Session Report 2: CPT Interpretation

J.A. Schneider
University of Wisconsin-Madison, Madison, Wisconsin, USA

1 OVERVIEW

Session 2 on interpretation of cone penetration test (CPT) data included nearly 60

papers covering a wide variety of sites, analysis methods, and engineering parame-
ters. This session report focuses on a summary of the different types of information
presented in these papers and will provide details from select papers that highlight
various aspects of soil behaviour. Papers within this session are referenced in italics,
with no year, to distinguish them from other references.
Measurements of soils data at a given site can be broken down into in situ testing,
laboratory index testing, and laboratory performance testing. Laboratory performance
tests are those which measure a specific soil engineering property, such as strength (su
or '), stiffness (G), compressibility (, , CC, CS, or M), and water flow characteris-
tics (k, cv, or ch). Conversion of laboratory or in situ test measurements to engineering
parameters that can be used for design requires some degree of interpretation. Inter-
pretation methods can include analytical solutions, numerical analyses, probabilistic
methods, experimental studies, as well as empiricism (judgment). Experimental stud-
ies are separated in this discussion of interpretation methods to highlight papers that
focused on differences in response due to changes in specific variables, rather than
comparison of different measurement techniques for standard test methods employed.
Interpretation of soil properties based on the results of laboratory or in situ tests
each have their own advantages and disadvantages. Laboratory testing has controlled
boundary conditions and an extensive and rational basis for interpretation, although it
is time consuming and only tests a limited volume of soil that has been disturbed to
varying degrees by sampling. On the contrary, in situ testing is rapid and tests a large
volume of soil, but there are poorly known boundary condition (particularly drainage,
rate effects, and horizontal effective stress), installation or performance of the test
disturbs the soil to varying degrees and the basis for interpretation is somewhat lim-
ited and often inconsistent between sites. Both laboratory and in situ test results are
presented for about 60% of the papers in this session to aid development and verifica-
tion of interpretation methods. A summary of the distribution of content in Session 2
is presented in Figures 1 and 2.
Figure 1. Distribution of papers within interpretation session (a) analysis methods; (b) soil
characteristic evaluated; (c) soil types discussed; and (d) in situ tests used

Figure 2. Distribution of (a) number of sites discussed in each paper; and (b) number of in
situ tests used for papers in Session 2

The largest proportion of papers in this session focus on increasing the number of
soils types and geological conditions with which the engineering community has
documented experience. Most papers focus on comparing results from one type of
testing to another. This may include comparisons between in situ test methods (i.e.,
CPT, DMT, PMT), as well as comparison of laboratory tests (index or performance)
with results from in situ tests. The use of probabilistic (Uzielli et al., Kurup et al., Ku-
rup & Griffin, van der Wal et al., and Ng & Zhou), numerical (Butlanska et al., Ghaf-
ghazi & Shuttle, Kumar et al.), and analytical (Larsson et al., Isaer & Ryzhkor,
Löfroth, Russell et al., Ghafghazi & Shuttle) models have a comparatively low pro-
portion of papers in this session. This may result from difficulties in accurately mod-
elling the penetration process using numerical methods with realistic soil models, as
many constitutive models do not account for soil structure and aging effects, high
stress compressibility and particle crushing, stress dependent dilation, as well as
stress and strain dependence on soil stiffness. These features of soil behaviour are be-
ing addressed in analytical and numerical modelling of in situ tests, and we should
expect to see more application in future conferences (e.g., Salgado et al. 1997, Fahey
2005, Zhou & Randolph 2009, Einav et al. 2010).
A wide variety of soil types are discussed, although most papers still focus on
normally consolidated (NC) clays, overconsolidated (OC) clays, or siliceous sands.
Unsurprisingly, over half of the in situ tests discussed were cone, piezocone, or seis-
mic piezocone tests. Other penetrometers discussed included temperature, resistivity,
environmental (ROST/UVOST), water injection penetrometers, as well as full flow
penetrometers (T-bar and Ball). Other in situ tests discussed included pressuremeter
testing, flat dilatometer (DMT) and seismic dilatometer (SDMT), vane shear testing
(VST), and plate load testing.
The remainder of the session report highlights certain aspects of soil response that
were observed during review of these papers. This discussion is written around inter-
pretation of (i) strength; (ii) stiffness; (iii) compressibility; (iv) water flow character-
istics; (v) soil state; and (vi) soil
classification. Interpretation of tem-
perature characteristics (Isaev &
Ryzhkov, Trevor et al.) and environ-
mental issues (Smith & Gleason)
made up a relatively minor portion of
this session and will not be addressed
further.

2 STRENGTH

Selection of appropriate strength pa-

rameters for design depends on the
rate of loading in relation to the size
of the loaded object and the rate of
water flow through the loaded soil.
For interpretation of CPT and CPTU
of standard diameters and standard
rates of penetration, most sandy soils
will behave as drained while most Figure 4. Comparison of vane shear strength
clayey soils will behave as undrained. and net cone tip resistance for Brazilian Marine
clays (Almeida et al.)
 Use of the undrained strength (su) implies that changes in mean effective stress dur-
ing loading are solely due to shear induced pore pressures. For this case, laboratory
measurements of undrained strength performed at the current in situ effective stress
(and state) are appropriate for design (f = su). There is no volume change during
undrained loading, therefore the dilation angle () is equal to zero. Cone tip resis-
tance is strongly controlled by soil strength such that:
q t   v 0 q cnet
su   (1)
N kt N kt
Interpretation of undrained strength from CPT and CPTU was the most commonly
discussed parameter in this session, and was the focus of 14 papers. Values of Nkt are
widely reported in the literature, and ranges on the order of 8 to 27 are commonly
discussed (e.g., Denver 1995). A large range of Nkt values was also presented in this
session. An example of this range is shown in Figure 3 for marine clays from coastal
Brazil (Almeida et al.). Typical reasons for variation in Nkt are associated with (i)
mode of strength (su,DSS, su,CIUC, su,UU, su,VST, su,PMT, etc); (ii) soil rigidity (G/su); (iii)
viscous rate effects; (iv) partial consolidation during penetration, testing, or between
installation and testing; (v) sample disturbance; (vi) spatial variability between CPT
and samples; (vii) CPTU area ratio correction factors; (viii) drift in CPT zero load
readings during testing; as well as (ix) soil structure and fabric, such as fissuring as
sensitivity. Use of full flow penetrometers, as presented in this session by Cargill &
Camp, may reduce some of the uncertainty in N (Nkt, Nt-bar, Nball) values (e.g.,
Randolph 2004).
Strength parameters for drained analysis are typically the peak friction angle ('pk),
which is a combination of the critical state friction angle ('cv) and the dilation angle
('). Due to stress dependent dilation of sands and crushing of particles at high
stresses typical of cone penetration testing, practice tends to rely on empirical correla-
tions to 'pk. For interpretation of strength of sands, 4 papers discussed friction angle
(Kumar et al., Aykin et al., McNulty & Harney, Robertson) and one discussed dilation
angle (Kumar et al.). Most interpretation in sands focused on state or relative density
(discussed later in this report).
An interesting case history was presented by McNulty & Harney, which included
CPTU, DMT, and triaxial test results. The focus of this discussion is the CPTU data
and triaxial test results. Figure 4 attempts to reproduce part of Figure 1 of McNulty &
Harney using data from their Table 1. The correlation between CPT data and friction
angle used by McNulty & Harney was not stated, and whether a ‘fines content correc-
tion’ was applied to CPTU data for their silty and clayey sands was also unclear. Re-
interpretation in Figure 4 used no fines content correction and the correlation between
stress normalized CPT data and friction angle that was published by Kulhawy &
Mayne (1990), and seems to agree with the authors’ original interpretation:
 q c p ref 
' pk  17.6  11  log  0.5 
 17.6  11  logq c1N  (2)
 ' v 0 p ref  
For additional comparison, peak friction angles calculated using the method of
Robertson are included in the figure. The method discussed by Robertson is based on
the concept of a state parameter for sands (Been & Jefferies 1985) and includes a
‘fines content correction’ estimated from normalized cone tip resistance and friction
ratio (Fr). A constant volume friction angle of 30o was assumed for all layers when
using the analysis procedures discussed by Robertson for this profile.
Two questions arise from comparison of the laboratory and in situ test data (i) is a
correlation developed from drained calibration chamber data on clean siliceous sands
applicable to silty and clayey sands; (ii) is the laboratory measured friction angle of
20o (at 15m depth) realistic? To address the first question, the degree of drainage dur-
ing cone penetration is assessed using standard CPTU classification charts (e.g.,
Robertson 1990, Schneider et al. 2008). In the Schneider et al. (2008) charts, it ap-
pears that 6 of the layers classify as drained sands (Zone 2) where the 'pk correlation
is likely applicable, 3 of the layers classify as ‘transitional soils’ (Zone 3) where pore
pressures during penetration likely have some effect on interpretation, and high pene-
tration pore pressures in 4 layers indicate the material is undrained (Zone 1) and an
interpretation of undrained strength is more applicable. The reasonable friction angles
of 30 to 35 degrees for the Zone 1 soils are considered to have little to do with pene-
tration resistance and soil response. Agreement is considered to be more likely a re-
sult of compensating effects than application of an appropriate correlation. Higher
friction angles in Zone 1 soils calculated using the method discussed by Robertson
resulted from an increase in normalized tip resistance due to the 'Kc' correction.
With respect to the low measured friction angle in the overconsolidated clays,
McNulty & Harney attributed this response to apparent cohesion at stress levels lower
than the preconsolidation stress (p'c). It is correct that adding an apparent cohesion in-
tercept will reduce the interpreted friction angle from laboratory tests, but this does
not reflect the actual soil behaviour in the laboratory test or in the field. An increase
in resistance and occurrence of peak friction angles at effective stresses below the
preconsolidation stress results from interlocking and resistance in addition to that
from friction. Misinterpretation of this behaviour as caused by ‘cohesion’ does not re-
flect the mechanical behaviour of the soil and may be unconservative for certain de-
sign situations (e.g., Santamarnia 1997, Schofield 2005).

Figure 4. Soil classification and friction angle correlations (data from McNulty & Harney, for
details on classification charts see Robertson 1990, Schneider et al. 2008)
3 STIFFNESS

To evaluate deformation of materials, two elastic parameters are needed; (i) the elas-
tic modulus (E) and Poisson ratio (); or (ii) the shear modulus (G) and bulk modulus
(K). It is more appropriate to consider G and K for soil mechanics applications, as
this separates the effects of shearing (change in shape due to application of a deviator
stress) from compression (change in size due to changes in mean effective stress)
(e.g., Atkinson 2007). This section will focus on evaluation of shear modulus, while
the next section will discuss compression characteristics of soils.
The stiffness of soil is characterized using the shear modulus, G. Our ability to
characterize stiffness of soils from cone penetration tests has greatly increased since
the development of the seismic cone (SCPT) and seismic piezocone penetration
(SCPTU) tests (e.g., Robertson et al. 1986). Measured values of shear wave velocity
(Vs) can be used to calculate stiffness, or more specifically the small strain shear
modulus (G0):
2
G 0    Vs (3)
where  is the total mass density of the soil. A majority of papers within this session
that relate to soil stiffness focus on shear wave velocity (Uzielli et al., Paoletti et al.,
Prakoso, Aykin et al., Trevor et al.), while 3 papers discuss small strain shear
modulus (Kavur et al., Schneider & Lehane, Kumar et al.), and 1 paper discusses lar-
ger strain stiffness measurements (Gsec) from a self boring pressuremeter test;
(Schneider & Lehane). It is common to quantify shear modulus as a multiple of CPT
tip resistance, and G/qcnet is presented in this section:
n
G  q c p ref  n
 KG  0.5 
 K G  q c1N (4)
q cnet  ' v 0 p ref  
where ‘KG’ is a soil deposit dependent coefficient and ‘n’ is a stress exponent.

Figure 5. Influence of measurement technique on soil stiffness (a) laboratory tests and design
situations (after Atkinson 2000); (b) in situ test method (for a calcareous sand, data from
Schneider & Lehane)
 Since shear modulus measured by dynamic methods is much larger than opera-
tional levels for typical design situations (e.g., Figure 5a) measurement of G0 is useful
but must be reduced appropriately. Modulus reduction schemes are not unique, and
measurement of stiffness from a number of different in situ tests can aid in the as-
sessment of modulus reduction for design (e.g., Lehane & Fahey 2004). Figure 5b
shows small strain shear modulus measured by seismic methods (SCPTU), as well as
that from self boring pressuremeter tests (SBPMT) and that estimated from flat plate
dilatometer [G ≈ ED/(2·(1+))]. Shear strain levels associated with each tests are ap-
proximately 10-4 % for SCPTU, 10-1 % for SBPMT, and 1% for DMT. It is clear that
the shear modulus reduces as characteristic strain levels associated with various test-
ing methods increase, but it is also interesting to see that the stress exponent ‘n’ ap-
pears to reduce with this increasing strain level. The implications of these trends in
stress exponents, and whether they are actually a function of soil behaviour rather
than scatter in the measurements, requires further study.

4 COMPRESSIBILITY

The change in size of an element of soil with increasing mean effective stress can be
considered as compression. Compressibility is essentially the inverse of the bulk
modulus, 1/K = v/p'. For geotechnical engineering purposes the one dimensional
(1D, no horizontal strain) case is typically considered and the stress dependence on
compressibility is handled using the compression index, CC = e/log'v. When de-
veloping correlations to in situ tests the 1D constrained modulus is typically quanti-
fied where, M = D' = 1/mv = 'v/v = 'v/a = 2.3·(1+e0)·'v,avg/CC. Within the pa-
pers presented as part of this session, three methods were used to evaluate constrained
modulus (i) oedometer tests (Lee, Choi, et al., Bihs et al., Ozer et al.); (ii) correlations
to DMT (Bihs et al., Kavur et al., Młynarek et al., Aykin et al.); and (iii) results from
a large scale test embankment (Tonni et al.). Most sites tested were clays and organic
clays, although data for a cemented sand (Lee, Choi, et al.) and layered sand, silt, and
clay site (Tonni et al.) were also presented.
Based on elastic theory, the constrained modulus should be approximately 0.9E
and therefore essentially equal to the dilatometer modulus ED (e.g., Marchetti 1980).
A unique correlation between ED and M has not been observed (e.g., Marchetti 1980,
Powell & Uglow 1988) highlighting the need to separate change in shape of soil ele-
ment (G or E) from change in size of a soil element (M or 1/K).
Difficulties also exist when attempting to select the appropriate stress range for M
when correlating to laboratory or field test data, as well as selecting whether cone tip
resistance is an appropriate parameter for correlations to M. Figure 6 shows two cor-
relations between CPT parameters and M (or D'). Typically, the in situ vertical effec-
tive stress level is used for correlations to M (e.g., Mayne 2007). Within this session
Ozer et al. utilized the value of M at the preconsolidation stress, while Tonni et al.
evaluated M for a stress range starting from 'v0 and increasing by up to 110kPa.
Since M increases with effective stress (for a given soil in an overconsolidated or
normally consolidated state), variations in effective stress range for assessment of M
will increase uncertainty in correlations. While incorporation of measured pore pres-
sures improved CPT-M correlations for the Treporti site, complex pore pressure re-
sponse that includes both shear and octahedral components as well as reduction in to-
tal stress behind the cone tip (e.g. Ha et al.) make application of this correlation to a
wider range of soil types and states uncertain.

Figure 6. CPT – constrained modulus correlations for loading of (a) sands silts and clays
loaded by 110kPa at the Treporti site (Tonni et al.); (b) global database for oedometer tests
on sand silts and clays with M taken at 'v0 (Mayne 2007)

5 WATER FLOW CHARACTERISTICS

Water flow characteristics were addressed in 7 papers, 3 focusing on coefficient of

consolidation (ch or cv; Kim, Prezzi, et al., Almeida et al., Jaeger et al.), 2 focusing on
coefficient of permeability (k; Vienken et al., Boone & Frietas), and 2 discussing both
parameters (Cai et al., Robertson). Direct methods for interpretation of ch and k in-
clude analytical solutions for piezocone dissipation tests (e.g., Teh & Houlsby 1991),
variable rate penetration testing (e.g., House et al. 2001, Randolph 2004), as well as
injection methods (e.g., Vienken et al.). Indirect methods typically involve correla-
tions to soil behaviour type (IC) and/or fines content (Boone & Frietas, Robertson).
While discussed since the early 1970’s, the prominence of variable rate penetration
testing has increased in research (and in practice) since about 2004 (e.g., Randolph
2004). Two similar experimental studies of variable rate penetration testing per-
formed independently but on similar materials (sand with 25% Kaolin) are presented
in this session; chamber tests of Kim, Prezzi et al, and centrifuge tests of Jaeger et al.
Normalized CPTU parameters (Q, u2/'v0, Bq) are plotted in Figure 7 against nor-
malized velocity (V = vd/cv, where v is the penetrometer velocity, d is the penetrome-
ter diameter, and cv is the vertical coefficient of consolidation). Similar response is
observed for both studies; (i) Drained behaviour occurs for V less than about 0.05 and
is characterized by relatively high Q values and essentially zero values of u2/'v0
and Bq; (ii) Undrained behaviour occurs for V greater than about 15 and is character-
ized by relatively low values of Q and ‘high’ values of u2/'v0 and Bq. Differences in
Bq are observed between the two studies during undrained penetration and seem to be
related to very low values of Q recorded by Jaeger et al.. Results from supplemental
undrained triaxial tests on the sand-clay mixtures would help clarify these differ-
ences.
It is difficult to characterize water flow characteristics in aquifers and some levees
since penetration in sands produces essentially no excess pore pressures (u2≈0).
Vieken et al. discuss a number of direct methods for aquifer characterization, includ-
ing active source penetrometers, while Boone & Frietas discuss application of an in-
direct method to a wide range of soil types.
A summary of indirect and direct methods for characterization of water flow char-
acteristics is given by Robertson. The importance of performing piezocone dissipa-
tion tests when the degree of drainage is unclear is highlighted in this paper, and it is
suggested that if t50 from a piezocone dissipation test takes longer than 30 seconds (t95
≈ 20 minutes) then tip resistance during penetration is essentially equal to the
undrained value. Recording piezocone dissipation response in questionable layers for
only one to three minutes can provide invaluable information for assessing degree of
drainage during penetration, and supplementing those results with variable rate pene-
tration tests (for rates of 0.1mm/s to 200 mm/s) can help quantify the change in resis-
tance to changing drainage conditions.

Figure 7. Influence of normalized velocity on penetration resistance in clayey sands (data

from Kim, Prezzi, et al., and Jaeger et al.)

6 SOIL STATE

A large portion of the papers in this session (18) focused on evaluation of soil state.
This typically revolved around OCR or preconsolidation stress (p'c) in clayey soil (9
papers) and relative density in sands (8 papers). Additionally, Ghafghazi & Shuttle
and Robertson discuss assessment of the state parameter for sands, 0. Soil state is an
indirect measure that can aid in prediction of soil response, and will have a strong in-
fluence of strength, stiffness, and compressibility.
While the state of sands is defined by current void ratio and effective stress in rela-
tion to the critical state line (e.g., Been & Jefferies 1985, Coop 2005), evaluation of
the state of sands typically revolves around relative density, Dr. Since CPT tip resis-
tance is strongly controlled by in situ horizontal effective stress, shear modulus, fric-
tion angle, (stress dependent) dilation angle, and high stress compressibility (crusha-
bility), there are large uncertainties in evaluation of Dr and 0. Two papers are fo-
cused on here in relation to relative density evaluation; (i) Butlanska et al. who at-
tempt to better understand the relationship between qc and Dr through DEM numeri-
cal modelling, and (ii) Senders who evaluates shallow embedment factors through
centrifuge testing. Selected results from the two studies are highlighted in Figure 8.
Results of the Butlanska et al. study may lead to a better understanding of CPT be-
haviour in sands, although additional study into scale effects between penetrometer
and particles as well as the near surface effects presented by Senders would be useful
in future work. It is noted that the ratio of penetrometer diameter (71.2mm) to particle
mean grain size (d50 = 28mm) appears to be 2.5 for the study of Butlanska et al.. Bol-
ton et al. (1999) recommend that this ratio be at least 20 such that the soil-
penetrometer interaction behaves closer to a continuum, which is more representative
of typical field conditions.

Figure 8. Comparison of sand relative density evaluations from (a) centrifuge testing (Send-
ers) and (b) DEM numerical modelling (Butlanska et al.)

Figure 9. Plot of normalized undrained strength based on CPT qcnet and estimated from stress
history (data from Liao et al.)

Soil state for clays, represented by the overconsolidation ratio (OCR = p'c/'v0) or
vertical yield stress ratio (YSR = 'vy/'v0), is a vital parameter for linking known
geologic history to anticipated mechanical behaviour at a given site. The relationship
between undrained shear strength ratio and OCR, based on the concepts of critical
state soil mechanics, is a powerful tool for understanding soil behaviour and predict-
ing engineering response.
su  s 
  u   OCR  (5)
' v 0  ' v 0  NC
where (su/'v0)NC is the normally consolidation undrained strength ratio [≈ (sin'cv)/2
for DSS] and is the plastic volumetric strain ratio (about 0.8) (e.g., Mayne 2001).
The application of this concept to a glacial soil site in Eastern Wisconsin discussed by
Liao et al. is presented in Figure 9.
Liao et al. present a case history where a 35m high lakeshore bluff was removed to
build a coal power plant. The underlying glacial clay and sandy clay till soils tended
to have a preconsolidation stress predicted by a range of empirical correlations that
was higher than the stress of the overburden removed. If the normalized undrained
strength ratio (using Nkt=12) is plotted for clayey layers with u2/'v0 greater than
about 1, trends match an estimated preconsolidation difference (p'c = p'c-'v0) of
1200kPa in the upper 21m and 2000kPa below 21m. These values of p'c are consis-
tent with the regional glacial geology and in excess of those related to bluff removal.
Knowledge of local geology is often invaluable for interpretation of in situ tests. Ad-
ditional validation of correlations through complementary laboratory oedometer test-
ing is recommended for these large projects, although, applied vertical stresses up to
10MPa may be necessary for a clear definition of p'c and the slope of the normal
compression line.

7 CLASSIFICATION

Soil classification was a major component of approximately 13 papers. Most papers

focused on cone and piezocone response in a range of soil conditions (Kurup et al.,
Dasenbrock et al., Lo Presti et al., Mateos, Robertson, Tonni et al.), while others fo-
cused primarily on characterization of clays (Bihs et al., Rømoen et al., Eslami et al.,
Ng & Zhou), calcareous sands (Schneider & Lehane), residual soils (Mio et al.), or
clayey sands (Jaeger et al.).
To rationally assess soil behaviour and classification from piezocone response
normalized parameters must be used (e.g., Wroth 1988, Robertson 1990). Robertson
points out that for typical stress ranges related to onshore projects (50 to 150kPa),
normalization does not significantly shift points within the classification charts and
an equivalent non-normalized chart (and ISBT parameter) can be used with similar
levels of reliability to normalized charts. When dealing with larger stress ranges and
normalized parameters, response due to effects of soil state must be separated from
those due to water flow characteristics. This can most easily be done using normal-
ized tip resistance (Q = qcnet/'v0) and penetration pore pressures (u2/'v0 = Bq·Q)
(e.g., Schneider et al. 2008), although use of Q and friction ratio (F = fs/qcnet) is com-
mon due to difficulties in obtaining high quality pore pressure data as well as when
testing soils above the water table.
Figure 10. Separation of partial consolidation during penetration from overconsolidation us-
ing classification charts of Schneider et al. (2008) for (a) natural soils; and (b) variable rate
penetration testing (values of Q below one not plotted)

Figure 10 highlights separation of soil state from water flow characteristics for an
OC Clay, a sensitive clay, and layered sands, silts, and clays. Site data comes from
papers within this session. It should be noted that soil state and partial consolidation
during penetration can not clearly be distinguished in Q - Bq classification charts, de-
spite using the same data as Figure 10 (u2/'v0 = Q·Bq). In Figure 10b, results from
variable rate penetration tests shown in Figure 7 are also plotted in a Q - u2/'v0
classification chart. Trends in behaviour are consistent with those for natural soils;
undrained penetration tends to plot in Zone 1 – undrained clays; as V decreases,
measurements move through Zone 3 – transitional soils towards Zone 2 – drained
sands; for V less than about 0.1 u2/'v0 is less than 0.1 and penetration is essentially
drained. The greater high stress compressibility of the clayey sands as compared to
soils that would have drained penetration for standard velocities and diameters lead to
relatively low Q values and data plotting below Zone 2. When comparing these vari-
able rate tests to others performed in Kaolin, it appears that the high stress com-
pressibility affects drained penetration resistance (e.g., Robertson & Campanella
1983) as well as the ratio of drained to undrained penetration resistance.

8 CONCLUSIONS

CPT data at a wide variety of new sites is presented in this session along with com-
plementary laboratory and in situ test data for interpreting a range of soil properties.
Most interpretations build on previously developed methods and discuss similarities
or differences due to specific site conditions. These studies are important for develop-
ing judgment for a wider range of soil conditions. It is noted that these data should be
viewed within a systematic engineering framework such that experiences from a sin-
gle site can be applied to differing soil conditions that will be encountered in the fu-
ture. Specifically, one must be sure to separate partial consolidation during loading
from increases in overconsolidation ratio, separate stiffness and compressibility, as
well as utilize critical state soil mechanics to couple strength with stress history. In
addition to comparison of data from new sites, new interpretation methods (analyti-
cal, numerical, probabilistic) still need to be developed to aid in consistent under-
standing of a wide range of soil conditions, as well as spatial variability of interpreted
engineering parameters.

REFERENCES

Italicized references in text presented without a year are papers from CPT10.

Atkinson, J.H. 2000. Non-linear soil stiffness in routine design, Géotechnique, 50(5): 487-508.
Atkinson, J.H. 2007. The Mechanics of Soils and Foundations, 2nd Ed., Taylor & Francis, London.
Been, K., and Jefferies, M.G. 1985. A state parameter for sands. Géotechnique, 35(2): 99-112.
Bolton, M. D., Gui, M. W., Garnier, J., Corte, J. F., Bagge, G., Laue, J., and Renzi, R. 1999. Centri-
fuge cone penetration tests in sand. Géotechnique, 49(4): 543-552.
Coop, M.R. 2005. On the mechanics of reconstituted and natural sands. Deformation Characteristics
of Geomaterials, Lyon, Taylor & Francis: 29-58.
Denver, H. 1995. Technical Report – Session 2, Proc., Int. Symp. on Cone Penetration Testing,
CPT’95, Report 3:95, Vol. 3, Swedish Geotechnical Society, Linköping: 105-118.
Einav, I., Vardoulakis, I., and Chan, A.H. 2010. Expanding perforations: comminution, permeability
reduction and sand production. Acta Geotechnica. submitted.
Fahey, M. 2005. Stiffness of sands of different ages from pressuremeter and seismic cone tests in
Perth. Proc. ISC’5, 50 Years of Pressuremeters, Vol. 1: 611-619.
House, A.R., Oliveira, J.R.M.S., and Randolph, M.F. 2001. Evaluating the coefficient of consolidation
using penetration tests. Int. J. of Physical Modelling in Geotechnics, 1(3): 17-25.
Kulhawy, F.H., and Mayne, P.W. 1990. Manual on estimating soil properties for foundation design.
Report EL-6800, Electric Power Research Institute, Palo Alta, CA, August: 306.
Lehane, B.M. and Fahey, M. 2004. Using SCPT and DMT data for settlement prediction in sand, Proc.
ISC-2 on Geotechnical and Geophysical Site Characterization, Millpress: 1673-1679.
Marchetti, S. 1980. In situ tests by flat dilatometer. Journal of the Geotechncial Engineering Division,
106(GT3): 299-321.
Mayne, P.W. 2001. Stress-Strain-Strength-Flow Parameters from Enhanced In-Situ Tests. Proc, In-
ternational Conference on In-Situ Measurement of Soil Properties & Case Histories (In-Situ 2001),
Bali, Indonesia: 27-47.
Mayne, P.W. 2007. Synthesis on Cone Penetration Test: State-of-Practice. NCHRP Report, Transpor-
tation Research Board, National Academies Press, Washington, D.C.
Powell, J.J.M., and Uglow, I.M. 1988. Marchetti dilatometer testing in U.K. soils. Penetration Testing
1988, Vol. 1, Balkema, Rotterdam: 555-562.
Randolph, M.F. 2004. Characterisation of soft sediments for offshore applications. Proc., Second Int.
Conf. on Site Characterization, ISC’2, Porto: 209-232.
Robertson, P.K. 1990. Soil classification using the cone penetration test. Can. Geotech. J., 27(1): 151 –
158.
Robertson, P.K., and Campanella, R.G. 1983. Interpretation of cone penetration tests, Part 1: Sand.
Canadian Geotechnical Journal, 20(4): 718-733.
Robertson. P.K., Campanella, R.G., Gillespie, D., and Rice, A. 1986. Seismic CPT to measure in situ
shear wave velocity, Journal of Geotechnical Engineering, 112(8): 791-803.
Salgado, R., Mitchell, J.K., and Jamiolkowski, M. 1997. Cavity expansion and penetration resistance
in sand.” Journal of Geotechnical and Geoenvironmental Engineering. 123(4): 344-354.
Santamarina, J.C. 1997. 'Cohesive Soil:' A Dangerous Oxymoron. The Electronic J. Geotechnical
Eng., August.
Schneider, J.A., Randolph, M.F., Mayne, P.W., and Ramsey, N. 2008. Analysis of factors influencing
soil classification using normalized piezocone tip resistance and pore pressure parameters, Journal
of Geotechnical and Geoenvironmental Engineering, 134(11): 1569-1586.
Schofield, A. 2005. Disturbed soil properties and geotechnical design, Thomas Telford, London.
Teh, C.I., and Houlsby, G.T. 1991. An analytical study of the cone penetration test in clay. Géotech-
nique. 41(1): 17-34.
Zhou, H. and Randolph, M. F. 2009. Numerical investigations into cycling of full-flow penetrometers
in soft clay. Géotechnique, 59(10): 801–812.
Wroth, C.P. 1988. Penetration testing – A more rigorous approach to interpretation. Penetration Test-
ing 1988, Balkema, Rotterdam: 303-311.