SlideShare a Scribd company logo
1 of 18
Download to read offline
ESTIMATION OF IN-SITU TEST UNCERTAINTY
Fred H. Kulhawy’ , F. ASCE and Charles H. Trautmann* , M. ASCE
ABSTRACT: In-situ tests have several primary sources of uncertainty, including
calibration errors, operator influence, sampling errors, and variations in test proce-
dure. This paper focuses on evaluating these sources and obtaining an overall esti-
mate of test uncertainty, as distinguished from soil variability, for the following test
methods: (a) standard penetration test, (b) cone penetration test, (c), dilatometer test,
(d) vane shear test, and (e) pressuremeter test. Previous studies are reviewed that
cover a wide range of goals, including: (a) qualitative discussions that emphasize the
various sources of error, (b) soil property correlations, which often make separation
of the sources of error difficult, and (c) studies specifically geared to evaluation of
test uncertainties. Laboratory calibration studies commonly are most useful in this
regard. Using data from prior investigations, this paper first places error sources in
order of importance. Next, estimates for the most important errors for each test are
obtained. Finally, the results are analyzed and summarized into a table of combined
uncertainties. This table represents the best-estimate uncertainty for each test proce-
dure, independent of soil variability or the correlation of test result to engineering
soil property. Basically, these results show that tests with more potential for equip-
ment and operator/procedural variations are likely to exhibit greater test uncertainty.
INTRODUCTION
Site characterization is performed to evaluate the subsurface conditions needed
for safe and economical geotechnical design. The information obtained during site
characterization includes the determination of the subsurface deposits and their geo-
metries and engineering properties. The level of detail required for this information
depends on the nature and economics of the specific project and can range from
general reconnaissance, which uses existing information, to detailed field sampling
1 - Prof., School of Civil & Environ. Eng., Cornell Univ., Hollister Hall, Ithaca, NY
14853-3501
2 - Exec. Dir., The Sciencenter, First Street, Ithaca, NY 14850-3507; Adjunct Assoc.
Prof., School of Civil & Environ. Eng., Cornell Univ.
BACK
210 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
and testing.
Most site characterization studies involve in-situ drilling, sampling, and testing
in one or more forms to provide geotechnical design information. Evaluation of the
uncertainties in this information commonly is done using judgment that is based on
experience. Two prominent sources of in-situ test variability that can be identified
include: (a) natural spatial variability of the soil deposits resulting from their method
of formation and the subsequent environmental changes that have taken place, and
(b) uncertainty in the test measurements resulting from equipment, test procedure,
operator, and random test effects. Quantification of spatial variability is discussed
elsewhere in these proceedings (Phoon & Kulhawy 1996), while test uncertainty is
discussed herein.
Wherever possible, it is preferable to use quantitative measures of uncertainty.
However, because it is often difficult to separate the various sources of uncertainty,
this approach has been used rarely. This paper evaluates pertinent quantitative infor-
mation on this issue and develops estimates of overall in-situ test uncertainty.
Most of the research reported herein was completed several years ago. The
results of this extensive study previously have been available only through a limited-
access report of the Electric Power Research Institute and were not in the general
literature. This paper presents the key results of this study in condensed form, along
with several important updates in ASTM standards. The detailed results and analy-
ses are provided in the original report (Orchant et al. 1988), while only illustrative
examples and best-estimate conclusions can be presented herein because of length
restrictions. Subsequent detailed studies that are not described and referenced herein
have only reinforced the key results of this study. We all await the detailed studies in
progress at the U. S. national test sites to delineate these issues in a (hopefully) broad
and definitive manner.
BACKGROUND
Many authors have discussed the various sources of errors and uncertainty with
in-situ test methods. Generally, these studies have focused on identification of
sources and whether they increase or decrease a particular test result. For example,
one of the developers of the split-spoon and hence the standard penetration test (SPT)
was among the first to describe the abuses of this test (Fletcher 1965). He was able
to identify 13 “important” factors that affect SPT results, including inadequate bore-
hole cleaning, failure to maintain sufficient hydrostatic head in the borehole, varia-
tions in the 760~mm (30-in) hammer drop height, extremely long drill rods, etc.
While some of these factors can bias the test results in either direction (e.g., varia-
tions in the hammer drop height), other factors lead to a bias in only one direction
(e.g., interference of the free fall of the hammer reduces the energy per blow and
therefore increases the blow count). This increase is unconservative, because it
makes the soil appear firmer than it actually is.
Most influences on test uncertainty can be divided into two groups: equipment
effects and procedural/operator effects. Both of these types of effects have received
BACK
IN-SITU TEST UNCERTAINTY 271
considerable attention in understanding the influence of testing parameters on the in-
situ test results, although the majority of early studies have been qualitative or semi-
quantitative. Increasingly, authors have focused on quantitative evaluation of in-situ
test accuracy and precision. Commonly, the most useful studies involve replicate
tests under highly controlled conditions, such as in the laboratory or at field sites
where soil conditions are consistent over wide horizontal and vertical distances. In
the remainder of this paper, these quantitative studies will be highlighted, and the
results will be combined to develop best-estimate uncertainties for the common in-
situ test methods addressed herein.
ANALYSIS OF UNCERTAINTIES
In this paper, the five most common in-situ tests are analyzed with respect to
quantitative uncertainties. The five tests include the standard penetration test (SPT),
cone penetration test (CPT), vane shear test (VST), dilatometer test (DMT), and
pressuremeter test (PMT). For the CPT, both the mechanical (MCPT) and electric
(ECPT) cones are included and, for the pressuremeter, both the pre-bored (PMT) and
self-boring (SBPMT) versions are addressed. In the following analyses, many effects
related to non-standard equipment and test procedures are identified, but they are not
analyzed rigorously, including such factors as non-standard samplers or failure to
clean the hole before testing. It is assumed that these error sources are best addressed
through proper training and inspection. On the other hand, random errors resulting
from best-effort performance of the test personnel or equipment variations that are
not identified in the standardized test procedure are addressed quantitatively where
possible, including such factors as variations in the SPT hammer drop height when
using a rope-and-cathead system.
STANDARD PENETRATION TEST (SPT) UNCERTAINTIES
Basic Testing Issues
The standard penetration test is described in ASTM D1586, and the major fac-
tors influencing the SPT results are listed in Table 1. The equipment employed in
the SPT includes the hammer, hammer drop system, drill rods, and sampler. Each of
these equipment variables can have a major influence on the resulting N-values from
the test. The rod weight and length have been analyzed extensively by field and
wave-equation methods and have been shown to have a relatively minor influence
(e.g., Gibbs & Holtz 1957, McLean et al. 1975, Matsumota & Matsubara 1982).
Safety and donut-type hammers deliver different energies to the SPT sampler, with
the safety hammer delivering more energy (Seed et al. 1984). The anvil, which
receives the hammer impact, can have a significant influence on the results. For
example, Schmertmann (1978) indicated that tests performed with a small anvil can
result in N-values up to 50% greater than those obtained with larger anvils.
Many geotechnical engineers have cited the high degree of operator dependency
of the SPT as its most important limitation. Although ASTM D1586 specifies some
BACK
272 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
TABLE 1. SPT Testing Variables
Group
Equipment
Procedural/
Operator
SPT Variable Relative Effect
Item on Test Results
Non-standard sampler moderate
Deformed or damaged sampler moderate
Rod diameter/weight minor
Rod length minor
Deformed drill rods minor
Hammer type moderate to significant
Hammer drop system significant
Hammer weight minor
Anvil size moderate to significant
Drill rig type minor
Borehole size moderate
Method of maintaining hole minor to significant
Borehole cleaning moderate to significant
Insufficient hydrostatic head moderate to significant
Seating of sampler moderate to significant
Hammer drop method moderate to significant
Error in counting blows minor
procedural requirements for this test, such as the hammer drop height and seating of
the sampler, other variables, including the method of borehole drilling and method of
hammer drop, are not specified completely. Furthermore, even where procedures are
mandated, minor operator variations that are difficult to control can influence the test
results significantly. Various authors have discussed the method of drilling (e.g.,
Fletcher 1965, Ireland et al. 1970, de Mello 1971, Schmertmann 1978, Riggs 1986).
For example, Fletcher (1965) noted that holes larger than 100 mm diameter can lead
to stress relaxation and lower N-values, and de Mello (197 1) reported that the reduc-
tion can be as large as 50% in boreholes larger than 100 mm. Riggs (1986) found no
significant difference between N-values in side-by-side holes drilled with hollow-
stem augers and with open-hole methods and drilling mud. Seating of the sampler in
the bottom of the hole has been discussed by various authors (e.g., Fletcher 1965,
Ireland et al. 1970, de Mello 1971, Schmertmann 1971) and, although few quanti-
tative results are available, the consensus among these authors is that these effects
can be minimized by careful drilling and adhering to existing standards of borehole
preparation and cleaning. Proper training and inspection are critical for ensuring
usable data. Nearly every author who has studied the SPT has commented on the
method of hammer release, the loss of energy through friction, or the variations in
energy that result from variations in drop height. Automated trip mechanisms, often
BACK
IN-SITU TEST UNCERTAINTY 273
referred to as trip monkeys, reduce energy loss significantly. In a study by Serota
and Lowther (1973), it was found that a trip monkey or hand release with one turn on
the cathead resulted in N-values 20 to 30% lower than release with two turns on the
cathead. Kovacs et al. (1975, 1981, 1983) showed that donut hammers cause N-
values to be 30 to 35% higher than with a safety hammer.
Analysis of Uncertainly
Random variations in test results could be evaluated using the laboratory data by
Bieganousky and Marcuson (1976, 1977). Using four sands and a conventional drill
rig fitted out with a trip monkey release system, they varied the sand type, dry den-
sity, overconsolidation ratio, relative density, vertical stress, and depth. The results
indicated that dry density and relative density were the most significant variables,
while vertical stress and sand type had a minor influence. Overconsolidation ratio
and depth had almost no influence within the values used in the study. Analysis of
their data suggests that, under highly controlled conditions, the coefficient of varia-
tion (COV) for the N-values in sand was on the order of 15 to 20%.
Kovacs et al. (1981) examined the energy delivered by various combinations of
rope, cathead, drill rig, cathead speed and rotation direction, etc. These studies, con-
ducted during routine actual field operations, indicated an average COV for hammer
drop height from 1 to 8%, with an average of about 4%, and an average COV for the
energy ratios (energy delivered to theoretical energy) from 4 to 2 I%, with an average
of about 8%.
Seven additional literature studies indicated a COV on total field variability
from 14 to 81%. None of these studies contained any information on equipment and
procedural controls, and therefore it is difficult to evaluate the cause of such large
variations. However, these results tend to suggest that typical field measurements
could exhibit COV values for N on the order of 40 to 45%.
Overall, these data and studies illustrated that a number of equipment and proce-
dural/operator parameters influence the SPT results. The COV for a particular site
investigation will depend to a large degree on the type and uniformity of equipment
and procedures employed by the contractor performing the tests. Therefore, “best”
and “worst” case scenarios must be considered. The “best” case is represented by
use of identical equipment and procedures for each test, including standard split-
spoon samplers in good condition with liners, constant drill rod diameters, consistent
and careful drilling and borehole preparation, and use of an automatic trip monkey
hammer drop system. These test conditions minimize the variability associated with
the testing equipment, and one can reasonably estimate a COV less than or equal to
5% associated with minor variations in the test equipment. Similarly, careful and
consistent drilling and borehole preparation methods, coupled with an automatic
hammer drop, should result in negligible procedural variations, again on the order of
5%. Available results (Serota and Lowther 1973) and analysis of the Bieganousky
and Marcuson (1976, 1977) data indicate that random errors of about 12% can be
expected for automatic hammer drop systems. The total measurement variability,
COV(total), can be estimated from the square root of the sum of the squares of the
BACK
274 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
equipment, procedural/operator, and random test errors, as given below, because the
individual errors are not strictly additive (e.g., Benjamin and Cornell 1970):
COV(tota1) = [COV(equipment)2 + COV(procedure)’ + COV(random)2J”2 (1)
Therefore, the total measurement COV (neglecting any effect of soil variability) for
the “best” case scenario is on the order of 14%.
Alternatively, the “worst” case would be realized if there was no control on the
equipment and procedures employed in the tests. In this instance, equipment varia-
bility resulting from use of inconsistent sampler design and condition, size of drill
rods, and type of hammer could result in COVs as large as 75%. Also, variable and
careless drilling methods, borehole preparation techniques, and methods of dropping
the hammer could result in procedural variability as high as 75% or so. Adding an
additional 15% for random measurement gives a “worst” ease total measurement
COV that could be 100%. This unrealistic COV is intended only to illustrate the
magnitude of error that can result from careless and poorly supervised SPT practices.
Typical COVs to be expected for routine field operations are between these extremes,
although they generally should be nearer the “best” case estimate. Previous statis-
tical analyses of field test results support this conclusion, but it must be recognized
that these results include the influence of soil variability.
CONE PENETRATION TEST (CPT) UNCERTAINTIES
Basic Testing Issues
The CPT, long established in northern Europe as the most common method of
subsurface profiling, has become popular in North America during the past two to
three decades, especially in offshore applications and softer or looser terrestrial soil
deposits. Electronic measurement and recording systems, now relatively common,
have minimized many of the operator influences associated with older mechanical
push-rod systems. Some of the basic correlations between cone readings and soil
properties, however, are based on mechanical cone devices, and it is useful to under-
stand how the two systems interact with various soil parameters.
The cone penetration test is described by ASTM D3441, and the major factors
affecting CPT readings are listed in Table 2, including the cone type, size, angle, and
wear, as well as rod compression (MCPT only) or leaky seals (ECPT only). Of these,
the cone angle and rod compression can have a significant affect, while cone wear
has only minor to moderate influence (Durgunoglu & Mitchell 1975, Baligh et al.
1979), and the other factors are relatively minor in comparison. Tests and theory
(Marsland & @rterman 1982) have shown that the standard 60” base angle leads to
maximum penetration for a given force, while more acute angles lead to significantly
higher penetration resistance (up to 30% for an 18’ cone angle, for example). Tests
comparing MCPT and ECPT results indicate that the differences are somewhat
material-dependent. In one study (Joustra 1974), for example, ECPT results were
about 3% lower than MCPT results. However, another study (Kok 1974) showed
BACK
IN-SITU TEST UNCERTAINTY 275
TABLE 2. CPT Testing Variables
Grouv
CPT Variable Relative Effect
Item on Test Results
Equipment
Procedural/
Operator
Cone type (MCPT or ECPT)
Cone size
Cone angle
Rod compression (MCPT)
Manufacturing defects
Leaky seals (ECPT)
Excessive cone wear
Telescoping vs. continuous penetration
Calibration error
Penetration rate
Inclined penetration
moderate to significant
minor
moderate to significant
significant
minor to moderate
minor
minor to moderate
moderate to significant
minor to moderate
minor
moderate to significant
that MCPT results were as much as 30% lower than the corresponding ECPT results.
Operator and procedural effects can have a marked influence on MCPT results,
while having a relatively minor influence on ECPT results (DeRuiter 1971, Schaap &
Zuidberg 1982). The MCPT telescoping method of advancement does not completely
separate side resistance from tip resistance measurements, which can be particularly
troublesome for side resistance measurements. Another problem is that the tip and
side resistances for the MCPT are not made at the same elevation. For a thinly
stratified deposit, this difference in elevation can lead to major dislocations.
The ECPT is virtually operator-independent, once it is calibrated and checked
(Schaap & Zuidberg 1982). The only major variable is the rate of penetration, which
has been shown to increase tip resistance readings approximately 10% for an order-
of-magnitude increase in penetration rate in soft clays (Baligh et al. 1979). This
study and others (e.g., Kok 1974, Marsland & Quarterman 1982, Jamiolkowski et al.
1985) indicate that the rates suggested in ASTM D3441 are reasonable and unlikely
to affect tip or side resistance significantly. Deviation from verticality can be a
major factor with either the mechanical or electric systems, because readings might
not correspond to the depth based on the rod length.
Analysis of Uncertainty
Several quantitative studies have evaluated the effects of various soil parameters
on CPT readings under laboratory and field conditions. These studies, while not
specifically performed to determine test variability, provide some insights. In one
study (Parkin et al. 1980), tests in dry sand resulted in COVs of 5% for tip resistance
(qc) measurements and 10% for side resistance (f,) measurements. Interestingly, these
results are equal to those cited in ASTM D3441 for the expected precision of CPT
BACK
276 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
measurements. Another study (Villet 1981) indicated a COV of about 7% for qc.
Two studies (Nottingham 1975, Boghrat 1982) reported comparative field CPT
test results. The COV for qc and f, ranged from 30 to 60% in general, but because no
attempt was made to separate soil variability from the total test variability, it is not
possible to determine the COV resulting from random test errors. Finally, assuming
that in cohesive soils the measured qc values are proportional to the undrained shear
strength sU,several studies indicate that a field COV for q, on the order of 10 to 15%
can be expected in homogeneous cohesive deposits for electric cone tests with good
equipment and procedural control.
Standards exist for both the MCPT and ECPT, and they address the primary
equipment and procedural variables that influence the CPT results. However, diffe-
rences exist in these tests, so they will be discussed separately.
The mechanical system, which requires two rods, leads to uncertainty in separa-
ting tip and side resistance measurements. These uncertainties can significantly
influence the interpretation of the test results but, providing that consistent proce-
dures are employed, the effect on measurement variability should not be as great.
However, the procedure required to advance the cone with this two rod system is
highly operator-dependent. Based on limited available data, the procedural/operator
COV for the MCPT is estimated at 10% for q, and 15% for f,. Equipment factors,
including damaged cone, clogging of the area between the friction sleeve and the
cone tip, and manufacturing defects, lead to an estimate of equipment COV of 5% or
less. ASTM 03441 indicates that the precision or random errors associated with the
MCPT are 10% for qc and 15% for f,. Therefore, total measurement errors of about
15% for qc and 22% for f, are to be expected. These estimates assume that standard
equipment and procedures are employed at all times.
The ECPT is a single rod system that allows for continuous penetration and
automatic data acquisition, and it is virtually operator-independent. Therefore, the
procedural/operator COV is estimated at 5% or less. Similarly, the electrical trans-
ducers and measuring devices are precise, so the associated equipment COV is likely
to be on the order of 3%. Results from laboratory calibration studies (e.g., Parkin et
al. 1980, Villet 1981) indicate random errors of about 5% for qc and 10% for f,.
Therefore, the total measurement COV for the ECPT is about 8% for qc and 12% for
f,. These values are supported by several previous statistical analyses of ECPT data
that include some limited effect of soil variability.
The piezocone test variability for q, and f, should be approximately equal to that
for the ECPT. However, a standard does not yet exist that specifies several of the
additional equipment and procedural variables for this test.
VANE SHEAR TEST (VST) UNCERTAINTIES
Basic Testing Issues
Since its inception, the vane shear test has been considered to be an efficient and
economical method of measuring the undrained shear strength of cohesive soils. In
recent years, there have been relatively few studies devoted to the precision of this
BACK
IN-SITU TEST UNCERTAINTY 277
test, which is probably due, in part, to ASTM D2.573 that addresses the fundamental
variables in vane shear testing. Key variables with the VST are given in Table 3.
The influence of equipment variables on test results has been minimized through
research (e.g., Osterberg 1956, Arman et al. 1975, Donald et al. 1977, Walker 1983,
Schmertmann & Morgenstem 1977), and ASTM D2573 now prescribes acceptable
vane sizes, shapes, and thicknesses. The principal source of equipment variability is
the torque measurement device, which ideally should be a geared drive, although a
torque wrench is allowed under the standard.
Several sources of uncertainty can result from the procedure followed by the test
operator. Insertion disturbance can decrease apparent soil strength, although distur-
bance is minimal if the vane is pushed at least five borehole diameters below the base
of the hole (e.g., Flaate 1966). Additionally, it is important to remove the vane and
clean it between tests, rather than simply pushing it deeper for the next test (Flaate
1966). In the latter case, soil can adhere to the vane and cause additional disturbance
(decreasing apparent strength) or increased effective vane diameter (increasing appa-
rent strength), Rate of torque application during the VST is well-known to affect
strength measurements (e.g., Ladd et al. 1971, Walker 1983,), but the rate has been
standardized for many years and, as with other variables listed above, is addressed
best through proper training and equipment specification.
Analysis of Uncertainty
There are relatively few controlled studies that specifically address random test
variability. In one such study (McCormack & Wilding 1979), replicate tests were
made in various soil horizons. The average COV was on the order of lo%, indica-
ting the VST is a highly reproducible test. The remainder of the studies compiled
(Ladd et al. 1971, Arman et al. 1975, Baligh et al. 1979, Asaoka & A.-Grivas 1982)
indicate that total field COV can range from 3 to 44%, depending on the soil in
TABLE 3. VST Testing Variables
VST Variable Relative Effect
Group Item on Test Results
Equipment Vane length
Height/diameter ratio
Blade thickness
Torque measuring device
Damaged vane
Procedural/
Operator
Vane insertion method
Rod friction calibration
Time delay between insertion & testing
Vane rotation rate
minor
moderate
moderate
moderate to significant
moderate to significant
moderate to significant
moderate to significant
minor to moderatea
moderate
a - perhaps significant in soft clays
BACK
218 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
which the test is performed.
Because of the clear equipment specifications in the VST standard, the varia-
bility associated with the test apparatus should be minimal, providing that the torque
is delivered and measured by a gear drive and proving ring, and not by a torque
wrench. Assuming that appropriate standards are maintained, the equipment COV
resulting from the torque measuring device and any vane damage is estimated at 5%.
The vane test procedure is very sensitive to soil disturbance caused by vane insertion.
Other procedural variables that are not standardized and can influence the test results
include the time delay between insertion and testing and whether the vane is cleaned
between tests. These test parameters suggest a COV of 8% for procedural/operator
variability. Limited data further indicate that the random error of the VST results is
about 10%. Therefore, the total measurement COV is about 14%. This value is
supported by the results of previous investigations in homogeneous deposits, which
indicate slightly higher variabilities resulting from soil heterogeneity.
DILATOMETER TEST @MT) UNCERTAINTIES
Basic Testing Issues
The DMT is a relatively recent addition to the available set of field in-situ tools.
It is still somewhat under development, although there is a suggested test standard
(Schmertmann 1986). Key variables are given in Table 4. The equipment used for
the test consists of the blade, push rods, and pressure gage control. This equipment
is relatively rugged and of simple design, but there are several features that can affect
test results. The blade must be kept sharp, and the stainless steel membrane on the
blade face must be inspected often for damage and potential leakage (Schmertmann
1984). The rods used to push the blade are subject to bending in stiff soils, and bent
rods can results in erroneous penetration readings. The pressure control box is rela-
tively accurate and is not considered a significant error source.
Many of the procedural variables that can affect the CPT also can affect the
DMT, including variations from the vertical, the method of advancing the blade
(driving vs. pushing and rate of advance), and the point at which thrust is measured
(above the blade or at the surface) (Campanella et al. 1985).
Analysis of Uncertainty
There have been relatively few published studies on the variability of DMT test
results because of the relative newness of the test. However, several limited labora-
tory calibration studies and field data are available. In one study (Schmertmann
1984), 30 DMTs were performed in loose, normally consolidated dry sand. COVs
for the A and B readings were 2 and 0.8%, respectively. Another study (Lacasse &
Lunne 1982) reported qualitatively that DMTs in homogeneous Norwegian clay
exhibited “nearly perfect repeatability.” These data suggest that DMT measurements
are highly repeatable and that random test errors are not large.
In one series of unpublished tests (Lutenegger 1985),
students (mostly inexperienced in DMT testing) performed
six different graduate
the same tests in one
BACK
IN-SITU TEST UNCERTAINTY
TABLE 4. DMT Testing Variables
279
DMT Variable Relative Effect
Group Item on Test Results
Equipment Leaking seals minor
Deformed membrane moderate
Bent or deformed push rods minor to moderate
Damaged blade minor
Procedural/
Operator
Push rod inclination
Testing rate
Driving method
Rod friction
Calibration error
minor to moderate
moderate to significant
minor to moderate
minor
minor to moderate
sounding within the same soil profile. Analysis of variance indicated that at the 1%
confidence level, only one person had a significantly different set of readings than
the other five for the A reading, while there was no difference between all six for the
B reading. This study indicates that the test is highly reproducible, even by
inexperienced persons.
Another study (Baldi et al. 1986) reported laboratory calibration measurements
of the DMT in dry sands. The tests do not represent purely replicate data, and there-
fore the variability is influenced by effects other than random test errors. Given
these constraints, the results of four small samples indicate COVs of the corrected
dilatometer readings of 5.4 to 18.5%, with mean values on the order of 10%. These
numbers likely represent upper-bound estimates of DMT precision.
The DMT is not yet standardized, but the equipment is patented and must be
obtained from one supplier. Therefore, variations in the dilatometer blade can result
only from manufacturing defects. Also, there is a suggested test procedure standard,
so everyone conducting the test is using the same equipment and procedure. The test
also is simple to perform, so operator effects are not substantial.
In terms of equipment variability, only the pressure gage sensitivity, deformation
of the expanding membrane, and severe damage to the blade should influence the test
results. Therefore, a COV of 5% is suggested as an estimate of equipment effects.
Procedural/operator variables influencing the DMT include the method of penetra-
tion, deviations from vertical, and rate of membrane expansion. Unless one of these
procedural variables is changed dramatically during a test, they will influence the
interpretation more than the variability of the test results. Accordingly, the proce-
dural/operator COV is estimated at 5%. Limited data from DMT calibration testing
indicate random errors on the order of 8%. Therefore, a total measurement COV of
about 11% is estimated for the DMT.
BACK
280 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
PRESSUREMETER TEST (PMT) UNCERTAINTIES
Basic Testing Issues
The PMT is a relatively complex test that, although standardized in ASTM
D4719, is subject to variations in result interpretation that are not associated with the
other common in-situ tests. Key testing variables are given in Table 5. The PMT has
been the subject of extensive research over the past few decades and, while a com-
plete discussion of the test is beyond the scope of this paper, several pertinent points
should be mentioned.
The equipment to perform the PMT is complex and consists of the probe, tubing,
and pressure-volume control unit. Each component has potential sources of uncer-
tainty. For the probe, the length/diameter (L/D) ratio has been shown to affect the
limit pressure both theoretically and experimentally (Laier 1973, Briaud 1986). An
L/D ratio of 6.5 has been recommended for all pressuremeters (Laier 1973), while an
earlier suggested procedure specifies a minimum L/D of 4. Membranes must be
matched to soil conditions and should be calibrated often to accommodate change in
elasticity with time and usage. One-cell and three-cell models are acceptable in the
standard test procedure, and research has not documented a consistent pattern of
results obtained with the two types (Briaud 1986). Expansion of the tubing during
inflation can affect volumetric measurements and, at high flow rates near the limit
pressure, head losses can affect measurement of inflation pressure (Baguelin et al.
1978). The pressure-volume control unit has gages or readouts that must be cali-
brated but, in general, readings are precise to 1% or better (Baguelin et al. 1978).
TABLE 5. PMT Testing Variables
Group
PMT Variable Relative Effect
Item on Test Results
Equipment
Procedural/
Operator
Gage error
Expansion of tubing
Frictional losses in tubing
Probe dimensions
Probe design (PMT)
Membrane aging
Cutting shoe size (SBPMT)
Cutter position (SBPMT)
Probe shape (SBPMT)
Drilling equipment (SBPMT)
Electrical sensor compliance (SBPMT)
Drilling method & borehole preparation (PMT)
Probe inflation rate
Relaxation time (SBPMT)
Probe advance rate (SBPMT)
minor
minor to moderate
minor
minor to moderate
minor
minor
minor to moderate
minor
minor to moderate
minor to moderate
minor
significant
minor to moderate
moderate to significant
moderate to significant
BACK
IN-SITU TEST UNCERTAINTY 281
Because of the complexity of the PMT, the results are sensitive to the test proce-
dure, including borehole preparation and rate of probe expansion (e.g., Baguelin et
al. 1978, Ervin 1983, Benoit 1983, Benoit & Clough 1986, Briaud 1986,) and, for
self-boring pressuremeters (SBPMT), the relaxation period after reaching the desired
test depth and the rate of probe advance (e.g., Ghionna et al. 1981, Benoit 1983,
Jamiolkowski et al. 1985).
Analysis of Uncertainties
Quantitative analysis of random test variations is complicated by the fact that
test results are presented as pressure-volume curves that must be interpreted to yield
three pressures: seating (p,), yield (pr), and limit (p,). Therefore, the variability in
these pressures depends not only on the test data but also on the method of interpre-
tation. In one study (Briaud & Tucker 1984), the soil modulus and limit pressure
were determined in a series of tests in sand and sandy gravel deposits. The results
indicated COVs for the PMT modulus of 28 to 68% and for p, ranging from 22 to
50%. Data from another study (Laier 1973) indicated that the COV for the PMT
modulus ranged from 16 to 24% and for p,, of 2 to 15%. Additional studies with
SBPMTs in San Francisco Bay Mud (Denby 1978) indicated COVs of 14 and 38%
for the inferred soil parameters s, (undrained shear strength) and Gi (shear modulus),
respectively. As above, these parameters require interpretation and are useful prima-
rily to illustrate the order of magnitude of field variability.
Laboratory tests in calibration chambers have shown that self-boring pressure-
meters can be highly reproducible under certain conditions. Tests in which the
unload-reload shear modulus was measured in four consecutive loading loops
yielded COVs in a range of 5 to 18%, with an average about 9%. In this case, skilled
operators were performing the tests under relatively well-controlled conditions.
Therefore, these values should be considered to represent the lower bound of random
errors associated with the SBPMT.
Although the PMT is recently standardized, there are equipment and procedural
variabilities that lead to uncertainty in the test results. Quantitative evaluation of
these uncertainties is complicated by two factors. First, PMT and SBPMT results
typically are expressed in terms of inferred soil parameters or characteristic pressures
interpreted from the volume-pressure curve. The variability of these measures is not
necessarily representative of the variability of the test data, because they include
some level of uncertainty associated with their interpretation. Second, previous
quantitative studies of pressuremeter test variability are virtually non-existent, and
therefore information necessary to estimate the sources of variability accurately is
not available. However, based on review of sources of error in these tests, an esti-
mate can be made of the magnitude of these sources of test uncertainty.
For routine testing with the pre-bored pressuremeter (PMT), variables that can
influence the test results include the volume and pressure measuring devices, expan-
sion of the tubing, and the pressuremeter dimensions. Proper calibration can account
for most of the variability with the first two, and previous research has shown that
the range of dimensions of commercially-available probes do not affect the results
BACK
282 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
significantly. Therefore, an approximate value of PMT equipment COV is 5%.
The self-boring equipment possesses several characteristics that can lead to
additional test uncertainty, including the cutting shoe position and size, shape of the
probe, and compliance of the electronic feeler arms. Therefore, a slightly larger
value of 8% is estimated for the COV of equipment variability for the SBPMT.
Procedural/operator influences affecting the degree of disturbance of the bore-
hole are a major factor in the uncertain@ of PMT results. The method of drilling,
density of the drilling fluid, and ratio of the borehole diameter to the probe diameter
all affect the quality of PMT results. The drilling operation is difficult to control,
because it depends on the care exercised by the operator. Therefore, the variability
of these procedural parameters will not be constant. Based on the limited data
available, an estimate of typical procedural/operator COV is 12%. The SBPMT has
several additional procedural variables and is a considerably more complex test to
perform. Hence, a COV of 15% is estimated for this test.
As mentioned previously, there have been few previous quantitative studies of
PMT and SBPMT results, and very little replicate test data are available to estimate
random measurement error. However, the statistical analyses performed for this
study indicate random test errors on the order of 10% for the PMT and 8% for the
SBPMT. Therefore, the total measurement COV for the PMT is about 16% and for
the SBPMT is about 19%.
SUMMARY AND CONCLUSIONS
This paper has reviewed the principal sources of error for five of the most
common m-situ soil tests. The errors fall into three classes: equipment variability,
procedural and operator influences, and random test errors. Tests with well-defined
specifications, accepted practices, and standardized equipment have relatively small
test variability. The SPT, while standardized to some degree, allows for enough
equipment variability that significant variations in test results are common between
different operators using different drilling equipment. Other tests generally have
lesser variability.
Table 6 summarizes the estimates developed for the in-situ test errors. For each
test, the first three columns give the uncertainties as COV for equipment, procedure,
and random test errors. The next column combines these errors as the square root of
the sum of the squares, since the errors are not strictly additive. Finally, because of
the limited data available and the need to use judgment to estimate these errors, the
last column represents the range of probable test measurement variability one can
expect in typical field in-situ tests.
ACKNOWLEDGMENTS
Much of the research cited herein was conducted by C. J. Orchant for his thesis
at Cornell under the writers’ supervision. He no longer practices engineering. This
study was supported by the Electric Power Research Institute under RP1493. V. J.
BACK
IN-SITU TEST UNCERTAINTY
TABLE 6. Uncertainty Estimates for Five Common In-Situ Tests
283
Test
(acronym)
Coefficient of Variation,COV (%)
Equipment Procedure Random Total” Rangeb
Standard Penetration Test 5” - 19 5f - 15d
(SPT)
Mechanical Cone Penetration 5 IO’- 15’
Test (MCPT)
Electric Cone Penetration Test 3 5
(ECPT)
Vane Shear Test 5 8
(VW
Dilatometer Test 5 5
(DMT)
Pressuremeter Test, Pre-Bored 5 12
U’M’U
Self-Boring Pressuremeter 8 15
Test (SBPMT)
12- 15
IO’- 15’
5e - 10’
IO
8
IO
8
14’- 1OOd
15’-22’
15-45
15-25
8’ - 12’ 5- 15
14 IO-20
11 5- 15
16 IO - 20g
19 15 - 25g
- COV(tota1) = [COV(equipment)* + COV(procedure)’ + COV(random)2]‘n
: - Because of limited data and judgment involved in estimating COVs, ranges represent probable
magnitudes of field test measurement error
c,d - Best to worst case scenarios, respectively, for SPT
e,f - Tip and side resistances, respectively, for CPT
g - Results may differ for p., pr, and p,, but data are insufficient to clarify this issue
Longo was the EPRI project manager. A. Hirany is the current project manager. We
appreciate contributions made by M. D. Grigoriu, K.-K. Phoon, and T. D. O’Rourke.
REFERENCES
American Society for Testing & Materials (1994), “Standard Test Method for Pene-
tration Test & Split-Barrel Sampling of Soils (D 1586-84),” Annual Book of
St&. 04.08, ASTM, Philadelphia, 129-133.
American Society for Testing & Materials (1994), “Standard Test Method for Field
Vane Shear Test in Cohesive Soil (D 2573-72),” Annual Book of St&. 04.08,
ASTM, Philadelphia, 228-230.
American Society for Testing & Materials (1994), “Standard Test Method for Deep,
Quasi-Static, Cone & Friction-Cone Penetration Tests of Soil (D 3441-86),”
Annual Book of Stcis.04.08, ASTM, Philadelphia, 338-343.
American Society for Testing & Materials (1994), “Standard Test Method for Pres-
suremeter Testing in Soils (D 4719-87),” Annual Book of Stds. 04.08, ASTM,
Philadelphia, 858-865.
Arman, A, Poplin, JK & Ahmad, N (1975), “Study of Vane Shear”, Proc., ASCE
BACK
284 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
Spec. Conf. In-Situ Measurement of Soil Properties (1), Raleigh, 93-120.
Asaoka, A & A-Grivas, D (1982), “Spatial Variability of Undrained Strength of
Clays”, J. Geofech. Eng. Div., ASCE, 108(GT5), 743-757.
Baguelin, F, Jezequel, JF & Shields, DH (1978), Pressuremeter & Fndn. Eng., Tram
Tech Pubs., Clausthal, 617 p.
Baldi, G, Bellotti, R, Ghionna, V, Jamiolkowski, M, Marchetti, S & Pasqualini, E
(1986) “Flat Dilatometer Tests in Calibration Chambers”, Use ofln-Situ Tests in
Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 43 l-446.
Baligh, MM, Vivatrat, V & Ladd, CC (1979), “Exploration & Evaluation of Eng.
Properties for Fndn. Design of Offshore Structures”, Rpt. MITSG 79-8, MIT,
Cambridge, 268 p.
Benjamin, JR & Cornell, CA (1970), Probability, Statistics & Decision for Civil
Engineers, McGraw-Hill, New York, 684 p.
Benoit, J (1983), “Analysis of Self-Boring Pressuremeter Tests in Soft Clay”, PhD
Thesis, Stanford Univ., Palo Alto, 361 p.
Benoit, J & Clough, GW (1986), “Self-Boring Pressuremeter Tests in Soft Clay”, J.
Geotech. Eng., ASCE, 112(l), 60-78.
Bieganousky, WA & Marcuson, WF, III (1976), “Liquefaction Potential of Dams &
Fndns.: Rpt. 1, Lab. Std Penetration Tests on Reid Bedford & Ottawa Sands”,
Rex Rpt. S-76-2, US Army Eng. Waterways Exper. Sta., Vicksburg, 96 p.
Bieganousky, WA & Marcuson, WF, III (1977), “Liquefaction Potential of Dams &
Fndns.: Rpt. 2, Lab. Std. Penetration Tests on Platte River & Concrete Sands”,
Res. Rpt. S-76-2, US Army Eng. Waterways Exper. Sta., Vicksburg, 47 p.
Boghrat, A (1982), “Design & Construction of Piezoblade & Evaluation of Marchetti
Dilatometer”, PhD Thesis, Univ. of Florida, Gainesville, 244 p.
Briaud, J-L (1986), “Pressuremeter & Foundation Design”, Use of In-Situ Tests in
Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 74-l 15.
Briaud, J-L & Tucker, L (1984), “Coefficient of Variation of In-Situ Tests in Sand”,
Probabilistic Characterization of Soil Properties: Bridge Between Theory &
Practice, Ed. DS Bowles & H-Y Ko, ASCE, New York, 119-139.
Campanella, RG, Robertson, PK, Gillespie, DG & Grieg, J (1985), “Recent Develop-
ments in In-Situ Testing of Soils”, Proc., 1lth Intl. Conf. Soil Mech. & Fndn.
Eng. (2), San Francisco, 849-854.
de Mello, VFB (1971), “Standard Penetration Test”, Proc., 4th Panam. Conf. Soil
Mech. & Fndn. Eng. (I), San Juan, l-86.
Denby, GM (1978), “Self-Boring Pressuremeter Study of San Francisco Bay Mud”,
PhD Thesis, Stanford Univ., Palo Alto, 270 p.
DeRuiter, J (1971), “Electric Penetrometer for Site Investigation”, J. Soil Mech. &
Fndns. Div., ASCE, 92(SM2), 457-472.
Donald, IB, Jordan, DO, Parker, RJ & Toh, CT (1977), “Vane Test - A Critical
Appraisal”, Proc., 9th Intl. Conf. Soil Mech. & Fndn. Eng. (1), Tokyo, 81-88.
Durgunoglu, HT & Mitchell, JK (1975), “Static Penetration Resistance of Soils: II -
Evaluation of Theory & Implications for Practice”, Proc., ASCE Spec. Conf. In-
Situ Measurement of Soil Properties (1), Raleigh, 172-189.
BACK
IN-SITU TEST UNCERTAINTY 285
Ervin, MC (1983), “Pressuremeter in Geotechnical Investigations”, In-Situ Testing
for Geotech. Investigations, Ed. MC Ervin, Balkema, Rotterdam, 49-64.
Flaate, K (1966), “Factors Influencing Results of Vane Tests”, Can. Geotech. .I.,
3(l), 18-31.
Fletcher, GFA (1965), “Standard Penetration Test, Its Uses and Abuses”, J. Soil
Mech. & Fndns. Div., ASCE, 91(SM4), 67-75.
Ghionna, V, Jamiolkowski, M, Lancellotta, R, Tordella, ML & Ladd, CC (1981)
“Performance of Self-Boring Pressuremeter Tests in Cohesive Deposits”, Rpt.
RD-H/173, Fed. Hwy. Admin., Washington, 182 p.
Gibbs, HJ & Holtz, WG (1957), “Research on Determining the Density of Sands by
Spoon Penetration Testing”, Proc., 4th Intl. Conf. Soil Mech. & Fndn. Eng. (I),
London, 35-39.
Ireland, HO, Moretto, 0 & Vargas, M (1970), “Dynamic Penetration Test: A Stan-
dard That Is Not Standardized”, Geotechnique, 20(2), 185-192.
Jamiolkowski, M, Ladd, CC, Germaine, JT & Lancellotta, R (1985), “New Develop-
ments in Field & Lab. Testing of Soils”, Proc., 1lth Intl. Conf. Soil Mech. &
Fndn. Eng. (l), San Francisco, 57-l 53.
Joustra, K (1974), “Comparative Measurements of Influence of Cone Shape on Re-
sults of Soundings”, Proc., 1st Eur. Symp. Penetration Testing (2.2), Stockholm,
199-204.
Kok, L (1974), “Effect of Penetration Speed & Cone Shape on Dutch Static Cone
Penetration Test Results”, Proc., 1st Eur. Symp. Penetration Testing (2.2),
Stockholm, 215-220.
Kovacs, WD, Evans, JC & Griffith, AH (1975), “Comparative Investigation of MO-
bile Drilling Company’s Safety-T-Driver w. Standard Cathead w. Manila Rope
for Performance of Standard Penetration Test”, Rpt, School of Civil Eng.,
Purdue Univ., Lafayette, 129 p. (As referenced in Kovacs et al. 1981)
Kovacs, WD, Salomone, LA & Yokel, FY (1981), “Energy Measurements in Stan-
dard Penetration Test”, Natl. Bldg. Science Series 135, Natl. Bureau Stds.,
Washington, 73 p.
Kovacs, WD, Salomone, LA & Yokel, FY (1983), “Comparison of Energy Measure-
ments in Standard Penetration Test Using Cathead & Rope Method”, NUREG/
CR-3545, Nuclear Reg. Comm., Washington, 69 p.
Lacasse, S & Lunne, T (1982), “Penetration Tests in Two Norwegian Clays”, Proc.,
2nd Eur. Symp. Penetration Testing (2), Amsterdam, 661-670.
Ladd, CC, Moh, Z-C & Gifford, DG (1971), “Statistical Analysis of Undrained
Strength of Bangkok Clay”, Proc., 1st Intl. Conf. Applications of Statistics &
Probability in Soil & Strut. Eng., Hong Kong, 3 14-322.
Laier, J (1973), “Effect of Pressuremeter Probe Length/Diameter Ratio & Borehole
Disturbance on Pressuremeter Test Results in Dry Sands”, PhD Thesis, Univ. of
Florida, Gainesville, 204 p.
Lutenegger, AJ (1985), “Personal Communication”.
Marsland, A & Quarterman, RST (1982), “Factors Affecting Measurements of Quasi-
Static Penetration Tests in Clays”, Proc., 2nd Eur. Symp. Penetration Testing
BACK
286 UNCERTAINTY IN GEOLOGIC ENVIRONMENT
(2), Amsterdam, 697-702.
Matsumota, M & Matsubara, K (1982), “Effect of Rod Diameter in Standard Penetra-
tion Test”, Proc., 2nd Eur. Symp. Penetration Testing (I), Amsterdam, 107-l 12.
McCormack, DE & Wilding, LP (1979), “Soil Properties Influencing Strength of
Cantield & Geeburg Soils”, J. Soil Science Sot. Ame&a, 43( 1), 167-I 73.
McLean, FG, Franklin, AG & Dahlstrand, TK (1975) “Influence of Mechanical Var-
iables on SPT”, Proc., ASCE Spec. Conf. In-Situ Measurement of Soil Proper-
ties (2), Raleigh, 287-3 18.
Nottingham, LC (1975) “Use of Quasi-Static Penetrometer Data to Predict Load Ca-
pacity of Piles”, PhD Thesis, Univ. of Florida, Gainesville, 553 p.
Orchant, CJ, Kulhawy, FH & Trautmann, CH (1988), “Reliability-Based Foundation
Design for Transmission Line Structures: Critical Evaluation of In-Situ Test
Methods,” Rpt. EL-5507(2), Electric Power Res. Inst., Palo Alto, 207 p.
Osterberg, JO (1956), “Introduction”, Symp. Vane Shear Testing of Soils (STP 193),
ASTM, Philadelphia, l-8.
Parkin, A, Holden, J, Aamot, K, Last, N & Lunne, T (1980) “Lab. Investigation of
CPTs in Sand”, Rpt 52108-9, Norwegian Geotech. Inst., Oslo, 3 1 p.
Phoon, KK & Kulhawy, FH (1996), “On Quantifying Inherent Soil Variability”, this
conference proceedings.
Riggs, CO (1986), “North American Standard Penetration Testing”, Use of In-Situ
Tests in Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 949-967.
Schaap, LHJ & Zuidberg, HM (1982) “Mechanical & Electrical Aspects of Electric
Cone Penetrometer Tip”, Proc., 2nd Eur. Symp. Penetration Testing (2)
Amsterdam, 84 1-85 1.
Schmertmann, JH (1971), Disc. on “Standard Penetration Test”, Proc., 4th Panam
Conf. Soil Mech. & Fndn. Eng. (3), San Juan, 90-98.
Schmertmann, JH (1975), “Measurement of In-Situ Shear Strength”, Proc., ASCE
Spec. Conf. In-Situ Measurement of Soil Properties (2), Raleigh, 57-138.
Schmertmamr, JH (1978), “Use of SPT to Measure Dynamic Soil Properties?... Yes
But!...“, Dynamic Geotech. Testing (STP 654), ASTM, Philadelphia, 341-355.
Schmertmann, JH (1984), Ed., DMT Digest 4, GPE Inc, Gainesville, 19 p.
Schmertmamr, JH (1986), “Suggested Method for Performing the Flat Dilatometer
Test,” Geotech.’ Test. J:, ASTM, 9(2), 93-101.
Schmertmann, JH & Morgenstern, NR (1977), Disc. to Session 1, Proc., 9th Intl.
Conf. Soil Mech. & Fndn Eng. (3), Tokyo, 356-359.
Seed, HB, Tokimatsu, K, Harder, LF & Chung, RM (1984), “Influence of SPT Proce-
dures in Soil Liquefaction Resistance Evaluations”, Rpt. UCB/EERC-84/15,
Univ. of California, Berkeley, 50 p.
Serota, S & Lowther, G (1973), Disc. on “Accuracy of Relative Density Measure-
ments”, Geotechnique, 23(2), 301-303.
Villet, WCB (1981), “Acoustic Emissions During Static Penetration of Soils”, PhD
Thesis, Univ. of California, Berkeley, 400 p.
Walker, BF (1983), “Vane Shear Strength Testing”, In-Situ Testing for Geotech.
Investigations, Ed. MC Ervin, Balkema, Rotterdam, 65-72.
BACK

More Related Content

What's hot

Q922+de2+l03 v1
Q922+de2+l03 v1Q922+de2+l03 v1
Q922+de2+l03 v1
AFATous
 
Bearing capacity estimation rocks for foundation
Bearing capacity estimation rocks for foundationBearing capacity estimation rocks for foundation
Bearing capacity estimation rocks for foundation
FajruSied
 
Drilling waste management_technology_1_
Drilling waste management_technology_1_Drilling waste management_technology_1_
Drilling waste management_technology_1_
kelvins_luv
 

What's hot (20)

Lecture 5 shear strength of soils
Lecture 5  shear strength of soilsLecture 5  shear strength of soils
Lecture 5 shear strength of soils
 
Geotechnical Engineering-I [Lec #18: Consolidation-II]
Geotechnical Engineering-I [Lec #18: Consolidation-II]Geotechnical Engineering-I [Lec #18: Consolidation-II]
Geotechnical Engineering-I [Lec #18: Consolidation-II]
 
Cementing
CementingCementing
Cementing
 
Q922+de2+l03 v1
Q922+de2+l03 v1Q922+de2+l03 v1
Q922+de2+l03 v1
 
Class 8 Triaxial Test ( Geotechnical Engineering )
Class 8    Triaxial Test ( Geotechnical Engineering )Class 8    Triaxial Test ( Geotechnical Engineering )
Class 8 Triaxial Test ( Geotechnical Engineering )
 
Well design and construction i
Well design and construction iWell design and construction i
Well design and construction i
 
SHEAR STRENGTH THEORY
SHEAR STRENGTH THEORYSHEAR STRENGTH THEORY
SHEAR STRENGTH THEORY
 
Drillstring & BHA Design
Drillstring & BHA DesignDrillstring & BHA Design
Drillstring & BHA Design
 
Drilling Engineering - Casing Design
Drilling Engineering - Casing DesignDrilling Engineering - Casing Design
Drilling Engineering - Casing Design
 
Stuck pipe manual
Stuck pipe manualStuck pipe manual
Stuck pipe manual
 
Well control during drilling operations
Well control during drilling operationsWell control during drilling operations
Well control during drilling operations
 
Bearing capacity estimation rocks for foundation
Bearing capacity estimation rocks for foundationBearing capacity estimation rocks for foundation
Bearing capacity estimation rocks for foundation
 
Static cone penetration test-basics
Static cone penetration test-basicsStatic cone penetration test-basics
Static cone penetration test-basics
 
Slope stability analysis methods
Slope stability analysis methodsSlope stability analysis methods
Slope stability analysis methods
 
Vane shear test
Vane shear testVane shear test
Vane shear test
 
Stuck pipe
Stuck pipeStuck pipe
Stuck pipe
 
Drilling waste management_technology_1_
Drilling waste management_technology_1_Drilling waste management_technology_1_
Drilling waste management_technology_1_
 
Drill stem test
Drill stem testDrill stem test
Drill stem test
 
MOHR COULOMB FAILURE CRITERION
MOHR COULOMB FAILURE CRITERIONMOHR COULOMB FAILURE CRITERION
MOHR COULOMB FAILURE CRITERION
 
Basic Well Control
Basic Well ControlBasic Well Control
Basic Well Control
 

Similar to In situ tests uncertainity

Paper id 23201426
Paper id 23201426Paper id 23201426
Paper id 23201426
IJRAT
 
Testing of Adhesives
Testing of AdhesivesTesting of Adhesives
Testing of Adhesives
Dung Lee
 
SPE 171517_Estimating Probability of Failure _2014_Final
SPE 171517_Estimating Probability of Failure _2014_FinalSPE 171517_Estimating Probability of Failure _2014_Final
SPE 171517_Estimating Probability of Failure _2014_Final
Katrina Carter-Journet
 
214 77 recommended practice for evaluation of strength tes
214 77   recommended practice for evaluation of strength tes214 77   recommended practice for evaluation of strength tes
214 77 recommended practice for evaluation of strength tes
MOHAMMED SABBAR
 
E8E8M.3597 Tension Testing of Metallic Materials.pdf
E8E8M.3597 Tension Testing of Metallic Materials.pdfE8E8M.3597 Tension Testing of Metallic Materials.pdf
E8E8M.3597 Tension Testing of Metallic Materials.pdf
mahmoodkhan77
 

Similar to In situ tests uncertainity (20)

ch01.pdf
ch01.pdfch01.pdf
ch01.pdf
 
C p test
C p testC p test
C p test
 
Sample rpts
Sample rptsSample rpts
Sample rpts
 
Thesis
ThesisThesis
Thesis
 
Paper id 23201426
Paper id 23201426Paper id 23201426
Paper id 23201426
 
Testing of Adhesives
Testing of AdhesivesTesting of Adhesives
Testing of Adhesives
 
Sopaqu
SopaquSopaqu
Sopaqu
 
SPE 171517_Estimating Probability of Failure _2014_Final
SPE 171517_Estimating Probability of Failure _2014_FinalSPE 171517_Estimating Probability of Failure _2014_Final
SPE 171517_Estimating Probability of Failure _2014_Final
 
Analytical chemistry handbook
Analytical chemistry handbookAnalytical chemistry handbook
Analytical chemistry handbook
 
R OUND-ROBIN VERIFICATION AND FINAL DEVELOPMENT OF THE IEC 62788-1-5 ENCAPSUL...
R OUND-ROBIN VERIFICATION AND FINAL DEVELOPMENT OF THE IEC 62788-1-5 ENCAPSUL...R OUND-ROBIN VERIFICATION AND FINAL DEVELOPMENT OF THE IEC 62788-1-5 ENCAPSUL...
R OUND-ROBIN VERIFICATION AND FINAL DEVELOPMENT OF THE IEC 62788-1-5 ENCAPSUL...
 
Preventative control measures of rock/coal bursts in underground mines. 5706671
Preventative control measures of rock/coal bursts in underground mines. 5706671Preventative control measures of rock/coal bursts in underground mines. 5706671
Preventative control measures of rock/coal bursts in underground mines. 5706671
 
Characterization of Differential Concrete Mix Designs by Ultrasonic Pulse Vel...
Characterization of Differential Concrete Mix Designs by Ultrasonic Pulse Vel...Characterization of Differential Concrete Mix Designs by Ultrasonic Pulse Vel...
Characterization of Differential Concrete Mix Designs by Ultrasonic Pulse Vel...
 
Failure Analysis of Polymer and Rubber Components
Failure Analysis of Polymer and Rubber ComponentsFailure Analysis of Polymer and Rubber Components
Failure Analysis of Polymer and Rubber Components
 
Structural health monitoring 2011-wei fan-83-111
Structural health monitoring 2011-wei fan-83-111Structural health monitoring 2011-wei fan-83-111
Structural health monitoring 2011-wei fan-83-111
 
214 77 recommended practice for evaluation of strength tes
214 77   recommended practice for evaluation of strength tes214 77   recommended practice for evaluation of strength tes
214 77 recommended practice for evaluation of strength tes
 
CPTu quality
CPTu qualityCPTu quality
CPTu quality
 
E8E8M.3597 Tension Testing of Metallic Materials.pdf
E8E8M.3597 Tension Testing of Metallic Materials.pdfE8E8M.3597 Tension Testing of Metallic Materials.pdf
E8E8M.3597 Tension Testing of Metallic Materials.pdf
 
Laboratory experimental study and elastic wave velocity on physical propertie...
Laboratory experimental study and elastic wave velocity on physical propertie...Laboratory experimental study and elastic wave velocity on physical propertie...
Laboratory experimental study and elastic wave velocity on physical propertie...
 
Jorc code table_1_report_template
Jorc code table_1_report_templateJorc code table_1_report_template
Jorc code table_1_report_template
 
Parameter Optimization of Shot Peening Process of PMG AL2024 Alloy Cover
Parameter Optimization of Shot Peening Process of PMG AL2024 Alloy CoverParameter Optimization of Shot Peening Process of PMG AL2024 Alloy Cover
Parameter Optimization of Shot Peening Process of PMG AL2024 Alloy Cover
 

Recently uploaded

scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
HenryBriggs2
 
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
Integrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - NeometrixIntegrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - Neometrix
Neometrix_Engineering_Pvt_Ltd
 

Recently uploaded (20)

COST-EFFETIVE and Energy Efficient BUILDINGS ptx
COST-EFFETIVE  and Energy Efficient BUILDINGS ptxCOST-EFFETIVE  and Energy Efficient BUILDINGS ptx
COST-EFFETIVE and Energy Efficient BUILDINGS ptx
 
Introduction to Serverless with AWS Lambda
Introduction to Serverless with AWS LambdaIntroduction to Serverless with AWS Lambda
Introduction to Serverless with AWS Lambda
 
A Study of Urban Area Plan for Pabna Municipality
A Study of Urban Area Plan for Pabna MunicipalityA Study of Urban Area Plan for Pabna Municipality
A Study of Urban Area Plan for Pabna Municipality
 
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
 
School management system project Report.pdf
School management system project Report.pdfSchool management system project Report.pdf
School management system project Report.pdf
 
HAND TOOLS USED AT ELECTRONICS WORK PRESENTED BY KOUSTAV SARKAR
HAND TOOLS USED AT ELECTRONICS WORK PRESENTED BY KOUSTAV SARKARHAND TOOLS USED AT ELECTRONICS WORK PRESENTED BY KOUSTAV SARKAR
HAND TOOLS USED AT ELECTRONICS WORK PRESENTED BY KOUSTAV SARKAR
 
Block diagram reduction techniques in control systems.ppt
Block diagram reduction techniques in control systems.pptBlock diagram reduction techniques in control systems.ppt
Block diagram reduction techniques in control systems.ppt
 
Rums floating Omkareshwar FSPV IM_16112021.pdf
Rums floating Omkareshwar FSPV IM_16112021.pdfRums floating Omkareshwar FSPV IM_16112021.pdf
Rums floating Omkareshwar FSPV IM_16112021.pdf
 
Work-Permit-Receiver-in-Saudi-Aramco.pptx
Work-Permit-Receiver-in-Saudi-Aramco.pptxWork-Permit-Receiver-in-Saudi-Aramco.pptx
Work-Permit-Receiver-in-Saudi-Aramco.pptx
 
FEA Based Level 3 Assessment of Deformed Tanks with Fluid Induced Loads
FEA Based Level 3 Assessment of Deformed Tanks with Fluid Induced LoadsFEA Based Level 3 Assessment of Deformed Tanks with Fluid Induced Loads
FEA Based Level 3 Assessment of Deformed Tanks with Fluid Induced Loads
 
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
 
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
Call Girls in South Ex (delhi) call me [🔝9953056974🔝] escort service 24X7
 
Computer Networks Basics of Network Devices
Computer Networks  Basics of Network DevicesComputer Networks  Basics of Network Devices
Computer Networks Basics of Network Devices
 
data_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdfdata_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdf
 
Minimum and Maximum Modes of microprocessor 8086
Minimum and Maximum Modes of microprocessor 8086Minimum and Maximum Modes of microprocessor 8086
Minimum and Maximum Modes of microprocessor 8086
 
Integrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - NeometrixIntegrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - Neometrix
 
Online food ordering system project report.pdf
Online food ordering system project report.pdfOnline food ordering system project report.pdf
Online food ordering system project report.pdf
 
Employee leave management system project.
Employee leave management system project.Employee leave management system project.
Employee leave management system project.
 
Thermal Engineering -unit - III & IV.ppt
Thermal Engineering -unit - III & IV.pptThermal Engineering -unit - III & IV.ppt
Thermal Engineering -unit - III & IV.ppt
 
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best ServiceTamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
 

In situ tests uncertainity

  • 1. ESTIMATION OF IN-SITU TEST UNCERTAINTY Fred H. Kulhawy’ , F. ASCE and Charles H. Trautmann* , M. ASCE ABSTRACT: In-situ tests have several primary sources of uncertainty, including calibration errors, operator influence, sampling errors, and variations in test proce- dure. This paper focuses on evaluating these sources and obtaining an overall esti- mate of test uncertainty, as distinguished from soil variability, for the following test methods: (a) standard penetration test, (b) cone penetration test, (c), dilatometer test, (d) vane shear test, and (e) pressuremeter test. Previous studies are reviewed that cover a wide range of goals, including: (a) qualitative discussions that emphasize the various sources of error, (b) soil property correlations, which often make separation of the sources of error difficult, and (c) studies specifically geared to evaluation of test uncertainties. Laboratory calibration studies commonly are most useful in this regard. Using data from prior investigations, this paper first places error sources in order of importance. Next, estimates for the most important errors for each test are obtained. Finally, the results are analyzed and summarized into a table of combined uncertainties. This table represents the best-estimate uncertainty for each test proce- dure, independent of soil variability or the correlation of test result to engineering soil property. Basically, these results show that tests with more potential for equip- ment and operator/procedural variations are likely to exhibit greater test uncertainty. INTRODUCTION Site characterization is performed to evaluate the subsurface conditions needed for safe and economical geotechnical design. The information obtained during site characterization includes the determination of the subsurface deposits and their geo- metries and engineering properties. The level of detail required for this information depends on the nature and economics of the specific project and can range from general reconnaissance, which uses existing information, to detailed field sampling 1 - Prof., School of Civil & Environ. Eng., Cornell Univ., Hollister Hall, Ithaca, NY 14853-3501 2 - Exec. Dir., The Sciencenter, First Street, Ithaca, NY 14850-3507; Adjunct Assoc. Prof., School of Civil & Environ. Eng., Cornell Univ. BACK
  • 2. 210 UNCERTAINTY IN GEOLOGIC ENVIRONMENT and testing. Most site characterization studies involve in-situ drilling, sampling, and testing in one or more forms to provide geotechnical design information. Evaluation of the uncertainties in this information commonly is done using judgment that is based on experience. Two prominent sources of in-situ test variability that can be identified include: (a) natural spatial variability of the soil deposits resulting from their method of formation and the subsequent environmental changes that have taken place, and (b) uncertainty in the test measurements resulting from equipment, test procedure, operator, and random test effects. Quantification of spatial variability is discussed elsewhere in these proceedings (Phoon & Kulhawy 1996), while test uncertainty is discussed herein. Wherever possible, it is preferable to use quantitative measures of uncertainty. However, because it is often difficult to separate the various sources of uncertainty, this approach has been used rarely. This paper evaluates pertinent quantitative infor- mation on this issue and develops estimates of overall in-situ test uncertainty. Most of the research reported herein was completed several years ago. The results of this extensive study previously have been available only through a limited- access report of the Electric Power Research Institute and were not in the general literature. This paper presents the key results of this study in condensed form, along with several important updates in ASTM standards. The detailed results and analy- ses are provided in the original report (Orchant et al. 1988), while only illustrative examples and best-estimate conclusions can be presented herein because of length restrictions. Subsequent detailed studies that are not described and referenced herein have only reinforced the key results of this study. We all await the detailed studies in progress at the U. S. national test sites to delineate these issues in a (hopefully) broad and definitive manner. BACKGROUND Many authors have discussed the various sources of errors and uncertainty with in-situ test methods. Generally, these studies have focused on identification of sources and whether they increase or decrease a particular test result. For example, one of the developers of the split-spoon and hence the standard penetration test (SPT) was among the first to describe the abuses of this test (Fletcher 1965). He was able to identify 13 “important” factors that affect SPT results, including inadequate bore- hole cleaning, failure to maintain sufficient hydrostatic head in the borehole, varia- tions in the 760~mm (30-in) hammer drop height, extremely long drill rods, etc. While some of these factors can bias the test results in either direction (e.g., varia- tions in the hammer drop height), other factors lead to a bias in only one direction (e.g., interference of the free fall of the hammer reduces the energy per blow and therefore increases the blow count). This increase is unconservative, because it makes the soil appear firmer than it actually is. Most influences on test uncertainty can be divided into two groups: equipment effects and procedural/operator effects. Both of these types of effects have received BACK
  • 3. IN-SITU TEST UNCERTAINTY 271 considerable attention in understanding the influence of testing parameters on the in- situ test results, although the majority of early studies have been qualitative or semi- quantitative. Increasingly, authors have focused on quantitative evaluation of in-situ test accuracy and precision. Commonly, the most useful studies involve replicate tests under highly controlled conditions, such as in the laboratory or at field sites where soil conditions are consistent over wide horizontal and vertical distances. In the remainder of this paper, these quantitative studies will be highlighted, and the results will be combined to develop best-estimate uncertainties for the common in- situ test methods addressed herein. ANALYSIS OF UNCERTAINTIES In this paper, the five most common in-situ tests are analyzed with respect to quantitative uncertainties. The five tests include the standard penetration test (SPT), cone penetration test (CPT), vane shear test (VST), dilatometer test (DMT), and pressuremeter test (PMT). For the CPT, both the mechanical (MCPT) and electric (ECPT) cones are included and, for the pressuremeter, both the pre-bored (PMT) and self-boring (SBPMT) versions are addressed. In the following analyses, many effects related to non-standard equipment and test procedures are identified, but they are not analyzed rigorously, including such factors as non-standard samplers or failure to clean the hole before testing. It is assumed that these error sources are best addressed through proper training and inspection. On the other hand, random errors resulting from best-effort performance of the test personnel or equipment variations that are not identified in the standardized test procedure are addressed quantitatively where possible, including such factors as variations in the SPT hammer drop height when using a rope-and-cathead system. STANDARD PENETRATION TEST (SPT) UNCERTAINTIES Basic Testing Issues The standard penetration test is described in ASTM D1586, and the major fac- tors influencing the SPT results are listed in Table 1. The equipment employed in the SPT includes the hammer, hammer drop system, drill rods, and sampler. Each of these equipment variables can have a major influence on the resulting N-values from the test. The rod weight and length have been analyzed extensively by field and wave-equation methods and have been shown to have a relatively minor influence (e.g., Gibbs & Holtz 1957, McLean et al. 1975, Matsumota & Matsubara 1982). Safety and donut-type hammers deliver different energies to the SPT sampler, with the safety hammer delivering more energy (Seed et al. 1984). The anvil, which receives the hammer impact, can have a significant influence on the results. For example, Schmertmann (1978) indicated that tests performed with a small anvil can result in N-values up to 50% greater than those obtained with larger anvils. Many geotechnical engineers have cited the high degree of operator dependency of the SPT as its most important limitation. Although ASTM D1586 specifies some BACK
  • 4. 272 UNCERTAINTY IN GEOLOGIC ENVIRONMENT TABLE 1. SPT Testing Variables Group Equipment Procedural/ Operator SPT Variable Relative Effect Item on Test Results Non-standard sampler moderate Deformed or damaged sampler moderate Rod diameter/weight minor Rod length minor Deformed drill rods minor Hammer type moderate to significant Hammer drop system significant Hammer weight minor Anvil size moderate to significant Drill rig type minor Borehole size moderate Method of maintaining hole minor to significant Borehole cleaning moderate to significant Insufficient hydrostatic head moderate to significant Seating of sampler moderate to significant Hammer drop method moderate to significant Error in counting blows minor procedural requirements for this test, such as the hammer drop height and seating of the sampler, other variables, including the method of borehole drilling and method of hammer drop, are not specified completely. Furthermore, even where procedures are mandated, minor operator variations that are difficult to control can influence the test results significantly. Various authors have discussed the method of drilling (e.g., Fletcher 1965, Ireland et al. 1970, de Mello 1971, Schmertmann 1978, Riggs 1986). For example, Fletcher (1965) noted that holes larger than 100 mm diameter can lead to stress relaxation and lower N-values, and de Mello (197 1) reported that the reduc- tion can be as large as 50% in boreholes larger than 100 mm. Riggs (1986) found no significant difference between N-values in side-by-side holes drilled with hollow- stem augers and with open-hole methods and drilling mud. Seating of the sampler in the bottom of the hole has been discussed by various authors (e.g., Fletcher 1965, Ireland et al. 1970, de Mello 1971, Schmertmann 1971) and, although few quanti- tative results are available, the consensus among these authors is that these effects can be minimized by careful drilling and adhering to existing standards of borehole preparation and cleaning. Proper training and inspection are critical for ensuring usable data. Nearly every author who has studied the SPT has commented on the method of hammer release, the loss of energy through friction, or the variations in energy that result from variations in drop height. Automated trip mechanisms, often BACK
  • 5. IN-SITU TEST UNCERTAINTY 273 referred to as trip monkeys, reduce energy loss significantly. In a study by Serota and Lowther (1973), it was found that a trip monkey or hand release with one turn on the cathead resulted in N-values 20 to 30% lower than release with two turns on the cathead. Kovacs et al. (1975, 1981, 1983) showed that donut hammers cause N- values to be 30 to 35% higher than with a safety hammer. Analysis of Uncertainly Random variations in test results could be evaluated using the laboratory data by Bieganousky and Marcuson (1976, 1977). Using four sands and a conventional drill rig fitted out with a trip monkey release system, they varied the sand type, dry den- sity, overconsolidation ratio, relative density, vertical stress, and depth. The results indicated that dry density and relative density were the most significant variables, while vertical stress and sand type had a minor influence. Overconsolidation ratio and depth had almost no influence within the values used in the study. Analysis of their data suggests that, under highly controlled conditions, the coefficient of varia- tion (COV) for the N-values in sand was on the order of 15 to 20%. Kovacs et al. (1981) examined the energy delivered by various combinations of rope, cathead, drill rig, cathead speed and rotation direction, etc. These studies, con- ducted during routine actual field operations, indicated an average COV for hammer drop height from 1 to 8%, with an average of about 4%, and an average COV for the energy ratios (energy delivered to theoretical energy) from 4 to 2 I%, with an average of about 8%. Seven additional literature studies indicated a COV on total field variability from 14 to 81%. None of these studies contained any information on equipment and procedural controls, and therefore it is difficult to evaluate the cause of such large variations. However, these results tend to suggest that typical field measurements could exhibit COV values for N on the order of 40 to 45%. Overall, these data and studies illustrated that a number of equipment and proce- dural/operator parameters influence the SPT results. The COV for a particular site investigation will depend to a large degree on the type and uniformity of equipment and procedures employed by the contractor performing the tests. Therefore, “best” and “worst” case scenarios must be considered. The “best” case is represented by use of identical equipment and procedures for each test, including standard split- spoon samplers in good condition with liners, constant drill rod diameters, consistent and careful drilling and borehole preparation, and use of an automatic trip monkey hammer drop system. These test conditions minimize the variability associated with the testing equipment, and one can reasonably estimate a COV less than or equal to 5% associated with minor variations in the test equipment. Similarly, careful and consistent drilling and borehole preparation methods, coupled with an automatic hammer drop, should result in negligible procedural variations, again on the order of 5%. Available results (Serota and Lowther 1973) and analysis of the Bieganousky and Marcuson (1976, 1977) data indicate that random errors of about 12% can be expected for automatic hammer drop systems. The total measurement variability, COV(total), can be estimated from the square root of the sum of the squares of the BACK
  • 6. 274 UNCERTAINTY IN GEOLOGIC ENVIRONMENT equipment, procedural/operator, and random test errors, as given below, because the individual errors are not strictly additive (e.g., Benjamin and Cornell 1970): COV(tota1) = [COV(equipment)2 + COV(procedure)’ + COV(random)2J”2 (1) Therefore, the total measurement COV (neglecting any effect of soil variability) for the “best” case scenario is on the order of 14%. Alternatively, the “worst” case would be realized if there was no control on the equipment and procedures employed in the tests. In this instance, equipment varia- bility resulting from use of inconsistent sampler design and condition, size of drill rods, and type of hammer could result in COVs as large as 75%. Also, variable and careless drilling methods, borehole preparation techniques, and methods of dropping the hammer could result in procedural variability as high as 75% or so. Adding an additional 15% for random measurement gives a “worst” ease total measurement COV that could be 100%. This unrealistic COV is intended only to illustrate the magnitude of error that can result from careless and poorly supervised SPT practices. Typical COVs to be expected for routine field operations are between these extremes, although they generally should be nearer the “best” case estimate. Previous statis- tical analyses of field test results support this conclusion, but it must be recognized that these results include the influence of soil variability. CONE PENETRATION TEST (CPT) UNCERTAINTIES Basic Testing Issues The CPT, long established in northern Europe as the most common method of subsurface profiling, has become popular in North America during the past two to three decades, especially in offshore applications and softer or looser terrestrial soil deposits. Electronic measurement and recording systems, now relatively common, have minimized many of the operator influences associated with older mechanical push-rod systems. Some of the basic correlations between cone readings and soil properties, however, are based on mechanical cone devices, and it is useful to under- stand how the two systems interact with various soil parameters. The cone penetration test is described by ASTM D3441, and the major factors affecting CPT readings are listed in Table 2, including the cone type, size, angle, and wear, as well as rod compression (MCPT only) or leaky seals (ECPT only). Of these, the cone angle and rod compression can have a significant affect, while cone wear has only minor to moderate influence (Durgunoglu & Mitchell 1975, Baligh et al. 1979), and the other factors are relatively minor in comparison. Tests and theory (Marsland & @rterman 1982) have shown that the standard 60” base angle leads to maximum penetration for a given force, while more acute angles lead to significantly higher penetration resistance (up to 30% for an 18’ cone angle, for example). Tests comparing MCPT and ECPT results indicate that the differences are somewhat material-dependent. In one study (Joustra 1974), for example, ECPT results were about 3% lower than MCPT results. However, another study (Kok 1974) showed BACK
  • 7. IN-SITU TEST UNCERTAINTY 275 TABLE 2. CPT Testing Variables Grouv CPT Variable Relative Effect Item on Test Results Equipment Procedural/ Operator Cone type (MCPT or ECPT) Cone size Cone angle Rod compression (MCPT) Manufacturing defects Leaky seals (ECPT) Excessive cone wear Telescoping vs. continuous penetration Calibration error Penetration rate Inclined penetration moderate to significant minor moderate to significant significant minor to moderate minor minor to moderate moderate to significant minor to moderate minor moderate to significant that MCPT results were as much as 30% lower than the corresponding ECPT results. Operator and procedural effects can have a marked influence on MCPT results, while having a relatively minor influence on ECPT results (DeRuiter 1971, Schaap & Zuidberg 1982). The MCPT telescoping method of advancement does not completely separate side resistance from tip resistance measurements, which can be particularly troublesome for side resistance measurements. Another problem is that the tip and side resistances for the MCPT are not made at the same elevation. For a thinly stratified deposit, this difference in elevation can lead to major dislocations. The ECPT is virtually operator-independent, once it is calibrated and checked (Schaap & Zuidberg 1982). The only major variable is the rate of penetration, which has been shown to increase tip resistance readings approximately 10% for an order- of-magnitude increase in penetration rate in soft clays (Baligh et al. 1979). This study and others (e.g., Kok 1974, Marsland & Quarterman 1982, Jamiolkowski et al. 1985) indicate that the rates suggested in ASTM D3441 are reasonable and unlikely to affect tip or side resistance significantly. Deviation from verticality can be a major factor with either the mechanical or electric systems, because readings might not correspond to the depth based on the rod length. Analysis of Uncertainty Several quantitative studies have evaluated the effects of various soil parameters on CPT readings under laboratory and field conditions. These studies, while not specifically performed to determine test variability, provide some insights. In one study (Parkin et al. 1980), tests in dry sand resulted in COVs of 5% for tip resistance (qc) measurements and 10% for side resistance (f,) measurements. Interestingly, these results are equal to those cited in ASTM D3441 for the expected precision of CPT BACK
  • 8. 276 UNCERTAINTY IN GEOLOGIC ENVIRONMENT measurements. Another study (Villet 1981) indicated a COV of about 7% for qc. Two studies (Nottingham 1975, Boghrat 1982) reported comparative field CPT test results. The COV for qc and f, ranged from 30 to 60% in general, but because no attempt was made to separate soil variability from the total test variability, it is not possible to determine the COV resulting from random test errors. Finally, assuming that in cohesive soils the measured qc values are proportional to the undrained shear strength sU,several studies indicate that a field COV for q, on the order of 10 to 15% can be expected in homogeneous cohesive deposits for electric cone tests with good equipment and procedural control. Standards exist for both the MCPT and ECPT, and they address the primary equipment and procedural variables that influence the CPT results. However, diffe- rences exist in these tests, so they will be discussed separately. The mechanical system, which requires two rods, leads to uncertainty in separa- ting tip and side resistance measurements. These uncertainties can significantly influence the interpretation of the test results but, providing that consistent proce- dures are employed, the effect on measurement variability should not be as great. However, the procedure required to advance the cone with this two rod system is highly operator-dependent. Based on limited available data, the procedural/operator COV for the MCPT is estimated at 10% for q, and 15% for f,. Equipment factors, including damaged cone, clogging of the area between the friction sleeve and the cone tip, and manufacturing defects, lead to an estimate of equipment COV of 5% or less. ASTM 03441 indicates that the precision or random errors associated with the MCPT are 10% for qc and 15% for f,. Therefore, total measurement errors of about 15% for qc and 22% for f, are to be expected. These estimates assume that standard equipment and procedures are employed at all times. The ECPT is a single rod system that allows for continuous penetration and automatic data acquisition, and it is virtually operator-independent. Therefore, the procedural/operator COV is estimated at 5% or less. Similarly, the electrical trans- ducers and measuring devices are precise, so the associated equipment COV is likely to be on the order of 3%. Results from laboratory calibration studies (e.g., Parkin et al. 1980, Villet 1981) indicate random errors of about 5% for qc and 10% for f,. Therefore, the total measurement COV for the ECPT is about 8% for qc and 12% for f,. These values are supported by several previous statistical analyses of ECPT data that include some limited effect of soil variability. The piezocone test variability for q, and f, should be approximately equal to that for the ECPT. However, a standard does not yet exist that specifies several of the additional equipment and procedural variables for this test. VANE SHEAR TEST (VST) UNCERTAINTIES Basic Testing Issues Since its inception, the vane shear test has been considered to be an efficient and economical method of measuring the undrained shear strength of cohesive soils. In recent years, there have been relatively few studies devoted to the precision of this BACK
  • 9. IN-SITU TEST UNCERTAINTY 277 test, which is probably due, in part, to ASTM D2.573 that addresses the fundamental variables in vane shear testing. Key variables with the VST are given in Table 3. The influence of equipment variables on test results has been minimized through research (e.g., Osterberg 1956, Arman et al. 1975, Donald et al. 1977, Walker 1983, Schmertmann & Morgenstem 1977), and ASTM D2573 now prescribes acceptable vane sizes, shapes, and thicknesses. The principal source of equipment variability is the torque measurement device, which ideally should be a geared drive, although a torque wrench is allowed under the standard. Several sources of uncertainty can result from the procedure followed by the test operator. Insertion disturbance can decrease apparent soil strength, although distur- bance is minimal if the vane is pushed at least five borehole diameters below the base of the hole (e.g., Flaate 1966). Additionally, it is important to remove the vane and clean it between tests, rather than simply pushing it deeper for the next test (Flaate 1966). In the latter case, soil can adhere to the vane and cause additional disturbance (decreasing apparent strength) or increased effective vane diameter (increasing appa- rent strength), Rate of torque application during the VST is well-known to affect strength measurements (e.g., Ladd et al. 1971, Walker 1983,), but the rate has been standardized for many years and, as with other variables listed above, is addressed best through proper training and equipment specification. Analysis of Uncertainty There are relatively few controlled studies that specifically address random test variability. In one such study (McCormack & Wilding 1979), replicate tests were made in various soil horizons. The average COV was on the order of lo%, indica- ting the VST is a highly reproducible test. The remainder of the studies compiled (Ladd et al. 1971, Arman et al. 1975, Baligh et al. 1979, Asaoka & A.-Grivas 1982) indicate that total field COV can range from 3 to 44%, depending on the soil in TABLE 3. VST Testing Variables VST Variable Relative Effect Group Item on Test Results Equipment Vane length Height/diameter ratio Blade thickness Torque measuring device Damaged vane Procedural/ Operator Vane insertion method Rod friction calibration Time delay between insertion & testing Vane rotation rate minor moderate moderate moderate to significant moderate to significant moderate to significant moderate to significant minor to moderatea moderate a - perhaps significant in soft clays BACK
  • 10. 218 UNCERTAINTY IN GEOLOGIC ENVIRONMENT which the test is performed. Because of the clear equipment specifications in the VST standard, the varia- bility associated with the test apparatus should be minimal, providing that the torque is delivered and measured by a gear drive and proving ring, and not by a torque wrench. Assuming that appropriate standards are maintained, the equipment COV resulting from the torque measuring device and any vane damage is estimated at 5%. The vane test procedure is very sensitive to soil disturbance caused by vane insertion. Other procedural variables that are not standardized and can influence the test results include the time delay between insertion and testing and whether the vane is cleaned between tests. These test parameters suggest a COV of 8% for procedural/operator variability. Limited data further indicate that the random error of the VST results is about 10%. Therefore, the total measurement COV is about 14%. This value is supported by the results of previous investigations in homogeneous deposits, which indicate slightly higher variabilities resulting from soil heterogeneity. DILATOMETER TEST @MT) UNCERTAINTIES Basic Testing Issues The DMT is a relatively recent addition to the available set of field in-situ tools. It is still somewhat under development, although there is a suggested test standard (Schmertmann 1986). Key variables are given in Table 4. The equipment used for the test consists of the blade, push rods, and pressure gage control. This equipment is relatively rugged and of simple design, but there are several features that can affect test results. The blade must be kept sharp, and the stainless steel membrane on the blade face must be inspected often for damage and potential leakage (Schmertmann 1984). The rods used to push the blade are subject to bending in stiff soils, and bent rods can results in erroneous penetration readings. The pressure control box is rela- tively accurate and is not considered a significant error source. Many of the procedural variables that can affect the CPT also can affect the DMT, including variations from the vertical, the method of advancing the blade (driving vs. pushing and rate of advance), and the point at which thrust is measured (above the blade or at the surface) (Campanella et al. 1985). Analysis of Uncertainty There have been relatively few published studies on the variability of DMT test results because of the relative newness of the test. However, several limited labora- tory calibration studies and field data are available. In one study (Schmertmann 1984), 30 DMTs were performed in loose, normally consolidated dry sand. COVs for the A and B readings were 2 and 0.8%, respectively. Another study (Lacasse & Lunne 1982) reported qualitatively that DMTs in homogeneous Norwegian clay exhibited “nearly perfect repeatability.” These data suggest that DMT measurements are highly repeatable and that random test errors are not large. In one series of unpublished tests (Lutenegger 1985), students (mostly inexperienced in DMT testing) performed six different graduate the same tests in one BACK
  • 11. IN-SITU TEST UNCERTAINTY TABLE 4. DMT Testing Variables 279 DMT Variable Relative Effect Group Item on Test Results Equipment Leaking seals minor Deformed membrane moderate Bent or deformed push rods minor to moderate Damaged blade minor Procedural/ Operator Push rod inclination Testing rate Driving method Rod friction Calibration error minor to moderate moderate to significant minor to moderate minor minor to moderate sounding within the same soil profile. Analysis of variance indicated that at the 1% confidence level, only one person had a significantly different set of readings than the other five for the A reading, while there was no difference between all six for the B reading. This study indicates that the test is highly reproducible, even by inexperienced persons. Another study (Baldi et al. 1986) reported laboratory calibration measurements of the DMT in dry sands. The tests do not represent purely replicate data, and there- fore the variability is influenced by effects other than random test errors. Given these constraints, the results of four small samples indicate COVs of the corrected dilatometer readings of 5.4 to 18.5%, with mean values on the order of 10%. These numbers likely represent upper-bound estimates of DMT precision. The DMT is not yet standardized, but the equipment is patented and must be obtained from one supplier. Therefore, variations in the dilatometer blade can result only from manufacturing defects. Also, there is a suggested test procedure standard, so everyone conducting the test is using the same equipment and procedure. The test also is simple to perform, so operator effects are not substantial. In terms of equipment variability, only the pressure gage sensitivity, deformation of the expanding membrane, and severe damage to the blade should influence the test results. Therefore, a COV of 5% is suggested as an estimate of equipment effects. Procedural/operator variables influencing the DMT include the method of penetra- tion, deviations from vertical, and rate of membrane expansion. Unless one of these procedural variables is changed dramatically during a test, they will influence the interpretation more than the variability of the test results. Accordingly, the proce- dural/operator COV is estimated at 5%. Limited data from DMT calibration testing indicate random errors on the order of 8%. Therefore, a total measurement COV of about 11% is estimated for the DMT. BACK
  • 12. 280 UNCERTAINTY IN GEOLOGIC ENVIRONMENT PRESSUREMETER TEST (PMT) UNCERTAINTIES Basic Testing Issues The PMT is a relatively complex test that, although standardized in ASTM D4719, is subject to variations in result interpretation that are not associated with the other common in-situ tests. Key testing variables are given in Table 5. The PMT has been the subject of extensive research over the past few decades and, while a com- plete discussion of the test is beyond the scope of this paper, several pertinent points should be mentioned. The equipment to perform the PMT is complex and consists of the probe, tubing, and pressure-volume control unit. Each component has potential sources of uncer- tainty. For the probe, the length/diameter (L/D) ratio has been shown to affect the limit pressure both theoretically and experimentally (Laier 1973, Briaud 1986). An L/D ratio of 6.5 has been recommended for all pressuremeters (Laier 1973), while an earlier suggested procedure specifies a minimum L/D of 4. Membranes must be matched to soil conditions and should be calibrated often to accommodate change in elasticity with time and usage. One-cell and three-cell models are acceptable in the standard test procedure, and research has not documented a consistent pattern of results obtained with the two types (Briaud 1986). Expansion of the tubing during inflation can affect volumetric measurements and, at high flow rates near the limit pressure, head losses can affect measurement of inflation pressure (Baguelin et al. 1978). The pressure-volume control unit has gages or readouts that must be cali- brated but, in general, readings are precise to 1% or better (Baguelin et al. 1978). TABLE 5. PMT Testing Variables Group PMT Variable Relative Effect Item on Test Results Equipment Procedural/ Operator Gage error Expansion of tubing Frictional losses in tubing Probe dimensions Probe design (PMT) Membrane aging Cutting shoe size (SBPMT) Cutter position (SBPMT) Probe shape (SBPMT) Drilling equipment (SBPMT) Electrical sensor compliance (SBPMT) Drilling method & borehole preparation (PMT) Probe inflation rate Relaxation time (SBPMT) Probe advance rate (SBPMT) minor minor to moderate minor minor to moderate minor minor minor to moderate minor minor to moderate minor to moderate minor significant minor to moderate moderate to significant moderate to significant BACK
  • 13. IN-SITU TEST UNCERTAINTY 281 Because of the complexity of the PMT, the results are sensitive to the test proce- dure, including borehole preparation and rate of probe expansion (e.g., Baguelin et al. 1978, Ervin 1983, Benoit 1983, Benoit & Clough 1986, Briaud 1986,) and, for self-boring pressuremeters (SBPMT), the relaxation period after reaching the desired test depth and the rate of probe advance (e.g., Ghionna et al. 1981, Benoit 1983, Jamiolkowski et al. 1985). Analysis of Uncertainties Quantitative analysis of random test variations is complicated by the fact that test results are presented as pressure-volume curves that must be interpreted to yield three pressures: seating (p,), yield (pr), and limit (p,). Therefore, the variability in these pressures depends not only on the test data but also on the method of interpre- tation. In one study (Briaud & Tucker 1984), the soil modulus and limit pressure were determined in a series of tests in sand and sandy gravel deposits. The results indicated COVs for the PMT modulus of 28 to 68% and for p, ranging from 22 to 50%. Data from another study (Laier 1973) indicated that the COV for the PMT modulus ranged from 16 to 24% and for p,, of 2 to 15%. Additional studies with SBPMTs in San Francisco Bay Mud (Denby 1978) indicated COVs of 14 and 38% for the inferred soil parameters s, (undrained shear strength) and Gi (shear modulus), respectively. As above, these parameters require interpretation and are useful prima- rily to illustrate the order of magnitude of field variability. Laboratory tests in calibration chambers have shown that self-boring pressure- meters can be highly reproducible under certain conditions. Tests in which the unload-reload shear modulus was measured in four consecutive loading loops yielded COVs in a range of 5 to 18%, with an average about 9%. In this case, skilled operators were performing the tests under relatively well-controlled conditions. Therefore, these values should be considered to represent the lower bound of random errors associated with the SBPMT. Although the PMT is recently standardized, there are equipment and procedural variabilities that lead to uncertainty in the test results. Quantitative evaluation of these uncertainties is complicated by two factors. First, PMT and SBPMT results typically are expressed in terms of inferred soil parameters or characteristic pressures interpreted from the volume-pressure curve. The variability of these measures is not necessarily representative of the variability of the test data, because they include some level of uncertainty associated with their interpretation. Second, previous quantitative studies of pressuremeter test variability are virtually non-existent, and therefore information necessary to estimate the sources of variability accurately is not available. However, based on review of sources of error in these tests, an esti- mate can be made of the magnitude of these sources of test uncertainty. For routine testing with the pre-bored pressuremeter (PMT), variables that can influence the test results include the volume and pressure measuring devices, expan- sion of the tubing, and the pressuremeter dimensions. Proper calibration can account for most of the variability with the first two, and previous research has shown that the range of dimensions of commercially-available probes do not affect the results BACK
  • 14. 282 UNCERTAINTY IN GEOLOGIC ENVIRONMENT significantly. Therefore, an approximate value of PMT equipment COV is 5%. The self-boring equipment possesses several characteristics that can lead to additional test uncertainty, including the cutting shoe position and size, shape of the probe, and compliance of the electronic feeler arms. Therefore, a slightly larger value of 8% is estimated for the COV of equipment variability for the SBPMT. Procedural/operator influences affecting the degree of disturbance of the bore- hole are a major factor in the uncertain@ of PMT results. The method of drilling, density of the drilling fluid, and ratio of the borehole diameter to the probe diameter all affect the quality of PMT results. The drilling operation is difficult to control, because it depends on the care exercised by the operator. Therefore, the variability of these procedural parameters will not be constant. Based on the limited data available, an estimate of typical procedural/operator COV is 12%. The SBPMT has several additional procedural variables and is a considerably more complex test to perform. Hence, a COV of 15% is estimated for this test. As mentioned previously, there have been few previous quantitative studies of PMT and SBPMT results, and very little replicate test data are available to estimate random measurement error. However, the statistical analyses performed for this study indicate random test errors on the order of 10% for the PMT and 8% for the SBPMT. Therefore, the total measurement COV for the PMT is about 16% and for the SBPMT is about 19%. SUMMARY AND CONCLUSIONS This paper has reviewed the principal sources of error for five of the most common m-situ soil tests. The errors fall into three classes: equipment variability, procedural and operator influences, and random test errors. Tests with well-defined specifications, accepted practices, and standardized equipment have relatively small test variability. The SPT, while standardized to some degree, allows for enough equipment variability that significant variations in test results are common between different operators using different drilling equipment. Other tests generally have lesser variability. Table 6 summarizes the estimates developed for the in-situ test errors. For each test, the first three columns give the uncertainties as COV for equipment, procedure, and random test errors. The next column combines these errors as the square root of the sum of the squares, since the errors are not strictly additive. Finally, because of the limited data available and the need to use judgment to estimate these errors, the last column represents the range of probable test measurement variability one can expect in typical field in-situ tests. ACKNOWLEDGMENTS Much of the research cited herein was conducted by C. J. Orchant for his thesis at Cornell under the writers’ supervision. He no longer practices engineering. This study was supported by the Electric Power Research Institute under RP1493. V. J. BACK
  • 15. IN-SITU TEST UNCERTAINTY TABLE 6. Uncertainty Estimates for Five Common In-Situ Tests 283 Test (acronym) Coefficient of Variation,COV (%) Equipment Procedure Random Total” Rangeb Standard Penetration Test 5” - 19 5f - 15d (SPT) Mechanical Cone Penetration 5 IO’- 15’ Test (MCPT) Electric Cone Penetration Test 3 5 (ECPT) Vane Shear Test 5 8 (VW Dilatometer Test 5 5 (DMT) Pressuremeter Test, Pre-Bored 5 12 U’M’U Self-Boring Pressuremeter 8 15 Test (SBPMT) 12- 15 IO’- 15’ 5e - 10’ IO 8 IO 8 14’- 1OOd 15’-22’ 15-45 15-25 8’ - 12’ 5- 15 14 IO-20 11 5- 15 16 IO - 20g 19 15 - 25g - COV(tota1) = [COV(equipment)* + COV(procedure)’ + COV(random)2]‘n : - Because of limited data and judgment involved in estimating COVs, ranges represent probable magnitudes of field test measurement error c,d - Best to worst case scenarios, respectively, for SPT e,f - Tip and side resistances, respectively, for CPT g - Results may differ for p., pr, and p,, but data are insufficient to clarify this issue Longo was the EPRI project manager. A. Hirany is the current project manager. We appreciate contributions made by M. D. Grigoriu, K.-K. Phoon, and T. D. O’Rourke. REFERENCES American Society for Testing & Materials (1994), “Standard Test Method for Pene- tration Test & Split-Barrel Sampling of Soils (D 1586-84),” Annual Book of St&. 04.08, ASTM, Philadelphia, 129-133. American Society for Testing & Materials (1994), “Standard Test Method for Field Vane Shear Test in Cohesive Soil (D 2573-72),” Annual Book of St&. 04.08, ASTM, Philadelphia, 228-230. American Society for Testing & Materials (1994), “Standard Test Method for Deep, Quasi-Static, Cone & Friction-Cone Penetration Tests of Soil (D 3441-86),” Annual Book of Stcis.04.08, ASTM, Philadelphia, 338-343. American Society for Testing & Materials (1994), “Standard Test Method for Pres- suremeter Testing in Soils (D 4719-87),” Annual Book of Stds. 04.08, ASTM, Philadelphia, 858-865. Arman, A, Poplin, JK & Ahmad, N (1975), “Study of Vane Shear”, Proc., ASCE BACK
  • 16. 284 UNCERTAINTY IN GEOLOGIC ENVIRONMENT Spec. Conf. In-Situ Measurement of Soil Properties (1), Raleigh, 93-120. Asaoka, A & A-Grivas, D (1982), “Spatial Variability of Undrained Strength of Clays”, J. Geofech. Eng. Div., ASCE, 108(GT5), 743-757. Baguelin, F, Jezequel, JF & Shields, DH (1978), Pressuremeter & Fndn. Eng., Tram Tech Pubs., Clausthal, 617 p. Baldi, G, Bellotti, R, Ghionna, V, Jamiolkowski, M, Marchetti, S & Pasqualini, E (1986) “Flat Dilatometer Tests in Calibration Chambers”, Use ofln-Situ Tests in Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 43 l-446. Baligh, MM, Vivatrat, V & Ladd, CC (1979), “Exploration & Evaluation of Eng. Properties for Fndn. Design of Offshore Structures”, Rpt. MITSG 79-8, MIT, Cambridge, 268 p. Benjamin, JR & Cornell, CA (1970), Probability, Statistics & Decision for Civil Engineers, McGraw-Hill, New York, 684 p. Benoit, J (1983), “Analysis of Self-Boring Pressuremeter Tests in Soft Clay”, PhD Thesis, Stanford Univ., Palo Alto, 361 p. Benoit, J & Clough, GW (1986), “Self-Boring Pressuremeter Tests in Soft Clay”, J. Geotech. Eng., ASCE, 112(l), 60-78. Bieganousky, WA & Marcuson, WF, III (1976), “Liquefaction Potential of Dams & Fndns.: Rpt. 1, Lab. Std Penetration Tests on Reid Bedford & Ottawa Sands”, Rex Rpt. S-76-2, US Army Eng. Waterways Exper. Sta., Vicksburg, 96 p. Bieganousky, WA & Marcuson, WF, III (1977), “Liquefaction Potential of Dams & Fndns.: Rpt. 2, Lab. Std. Penetration Tests on Platte River & Concrete Sands”, Res. Rpt. S-76-2, US Army Eng. Waterways Exper. Sta., Vicksburg, 47 p. Boghrat, A (1982), “Design & Construction of Piezoblade & Evaluation of Marchetti Dilatometer”, PhD Thesis, Univ. of Florida, Gainesville, 244 p. Briaud, J-L (1986), “Pressuremeter & Foundation Design”, Use of In-Situ Tests in Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 74-l 15. Briaud, J-L & Tucker, L (1984), “Coefficient of Variation of In-Situ Tests in Sand”, Probabilistic Characterization of Soil Properties: Bridge Between Theory & Practice, Ed. DS Bowles & H-Y Ko, ASCE, New York, 119-139. Campanella, RG, Robertson, PK, Gillespie, DG & Grieg, J (1985), “Recent Develop- ments in In-Situ Testing of Soils”, Proc., 1lth Intl. Conf. Soil Mech. & Fndn. Eng. (2), San Francisco, 849-854. de Mello, VFB (1971), “Standard Penetration Test”, Proc., 4th Panam. Conf. Soil Mech. & Fndn. Eng. (I), San Juan, l-86. Denby, GM (1978), “Self-Boring Pressuremeter Study of San Francisco Bay Mud”, PhD Thesis, Stanford Univ., Palo Alto, 270 p. DeRuiter, J (1971), “Electric Penetrometer for Site Investigation”, J. Soil Mech. & Fndns. Div., ASCE, 92(SM2), 457-472. Donald, IB, Jordan, DO, Parker, RJ & Toh, CT (1977), “Vane Test - A Critical Appraisal”, Proc., 9th Intl. Conf. Soil Mech. & Fndn. Eng. (1), Tokyo, 81-88. Durgunoglu, HT & Mitchell, JK (1975), “Static Penetration Resistance of Soils: II - Evaluation of Theory & Implications for Practice”, Proc., ASCE Spec. Conf. In- Situ Measurement of Soil Properties (1), Raleigh, 172-189. BACK
  • 17. IN-SITU TEST UNCERTAINTY 285 Ervin, MC (1983), “Pressuremeter in Geotechnical Investigations”, In-Situ Testing for Geotech. Investigations, Ed. MC Ervin, Balkema, Rotterdam, 49-64. Flaate, K (1966), “Factors Influencing Results of Vane Tests”, Can. Geotech. .I., 3(l), 18-31. Fletcher, GFA (1965), “Standard Penetration Test, Its Uses and Abuses”, J. Soil Mech. & Fndns. Div., ASCE, 91(SM4), 67-75. Ghionna, V, Jamiolkowski, M, Lancellotta, R, Tordella, ML & Ladd, CC (1981) “Performance of Self-Boring Pressuremeter Tests in Cohesive Deposits”, Rpt. RD-H/173, Fed. Hwy. Admin., Washington, 182 p. Gibbs, HJ & Holtz, WG (1957), “Research on Determining the Density of Sands by Spoon Penetration Testing”, Proc., 4th Intl. Conf. Soil Mech. & Fndn. Eng. (I), London, 35-39. Ireland, HO, Moretto, 0 & Vargas, M (1970), “Dynamic Penetration Test: A Stan- dard That Is Not Standardized”, Geotechnique, 20(2), 185-192. Jamiolkowski, M, Ladd, CC, Germaine, JT & Lancellotta, R (1985), “New Develop- ments in Field & Lab. Testing of Soils”, Proc., 1lth Intl. Conf. Soil Mech. & Fndn. Eng. (l), San Francisco, 57-l 53. Joustra, K (1974), “Comparative Measurements of Influence of Cone Shape on Re- sults of Soundings”, Proc., 1st Eur. Symp. Penetration Testing (2.2), Stockholm, 199-204. Kok, L (1974), “Effect of Penetration Speed & Cone Shape on Dutch Static Cone Penetration Test Results”, Proc., 1st Eur. Symp. Penetration Testing (2.2), Stockholm, 215-220. Kovacs, WD, Evans, JC & Griffith, AH (1975), “Comparative Investigation of MO- bile Drilling Company’s Safety-T-Driver w. Standard Cathead w. Manila Rope for Performance of Standard Penetration Test”, Rpt, School of Civil Eng., Purdue Univ., Lafayette, 129 p. (As referenced in Kovacs et al. 1981) Kovacs, WD, Salomone, LA & Yokel, FY (1981), “Energy Measurements in Stan- dard Penetration Test”, Natl. Bldg. Science Series 135, Natl. Bureau Stds., Washington, 73 p. Kovacs, WD, Salomone, LA & Yokel, FY (1983), “Comparison of Energy Measure- ments in Standard Penetration Test Using Cathead & Rope Method”, NUREG/ CR-3545, Nuclear Reg. Comm., Washington, 69 p. Lacasse, S & Lunne, T (1982), “Penetration Tests in Two Norwegian Clays”, Proc., 2nd Eur. Symp. Penetration Testing (2), Amsterdam, 661-670. Ladd, CC, Moh, Z-C & Gifford, DG (1971), “Statistical Analysis of Undrained Strength of Bangkok Clay”, Proc., 1st Intl. Conf. Applications of Statistics & Probability in Soil & Strut. Eng., Hong Kong, 3 14-322. Laier, J (1973), “Effect of Pressuremeter Probe Length/Diameter Ratio & Borehole Disturbance on Pressuremeter Test Results in Dry Sands”, PhD Thesis, Univ. of Florida, Gainesville, 204 p. Lutenegger, AJ (1985), “Personal Communication”. Marsland, A & Quarterman, RST (1982), “Factors Affecting Measurements of Quasi- Static Penetration Tests in Clays”, Proc., 2nd Eur. Symp. Penetration Testing BACK
  • 18. 286 UNCERTAINTY IN GEOLOGIC ENVIRONMENT (2), Amsterdam, 697-702. Matsumota, M & Matsubara, K (1982), “Effect of Rod Diameter in Standard Penetra- tion Test”, Proc., 2nd Eur. Symp. Penetration Testing (I), Amsterdam, 107-l 12. McCormack, DE & Wilding, LP (1979), “Soil Properties Influencing Strength of Cantield & Geeburg Soils”, J. Soil Science Sot. Ame&a, 43( 1), 167-I 73. McLean, FG, Franklin, AG & Dahlstrand, TK (1975) “Influence of Mechanical Var- iables on SPT”, Proc., ASCE Spec. Conf. In-Situ Measurement of Soil Proper- ties (2), Raleigh, 287-3 18. Nottingham, LC (1975) “Use of Quasi-Static Penetrometer Data to Predict Load Ca- pacity of Piles”, PhD Thesis, Univ. of Florida, Gainesville, 553 p. Orchant, CJ, Kulhawy, FH & Trautmann, CH (1988), “Reliability-Based Foundation Design for Transmission Line Structures: Critical Evaluation of In-Situ Test Methods,” Rpt. EL-5507(2), Electric Power Res. Inst., Palo Alto, 207 p. Osterberg, JO (1956), “Introduction”, Symp. Vane Shear Testing of Soils (STP 193), ASTM, Philadelphia, l-8. Parkin, A, Holden, J, Aamot, K, Last, N & Lunne, T (1980) “Lab. Investigation of CPTs in Sand”, Rpt 52108-9, Norwegian Geotech. Inst., Oslo, 3 1 p. Phoon, KK & Kulhawy, FH (1996), “On Quantifying Inherent Soil Variability”, this conference proceedings. Riggs, CO (1986), “North American Standard Penetration Testing”, Use of In-Situ Tests in Geotech. Eng. (GSP 6), Ed. SP Clemence, ASCE, New York, 949-967. Schaap, LHJ & Zuidberg, HM (1982) “Mechanical & Electrical Aspects of Electric Cone Penetrometer Tip”, Proc., 2nd Eur. Symp. Penetration Testing (2) Amsterdam, 84 1-85 1. Schmertmann, JH (1971), Disc. on “Standard Penetration Test”, Proc., 4th Panam Conf. Soil Mech. & Fndn. Eng. (3), San Juan, 90-98. Schmertmann, JH (1975), “Measurement of In-Situ Shear Strength”, Proc., ASCE Spec. Conf. In-Situ Measurement of Soil Properties (2), Raleigh, 57-138. Schmertmamr, JH (1978), “Use of SPT to Measure Dynamic Soil Properties?... Yes But!...“, Dynamic Geotech. Testing (STP 654), ASTM, Philadelphia, 341-355. Schmertmann, JH (1984), Ed., DMT Digest 4, GPE Inc, Gainesville, 19 p. Schmertmamr, JH (1986), “Suggested Method for Performing the Flat Dilatometer Test,” Geotech.’ Test. J:, ASTM, 9(2), 93-101. Schmertmann, JH & Morgenstern, NR (1977), Disc. to Session 1, Proc., 9th Intl. Conf. Soil Mech. & Fndn Eng. (3), Tokyo, 356-359. Seed, HB, Tokimatsu, K, Harder, LF & Chung, RM (1984), “Influence of SPT Proce- dures in Soil Liquefaction Resistance Evaluations”, Rpt. UCB/EERC-84/15, Univ. of California, Berkeley, 50 p. Serota, S & Lowther, G (1973), Disc. on “Accuracy of Relative Density Measure- ments”, Geotechnique, 23(2), 301-303. Villet, WCB (1981), “Acoustic Emissions During Static Penetration of Soils”, PhD Thesis, Univ. of California, Berkeley, 400 p. Walker, BF (1983), “Vane Shear Strength Testing”, In-Situ Testing for Geotech. Investigations, Ed. MC Ervin, Balkema, Rotterdam, 65-72. BACK