OCR. Grozic, J.L.H., Lunne, T. & Pande, S. (2005).
In practice, we recommend to employ at Reply to the discussion by Clementino on: "An
least three different methods for evaluating oedometer test study on the preconsolidation stress of
glaciomarine clays". Can. Geotech. J 42, 975-976.
pc' and OCR. This will guarantee a more Holtz, R.D. & Kovacs, W.D. (1981). An
consistent choice of parameters for a given Introduction to Geotechnical Engineering. Prentice-
user and project. In case of doubts and to Hall, Inc., New Jersey, 733.
reduce uncertainties, evaluation of pc' and Jacobsen, H.M. (1992). Bestemmelse af
OCR should be performed and/or checked by forbelastningstryk i laboratoriet. Proceedings of
Nordiske Geotechnikermonde NGM-92. Danish
a colleague. Geotechnical Society, Bulletin 9, 455-460.
This work does no pretend to conclude which Janbu, N. (1963). Soil compressibility as
method is correct. As Becker et al. (1988) determined by oedometer and triaxial tests. Proc.
mentioned "... the issue is which technique Euro. Conf. on Soil Mech. and Found. Eng. 1, 19-25.
provides the most repeatable result and is Janbu, N. (1989). Grunnlag i Geoteknikk,
Trondheim: Tapir.
least ambiguous."
8
 Modelling, analysis and design - Study on the practices for pc' evaluation from oedometer tests

Karlsrud, K. & Hernandez-Martinez, F.G. (2013). NGF (2013). Veiledning for prøvetaking, melding
Strength and deformation properties of Norwegian nr. 11.
clays from laboratory tests on high-quality block Oikawa, H. (1987). Compression curve of soft
samples. Can. Geotech. J. 50, 1273-1293. soils. Soils and Foundations 27, 99–104.
Karlsrud, K. (1991). Sammenstilling av noen Onitsuka, K., Z. Hong, Y. Hara, & S. Yoshitake.
erfaringer med prøvetaking og effekt av (1995). Interpretation of oedometer test data for
prøveforstyrrelse i norske marine leire. NGI report natural clays. Soils and Foundations 35, 61–70.
521500-6. Pacheco Silva, F. (1970). A new graphical
Lunne, T., Berre, T. & Strandvik, S. (1998). construction for determination of the pre-consolidation
Sample disturbance effects in deepwater soil stress of a soil sample. Proceedings 4th Brazilian
investigations. Conf. Soil Invest. Found. Behav., Conference Soil Mechanincs and Foundation
London. Engineering, Rio de Janeiro, 225-232.
Lunne, T., Berre, T., Andersen, K.H., Sjursen, M., Pinto, C.S. (1992). First Pacheco’s Silva
& Mortensen, N. (2008). Effects of sample conference – Part I: Distinguished aspect of Pacheco
disturbance on consolidation behaviour of soft marine Silva’s activities. Brazilian Geotechnical Journal
Norwegian clays. Proc. 3rd Int. Conf. on Site Charact. (Solos e Rochas), ABMS-ABGE. 15 (2), 49–87.
ISC'3, Taipei, 1471-1479.

View publication stats