PAPER
I:one seneI:ra~:ion
I:es:inc in
jo sica
resir, ua soi
I
J Peuchen, SJ Plasman
s
8 R van Steveninck, Fugro
Engineers BV.
1. lahodIclon
Tropical regions include soil conditions not commonly found in the
more temperate areas of Europe and North America where electrical
cone penetration testing (CPT) apparatus was originally developed.
The more complex of these tropical conditions are residual soils and
calcareous soils, both of which can be characterised by highly variable
and erratic strength. They offer a harsh environment for CPT apparatus.
This paper presents examples of commercial CPT experience for
tropical residual soils. Three sites are considered, covering more than
800 tests including 500 piezocone tests. The sites are situated in
Thailand, Malaysia and Singapore. They represent residual soils
derived from the weathering of igneous, metamorphic and sedimentary rocks respectively.
2. 1tophal nsklual solh
2.1 Chararaarlalaa
Residual soils are weathering products of a parent rock. The soils have
not been significantly transported. Residual soils are defined by the
British Standards Institution (1981) as rock mass with the highest
weathering grade (Grade VI). All rock material is converted to soil, and
mass structure and material fabric are destroyed. The Geological
Society (1990) also includes weathering Grades IV (moderately weathered) and V (extremely weathered) and defines residual soils as rock
mass whereby more than half of the rock material is converted to soil.
The latter definition is used in this paper.
Processes forming residual soils include physical and chemical
weathering. Weathering processes in the tropics are generally much
faster than in regions of moderate climates. Apart from the nature and
composition of the parent rock, wetting and drying cycles of the soils
as a result of precipitation are the main factors of influence on structure and physical behaviour.
22 Taal glaa
The examples of CPT experience apply to three different sites in tropical regions: (1) Southeast Thailand, (2) West Coast of Peninsular
Malaysia and (3) Singapore.
(1) For the Thailand site, the residual soils consist of weathered granites of Carboniferous age. The granites are coarse to medium grained
biotite muscovite granites with pegmatite and quartz veins. The residual soils are overlain by 6 to 8 m thick alluvial deposits consisting
mainly of sands and clayey sands.
(2) For the Malaysia site, the residual soils consist of weathered schists
of Lower Palaeozoic age. The schists are strongly schistose with steep
bedding. Quartz veins are common. Laterites, a product of intense
weathering of schist, are also present. There are no or thin overburden
deposits present.
(3) For the Singapore site, the residual soils consist
of weathered rocks of Triassic/Jurassic igneous
breccia rocks and sedimentary sandstone and mudstone. Overlying soil are 10m to 15m thick sedimentary clay and 3m to 5m thick made-ground.
The groundwater table is reasonably close to ground
surface for each of the test sites.
3. CPT Agueltul
3.1 "=:=„===::~
A typical CPT set-up (Figure 1) consists of a
hydraulic thrust machine which pushes a penetrometer fixed to a string of pushrods into the ground at
a constant rate of penetration of 200mm/s.
A basic electrical penetrometer (Figure 2) has a
diameter of 36mm and a cross sectional area of
Rffraa 1 fyfaaal
CPT
aat~
Part
Position
Shaft
Gap between friction
sleeve and shaft
~
~
—
~
~
~
Push rod connector
Soil seal
Electric cable for signal
transmission
Water seal
Amplifier unit
E
0
Friction sleeve sensor
Friction sleeve
lnclinometer
c
J ~Friction
t
~ ~
sleeve
Cone sensor
Gap between friction
sleeve and cone
Water seal
Cylindrical extension
above base of cone
Base of cone
Face of cone
~Soil seal
~
Pressure sensor
Filter
Cone
Apex of cone
GROUND ENGINEERING JANUARY/FEBRUARY
1996
C
37
PAPER CONE PENETRATION TESTING
1000mm2. It includes load sensors for measuring
axial load on the cone tip and axial load on the friction sleeve. Often a penetrometer is also equipped
with a pressure sensor: a piezocone penetrometer.
Data processing facilities allow presentation of cone
resistance, sleeve friction and pore water pressure
versus penetration.
Test apparatus and test procedures are standardised, for example by the British Standards
Institution (1990).
32 Hlshrhel
IN
RESIDUAL SOILS
Rgme3.
0
8
4
6
Cone resistance qc MN/ms
12
16
20
Rgnre 3.
Examph CPT la
rnskhal segs ol
Fnctio ratio fs/qc a 100
12
4
8
0
grenHL
Rgme 4.
Exemple CPT In
rnskhel segs of
srfdsL
Rgmn 5.
Example CPT In
C
Cone penetration test apparatus was originally
developed and refined for sedimentary soils of temperate areas. Since the early 1970s, Fugro has used
mechanical CPTs in tropical soils. In the 1970s,
North Sea offshore hydrocarbon prospects promoted
advancements in CPT apparatus for harsh environmental and ground conditions. Robust and accurate
systems with electrical downhole sensors were
developed. This development was followed by offshore experience with CPTs in calcareous soils.
Since 1989, electrical CPTs are routinely used by
Fugro for testing of onshore and nearshore tropical
residual soils. Important promoting factors are the
robustness, accuracy and data management capabilities of currently available systems. This has obvious consequences for economics.
ai
s
al
s
rnsahel sogs of
sainl- eml nmd-
'2
shna
al
1
E
3
~ —--
tx
TOF OF RESOUAL 80tL
8
l, 48-
12
14
.as
18
0
0.1
0.2
0.3
0.4
0.6
Local fnction fs MN/me
34 Crnrwt peneL~s
Features of current Fugro penetrometers (Figure 2)
include:
~load sensors placed in series for cone resistance and sleeve friction
absolute pressure sensors for water pressure measurements with
piezocone penetrometers
~omni-directional inclinometers.
A range of 100kN applies to the load sensor for measurement of cone
resistance. This corresponds with a maximum range for cone resistance of OMPa to 100MPa. Measurement of sleeve friction is by a second 100kN load sensor placed in series with the load sensor for cone
resistance. The load on the second load sensor is thus the sum of the
load on the cone tip and the friction sleeve. In practice, sleeve friction
values of up to about 3MPa are permissible.
The advantage of load sensors placed in series is robustness within
the limitations of the geometry of cone penetrometers (36mm diameter). Practical accuracy of available large-range load sensors is typically linear at 0.1% of full range (cone resistance of 100kPa). This accuracy can also be handled by conventional analogue to digital (A/D) converters. The load sensing accuracy is adequate for detailed engineering
interpretation of residual soils and in most cases also adequate for any
softer sedimentary deposits overlying residual soil. Studies have indicated that large-range load sensors can offer better performance in soft
soils than more sensitive systems (Zuidberg, 1988). This is due to factors such as in situ non-symmetrical loading, temperature effects, load
transfer due to minute movements of penetrometer components under
load, and water/soil seal effects.
Standard Fugro piezocone penetrometers are equipped with a 10MPa
pressure sensor. A ceramic or HDPE filter element is fitted either in the
conical part of the cone tip or in the cylindrical extension above the
base of the cone (Figure 2).
3A Spechl penegnmshr sensors
Additional measuring sensors can be incorporated for specific studies
at relatively little additional cost. Examples for the test sites include
downhole electrical conductivity measurements for cathodic protection studies and seismic downhole measurements to obtain (dynamic)
machine foundation parameters. These measurements will not be discussed further in this paper.
*.I Operalenal
ttgllerleftee
Sehchd prtisedmns
CPTs at the test sites were mostly
performed with track-mounted
vehicles ballasted to 20 tonnes and
fitted with hydraulic levelling
devices. Anchored or ballasted
trailers were used for some tests.
Conventional 20t thrust machines
and push rods were used. A friction reducer was incorporated in
the push rods. The friction reducer comprised a local widening
~
lagh 1 cpr
Sae
~
~~
Amount of Aenrage
(P)CPT
wheal
tech
—
~
Aearage
deptg In
s~l
doper hdow
eeg (m)
emend
snrhce (m)
n
n
Mahfnh
approximately 1m above the cone tip. It serves to reduce soil friction
along the push rods. CPT casing was used in near-surface soft soils to
avoid buckling of push rods under high loads. CPT casing is pushed
into the ground by the thrust machine simultaneously with the push
rods. Additional time required is minimaL
Piezocone penetrometers were equipped with a filter element mounted in the cylindrical extension above the base of the cone. The advantages and disadvantages of filter position with regard to correlations
for soil type, shear strength and overconsolidation ratio are discussed
by Robertson (1990). Practically, the harsh environment offered by
residual soils dictates a filter element mounted in the cylindrical extension above the base of the cone. This location reduces effects of wear
and filter compression.
A new, saturated, HDPE filter element was fitted before each piezocone test. Internal connections between the filter element and the pressure sensor were saturated by injection of glycerin or silicon oil before
each test. Subsequently, the cone tip was fitted with a thin rubber membrane. For commercial reasons, no predrilling to the groundwater table
was adopted. For most tests, pore pressure signatures indicated adequate maintenance of saturation during penetration of the partially
saturated zone.
The selected criteria for test termination in tropical residual soil
were one or more of the following:
~overloading of a load sensor (100kN),
~maximum thrust (20t),
~selected maximum penetrometer inclination (usually 15 ),
~rapid increase of penetrometer inclination.
These criteria were developed to minimise loss of penetrometers and
thrust rods when testing in harsh environments.
Computer-based data processing and presentation was performed on
site. This included processing of sensor calibration factors, zero-oad
offsets, and joining of records between add-ons of push rods.
4.2 SH+efxxdac inqPenetration depths and applied CPT termination criteria are listed in
Table 1 for each of the test sites.
In the residual soils of the Thailand site, cone resistances varied typically between 4MPa and 10MPa. Sleeve frictions varied between
0.1MPa to over IMPa while pore pressures varied between 0.5 MPa to
21
6
14
16
~
ghp crftwla append
Prhaon
load cell
Maxhnm
seems mmad
(%)
~~~
ee
erwet (%) of faetow (%)
(%)
51
21
11
15
16
9
2
21
15
20
25
16
15
35
17
GROUND ENGINEERING JANUARY/EEBRUARY
1996
PAPER CONE PENETRATION TESTING
IN
4
COne reeiSIanoe
0
18
4
8
~
12
14
E
5
—- - TOP OF
4
F notion rane IIA/c r 100
12
4
4
4
Cone resistance oc MN/mt
I2
Is
10
Fr ct:on rat 4 tt, 4» s 100
4
12
8
0
OC M/t/m
18
I
RESIDUAL SOILS
RES/DU/tL SOIL
0
12
0
3
Z'
10
3 8
8
-8
2
s-t 2
IO
1
5'
ts
c'-
;E, ~na
n
.)
O
14
e
I~
sl
CI
e
-IS
4I
E
E.
!
!
I
--!I— !
!
I
Ql
~~~
m
—
—--TOP
OF RES/DUAL SOIL
5-20
2'
5
ttI
0
0.1
0.2
0.3
---- —— Local
-—
0.4
frrction ts
22
0.5
MN/m'—
I
"24
more than 1MPa. Figure 3 shows a typical CPT signature. Within a
part of the test site, CPTs were locally terminated at 6m to 8m depth.
This depth corresponded with the top of the residual soil stratum.
Nearby CPTs could reach a depth of 18m to 23m. This phenomenon is
probably related to the presence of a duricrust.
For the Malaysia site, cone resistances in the residual soils increased
with depth from between 2MPa to 4MPa to between 10MPa to 20MPa.
Sleeve frictions varied between 0.1MPa to 0.2MPa increasing with
depth to over 1MPa. Figure 4 shows typical CPT results. High bedding
inclination and a succession of stronger and weaker schist strata
resulted in highly variable CPT penetrations over relatively short lateral distances. In addition, CPTs appeared to follow weaker zones along
the bedding inclination, which led to increasing deviation from vertical for some tests.
For the Singapore site, cone resistances varied typically between
2MPa and 4MPa and between 15MPa and 20MPa for the residual soils
derived from mudstone and sandstone respectively. Sleeve frictions for
the mudstone residual soil varied between 0.1MPa and 0.4MPa while
values in excess of 1MPa were recorded for the sandstone residual
soils. Cone resitances for the residual soils of the breccia varied bewteen 4MPa and 20 MPa with sleeve frictions up to 1MPa. Pore pressures
were typically more than 1.5MPa in the mudstone residual soils, about
0.1MPa to 0.2MPa for the sandstone residual soils and about 0.5MPa in
the breccia residual soils. Figure 5 shows a typical CPT signature for
an alternating residual mudstone/sandstone
sequence. Table 1 shows
an average penetration in the residual soil of about 6m, primarily related to overloading of the cone and friction sleeve load sensors. This is
attributed to a limited weathering depth of the sandstone strata.
Penetrations into residual soil of up to 30m were reached for test locations with weathered mudstone and breccia. It can also be noted that
the Singapore site includes a relatively thick overburden in comparison with the Thailand and Malaysia sites. The overburden soils cause
build-up of friction along the push rods.
The maximum CPT penetrations (Table 1) depend on parent rock
type as well as weathering grade. For example, comparison of CPTs
with nearby borings indicates that CPT penetration into sandstone is
estimated to be feasible for a weathering grade equal to or higher than
Grade V (extremely weathered). Similarly, CPT penetration into mudstone of weathering Grades III and IV (highly and moderately weathered) is estimated to be feasible. These estimates apply to the selected
test apparatus and procedures.
5. Data
Inteyt1yetatlon
5.1 Approach
CPT data interpretation was carried out within the framework of an
engineering geological model. The models for the test sites were developed by integration of investigation and interpretation techniques.
Data acquisition included geological desk studies and ground investigation. In addition to CPTs, ground investigation included drilling,
sampling, Standard Penetration Testing (SPT) and triple-tube rock corGROUND ENGINEERING JANUARY/FEBRUARY
1996
-28
Mt/DSTONE
3
22
'ANDSTONE
P -::-'"
M2
as
Ist
os
ae
02
—————
Local tnction
4
fs
MN/m'—
ing of boreholes. In case of pile engineering requirements, boreholes
were generally continued below CPT termination depths. The ratio of
CPTs to borings ranged from 10:1 to 3:1.
Interpretation, handling and presentation of the large amount of
data were facilitated by use of a computerised data base system.
52 5hTE55raphy ail
sall type
In comparison with other in situ test methods electrical cone penetration tests have the highest applicability for identification of soil
stratigraphy and soil type (Campanella & Robertson, 1983). The examples of CPT experience presented in this paper indicate that the high
applicability for identification can be extended to tropical residual
soils.
In particular, the following features are of interest:
~differentiation between residual soils and any overlying sedimentary
deposits;
~identification of interlayered parent rock types;
~characteristics of non-homogeneous weathering sequences and
resulting residual soil types.
5.3 fshysiaal
hahashiar
Penetration of a cone penetrometer into soil is generally compared
with bearing capacity and flow phenomena. Large-strain soil behaviour dominates CPT parameters. Thus, repeatable correlations with
soil stiffness are unlikely, except under a limited range of conditions.
More general correlations with soil strength are generally feasible.
This is further discussed below.
The fixed rate of CPT penetration and soil permeability determine
whether a particular soil will show a predominantly undrained or a
predominantly drained response. Inspection of CPT signatures of tropical residual soils described in this paper indicates the following soil
behaviour characteristics:
~undrained CPT behaviour applies;
PAPER CONE PENETRATION TESTING
R$nre
RESIDUAL SOILS
L
Examtds
seee
IN
~Horizontal
si
distance between CPT location and
adjacent borehole is less than 10m.
~N/qc ratios of less than 0.5 x N/qc and greater
than 2 x N/qc are omitted.
~The linear fit applies to the least square method
forced through the origin.
The presented correlation for the Singapore breccia, sandstone and mudstone should be viewed with
caution, as the number of data points is limited.
Furthermore, no distinction between rock types is
made. The Singapore correlation is quite similar to
that reported by Poh et al. (1985) and Chang (1988).
Chang reported N/qc = 5.6, whereby no separation
was made between rock types. Unclear are also correlation details in terms of statistics.
rr
6. Summary
st
eeeeenhd
nndrahsd
trhxhl test
retndL
5 300-
I
I
I
100
200
300
p'kPa)
~soil behaviour classification cor- tate 2 thrrelathrm ct Wttc.
responds with overconsolidated
and/or cemented sedimentary
clay or clayey sand in accordance
with Robertson (1990).
For the residual soil derived from
2A
233
$ctdst
Singapore sandstone layers, the
$.$
14$
soil behaviour
characteristics
Breach/
suggest drained CPT behaviour.
sandstann/
$.$
32
The data base for this residual soil
mmhtcna
is limited and will not be considered further.
Detailed interpretation of soil strength from CPT results is often
baSed On COrrelatianS With labOratOry Strength teStS. Far trOpiCal reSid
ual soils, this practice warrants caution.
Electrical CPTs are in situ tests. They are repeatable if properly calibrated apparatus is operated by trained personnel. Laboratory test
reSultS depend On SeVeral detallS SUCh aS Sampling Standard, Sample
handling inCluding SOil mOiSture retentiOn, and SpeCimen SeleCtian
and trimming methOdS. FurthermOre, Sampling and labOratOry teSting
needS tO take aCCOunt Of the StruCture Of trapiCal reSidual SOIIS and the
inhomogeneous
anisotropic consistency In commercial practice, signifICant Sample diSturbanCe iS uSually inCurred. ThiS iS eVident frOm
initial laboratory consolidation strains (Lacasse 3r Berre, 1988).
FOr the eXampleS preSented in thiS paper, it haS been fOund that nO reasonable strength measurements
are obtained by unconsolidated
undrained triaXial teStS. COnSOlidated undrained triaXial teStS Can prOvide acceptable strength measurements, in particular for drained param terSC'and 'neXam leiS VeninFi
6 Th
I a li t .
percussion samp mg ec niques;
~thin-walled sampling tubes of 75mm diameter;
~careful sample handling practice including on-site air conditioned
temperature controlled storage;
~test specimens of 50mm diameter.
$.4 Wq rehdsnshit)s
Apart from academic interest, correlation of CPT with SPT results is
often carried out for "calibration" of local or regional geotechnical
practice. Results for the three example sites are presented in Table 2.
The following comments apply:
~Cone resistance qc is averaged over length of SPT at same elevation
in adjacent borehole.
~N-values are corrected to N60 in accordance with British Standards
(1990).
~SPT blowcounts in excess of 50 are a result of extrapolation.
e
Ground behaviour characteristics of tropical residual soils include highly variable and erratic
strength. This leads to a harsh environment for
Cone Penetration Test (CPT) apparatus.
Commercial employment of electrical (piezo)cone
penetrometer apparatus is now feasible for routine
testing of onshore and nearshore tropical residual
soils. Important promoting factors are the robustness, accuracy and data management capabilities of
the available systems.
Project examples illustrate practical CPT applications for tropical residual soils derived from the
weathering of igneous, metamorphic and sedimenI
tary rocks. The examples cover more than 800 tests
400
including 500 piezocone tests.
CPT penetration capabilities depend on parent
rock type as well as weathering grade. Comparison
of CPTs with nearby borings indicates that CPT penetration into sandy residual soil is feasible for Grade V (extremely
weathered). Similarly, CPT penetration into clayey residual soil of
weathering Grades III and IV (highly and moderately weathered) is feasible.
CPT data interpretation practice for tropical residual soils largely
follows conventional
interpretation
for sedimentary
deposits.
Difficulties in correlations between CPT results and laboratory test
results occur in practice due to laboratory sample disturbance.
The penetration capabilities together with the data interpretation
options allow prominence of electrical CPT techniques in high-quality,
economic and fast-track ground investigation strategies.
7. Refwences
Standards Institution (1981). "BS5930, Code of practice for Site Investigations",
Standards Institution, London, pp111-120.
Standards Institution (1990). BS 1377:Part 9:1990, "British Standard Test Methods
for sous for civu engmeering purposes: part 9. Insitu tests", British Standards
hrstitution, London, pp1825.
CamPaneua, RG & Robertson, PK (1983). "Flat Plate Duatometer Testmg:
and
Development", Soil Mech. Report 88, University of British Colombia.
Geologicaigociety(1990)."Engineering GroupWorkingpartyReport:
TropicalResidual
Soils." The Quarterly Journal of Engineering Geology. Vol23, Nol, 101pp.
Lacasse, S 8 Berre T 0988). "Triaxial Testing Methods for Soils.", Advanced Triaxial
British
British
British
of test
Reset
American Society for Testing and Materials, Philadelphia, pp284-289.
Poh, KB, Chuah, HL & Tan, SB (1985). "Residual Granite Soil of Singapore." Proceedings
8th Southeast Asian Geotechnical Conference, Kuala Lumpur, Voll, pp3.1-3.9.
Rad, NS &
T (1988).
Co~latlons b t"een Piezocone Test Res~t'nd
Undrained Shear Strength of Clay." Penetration Testing, ISOPT-1, De Ruiter (ed.), pp911917
Robertson, PK (1990). "Soil Classification using the Cone Penetration Test", Can Geotech
L~e,
-D~t
Jni No 2» Pp»'158
Zuidberg,'H (1988). -Piezocone Penetration Testmg, P be D'veloPment-,
Speciality Session No13, March 24, Orlando, Florida.
ISOPT-I,
GROUND ENGINEERING JANUARY/FEBRUARY
1996