Variability in Cone Penetration Testing
Variability in Cone Penetration Testing
“[Le manifestazioni dei fenomeni possono] servire a fondare una seducente, fallace e dogmatica teoria;
chi ricerca il vero si arresta perplesso davanti la complessità del problema.”
CHAPTER 1 5
set of site conditions, measurement techniques and to measurement techniques and procedures). While,
procedures, and correlation models for which the in fact, a variety of well established mathematical
design soil properties were derived. Moreover, the procedures (e.g. random field modeling, geostatistics,
estimated coefficients of variation may considerably wavelet theory, maximum likelihood methods) are
overestimate real variability (Fenton 1999a; Phoon & available to pursue the estimation of inherent soil
Kulhawy 1999a). Possible sources of overestimation variability, literature results regarding the direct
include: a) analysis on non-homogeneous soil units; assessment of epistemic uncertainty are less
b) insufficient equipment and procedural controls; c) numerous and of limited reliability, as such
neglection of deterministic trends in the soil data; and procedures require comparative laboratory tests in
d) excessive chronological span of data acquisition. replicate conditions (Orchant et al. 1988; Phoon &
Separating and independently quantifying the sources Kulhawy 1999a), and as documentation on equipment
of test variability provides data that can be used and procedural controls during in situ testing is
generally, rather than only in site-specific usually not detailed sufficiently to allow for a
applications (Orchant et al. 1988). The separation of quantitative evaluation of measurement errors (Phoon
the components of uncertainty is a complex process. & Kulhawy 1999a).
Firstly, the effects of each source of uncertainty are While inherent variability will be addressed in detail
variable, as is the way each propagates in a given in Chapter 4, a synthetic review of previous studies
engineering model (Baecher 1986). Moreover, the focusing on the direct estimation of CPT
relative contribution of the three sources to the overall measurement uncertainty is provided in the following.
uncertainty in the design soil property is highly
variable, and depends on site conditions, degree of 1.3. PRECISION, ACCURACY AND BIAS
equipment and procedural control, and precision of Past research (e.g. Agterberg 1970; Baecher 1982,
the correlation model (Phoon et al. 1999a). 1986; Cividini 1981; Handy 1980; Jaksa 1995; Lumb
The most serious complication in the process of 1971,1975; Murphy 1961; Orchant et al. 1988; Phoon
separating the sources of geotechnical uncertainty is & Kulhawy 1999a; Vanmarcke 1978) has made use
possibly the fractal nature of soil properties and the of a variety of terms to describe measurement
dependence of the correlation structure from the uncertainty. This lack of common terminology in the
sampling scheme, i.e. the influence of the proportion geotechnical literature has led to occasional
between the spatial magnitudes of the sampled aera misinterpretation and ambiguity.
and the test sampling interval. Such issue was initially Here, the definitions proposed by Orchant et al.
addressed in the context of geologic materials by (1988) will be used. Orchant et al. (1988) used the
Agterberg (1970). Other researchers have definition of accuracy by Bainbridge (1985) as “the
ackowledged the issue, among which Soulié et al. closeness of agreement between any one measured
(1990) pointed out that a scatter due to small-scale value and an accepted reference value.” Such concept
but real variations in the soil is frequently taken as of accuracy, when applied to individual observations,
coming from measurement error. contains elements of both precision and bias, with
In the context of CPT variability, however, such precision referring to random testing error and bias
distinction is possibly of limited relevance. Cone referrring to variability deriving from systematic error
penetration results in soil failure and plastic (i.e. equipment, operator and procedeural errors).
deformation of a spatial volume of soil. Thus, CPT Systematic error, or bias, was defined by Orchant et
measurements are thus not point measurements, but al. (1988) as the difference between a population
are representative parameters of the extent of the zone mean of test results and an accepted reference value.
in which such failure occurs. In such cases, small- Orchant et al. (1988) used the term “precision” to
scale variation is quickly lost in spatial averages describe random error, and defined it as the closeness
(Vanmarcke 1983). Hence, the separation of of agreement between randomly selected individual
uncertainty into unrelated components is conceptually measurements or test results. It describes the
justified. consistency or repeatability of the test.
Results of past research focusing on the estimation of Collectively, systematic and random errors fully
the magnitudes of the sources of CPT variability constitute measurement error.
suggest that the direct determination of aleatory The conceptual distinction between accuracy and
uncertainty (i.e. pertaining to inherent soil variability) precision may be appreciated in Figure 1-1.
is significantly more reliable than the direct
quantification of epistemic uncertainty (i.e. pertaining
6 CHAPTER 1
be random in terms of its effect on test variability
because, in general, insufficient information
regarding the equipment and operator/procedural
variables is available to assess their relative impact on
the test results.
precise but accurate but precise and
inaccurate imprecise accurate All standards for CPT testing specify the degree of
measurements measurements measurements
accuracy and precision required in the measured
values. The International Reference Test Procedure,
Figure 1-1. Target analogy for the conceptual
distinction between accuracy and precision (adapted or IRTP (ISSMGE 2001), states that “taking account
from Orchant et al. 1988) of all possible sources of error (parasitic frictions,
errors of the measuring devices, eccentricity of the
load on the cone with respect to the sleeve,
Assessments of the accuracy of the CPT have also
temperature effects, etc.), the precision of the
been based on a procedure proposed by Baecher
measurement shall not be worse than [the greater of]
(1982; 1986): several studies (e.g. Christian et al.
5% of the measured value and 1% of the maximum
1994; De Groot & Baecher 1993; Filippas et al. 1988;
value of the measured resistance in the layer under
Jaksa 1995; Kay 1990; Kay et al. 1991; Lacasse &
consideration. The precision shall be verified in the
Nadim 1996; Spry et al. 1988; Tang 1984) have used
laboratory or in the field taking into account all
this method, or results based on it, to postulate
possible disturbing influences.”
various aspects relating to CPT testing variability.
However, the method has received several criticisms
(e.g. Jaksa et al. 1997), and will not be treated herein. 1.4.1. Systematic error
The terms precision, bias and accuracy are useful Systematic error, as previously stated, comprises
qualitative descriptors of the variability of test results. equipment error and operator/procedural error. The
However, to apply the results of an analysis of components of systematic error are addressed
variability to practical problems, quantitative synthetically hereinafter.
estimates must be made. In the following, an
overview of selected results from the literature is 1.4.1.1. Equipment error
provided. While main concepts and inferences are
Equipment error may be caused, among other factors,
reported in the text, numerical values and specific
by poor selection of installation procedures and by
data pertaining to individual studies are presented in
limitations in the instrument design. The type and
footnotes.
magnitude of equipment error is dependent from,
1.4. COMPONENTS OF CPT MEASUREMENT
among other things: a) type of test (mechanical or
UNCERTAINTY electrical cone)1; b) location of the filter for pore
pressure measurement [in piezocone testing]2; c) axial
The electric CPT and piezocone test have been
reported to have the lowest total measurement error of 1
The equipment in mechanical cone tests differs in several
any in-situ test in current practice (e.g. Kulhawy &
aspects from that used in electrical tests. The mechanical
Trautmann 1996). As previously stated, CPT CPT commonly uses a two rod system in which the cone tip
measurement error represents a combination of and the friction sleeve move indipendently. Such system
several sources of variability including the leads to relevant measurement errors, discussed in detail by
uncertainty associated with systematic testing error De Ruiter (1971) and Orchant et al. (1988): friction may
develop between the cone tip and the outer sleeve because
(i.e. equipment and operator/procedural effects), and
soil fills the space when the tip moves relatively to the
random testing error which is not assignable to sleeve. Also, the lower edge friction sleeve may develop tip
specific testing parameters. resistance as high as two thirds of side resistance. Such
Some researchers (e.g. Baecher 1983; Vanmarcke phenomenon may determine relevant errors in side
1978) have considered measurement error in its resistance measurements, especially in sands (De Ruiter
entirety as a random variable, considering errors 1971). Joustra (1974) obtained tip resistances only about 3%
lower using electrical CPT cones than using mechanical
involved in the measuring processes as being non-
CPT cones. However, Kok (1974) noted that, in a series of
repeatable. However, Orchant et al. (1988) suggest tests conducted, the electrical cone produced tip resistances
that there are components of test variability which can 30% greater, on average, than the mechanical cone.
2
be modeled as systematic variables, and which can be The measured pore pressure varies with the soil type and
assigned directly to specific causes. Measurement the location of the filter element. Pore pressures may be
measured in one or more of the following three locations:
error, according to Orchant et al. (1988), appears to
on the cone (u1), behind the cone (u2) and behind the friction
CHAPTER 1 7
load on the cone3; d) wear of the cone and
manufacturing defects4; e) cone size and geometry5;
sleeve (u3). Presently, the location of the filter element is not cm2 cone diameter, is 5% (Lunne et Al. 1997). The error
standardized. The IRTP refers to the location behind the may become significantly greater if regular checks are not
cone (u2) as the preferred filter location, as it has the made for wear of the cone tip and friction sleeve. Schaap &
advantage of good protection from damage due to abrasions Zuidberg (1982) investigated the potential wear of the cone
and smearing and generally easier saturation procedures. and friction sleeve as a function of meters of penetration
The location behind the cone is also the required location to testing. De Ruiter (1982) recognized that minor
correct the measured cone resistance. While the availability manufacturing defects may lead to in accurate side
of more than one pore pressure measurement for the same resistance measurement. Orchant et Al. (1988) added that
sounding is beneficial, as it allows for enhanced profiling faulty or leaky seals in the electrical cone may damage the
capability, improved interpretation in terms of OCR and K0, measuring devices and lead to erroneous results.
5
the location immediately behind the cone (u2) is to be chosen Several authors (e.g. Baligh et al. 1979; de Lima & Tumay
if only one filter location is to be included (Lunne et al. 1991; De Ruiter 1982; Durguno÷lu & Mitchell 1975; Holden
1997). For theoretical and practical studies regarding the 1971; Joustra 1974; Juran & Tumay 1989; Kok 1974; Lunne
relationships between u1, u2 and u3, see, for example, et al. 1986; Marsland & Quarterman 1982; Parkin & Lunne
Houlsby & Teh (1988); Levadoux & Baligh (1986); Lunne 1982; Rol 1982; Sanglerat 1972; Schmertmann 1978;
et al. (1986); Powell & Quarterman (1991); Robertson & Shields 1981; Smits 1982; Sweeney 1987; Wang & Lu
Campanella (1988); Whittle & Aubeny (1991). During 1982) have investigated the effects of cone geometry, size
penetration, the largest pore pressures are generally and apex angle on CPT results. Sanglerat (1972) stated that
measured in the zone beneath the cone, where the cones with different cross-section aeras give almost the same
compressive stresses are at a maximum. The cylindrical penetration resistance in all soils. Durguno÷lu & Mitchell
section immediately behind the cone will be in a zone of (1975) and Baligh et al. (1979) found that the cone apex
normal stress relief, while both zones undergo large shear angle has a relevant importance on CPT measurements, and
stresses. Thus, the pore pressure response under the cone is that a 60° apex angle favors maximum penetration for a
influenced mostly by large normal stresses, while the pore given axial load. Holden (1977) suggested that a small size
pressure response along the cone shaft is dominated by large cone penetrometer could give higher penetration resistance
shear stresses. Interpretation of pore water pressures is than the 10 cm2 cone penetrometer in sands. Schmertmann
especially beneficial in fine-grained soils in which (1978) reported no significant variations in the measured
penetration in essentially undrained. In a saturated soil, an penetration resistance with cross-section areas in the range
increase in normal stresses will yield positive pore pressures, 5-20 cm2, and suggested that the differences in the tip
whereas an increase in shear stresses may induce either resistances measured using electrical and mechanical cones
positive or negative pore pressure variations, depending on may be a function of material consistency, with mechanical
the dilatancy properties and on the mobilized shear stress cones producing higher values in loose sands and NC clays,
level in the soil. Soft, NC to moderately OC fine-grained and electrical cones producing significantly higher values in
soils will, because of their contractive behavior, yield dense sands. Shields (1981) asserted that the influence of
positive pore pressures both on the cone and along the cone diameter in measured penetration resistance should be
cylindrical shaft. In such soils, the measured cone resistance small in an isotropic, homogeneous and uniform soil, while
qc is generally small, whereas the pore pressure u2 is such influence could increase in layered soil systems. Smits
generally large in comparison. Thus, the correction to qc is (1982) confirmed Schmertmann’s (1978) observations
significant. In dilative soils, such as dense sands, silts and regarding higher resistance measurements by electrical
heavily OC clays, larger positive excess pore pressures will cones in dense sands. De Ruiter (1982) reported no
be developed on the cone, whereas very low or even significant variations in resistance measurements with cone
negative pore pressures will be developed along the cone cross-section aeras ranging from 5 to 15 cm2, and suggested
shaft. Hence, In such soils, the pore pressures measured that cone size has a less significant effect on resistance
behind the cone (u2) can be less than the equilibrium pore values than the apex angle. Lunne et al. (1986) also reported
water pressure, and cavitation can occur. Because the qc no significant variations in resistance measurements between
values are generally large and u2 values are generally small resistances measured using 10 cm2 and 15 cm2 cones.
in comparison, the correction for qt is often negligible. Thus, Sweeney (1987) noted the existence of a scale effect
it is preferrable to use the measured cone resistance qc for between a 4.1 cm2 and a 10 cm2 cone, more pronounced for
interpretation in stiff, highly OC fine-grained saturated soils. cone tip resistance readings than for sleeve friction readings,
3
Bruzzi & Battaglio (1987) observed that the penetration and for higher penetration resistances. Juran & Tumay
pore pressure could be influenced by the axial load applied (1989) compared results of piezocone tests performed with
to the cone as a result of the deformations in the components 10 cm2 and 15 cm2 cones, and reported no significant
of the probe, especially the pore pressure transducers and the difference in tip resistance and excess pore pressure, though
filter elements. referring to the fact that the 10 cm2 cone systematically gave
4
Progressive wear of the cone may lead to decreased local side friction resistance 20% higher than the 15 cm2
accuracy of measurements. The IRTP (ISSMGE 2001) sets cone. De Lima & Tumay (1991) conducted a statistical
tolerances for changes in cone geometry due to wear. If the evaluation of scale effects, and provided corrective
allowable tolerances set by IRTP for cone diameter are equations for proper cross-correlation for converting cone
adhered to, then the maximum error in the measured cone tip tip and sleeve friction measurements obtained with a 15 cm2
resistance that can be obtained , simply from wear of a 10 cone to a 10 cm2 cone.
8 CHAPTER 1
f) compression of rods (in mechanical CPT tests) 6; g) upon: a) type of test (electrical or mechanical)9; b)
hysteresis;7 and h) nonlinearity8 of testing inclination10; c) rate of penetration11; d) calibration
instrumentation. errors12; and e) temperature changes13. Based on the
Based on the results from the geotechnical literature, results from the geotechnical literature, Orchant et al.
Orchant et al. (1988) classified the relative effect of (1988) classified the relative effect of the major
the major components of equipment-related error on components of operator/procedural error on CPT
CPT results as: “moderate to significant” for the type results as: “moderate to significant” for telescoping
of test performed (electrical or mechanical); “minor”
9
for cone size; “moderate to significant” for cone Several authors (e.g. De Ruiter 1971, 1982; Kulhawy &
angle; “significant” for rod compression in Trautmann 1996; Schmertmann 1978; Smits 1982) have
mechanical CPT tests; “minor to moderate” for highlighted the higher dependence on operator/procedural
errors of mechanical CPT tests than electrical tests. The
manufacturing defects and excessive cone wear; and
“telescoping” advancement in mechanical CPT testing, for
“minor” for leaky seals in electrical CPT testing. instance, does not completely separate tip and side resistance
measurements; thus, the expertise of personnel has been
1.4.1.2. Operator/procedural errors observed to play a relevant role in the quality of ouput data
(De Ruiter 1971; Kulhawy & Trautmann 1996; Smits 1982).
Orchant et al. (1988) showed that the degree of
Electrical cone penetration testing has been recognized to be
uncertainty associated with the other component of almost operator-independent (e.g. De Ruiter 1981; Kulhawy
systematic error, i.e. equipment and & Trautmann 1996; Schaap & Zuidberg 1982) provided that
operator/procedural parameters, is also closely related the required calibration procedures are observed strictly
to the level of standardization of the test procedure. before and during testing.
10
Tests with well-defined specifications, such as the The issue of deviation from verticality in CPT testing has
been addressed by various authors (e.g. Baligh et al. 1979;
CPT, show less total measurement variability than
De Ruiter 1971; Schmertmann 1978). Deviations from
tests with less stringent standards, such as the SPT. verticality can lead to substantial errors in the interpretation
Operator/procedural influences are considered of results as the depth of the measured resistance values
together because they are closely related. In most becomes uncertain. The IRTP (ISSMGE 2001) states that
cases, the operator controls the testing procedure, and “the thrust machine shall be set up so as to obtain a thrust
direction as near vertical as practicable. The deviation from
deviations from normal or standard procedures are
vertical of the thrust direction shall not exceed 2%.”
directly assignable to the operator. Possible sources of 11
The effect of penetration rate in CPT testing has been
operator/procedural errors include carelessness, addressed by several authors (e.g. Baligh et al. 1979;
fatigue, inexperience, misreading, misrecording, Campanella et al. 1982; De Ruiter 1981; Jamiolkowski et al.
computational errors, failure to operate the readout 1985; Kok 1974; Marsland & Quarterman 1982; Powell &
instrument correctly, incorrect installation, improper Quarterman 1988; Sanglerat et al. 1982; Te Kamp 1982;
Villet & Mitchell 1981). Rate effects can be caused by creep
electrical connections, and wrong switch positions
and particle crushing, though pore pressure effects usually
(Dunnicliff 1993). predominate (Lunne et al. 1997). The effect of penetration
Orchant et al. (1988) suggested that the magnitude has been shown to be more pronounced in clays (Baligh et
and type of operator/procedural error may depend al. 1979) and silts (Campanella et al. 1982; Sanglerat et al.
1982) than in sands (De Ruiter 1981; Kok 1974). Te Kamp
(1982) investigated the effect of the rate of penetration in
offshore tests. According to the IRTP (ISSMGE 2001), “the
6
When penetrating through dense sands, at depths greater rate of penetration shall be 20 mm/s with a tolerance of ± 5
than 30 m, the elastic compression of the smaller inner rods mm/s. In case of a piezocone this tolerance shall be
can become considerably large, and may impede the forward narrowed.” Lunne et al. (1997) report a tolerance of ± 2
movement of the cone tip ahead of the friction sleeve, thus mm/s for piezocone testing. Campanella et al. (1982) and
leading to misinterpretations of soil parameters deriving Jamiolkowski et al. (1985) suggested that the rate of
from test results. The above complications are eliminated in penetration of 20 mm/s should be maintained as strictly as
electrical CPT tests, as the latter is a single-rod system possible.
12
(Orchant et al. 1988). According to Lunne et al. (1997), the calibration of
7
When the quantity being measured is subject to cyclic sensors is fundamental to the successful and reliable use of
change, the indicated measured value sometimes depends on CPT testing. Calibration should evaluate - and may
whether the measurement is increasing or decreasing. The minimize, if properly conducted - repeatability, nonlinearity
separation, if existing, betweeen the curves of increasing and and hysteresis effects. The IRTP (ISSMGE 2001) states that
decreasing measurements is called hysteresis. Instruments “manometers shall be calibrated at least every 6 months”,
with large hysteresis are not suitable for measurement of while “load cells or proving rings shall be calibrated at least
rapidly changing parameters. every 3 months.”
8 13
An instrument is said to be linear when its indicated Lunne et al. (1986) showed that temperature changes may
measured values are directly proportional to the quantity influence measurements significantly, as a change in
being measured. temperature can cause a shift in load cell output at zero load.
CHAPTER 1 9
versus continuous penetration; “minor to moderate” test effects may contribute significantly to the total
for calibration error; “minor” for the rate of measurement error.
penetration; and “moderate to significant” for Jaksa (1995) asserted that it is likely that the random
deviation from verticality during testing. measurement error associated with the CPT may be
Systematic error, according to Baecher (1982; 1986) less than, or equal to, 1.6 %. This result compares
is the most difficult source of uncertainty to estimate. well with the conclusion made by Campanella et al.
The only way to estimate this component may be (1987), that the random measurement error of the
through comparison of predicted with observed CPT may be as low as 1%.
performance or by field-scale experiments (e.g.
Bjerrum 1972). In such an approach, however, a large 1.4.2.1. Statistical estimation error
number of uncertainties and biases are aggregated Statistical estimation error is commonly included in
together, including those due to inaccuracies in theory random testing error (Kulhawy et al. 1992). Such
and method of analysis. Thus, systematic error and source of uncertainty arises because of limited
model bias are usually inseparable. numerosity of data and, consequently, poor estimation
1.4.2. Random testing error of data statistics (i.e. mean, standard deviation) even
in absence of measurement bias. Statistical error may
Random testing error represents the zero-mean scatter
be reduced by increasing the number of
of test results, which is not directly assignable to
measurements. Systematic error in the statistical
specific operator/procedural or equipment parameters
estimation of soil parameters may be calculated
and is not caused by spatial variability. Random
directly from statistical theory. Here, the issue of
testing error affects test results above and below the
statistical uncertainty will not be addressed
test mean equally (e.g. Baecher 1982, 1986; Orchant
specifically. A detailed treatment of statistical
et al. 1988). Dunnicliff (1993) identified the
estimation error is provided in Spry et al. (1988).
following possible sources of random error: noise,
internal friction, hysteresis, and environmental
effects, the latter arising because of the influence of soil variability, the above values indicate total CPT
heat, humidity, vibration, shock waves, moisture, variability. However, as the locations of the soundings were
pressure, etc. The same Author suggested two adjacent to each other, the effect of inherent variability
possible approaches for minimizing environmental should be nearly equal for electrical and mechanical tests.
Among the most significant deductions from the
errors: a) measuring the extent of the influence and
Nottingham’s (1975) results, the following may be
applying suitable correction factors; and b) choosing highlighted: a) side resistance measurements obtained with
an instrument which is not adversely affected by the mechanical cones were generally less variable than the side
environment. measurements obtained with electrical cones; and b) the
Numerically, random error may be evaluated by differences in side resistance measurements between
analyzing series of tests performed under identical electrical and mechanical equipments depends on the type of
soil penetrated.
conditions. Orchant et al. (1988) reviewed and
Baligh et al. (1979) conducted series of electrical cone tests
presented the results of several studies focusing on in homogeneous clay deposits in conditions of equipment
the quantification of random testing error for cone and procedural control, obtaining coefficients of variation
penetration testing14. Such data indicate that random for qc ranging from 0.05 to 0.15.
Parkin et al. (1980) conducted a series of electrical CPT
tests, obtaining coefficients of variation of 0.05 for qc and
14
Poplin (1965) performed a series of laboratory CPT tests 0.10 for fs. Orchant et al. (1988) assert that, as testing
in controlled dry sand deposits using a mechanical cone with conditions were nearly replicate, such values may indicate
non-standard geometry; the obtained coefficients of the random error associated with the tests. Villet (1981)
variation for qc ranged from 0.04 to 0.07. Joustra (1974) conducted a series of laboratory CPT tests in four sands;
performed tests in medium to dense sand with equipment among the results obtained, average coefficients of variation
and procedural control, using: a) standard electrical cone; b) of 0.07 for qc, regardless of penetration rate.
mechanical mantle cone; and c) “necked-down” electrical Rol (1982) performed three series of tests using: a) standard
cone, and obtaining ranges of the coefficient of variation for mechanical cone; b) standard electrical cone; and c) Delft
qc of 0.14-0.45, 0.12-0.43 and 0.15-0.44, respectively. electrical cone, and obtaining coefficients of variation for qc
Nottingham (1975) reported the results of a series of of 0.24, 0.17 and 0.21, respectively.
mechanical and electrical tests performed side by side in Boghrat (1982) presented the results of a series of
silty sands. Coefficients of variations ranged from 0.28 to Begemann friction CPT tests in heterogeneous sand to
0.59 for qc and from 0.30 to 0.88 for fs for electrical tests, clayey sand deposits. Values from 0.17 to 0.81 were
and from 0.21 to 0.52 for qc and from 0.24 to 0.80 for fs for obtained for the coefficient of variation of qc measurements;
mechanical tests. As no attempt was made to separate the Orchant et al. (1988) described such value as pertaining to
various components of variability and to quantify inherent total CPT variability.
10 CHAPTER 1
1.5. REFERENCE VALUES OF EPISTEMIC Proceedings of the Seminar on Probabilistic Methods in
UNCERTAINTY USED IN THE PRESENT STUDY Geotechnical Engineering, M.E. Hynes-Griffin and L.L.
Buege (eds.), Miscellaneous Paper GL-83-26, U.S.
Kulhawy & Trautmann (1996) extensively reviewed Army Engineer Waterways Experiment Station,
existing literature and investigated the magnitudes of Vicksburg, 65 p.
the various components of epistemic uncertainty for BAECHER, G.B. (1986). “Geotechnical Error Analysis.”
mechanical cone testing (MCPT), electric cone Transportation Researche Record No. 1105, 23-31.
testing (ECPT) and piezocone testing (CPTU). An BAINBRIDGE, T.R. (1985). “The Committee on Standards:
additive model was proposed for the total epistemic Precision and Bias.” ASTM Standardization News, Vol.
13, No. 1, 44-46.
uncertainty:
BALIGH, M.M., VIVATRAT, V., LADD, C.C. (1979).
(1-1) K epi
2
K eqp
2
K 2prc K rnd
2
“Exploration and Evaluation of Engineering Properties
for Foundation Design of Offshore Structures.” Report
in which Kepi is the coefficient of variation of total MITSG 79-8, Massachusetts Institute of Technology,
Cambridge, 268 p.
epistemic uncertainty; Kprc is the coefficient of
variation of equipment error; Kprc is the coefficient of BJERRUM, L. (1972). “Embankments on Soft Ground.”
Proceedings of the ASCE Specialty Conference on
variation of operator/procedural errors, and Krnd is the
Earth and Earth-Supported Structures, Boulder,
coefficient of variation of random testing error. The Colorado, 111-159.
synthetic results, which include the studies presented
BOGHRAT, A. (1982). “The Design and Construction of of a
herein, are shown in Table 1-1. Piezoblade and an Evaluation of the Marchetti
Dilatometer.” Ph.D. Thesis, University of Florida,
Gainesville, Department of Civil Engineering, 244 p.
Table 1-1. Estimated coefficients of variation for
mechanical (MCPT), electric (ECPT) and piezocone BRUZZI, D., BATTAGLIO, M. (1987). “Pore Pressure
(CPTU) testing (from Kulhawy & Trautmann 1996) Measurement During Cone Penetration Test.” ISMES
Research Report No. 229.
test Keqp Kprc Krnd Kepi
qc , f s qc fs qc fs qc fs CAMPANELLA, R.G., GILLESPIE, D., ROBERTSON, P.K.
MCPT 0.05 0.10 0.15 0.10 0.15 0.15 0.22 (1982). “Pore Pressure During Cone Penetration
ECPT 0.03 0.05 0.05 0.05 0.10 0.08 0.12 Testing.” Proceedings of the 2nd European Symposium
CPTU 0.03 0.05 0.05 0.05 0.10 0.08 0.12 on Penetration Testing, ESOPT II, Amsterdam,
Netherlands, 507-512, Balkema, Rotterdam,
Netherlands.
As may be seen in Table 1-1, the magnitudes of all
CARDOSO, A.S., FERNANDES, M.M. (2001). “Characteristic
components of epistemic uncertainty are greater in Values of Ground Parameters and Probability of Failure
MCPT tests than in ECPT and CPTU. The numerical in Design According to Eurocode 7.” Géotechnique,
values attributed to ECPT and CPTU coincide. The Vol. 51, No. 6, 519-531.
coefficients of variation representative of epistemic CHERUBINI, C., GIASI, C.I., RÉTHÁTI, L. (1993). “The
CPT uncertainty reported in Table 1-1 will be referred Coefficients of Variations of Some Geotechnical
Properties.” In Probabilistic methods in geotechnical
to in Part B of the present dissertation. engineering (eds. K.S. Li, and S.C.R. Lo), pp. 179-184,
A.A. Balkema, Rotterdam, Netherlands.
CHAPTER 1 11
CRESPELLANI, T., MADIAI, C., VANNUCCHI, G. (2000). HATAKEYAMA, A., YASUDA, N. (1987). “Optimization of
“Analysis of the Uncertainties in Geotechnical Braced Excavation Using Sheet Piles by Reliability-
Applications of Newmark’s Rigid Block.” Proceedings Based Design.” Proceedings of the 5th International
of the 8th International Conference on Applications of Conference on Applications of Statistics and Probability
Statistics and Probability in Soil and Structural in Soil and Structural Engineering, Vancouver, BC,
Engineering (Melchers & Stewart, eds.), Balkema, Canada, Vol. 2, 893-900.
Rotterdam, 431-437.
HOLDEN, J.C. (1977). “The Calibration of Electrical Cone
DEGROOT, D.J., BAECHER, G.B. (1993). “Estimating Penetrometers in Sand.” Norwegian Geotechnical
Autocovariances of In-Situ Soil Properties.” Journal of Institute, Internal Report No. 52108-2, 29, 3.
Geotechnical Engineering, Vol. 119, No. 1, 147-166.
HOLTZ, R.D., KRIZEK, R.J. (1971). “Statistical Evaluation of
Soil Test Data.” Proceedings of the 1st International
DE LIMA, D.C., TUMAY, M.T. (1991). “Scale Effects in
Conference on Applications of Statistics and Probability
Cone Penetration Tests.” Proceedings of the
Geotechnical Engineering Congress 1991, ASCE, in Soil and Structural Engineering, Hong Kong, 229-
Boulder, CO, U.S.A., Vol. 1, 38-51. 266.
DE RUITER, J. (1971). “Electric Penetrometer for Site HOULSBY, G.T., TEH, C.I. (1988). “Analysis of the
Investigation.” Journal of the Soil Mechanics and Piezocone in Clay.” Proceedings of the International
Foundations Division, ASCE, Vol. 92, No. SM2, 457- Symposium on Penetration Testing, ISOPT-1, Orlando,
472. FL, U.S.A., 777-783, Balkema, Rotterdam, Netherlands.
DE RUITER, J. (1981). “Current Penetrometer Practice: Cone JAKSA, M.B. (1995). “The Influence of Spatial Variability
Penetration Testing and Experience.” Proceedings of on the Geotechnical Design Properties of a Stiff,
the ASCE National Convention, St. Louis, MI, U.S.A., Overconsolidated Clay.” Ph.D. Tehsis, University of
1-48. Adelaide, Australia.
DE RUITER, J. (1982). “The Static Cone Penetration Test.”
JAKSA, M.B., BROOKER, P.I., KAGGWA, W.S. (1997).
Proceedings of the 2nd European Symposium on
“Inaccuracies Associated with Estimating Random
Penetration Testing, ESOPT II, Amsterdam, Measurement Errors.” Journal of Geotechnical and
Netherlands, 389-405, Balkema, Rotterdam, Geoenvironmental Engineering, ASCE, Vol. 123, No. 5,
Netherlands. 393-401; discussion by BAECHER, G.B. (1999), Vol.
125, No. 1, 79-81.
DUNNICLIFF, J. (1993). Geotechnical instrumentation for
monitoring field performance, John Wiley & Sons, Inc.,
New York, NY. JAMIOLKOWSKI, M., LADD, C.C., GERMAINE, J.T.,
LANCELLOTTA, R. (1985). “New Developments in Field
DURGUNOöLU, H.T., MITCHELL, J.K. (1975). “Static and Laboratory Testing of Soils.” Proceednigs of the
Penetration Resistance of Soils.” I-II. Proceedings of 11th International Conference on Soil Mechanics and
the ASCE Specialty Conference on In Situ Measurement Foundation Engineering, San Francisco, CA, Balkema,
of Soil Properties, Raleigh, NC, U.S.A., Vol. 1, 151- Rotterdam, Vol. 1, 57-153.
189.
JOUSTRA, K. (1974). “Comparative Measures on Influence
FENTON, G. (1999). “Estimation for Stochastic Soil of Cone Shape on Results of Soundings.” Proceedings
Models.” Journal of Geotechnical and of the 1st European Symposium on Penetration Testing,
Geoenvironmental Engineering, ASCE, Vol. 125, No. 6, Stockholm, Sweden, Vol. 2.2, 199-204.
470-485.
JURAN, I., TUMAY, M.T. (1989). “Soil Stratification Using
FILIPPAS, O.B., KULHAWY, F.H., GRIGORIU, M.D. (1988). the Dual Pore-Pressure Piezocone Test.” Transportation
“Reliability-Based Foundation Design for Transmission Research Record No. 1235, 68-78.
Line Structures: Uncertainties in Soil Property
Measurements.” Electric Power Research Institute Rpt. KAY, J.N. (1990). “Approximate Framework for
EL-5507(3), EPRI, Palo Alto. Probabilistic Evaluation of Soil Properties.”
Proceedings of the University of Adelaide Special
FREDLUND, D.G., DAHLMAN, A.E. (1971). “Statistical Symposium on the Occasion of George Sved’s 80th
Geotechnical Properties of Glacial Lake Edmonton Brithday, University of Adelaide, Adelaide, South
Sediments.” Proceedings of the 1st International Australia, 184-197.
Conference on Applications of Statistics and Probability
in Soil and Structural Engineering, Hong Kong, 204- KAY, J.N., KULHAWY, F.H., GRIGORIU, M.D. (1991).
228. “Assessment of Uncertainties in Geotechnical Design
Parameters.” Proceedings of the 6th International
GHINELLI, A., VANNUCCHI, G. (1983). “Statistical Conference on Applications of Statistics and Probability
Evaluation of Geotechnical Properties of Florence’s in Soil and Structural Engineering, Mexico City,
Clay by Means of Factor Analysis.” Proceedings of the Mexico, 683-692.
4th International Conference on Applications of
Statistics and Probability in Soil and Structural KOK, L. (1974). “Effect of Penetration Speed & Cone Shape
Engineering, Florence, Italy, Pitagora Editrice, 987- on Dutch Static Cone Penetration Test Results.”
1002. Proceedings of the 1st European Symposium on
Penetration Testing, Stockholm, Sweden, Vol. 2.2, 215-
HANDY, R.L. (1980). “Realism in Site Exploration: Past, 220.
Present and Future, and Then Some All-Inclusive.” in
Site Exploration in Soft Ground Using In-Situ
KULHAWY, F.H., ROTH, N.J.S., GRIGORIU, N.B. (1991).
Techniques, Report FHWA-TS-80-202, Federal “Some Statistical Evaluations of Geotechnical
Highway Administration, Alexandria, 239-248. Properties.” Proceedings of the 6th International
12 CHAPTER 1
Conference on Applications of Statistics and Probability MAGNAN, J.P. (1982). Les méthodes statistiques et
in Soil and Structural Engineering, Mexico City, probabilistes en mécanique des sols, Presses de l’Ecole
Mexico, 705-712. Nationale del Ponts et Chaussées.
KULHAWY, F.H., TRAUTMANN, C.H. (1996). “Estimation of MARSLAND, A., QUARTERMAN, R.S.T. (1982). “Factors
In-Situ Test Uncertainty.” Uncertainty in the Geologic Affecting the Measurement and Interpretation of Quasi
Environment: From Theory to Practice, ed. C.D. Static Penetration Tests in Clays.” Proceedings of the
Shackleford, P.P. Nelson and M.J.S. Roth, ASCE 2nd European Symposium on Penetration Testing,
Geotechnical Special Publication No. 58, Madison, WI, ESOPT II, Amsterdam, Netherlands, Vol. 2, 697-702,
U.S.A., 269-286. Balkema, Rotterdam, Netherlands.
KRAHN, J., FREDLUND, D.G. (1983). “Variability in the MURPHY, R.B. (1961). “On the Meaning of Precision and
Engineering Properties of Natural Soil Deposits.” Accuracy.” Materials Research and Standards, Vol. 1,
Proceedings of the 4th International Conference on No. 4, 264-267.
Applications of Statistics and Probability in Soil and
Structural Engineering, Florence, Italy, Pitagora NOTTINGHAM, L.C. (1975). “Use of Quasi-Static
Editrice, 1017-1029. Penetrometer Data to Predict Load Capacity of Piles.”
Ph.D. Thesis, University of Florida, Gainesville,
LACASSE, S., DE LAMBALLERIE, J.Y.N. (1995). “Statistical Department of Civil Engineering, 553 p.
Treatment of CPT Data.” Proceedings of the OBERGUGGENBERGER, M., FELLIN, W. (2002). “From
International Symposium on Cone Penetration Testing,
Probability to Fuzzy Sets: The Struggle for Meaning in
CPT ’95, Linköping, Sweden, Vol. 2, 369-380, Swedish
Geotechnical Society. Geotechnical Risk Analysis.” Proceedings of the
International Conference on Probabilistics in
LACASSE, S., NADIM, F. (1996). “Uncertainties in Geotechnics – Technical and Economic Risk
Characterizing Soil Properties.” Uncertainty in the Estimation, Graz, Austria, September 15-19, 2002,
Geologic Environment: From Theory to Practice, ed. VGE, 29-38.
C.D. Shackleford, P.P. Nelson and M.J.S. Roth, ASCE
Geotechnical Special Publication No. 58, Madison, WI, ORCHANT, C.J., TRAUTMANN, C.H., KULHAWY, F.H.,
U.S.A., 49-75. (1986). “In-Situ Testing to Characterize Electric
Transmission Line Routes.” Proceedings of the ASCE
LEE, I.K., WHITE, W., INGLES, O.G. (1983). Geotechnical Specialty Conference In Situ ’86: Use of In Situ Tests in
Engineering, Pitman Publishing Inc., Marshfield, MA, Geotechnical Engineering, Blacksburg, VA, U.S.A.,
U.S.A. 869-886.
LEVADOUX, J.N., BALIGH, M.M. (1986). “Consolidation ORCHANT, C.J., KULHAWY, F.H., TRAUTMANN, C.H.
After Undrained Piezocone Penetration. I: Prediction.” (1988). “Reliability-Based Foundation Design for
Journal of Geotechnical Engineering, ASCE, Vol. 112, Transmission Line Structures: Critical Evaluation of In-
No. 7, 707-726. Situ Test Methods.” Report EL-5507(2), Electric Power
Research Institute, Palo Alto, California, U.S.A., 207 p.
LUMB, P. (1966). “The Variability of Natural Soils.”
PARKIN, A.K., HOLDEN, J., AAMOT, K., LAST, N., LUNNE,
Canadian Geotechnical Journal, Vol. 3, No. 2, 74-97.
T. (1980). “Laboratory Investigation of CPT’s in Sand.”
Report 52108-9, Norwegian Geotechnical Institute,
LUMB, P. (1970). “Safety Factors and the Probability
Oslo, Norway, 31 p.
Distribution of Soil Strength.” Canadian Geotechnical
Journal, Vol. 7, 225-242. PARKIN, A.K., LUNNE, T. (1982). “Boundary Effects in the
Laboratory Calibration of a Cone Penetrometer in
LUMB, P. (1971). “Precision and Accuracy of Soil Tests.” Sand.” Proceedings of the 2nd European Symposium on
Proceedings of the 1st International Conference on Penetration Testing, ESOPT II, Amsterdam,
Applications of Statistics and Probability in Soil and Netherlands, Vol. 2, 761-768, Balkema, Rotterdam,
Structural Engineering, Hong Kong, 329-345. Netherlands.
LUMB, P. (1974). “Application of Statistics in Soil
Mechanics”, chapter 3 in Soil Mechanics – New PHOON, K.-K., KULHAWY, F.H. (1999a). “Characterization
Horizons, I.K. Lee (ed.), Newnes-Butterworths, of Geotechnical Variability.” Canadian Geotechnical
London, U.K., 44-111. Journal, Vol. 36, No. 4, 612-624.
LUMB, P. (1975). “Spatial Variability of Soil Properties.” PHOON, K.-K., KULHAWY, F.H. (1999b). “Evaluation of
Proceedings of the 2nd Conference on Application of Geotechnical Property Variability.” Canadian
Probability and Statistics to Soil and Structural Geotechnical Journal, Vol. 36, No. 4, 625-639.
Engineering, Aachen, Germany, Vol. 2, 397-421. POPESCU, R., PREVOST, J.H., DEODATIS, G. (1996).
LUNNE, T., EIDSMOEN, T., GILLESPIE, D., HOWLAND, J.D. “Influence of Spatial Variability of Soil Properties on
(1986). “Laboratory and Field Evaluation of Cone Seismically Induced Soil Liquefaction.” Uncertainty in
Penetrometers.” Proceedings of the ASCE Specialty the Geologic Environment: From Theory to Practice,
Conference In Situ ’86: Use of In Situ Tests in ed. C.D. Shackleford, P.P. Nelson and M.J.S. Roth,
Geotechnical Engineering, Blacksburg, VA, U.S.A., ASCE Geotechnical Special Publication No. 58,
714-729. Madison, WI, 1098-1112.
LUNNE, T., ROBERTSON, P.K., POWELL, J.J.M. (1997). Cone POPLIN, J.K. (1965). “Statistical Evaluation of Cone
Penetration Testing in Geotechnical Practice, E & FN Penetrometer Test Data.” Miscellaneous PaperNo.3-
Spon, London. 749¸U.S. Army Engineer Waterways Experiment
Station, Vicksburg, 16 p.
CHAPTER 1 13
POST, M.L., NEBBELING, H. (1995). “Uncertanities in Cone SCHULTZE, E. (1975). “Some Aspects Concerning the
Penetration Testing.” Proceedings of the International Application of Statistics and Probability to Foundation
Symposium on Cone Penetration Testing, CPT ’95, Structures.” Proceedings of the 2nd International
Linköping, Sweden, Vol. 2, 73-78, Swedish Conference on Applications of Statistics and Probability
Geotechnical Society. in Soil and Structural Engineering, Aachen, Germany,
Vol. 2, 457-494.
POWELL, J.J.M., QUARTERMAN, R.S.T. (1988). “The
Interpretation of Cone Penetration Tests in Clays, With SEMAAN, I., BOULEMIA, C., BOISSIER, D. (2000).
Particular Reference to Rate Effects.” Proceedings of “Possibilistic Modelling of Data from Site
the 1st International Symposium on Penetration Testing Investigation.” Applications of Statistics and Probability
(ISOPT-1), De Ruiter, J. (ed.), March 20-24, 1988, (Melcher & Stewart, eds.), Balkema, Rotterdam,
Orlando, FL, U.S.A., A.A. Balkema, Rotterdam, Netherlands, 1115-1122.
Netherlands, 903-910.
SHIELDS, D.H. (1981). “Should ASTM Adopt the European
POWELL, J.J.M., QUARTERMAN, R.S.T. (1991). “The Effect Standard CPT?” Proceedings of the ASCE National
of Filter Position on the Measurement of Pore Water Convention, St. Louis, MI, October 26-30, 1981, 383-
Pressures in Piezocone Tests.” BRE Report. 393.
PUàA, W., TRACZYK, R. (1987). “Statistical Method of SHOU, K.-J., WANG, C.-F., CHEN, Y.-L. (2002). “Risk
Evaluating of the Strength of a Dumping.” Proceedings Estimation of the Earthquake-Induced Chiufengershan
of the 5th International Conference on Applications of Landslide in Taiwan.” Proceedings of the International
Statistics and Probability in Soil and Structural Conference on Probabilistics in Geotechnics –
Engineering, Vancouver, BC, Canada, Vol. 2, 1047- Technical and Economic Risk Estimation, Graz, Austria,
1053. September 15-19, 2002, VGE, 313-320.
ROBERTSON, P.K., CAMPANELLA, R.G. (1988). “Guidelines SPRY, M.J., KULHAWY, F.H., GRIGORIU, M.D. (1988).
for Geotechnical Design Using CPT and CPTU.” “Reliability-Based Foundation Design for Transmission
University of British Columbia, Vancouver, Department Line Structures: Geotechnical Site Characterization
of Civil Engineering, Soil Mechanics Series 120. Strategy.” Electric Power Research Institute Rpt. EL-
5507(1), EPRI, Palo Alto.
ROL, A.H. (1982). “Comparative Study on Cone Resistance STAMATOPOULOS, A.C., KOTZIAS, P.C. (1975). “The
Measured With Three Types of CPT Tips.” Proceedings
Relative Value of Increasing Number of Observations.”
of the 2nd European Symposium on Penetration Testing,
Proceedings of the 2nd International Conference on
ESOPT II, Amsterdam, Netherlands, Vol. 2, 813-819,
Balkema, Rotterdam, Netherlands. Applications of Statistics and Probability in Soil and
Structural Engineering, Aachen, Germany, Vol. 2, 495-
510.
SANGLERAT, G. (1972). The Penetrometer and Soil
Exploration.Elsevier, Amsterdam, 464 p. SWEENEY, B.P. (1987). “Liquefaction Evaluation Using a
Miniature Cone Penetrometer and Large Scale
SCHAAP, L.H.J. ZUIDBERG, H.M. (1982). “Mechanical and Calibration Chamber.” Ph.D. Thesis, Stanford
Electrical Aspects of the Electric Cone Penetrometer University, 281 pp.
Tip.” Proceedings of the 2nd European Symposium on
Penetration Testing, ESOPT II, Amsterdam,
TABBA, M.M., YONG, R.N. (1981a). “Mapping and
Netherlands, 841-851, Balkema, Rotterdam,
Predicting Soil Properties: Theory.” Journal of the
Netherlands.
Engineering Mechanics Division, ASCE, Vol. 107, No.
5, 773-793.
SCHMERTMANN, J.H. (1978). “Guidelines for Cone
Penetration Test, Performance and Design.” US Federal TABBA, M.M., YONG, R.N. (1981b). “Mapping and
Highway Administration, Report FHWA-TS-78-209. Predicting Soil Properties: Applications.” Journal of the
Engineering Mechanics Division, ASCE, Vol. 107, No.
SCHULTZE, E. (1971). “Frequency Distributions and 5, 795-811.
Correlations of Soil Properties.” Proceedings of the 1st
International Conference on Applications of Statistics TANG, W.H. (1984). “Principles of Probabilistic
and Probability in Soil and Structural Engineering, Characterization of Soil Properties.” Probabilistic
Hong Kong, 371-387. Characterization of Soil Properties: Bridge Between
Theory and Practice, D.S. Bowles and H.Y. Ko (eds.),
ASCE, New York, 74-89.
14 CHAPTER 1
TE KAMP, W.G.B. (1982). “The Influence of the Rate of
Penetration on the Cone Resistance qc in Sand.”
Proceedings of the 2nd European Symposium on
Penetration Testing, ESOPT II, Amsterdam,
Netherlands, 627-633, Balkema, Rotterdam,
Netherlands.
CHAPTER 1 15
intentionally blank
16 CHAPTER 1