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

net/publication/245407959

Evaluating stiffness and strength of pavement materials

Article  in  ICE Proceedings Geotechnical Engineering · January 2005

DOI: 10.1680/geng.2005.158.4.217

CITATIONS READS

23 3,976

2 authors:

Tuncer B. Edil Auckpath Sawangsuriya

University of Wisconsin–Madison Government Organization
371 PUBLICATIONS   10,278 CITATIONS    32 PUBLICATIONS   492 CITATIONS   

SEE PROFILE SEE PROFILE

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

part of PhD Dissertation project View project

sustainability of highway construction View project

All content following this page was uploaded by Tuncer B. Edil on 12 February 2014.

The user has requested enhancement of the downloaded file.

Proceedings of the Institution of
Civil Engineers
Geotechnical Engineering 158
October 2005 Issue GE4
Pages 217–230

Paper 13910
Received 19/07/2004
Accepted 12/01/2005
Keywords: Auckpath Sawangsuriya Tuncer B. Edil
pavement design/quality control/ Graduate Research Assistant, Professor, Department of Civil and
strength & testing of materials Department of Civil and Environmental Engineering, University
Environmental Engineering, University of Wisconsin-Madison, USA
of Wisconsin-Madison, USA

Evaluating stiffness and strength of pavement materials

A. Sawangsuriya MSc and T. B. Edil PhD, PE

Current mechanical empirical-based pavement design DPI dynamic cone penetrometer (DCP) penetration index
requires use of the mechanical properties of pavement DPIavg arithmetic average of DPI value
materials. For quantitative evaluations of the DPIwt avg weighted average of DPI value
mechanical properties (i.e. stiffness and strength), field E modulus
tests are emphasised. In this paper, a recently EDCP modulus from the DCP test
developed instrument called the soil stiffness gauge EFWD back-calculated modulus from the FWD test
(SSG) and the dynamic cone penetrometer (DCP) have ELDW modulus from the light drop weight test
been used to assess respectively the in-situ stiffness and EPLT modulus from the plate load test
strength of natural earthen materials, industrial by- ERM modulus from the resilient modulus test
products, and chemically stabilised soils from ten ESEIS modulus from the seismic test
highway construction sites around the state of ESEIS-MOD modulus from the seismic test after adjustment to
Wisconsin, USA. The SSG and DCP survey data were realistic stress and strain levels for the pavement
analysed to develop a relationship between the SSG ESSG modulus from the SSG test
stiffness and DCP penetration index (DPI) values for KSSG SSG stiffness
individual material types and for all materials combined. LL liquid limit
A simple linear semi-logarithmic model is obtained PI plasticity index
between the SSG stiffness and DPI values, with the R2 coefficient of determination
coefficient of determination, R2 , ranging from 0.47 to Su1:0% stress causing 1% axial strain
0.75 for individual material types and an R2 value of 0.72 WN natural water content
for all materials combined. The SSG stiffness and DPI WOPT optimum water content
values can be also correlated with the modulus (E) and ªdmax maximum dry unit weight
California bearing ratio (CBR) of the materials
Poisson’s ratio of the materials
respectively. A good relationship is obtained between E  1 – 3 deviator stress
from the SSG and CBR from the DCP, and is compared 3 confining stress
with the well-known equations developed by Powell et
al. and AASHTO as well as other available correlations 1. INTRODUCTION
from different in-situ tests: the falling weight Given the importance of the mechanical properties (i.e. stiffness
deflectometer, German light drop weight, and plate and strength) in pavement materials evaluation, there has been
load tests. Finally, the proposed power model is a concerted effort in recent years to develop methods for
validated with a data set from two other test sites. quantitative evaluations of these properties. Direct monitoring
Either or both devices show good potential for future of stiffness and strength is consistent with the transition from
use in pavement and subgrade materials evaluation. The empirical to current mechanistic-empirical pavement design
in-situ stiffness and strength properties of various procedures for structural design of flexible pavements. To
materials can be rapidly and directly monitored in successfully implement mechanistic-empirical pavement design
companion with the conventional compaction control procedures, and to move toward the performance-based
tests during pavement construction. The modulus of specifications that are required to control the long-term
pavement and subgrade materials is uniquely related to functional and structural performance, additional in-situ
CBRs regardless of soil type and site, and their stiffness and strength measurements should be included along
relationship is also applicable to both as-compacted and with the conventional compaction control tests (i.e. nuclear
post-construction states. density or laboratory moisture content samples), which do not
give the mechanical properties of pavement materials
NOTATION directly. 1,2 Direct monitoring of stiffness and strength would
CBR California bearing ratio facilitate quantitative evaluations of alternative construction
Cu coefficient of uniformity practices and materials, such as recycled and reclaimed
D10 particle diameter corresponding to 10% finer materials, that result in cost savings and environmental
D60 particle diameter corresponding to 60% finer benefits. 2 For instance, the use of recycled and reclaimed

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 217
 materials, both as working platform over poor subgrade and as measurement in granular materials. A finite element analysis
a subbase in pavement structure, is being explored by the and the SSG measurements in a test box indicated that the
transportation community. Evaluation of these new materials radius of measurement influence extends to 300 mm. For two-
on the basis of index property measurements such as moisture layer materials with different stiffness, the SSG starts to
density or past subjective experience based on natural soil register the stiffness of an upper-layer material of 125 mm or
behaviour is also severely limited. Most of the correlations for thicker. The effect of the lower layer may continue to be
modulus (E) are based on water content, dry density, and present even at an upper-layer material thickness of 275 mm,
degree of saturation, and were developed from tests on depending on the relative stiffness (or contrast) of the layer
laboratory-compacted specimens. Because field compaction materials. A comparison of moduli of granular soils obtained
curves and the associated lines of optimum are often different from the SSG with moduli obtained from other tests on the
from those of laboratory compaction, and also the moisture basis of comparable stress levels indicates that the SSG
condition of pavement materials changes with time after measures moduli in the very small strain amplitude range (i.e.
construction, the use of these laboratory-based correlations 2.7 3 10
4 to 4.3 3 10
4 %, which is less than 10
3 %). The
may cause significant error in estimating the operating SSG-induced strain amplitudes are lower than the strain
pavement and subgrade moduli. Empirical correlations based amplitudes induced in the resilient modulus test, but are larger
on California bearing ratio (CBR) 3,4 or stress causing 1% strain than the strain amplitude of the seismic test. 6
(Su1:0% ) in the unconfined compression test 5 have been
successfully used in evaluating pavement and subgrade moduli. The measured soil stiffness from the SSG can be used to
Moreover, the relationship between E and CBR or E and Su1:0% calculate the modulus of the materials near the surface. For a
is not affected by the changes in subgrade condition after rigid ring-shaped foot resting on a linear-elastic,
construction and therefore is applicable to both as-compacted homogeneous, and isotropic infinite half-space, the stiffness
and post-construction states. (KSSG ) is related to the modulus of the soil (ESSG ): 9

The soil stiffness gauge (SSG), a recently developed instrument, 1:77ESSG R

and the dynamic cone penetrometer (DCP), which is commonly 1 K SSG ¼
1

2
used by the pavement community, offer a means of directly
monitoring the in-situ stiffness and index of strength of surficial
materials. SSG stiffness and DCP penetrometer index (DPI) have where
is the Poisson’s ratio of the materials, and R is the
been also correlated with modulus and CBR respectively. 6,7 In outside radius of the ring (57.2 mm). Note that KSSG and ESSG are
this study, these two devices will be implemented in a number of expressed in MN/m and MPa respectively. Because of the very
projects over two construction seasons in the state of Wisconsin, small strain amplitudes induced by the SSG, an elastic response
USA. The objectives of this study are: of the soils is assumed, and the use of equation (1) is justified.

(a) to examine the use of the SSG and DCP for pavement 2.2. Correlation with other moduli
materials evaluation Back-calculated moduli of base and subgrade soils from the
(b) to explore the degree of correlation and appropriate falling weight deflectometer (FWD) test have been used
effective depth zone for statistically significant empirical extensively in pavement design, construction and maintenance.
correlations between the two devices Wu et al. 10 found that the relationship between the SSG
(c) to develop the modulus and CBR relationship obtained stiffness (KSSG ) and the back-calculated modulus from the FWD
from these devices and compare it with the other well- (EFWD ) can be presented in the following form:
known relationships as well as with different in-situ tests.
:
2 EFWD ¼ 22:96e0 12 KSSG R2 ¼ 0:66
2. SOIL STIFFNESS GAUGE (SSG)

2.1. Description Note that KSSG and EFWD are expressed in MN/m and MPa
The soil stiffness gauge (SSG), which is currently marketed as respectively. Wu et al. 10 also noted that the difference between
the Humboldt GeoGaugeTM (Fig. 1(a)), is a recently developed these methods can be ascribed to the in-situ variability of the
instrument for directly measuring the in-situ stiffness of soils. material properties. As EFWD is obtained through inversion
The SSG measures near-surface stiffness by imparting a small based on all seven deflection measurements, which cover a
dynamic force to the soil though a ring-shaped foot at 25 distance of about 2 m, EFWD of pavement layers is therefore a
steady-state frequencies between 100 and 196 Hz. Based upon weighted average value over 2 m. By contrast, the SSG
the force and displacement–time history, the stiffness is measures only the near-surface soil stiffness right underneath
calculated internally as the average force per unit displacement its ring foot, with a measurement influence of less than 0.3 m.
over the measured frequencies and reported. In a previous Chen et al. 11 suggested that a general linear relationship
investigation, the acceleration and corresponding displacement between KSSG and EFWD is discernible as the following:
were measured.6 Given knowledge of the soil properties, the
force induced by the SSG was estimated based on finite 3 EFWD ¼ 37:65K SSG
261:96 R2 ¼ 0:82
element analysis. The maximum single amplitude dynamic
force produced during the SSG measurement is determined to
be 10 to 17 N. A measurement takes only about 1.5 min. Again, KSSG and EFWD are expressed in MN/m and MPa
Sawangsuriya et al.8 studied the zone of measurement respectively. Wu et al. 10 provided the correlation between SSG
influence and the effects of layered materials on the SSG stiffness and modulus from seismic tests, including dirt-seismic

218 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
pavement analyser (D-SPA) and spectral analysis of surface
waves (SASW), for soft to medium-stiff subgrades to very stiff
bases. They indicated that the modulus obtained from the SSG
is about a quarter of that obtained from seismic tests. A linear
relationship is obtained between the SSG stiffness (KSSG ) and
the seismic modulus (ESEIS ):

4 ESEIS ¼ 47:53K SSG þ 79:05 R2 ¼ 0:62

Note that KSSG and ESEIS are expressed in MN/m and MPa
respectively. The discrepancy between these two tests is
explained by the difference in the stress–strain levels used, as
well as by the uncertainty of the effective depth of the SSG,
which varies with stiffness, density, and types of materials
being tested. 10 Chen et al. 11 also conducted a similar study,
and indicated that the relationship between KSSG and ESEIS from
D-SPA and SASW for soft to medium-stiff subgrades to very (a)

stiff bases can be expressed as follows:

Handle
5 ESEIS ¼ 55:42K SSG
162:94 R2 ¼ 0:81 Upper stop
Hammer, 8 kg
(17·6 lb) or 4·6 kg
Again, the units of KSSG and ESEIS are in MN/m and MPa (10·1 lb)

respectively. The relationship between KSSG and ESEIS is found

to be obvious and convincing. The operation of the SSG is very
simple and feasible for the purpose of quality control; however,

575 mm (22·6 in)

it yields a stiffness value only near the surface. The seismic
tests can generate a stiffness–depth profile, but its operation is
more complicated. 11

3. DYNAMIC CONE PENETROMETER (DCP)

Anvil/coupler
3.1. Description
Assembly
Scala 12 developed the Scala penetrometer for assessing the in-
situ CBR of cohesive soils. In the last decade, the Scala
penetrometer has evolved into the DCP test for determining in-
1000 mm (39·4 in)

situ CBR and modulus. The DCP is now being used extensively Graduated drive
rod or vertical
in South Africa, the United Kingdom, the United States,
scale
Australia, and other countries because it is simple, rugged,
economical, and able to provide a rapid in-situ index of 16 mm (5/8 in)
Vertical side diameter drive
strength and, more indirectly, the modulus of the subgrade as 3 mm (1/8 in) rod
in length
well as the pavement structure.

The DCP is used for measuring the material resistance to

Cone angle 60°
penetration in terms of millimetres per blow while the cone of Tip
the device is being driven into the pavement structure or
subgrade. The typical DCP consists of an 8 kg hammer that 20 mm (0·79 in)
drops over a height of 575 mm, which yields a theoretical (b)
driving energy of 45 J or 14.3 J/cm2 , and drives a 608 20 mm
base diameter cone tip vertically into the pavement structure or Fig. 1. (a) Humboldt soil stiffness gauge (SSG); (b) dynamic
subgrade (Fig. 1(b)). The steel rod to which the cone is attached cone penetrometer (DCP)
has a smaller diameter than the cone (16 mm) to reduce skin
friction. The number of blows during operation is recorded
with the depth of penetration. The slope of the relationship been available longer than the SSG, and has been used as a
between number of blows and depth of penetration (in convenient field tool; however, it is not a direct property test
millimetres per blow) at a given linear depth segment is but an index test based on dynamic impact loading.
recorded as the DCP penetration index (DPI). In addition to soil
profiling (i.e. the thickness and nature of each layer in a given As DCP testing is basically a measure of penetration resistance,
pavement), DCP data are correlated with various pavement expressed as DCP penetration index (DPI), the analysis of the
design parameters, including CBR, shear strength, modulus, and DCP data must be interpreted, following a standardised
back-calculated modulus from the FWD. 11,13–17 The DCP has procedure, to generate a representative value of penetration per

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 219
 blow for the material being tested. This representative value can where z is the penetration distance per blow set and H is the
be obtained by averaging the DPI across the entire penetration overall penetration depth of interest. These two methods are
depth at each test location. Two methods of calculating the graphically presented in Fig. 2 for a lean clay with sand (STH
representative DPI value for a given penetration depth of 100 in Table 1). Allbright 19 reported that the weighted average
interest are considered: (a) arithmetic average and (b) weighted method yielded a narrower standard deviation for the
average. 18 The arithmetic average can be obtained as follows: representative DPI value and provided better correlations with
other field tests than the arithmetic average method based on
available field data. In this study, the weighted average method
X
N
ðDPIÞ i is employed to calculate the representative DPI value.
6
i
DPIavg ¼
N The influence of layers below and above the cone tip must also
be considered in the analysis of the DCP data. As cone
penetration is associated with the development of a failure
surface, the penetration resistance is influenced by the presence
where N is the total number of DPI recorded in a given
of a layer if the cone tip is located within a few cone diameters
penetration depth of interest. The weighted average technique
of the interface of two highly contrasting layers. Little effect on
uses the following formula:
penetration resistance is noticed as the cone tip approaches the
interface if both layers have similar properties. The extent of
1XN
 the zone of influence depends on the size of the cone, soil type,
7 DPIwt avg ¼ ðDPIÞ i ð zÞ i soil density, stress state, and the contrast in properties of
H i
adjacent layers. 20 The weighted average DPI was calculated

100

200
Penetration depth: mm

300

400

500

600

700
0 5 10 15 20 25 30 35
Number of blows

0 0
Arithmetic average Weighted average
100 100
Summation of all
shaded area
200 200
Penetration depth: mm

Penetration depth: mm

Total penetration
depth
300 300

400 400

500 500

600 600

700 700
0 10 20 30 40 50 0 10 20 30 40 50
DCP penetration index, DPI: mm/blow DCP penetration index, DPI: mm/blow

Fig. 2. Two methods for calculating the representative DPI value

220 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
 Site Soil name Specific Liquid Plasticity Classification WN : WOPT : ªdmax :
gravity limit index % % kN/m3
USCS AASHTO

STH 60 (test section) Joy silt loam 2.70 39 15 CL-ML A-6(16) 25.0 19.0 16.5
Scenic Edge development Plano silt loam 2.71 44 20 CL A-7-6(20) 27.0 20.0 16.2
Gils Way development Plano silt loam 2.71 46 20 CL A-7-6(20) 23.4 19.5 16.3
STH 26 Lean clay with sand 2.64 32 11 CL A-6(7) 20.7 13.5 19.2
STH 100 Lean clay with sand 2.74 29 14 CL A-6(9) 14.2 14.4 18.2
STH 44 Silty, clayey sand 2.70 23 7 SM-SC A-4(0) 9.8 11.7 19.8

Table 1. Properties of natural earthen materials and their classification: fine-grained soils (USCS, unified soil classification system;
AASHTO, American Association of State Highway and Transportation Officials)

over various DCP penetration depths, from the surface to 76, 4. FIELD TESTING PROGRAMME
152, 229, 305 and 381 mm, in the analyses to identify the most
representative depth in correlating DPI with SSG stiffness. The 4.1. Data collection
only exception was the chemically stabilised soils, where the A Humboldt SSG manufactured by Humboldt Mfg Co. was used
maximum penetration depth was limited by the thickness of to measure the in-situ stiffness properties of the pavement
the stabilised layer (i.e. 305 mm). As these selected penetration materials in this study. The SSG stiffness measurements were
depths over which the DPI was calculated were generally well made in accordance with ASTM D6758. 30 The SSG assesses
within a layer, the DPI obtained is considered to have near-surface stiffness with a maximum measurement depth of
negligible influence from an interface. It has been shown that approximately 300–380 mm. Sawangsuriya et al.6,8 reported
vertical confinement (i.e. due to rigid pavement structure or that the depth of measurement significance ranges from 125 to
upper granular/cohesive layers) and rod friction (i.e. due to a 178 mm, where the higher stress–strain conditions occur
collapse of the granular material on the rod surface during within the measurement zone (i.e. 125–300 mm), and this is
penetration) may affect DPI values. 21 These effects were not an also beyond a blind zone that exists at less than 125 mm in
issue in this investigation, because the current study involved SSG measurements. These findings were based on granular
only subgrade and subbase evaluation during construction soils, but similar conditions can also be expected in fine-
such that the DCP tests were performed directly on the exposed grained (cohesive) soils.
surface of these materials.
A DCP manufactured by Kessler Soils Engineering Products,
3.2. Correlation with California bearing ratio (CBR) Inc. was used to measure the in-situ strength index properties
To assess the structural properties of the pavement materials, of the pavement materials in this study. DCP penetration index
the DCP penetration index (DPI) values are usually correlated (DPI), in millimetres per blow, which can be used to estimate
with the California bearing ratio (CBR) of the pavement the shear strength characteristics of soils, was calculated in
materials. Extensive research has been conducted to develop an accordance with ASTM D6951. 31 The DCP is typically used to
empirical relationship between DPI and CBR for a wide range assess material properties to a depth of 1 m below the ground
of pavement and subgrade materials. This include research by surface. The size of the cone tip relative to the average grain
Livneh et al., 21 Kleyn, 22 Harison, 23 Livneh, 24 McElvaney and size of the material that is penetrated is found to influence the
Djatnika, 25 Webster et al., 26 and Livneh and Livneh. 27 Based penetration resistance. 20 This is because of the number of
on their researches, many of the relationships between DPI and grains that come into contact with the face of the cone and the
CBR can be quantitatively presented in the following form: failure surface. Therefore the DCP cannot be used in very
coarse-grained materials containing a large percentage of
logðCBRÞ ¼ Æ þ  logðDPIÞ aggregates greater than 50 mm, or in highly stabilised or
8
cemented materials.

where Æ and  are coefficients ranging from 2.44 to 2.56 and 4.2. Site description

1.07 to
1.16 respectively, which are valid for a wide range SSG and DCP measurements were made at 10 highway
of pavement and subgrade materials. Note also that CBR is in construction sites around the state of Wisconsin, USA. Test
percent and DPI is in millimetres per blow (mm/blow). For a section STH 60 is located approximately 40 km north of the
wide range of granular and cohesive materials, the US Army city of Madison, and consists of a 1.4 km segment of the
Corps of Engineers use the coefficients Æ and  of 2.46 and highway. This project consisted of a field demonstration of

1.12, which have been also adopted by several agencies and alternative soft subgrade reinforcement methods. Detailed
researchers 20,26,28,29 and are in general agreement with the descriptions of each test section are given in Edil et al. 32 The
various sources of information. Livneh et al. 21 also show that SSG and DCP tests were conducted in each test section.
there exists a universal correlation between DPI and CBR for a
wide range of pavement and subgrade materials, testing Scenic Edge development project is a 0.7 km city street
conditions, and technologies. In addition, the relationship constructed as a residential subdivision in Cross Plains by
between DPI and CBR is independent of water content and dry stabilising the soft subgrade in place with fly ash. Both the
unit weight, because water content and dry unit weight subgrade soil and the fly-ash-stabilised subgrade layer were
influence DPI and CBR equally. tested using the SSG and DCP.

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 221
 Gils Way development project, which uses a soil–lime mixture, earthen materials encountered and their properties, together with
is also located in Cross Plains. The construction section was their classification. Compaction curves corresponding to the
approximately 400 m long. Because of the time constraint, only standard compaction effort described in ASTM D698 33 were
the SSG was performed before the liming process—that is, on the developed, except for breaker run. Note that breaker run is the
untreated subgrade. The SSG and DCP were performed after the excavated and crushed rock including cobbles (75–350 mm in
liming process—that is, on the lime-stabilised subgrade layer. diameter) with a soil fraction. It was retrieved from the cuts in
parts of the project route. Its soil fraction consisted of
Seven highway construction sites that involved the use of only approximately 30% gravel, 65% sand, and 5% fines.
natural earthen materials were from different soil regions of
Wisconsin. 18 The SSG and DCP were performed on the exposed The properties of the processed construction materials (i.e.
subgrade soils that were either compacted (five sites) or had other than the natural earthen materials), along with their
not been re-compacted (two sites). classification, are summarised in Tables 3 and 4. These
materials are subdivided into two main categories: (a)
5. MATERIAL CLASSIFICATION AND PROPERTIES industrial by-products and (b) chemically stabilised soils. The
Samples were collected either along the centreline or near the by-products consisted of bottom ash, foundry slag, and
shoulder of the roadway from 10 highway construction sites to foundry sand. Bottom ash and foundry slag are well-graded
determine the index properties, soil classification, and coarse-grained sand-like materials, and thus are insensitive to
compaction characteristics. Tables 1 and 2 summarise the natural moisture content during compaction. Foundry sand is primarily

Site Soil name Specific D10 : D60 : Cu % Fines Classification WOPT : ªdmax :
gravity mm mm % kN/m3
USCS AASHTO

USH 12 Clayey sand* 2.69 NA 0.2 NA 34.45 SC A-2-4(0) 10.0 18.9

STH 131 Poorly graded sand with silt 2.62 0.09 0.35 3.9 6.54 SP-SM A-3(0) 8.0 18.0
STH 58 Clayey sand† 2.62 NA 0.25 NA 25.50 SC A-2-4(0) 8.5 19.8
STH 154 Poorly graded sand with silt 2.63 0.07 0.4 5.7 11.46 SP-SM A-3(0) 9.0 18.7
STH 60 (test section) Breaker run NM 0.25 29.0 116 3.12 GW A-1-a(0) None NM

Notes: NM ¼ not measured, NA ¼ not applicable.

* LL ¼ 24, PI ¼ 7.
† LL ¼ 24, PI ¼ 8.

Table 2. Properties of natural earthen materials and their classification: predominantly granular soils (USCS, unified soil classification
system; AASHTO, American Association of State Highway and Transportation Officials)

Site Soil name Specific D10 : D60 : Cu % Fines Classification WOPT : ªdmax :
gravity mm mm % kN/m3
USCS AASHTO

STH 60 (test section) Bottom ash* 2.65 0.06 1.9 31.7 13.23 SW A-1-b(0) None 15.1
Foundry sand† 2.55 0.0002 0.23 1150 28.92 SC A-2–7(2) 16.0 16.1
Foundry slag* 2.29 0.13 2.0 15.4 5.27 SW A-1-b(0) None 10.0

* Non-plastic.
† LL ¼ 44, PI ¼ 25.

Table 3. Properties of processed construction materials (industrial by-products) and their classification (USCS, unified soil
classification system; AASHTO, American Association of State Highway and Transportation Officials)

Site Soil name Chemical Stabilised soil Stabilised soil Soil

stabilised (no delay) (2 h delay) component
content:
% WOPT : ªdmax : WOPT : ªdmax : WOPT : ªdmax :
% kN/m3 % kN/m3 % kN/m3

STH 60 (test section) Fly-ash-stabilised soils 10 20.0 16.6 21.0 16.1 19.0 16.5
Scenic Edge development Fly-ash-stabilised soils 12 21.0 16.2 21.0 15.6 20.0 16.2
Gils Way development Lime-stabilised soils 5 NM NM NM NM 19.5 16.3

Note: NM = not measured.

Table 4. Properties of processed construction materials (chemically stabilised soils) and their classification

222 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
a mixture of fine sand and sodium bentonite (10% by weight) 6. SSG STIFFNESS AND DCP PENETRATION INDEX
that also contains small percentages of other additives. The (DPI)
foundry sand is sensitive to water content when compacted, Tables 5 and 6 summarise the results of the SSG and DCP
and exhibits a conventional compaction curve. measurements made. The mean SSG stiffness and DPI of
various materials are also illustrated graphically in Fig. 3.
Chemical stabilisation involved a mixture of natural soil and Fly-ash-stabilised soils have the highest mean stiffness, which
either fly ash or lime. The fly-ash-stabilised soil at the STH 60 increases with time of curing. For the lime-stabilised soil, the
test section and Scenic Edge development sites was prepared by mean stiffness after liming is nearly twice that of the untreated
mixing Class C fly ash with subgrade soil at its natural water subgrade. These results clearly indicate that the SSG can be
content (wetter than the optimum water content). Analysis from used to monitor increase in stiffness due to stabilisation
a series of mix designs evaluated in the laboratory indicated that reactions. In general, the granular earthen materials including
the subgrade soil stabilised using a fly ash content of 10% for breaker run are stiffer than fine-grained earthen materials.
the STH 60 test section site and 12% for the Scenic Edge Among three types of industrial by-product, foundry sand has
development site (on the basis of dry weight) provided sufficient the highest stiffness, with bottom ash and foundry slag having
strength and hence was adopted for field construction. The lime- nearly half its stiffness.
stabilised soil at Gils Way development in Cross Plains was
prepared by mixing 5% lime with subgrade soil at its natural Dynamic cone penetration is controlled primarily by the
water content (wetter than the optimum water content). strength of a material, and therefore DPI (amount of

Material type No. of Mean† Standard Coefficient of Standard Maximum Minimum

tests* deviation variation: % error

Natural earthen
Granular 18 7.8 1.88 24 0.44 12.1 5.4
Fine-grained 90 5.6 1.98 35 0.21 11.0 1.6
Industrial by-products
Bottom ash 4 3.9 0.20 5 0.10 4.1 3.7
Foundry sand 4 7.7 1.07 14 0.54 9.0 6.4
Foundry slag 18 3.1 1.03 33 0.24 4.8 1.5
Fly-ash-stabilised soils
24 hs 22 12.9 2.85 22 0.67 21.1 7.9
7–8 days 15 15.1 3.75 25 0.97 21.7 7.1
Lime-stabilised soils
Before liming 15 5.3 1.44 27 0.37 7.0 2.7
After liming 15 9.5 1.59 17 0.41 12.9 6.8
Other
Breaker run 8 6.7 1.19 18 0.42 8.8 4.5

* Corresponding to total number of test locations in the material category.

† Mean of SSG stiffness for total number of test locations in the material category.

Table 5. Results of soil stiffness gauge (SSG) measurements: stiffness (MN/m)

Material type No. of Mean† Standard Coefficient of Standard Maximum Minimum

tests* deviation variation: % error

Natural earthen
Granular 18 33.9 20.44 60 4.82 93.3 12.6
Fine-grained 81 44.7 27.89 62 3.10 170.0 12.2
Industrial by-products
Bottom ash 4 52.7 9.99 19 4.99 63.1 43.8
Foundry sand 3 59.1 16.59 28 9.58 69.8 40.0
Foundry slag 5 49.8 15.92 32 7.12 72.0 32.1
Fly-ash-stabilised soils
24 h 13 18.4 3.60 20 1.00 26.5 13.4
7–8 days 15 15.2 8.05 53 2.08 39.1 7.1
Lime-stabilised soils
After liming 15 13.7 1.94 14 0.50 17.3 11.6

Notes: DCP tests were not performed on lime-stabilised soils before liming and breaker run.
DPI was calculated by weighted average over a penetration depth of 152 mm.
* Corresponding to total number of test locations in the material category.
† Mean of weighted average DPI for total number of test locations in the material category.

Table 6. Results of DCP measurements: DCP penetration index (DPI) (mm/blow)

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 223
 16 60

SSG
14 DCP
50

12

DCP penetration index, DPI: mm/blow

40
SSG stiffness, KSSG: MN/m

10

8 30

6
20

10
2

0 0
Granular Fine- Bottom Foundry Foundry 24 .7 Before After Breaker
grained ash sand slag hrs days liming liming run

Natural Industrial Fly ash- Lime-

earthen by-products stabilised stabilised
soils soils

Fig. 3. Mean soil stiffness gauge stiffness and dynamic cone penetrometer penetration index (DPI) values of various materials

penetration per blow) is inversely proportional to shear (a) natural earthen materials (both granular and fine-grained
strength. The patterns exhibited by the DPI, in general, soils)
parallel those of the SSG stiffness in Fig. 3 with some (b) granular materials (natural soils, bottom ash and foundry
exceptions (for example, compare the relative stiffness and slag)
strength of industrial by-products and fine-grained soils). The (c) fine-grained (cohesive) soils
data in Tables 5 and 6 indicate that the standard error (d) fly-ash-stabilised soils
associated with DPI is considerably larger than that of the (e) fine-grained materials including fly-ash-stabilised soils
SSG stiffness. ( f ) all materials combined, including foundry sand.

7. CORRELATION BETWEEN SSG STIFFNESS AND Breaker run is not included because the DCP cannot be
DPI VALUES performed on this material. Foundry sand exhibits both
The correlation of SSG stiffness with DPI is examined based on granular and fine-grained material behaviour, so it is only
six material categories: included only in category ( f ). Table 7 summarises the results

Material category No. of Intercept Slope R2 Standard p-Value

points error

Natural earthen 79 17.9
7.5 0.60 1.44 0.00

Granular 27 19.3
8.3 0.55 1.73 0.00
Fine-grained soils 52 17.1
7.1 0.64 1.28 0.00
Fly-ash-stabilised soils 37 24.8
10.0 0.40 2.05 0.00
Fine-grained + fly-ash-stabilised soils 89 26.7
12.7 0.75 2.16 0.00
All combined (includes foundry sand) 119 25.6
12.0 0.72 2.15 0.00

Note: DPI was calculated by weighted average over a penetration depth of 152 mm from the surface.

Table 7. Parameters of linear regression analysis for relationship of SSG stiffness (MN/m) to log DPI (mm/blow)

224 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
of the linear regression analyses between SSG stiffness and log seismic test was adjusted, and the corresponding equation is
DPI in these material categories. Only those tests that were expressed as follows: 3
conducted at the same location are included in the analysis.
For all material categories in Table 7, the best correlations (i.e. 9
:
ESEIS
MOD ¼ 17:6 3 CBR0 64
highest R2 ) were obtained when DPI was averaged over a DCP
penetration depth of 152 mm after examining correlations of
average DPI calculated over varying DCP penetration depths. 18 Note that, for the sake of clarity, the modulus used in equation
It is also noted that, for natural subgrade soils, a better (9) is denoted as the modulus from the seismic test after
correlation was obtained after they were compacted, 18 adjustment to realistic stress and strain levels for the pavement
probably because of the more uniform conditions, which result (ESEIS-MOD ). ESEIS-MOD and CBR units are in MPa and percent
in a reduction of the dispersion of the data. Fig. 4 illustrates respectively.
the correlation between stiffness and DPI for all materials
combined. The SSG stiffness is related to DPI in a simple linear Another well-known relationship, which is widely used in
semi-logarithmic relationship. The dispersion of the data is North America, is the one proposed by Heukelom and Foster. 35
explainable to a degree by the fact that, although stiffness and It has been adopted by the American Association of State
strength are related in a general sense, there is not always a Highway and Transportation Officials (AASHTO) in the Guide
one-to-one relationship, as demonstrated in Fig. 3. for Design of Pavement Structures: 36
Nonetheless, the SSG stiffness and DPI correlate well, with an
R2 of 0.72. 10 ERM ¼ 10 3 CBR

8. CORRELATION BETWEEN MODULUS AND CBR

(ALSO DPI) where ERM is the modulus from the resilient modulus test, in
MPa.
8.1. Previous research
Empirical correlations between modulus (E) and CBR (also DPI) In addition to these two well-known relationships, Chen et al. 37
have been proposed by a number of researchers. A well-known suggested the following relationship between back-calculated
UK Transportation Research Laboratory (TRL) equation moduli from the FWD (EFWD ) and DPI:
developed between modulus (E) and CBR of the subgrade soil
:
has been given by Powell et al. 3 This equation has been 11 EFWD ¼ 338 3 DPI
0 39
established primarily from the comprehensive data relating
modulus measured by wave propagation to in situ CBR tests on
both remoulded and undisturbed subgrade soils. 34 After taking where EFWD and DPI are in units of MPa and millimetres per
into account the effects of the very low strain levels generated blow (mm/blow) respectively.
in the wave propagation technique, and other information
obtained from repeated load triaxial tests conducted at realistic Konrad and Lachance 20 presented a relationship between DPI
stress levels and in situ measurements of transient stress and using a 51 mm diameter cone and modulus of unbound
strain in experimental pavements, the modulus from the aggregates, gravelly, and sandy soils back-calculated from
plate load tests (EPLT ) by the following equation:

20 12 logð EPLT Þ ¼
0:884 logðDPIÞ þ 2:906

DPI depth: 0–152 mm

where DPI is the DCP penetration index in millimetres per blow

(mm/blow) using a 51 mm diameter cone and a 63.5 kg
15 hammer dropping 760 mm, and EPLT is expressed in MPa.
SSG stiffness, KSSG: MN/m

KSSG 5 25·6–12 log DPI Livneh and Goldberg 38 carried out comparative German light
drop weight (LDW) and DCP tests. The relationship between the
R2 5 0·72
10 modulus measured by the LDW (ELDW ) and the in situ CBR
values obtained from the DCP is expressed as follows for
clayey and sandy soils respectively:

5 300
13 ELDW ¼ 600 3 ln
300
6:019 3 CBR(1=1:41)

All materials combined

0 300
1 10 100 1000 14 ELDW ¼ 600 3 ln
DPI: mm/blow
300
4:035 3 CBR(1=1:41)

Fig. 4. Relationship between SSG stiffness and DPI for all

materials combined where ELDW and CBR are in units of MPa and percent
respectively.

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 225
 8.2. Development of empirical correlations
200
After a simple linear semi-logarithmic relationship between y 5 1·75 1 1·06x R2 5 0·740
SSG stiffness (KSSG ) and DPI values was determined (i.e. KSSG ¼
25.6
12 log DPI) based on direct regression from the actual
measured data for all materials combined (Fig. 4), such a
150
correlation can be further developed to become a more

SSG modulus, ESSG: MPa

meaningful and useful equation, which can be used in the
design of pavements. To accomplish this, the measured SSG
stiffness is converted to the SSG modulus (ESSG ) of the 100
pavement materials using equation (1). Values for Poisson’s
ratio (
) of the granular soils including foundry slag and
bottom ash were assumed to be 0.40. For fly-ash-stabilised
soils and fine-grained soils including foundry sand,
values of 50
0.25 and 0.35 were used. These values were selected according
to their typical ranges suggested by Huang. 39 Similarly, the
weighted average DPI value obtained from the DCP can be
converted to the CBR of the pavement materials using the well- 0
established correlation given in equation (8) with the 0 50 100 150 200
DCP modulus, EDCP: MPa
coefficients Æ and  of 2.46 and
1.12 respectively. Fig. 5
illustrates a plot of calculated SSG modulus (ESSG ) against CBR
from the DCP. The regression equation obtained is expressed as Fig. 6. Comparison between moduli from the SSG and those
from the DCP
follows:

:
15 ESSG ¼ 18:77 3 CBR0 63 R2 ¼ 0:74 (EDCP ) using the regression equations given in equations (8) and
(9). Remarkably good agreement is obtained with an
independent, widely used approach.
where the units of ESSG and CBR are MPa and percent
respectively. It can be seen that a unique relationship exists A regression equation obtained in this study is also compared
between ESSG and CBR, regardless of soil type and site, with that obtained from different in-situ tests including the
although the coefficient of correlation of equation (15) is FWD, LWD and plate load tests, as shown in Fig. 7. The
inherently dependent on the coefficient of correlation of equations obtained for the FWD, LWD and plate load test are
equation (8). The results of SSG tests and DCP tests are given by Chen et al., 37 Livneh and Goldberg, 38 and Konrad and
expected to be affected by the same factors (i.e. relating the Lachance 20 respectively. Note also that the parameter DPI in
two test results directly excludes the influence of water equations (11) and (12) can be converted to CBR using
content, dry density, and other basic indices). Furthermore, equation (8) with the coefficients Æ and  of 2.46 and
1.12
such a relationship is not affected by change in pavement respectively. The comparison results suggest that the regression
condition, and is also applicable to both the as-compacted and equation obtained in this study is best correlated to the
post-construction states. This equation yields almost identical equation given by Powell et al. (i.e. equation (9)), established
values to those obtained from the equation given by Powell et between modulus and CBR. 3 Within CBR values ranging from
al.3 —that is, equation (9). Fig. 6 illustrates the comparison 0% to 20%, the FWD and LWD (on sand) tests, respectively,
between moduli from the SSG (ESSG ) and those from the DCP provide the highest and lowest E for a given CBR. The

200
175
AASHTO36
Powell et al.3
150 SSG
150 FWD-clay37
SSG modulus, ESSG: MPa

LDW-clay }38
125 LDW-sand
Plate load t est:
Modulus, E: MPa

aggregates and sand20

100 100
0·63
ESSG 5 18·77 CBR
75
R2 5 0·74
50
50

25
0
0 5 10 15 20 25 30 35
0
CBR from DCP: % 0 5 10 15 20
CBR: %

Fig. 5. Correlation between modulus from the SSG and CBR

from the DCP Fig. 7. E–CBR relationship of different methods

226 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
relationship from the plate load test is in between the curve of equation fit reasonably well to the data set from STH 12 and
equation (9) and the LDW test (on sand). The suggested STH 32, and thus equation (15) appears to be useful for
equation by AASHTO 36 gives the highest modulus when the estimating the subgrade modulus. Additionally, the equation
CBRs are greater than 10%. Large deviation from the other suggested by AASHTO 36 overestimates the subgrade modulus
modulus tests is observed at relatively high CBRs as well. Using for soils with relatively high CBRs.
Fig. 7, the modulus can be estimated for any pavement
condition if CBR is obtained for the corresponding condition. It 9. CORRELATION BETWEEN MODULUS TEST AND
is perhaps not surprising that the modulus given in equation UNCONFINED COMPRESSION TEST
(9) by Powell et al. and the SSG modulus given in equation In addition to the correlation with CBR from the DCP, the
(15) as developed in this study as a function of CBR agree well. modulus from the SSG can be correlated with the strength from
The reason might be that both moduli were adjusted to realistic the conventional unconfined compression test. Lee et al. 5
strain amplitude and stress levels from those corresponding to suggested an empirical correlation between the modulus from
the wave propagation technique. According to Powell et al., the resilient modulus test (AASHTO T274-82) 40 (ERM ) and the
their moduli were adjusted to strain amplitude and stress level stress causing 1% axial strain (Su1:0% ) in the conventional
in pavements from those corresponding to the wave unconfined compressive test (ASTM D2166) 41 for the cohesive
propagation technique. In the case of the SSG, it was shown by soils sampled from five in-service subgrades. The relationship
Sawangsuriya et al.6 that the SSG modulus corresponds to is as follows:
strain amplitudes larger than the strain amplitudes of the wave
propagation technique, even though the SSG induces strain ERM ¼ 10 748:4 þ 5744:9Su1:0%
48S2u1:0%
16
amplitudes comparable to that of the wave propagation
technique. In fact, the stress and strain levels induced by the
SSG are 2 kPa and less than 10
3 % respectively. It has also where the units of ERM and Su1:0% are both kPa. Note that the
been shown that a somewhat reduced modulus is reported by ERM values used to develop equation (16) are at an axial
the internal computation of the SSG device.6 deviator stress (1
3 ) of 41.4 kPa and a confining stress (3 )
of 20.7 kPa. Su1:0% was found to have the best correlation with
8.3. Model validation ERM , compared with other variables (i.e. in-service water
To verify the power model given in equation (15), field content and dry density), and was chosen as a predictor
measurement data from another two test sites were plotted variable instead of the unconfined compressive strength
onto the developed power model and Powell’s equation, as because its strain level is comparable to those of the resilient
shown in Fig. 8. The test sites were a section of State Trunk modulus test, and the stresses at smaller axial strains may have
Highway (STH) 32 located in Port Washington, and a section of a larger error due to incorrect readings or imperfect contact
STH 12 located between Cambridge and Fort Atkinson, both in between the specimen and top cap.5
Wisconsin, USA. The SSG and DCP data from these sites were
obtained recently, and were not included in the development of In this study, the moduli from the SSG (ESSG ) conducted on
equation (15). The predominant subgrade soil of STH 32 is subgrade soils at STH 60 (test section) are correlated with the
clayey sand classified as SC and A-4(0) according to the Su1:0% values from the conventional unconfined compression
Unified Soil Classification System (USCS) and the AASHTO test. Unconfined compression tests were conducted following
classification system respectively. The predominant subgrade ASTM D2166 41 on undisturbed specimens (50 mm in diameter
soil of STH 12 comprises lean clay with sand (CL or A-6(15)) and 100 mm high) trimmed from the tube samples. The test
and clayey sand with gravel (SC or A-2–6(0)). Results show was performed using a strain rate of 2%/min, and the stress at
that both the developed model (equation (15)) and Powell’s about 1% axial strain was reported. A regression analysis was
conducted to obtain a relationship between ESSG and Su1:0% ,
and the coefficient of determination (R2 ) for this relationship is
250 0.64. A plot of E against Su1:0% was made, as shown in Fig. 9,
This study in order to compare the correlation results obtained with
Powell et al.3
STH 32-Subgrade equation (16). At a similar Su1:0% , the ESSG value is higher than
200 STH 12-Subgrade ERM obtained from equation (16), and, as Su1:0% increases, the
difference between ESSG and ERM also increases. This difference
AASHTO36
Modulus, E: MPa

150
may be attributed to the fact that the ERM values used in
equation (16) are at higher stress and hence higher
corresponding strain levels, whereas the ESSG are measured at
100 much lower stress–strain levels. 6

10. SUMMARY AND CONCLUSIONS

50
Soil stiffness gauge and DCP survey data of natural earthen
materials, industrial by-products, chemically stabilised soils,
0 and other materials from ten construction sites around the state
0 5 10 15 20 25 30
CBR: %
of Wisconsin are presented, along with their correlation with
each other. The data display considerable dispersion,
characteristic of the conditions and sampling sizes typical
Fig. 8. Comparison of predicted subgrade moduli at STH 32,
Wisconsin during construction. SSG provides in-situ, near-surface soil
stiffness averaged over a limited zone, whereas DCP provides

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 227
 The study indicates that either or both devices show good
100
potential for future use in pavement and subgrade property
evaluation during the construction phase. The in-situ stiffness
80 and strength properties of various materials can be rapidly and
ESSG 5 9·60 1 1·39(Su1·0%) directly monitored in companion with the conventional
R2 5 0·64 compaction control tests (i.e. nuclear density or laboratory
Modulus: MPa

60 moisture content samples) during earthwork construction.

Stiffness and strength are material properties that are needed in
different phases of highway design, both for long-term
40
pavement performance and for working platform support and
Lee et al .5
.

stability during construction. For post-construction conditions,

20 the modulus can be monitored using DCP and the correlations
provided, as direct access for SSG is not convenient. The
experience both with recycled and reclaimed materials and
0 with chemically stabilised soils is limited compared with
0 10 20 30 40 50 60 70
Su at 1·0% strain: kPa natural earthen materials in terms of moisture–density
relationships and the related mechanical behaviour. Direct
Fig. 9. Relationship between E and Su1:0% for undisturbed monitoring of stiffness and strength of these new materials
subgrade soils at STH 60 (test section), Wisconsin using these two devices also appears to be as effective as in
natural earthen materials.

individual points of an index of in-situ shear strength

11. ACKNOWLEDGEMENTS
expressed as DCP penetration index (DPI) as a function of
This research was funded by the Wisconsin Department of
depth. Therefore each relates to a fundamentally different
Transportation (WisDOT), USA. The contents of this paper are
material property, and is presented in a spatially different
those of the authors, and do not reflect the opinions or policies
manner. To deal with the latter issue, a weighted average of
of WisDOT. The senior author gratefully acknowledges
DPI over the depth of measurement is employed to obtain a
financial support through a scholarship from the Royal Thai
representative strength index of the material. This approach
Government. The authors acknowledge Robert Allbright, David
provided better correlations than the arithmetic average. It is
Staab, Mark Fredrickson, and Bert Trzebiatowski for their
also noted that the standard error associated with DPI is
assistances with the laboratory and field tests, and Drs Craig
considerably larger than that of the SSG stiffness, which
Benson and Peter Bosscher for their suggestions. Thanks are
reflects the nature of the two tests.
due to Dr Aykut Senol for providing the unconfined
compression test data. Finally, the authors also express their
A simple linear semi-logarithmic relationship is observed
sincere gratitude to Robert Arndorfer and other WisDOT
between SSG stiffness and DPI. DPI weighted average over a
personnel for their cooperation in arranging access to the sites.
DCP penetration depth of 152 mm yields the highest coefficient
of determination, and also yields a statistically significant
relationship between the SSG stiffness and DPI for most REFERENCES
materials. This depth is consistent with the significant depth of 1. PINARD M. I. Innovative compaction techniques for
measurement for the SSG, as shown in previous studies. The improving the bearing capacity of roads and airfields.
relationship is indicative of the fact that, although there is not Proceedings of the 5th International Conference on the
always a one-to-one relationship, stiffness and strength are Bearing Capacity of Roads and Airfields, Trondheim, 1998,
related in a general sense. 3, 1471–1480.
2. FLEMING P. R. Recycled bituminous planings as unbound
The results of the regression analysis show that there is a granular materials for road foundations in the UK.
significant correlation between the CBR obtained from the DCP Proceedings of the 5th International Conference on the
and the modulus obtained from the SSG. A power model Bearing Capacity of Roads and Airfields, Trondheim, 1998,
developed between these two properties is found to be in good 3, 1581–1590.
agreement with the well-known equation given by Powell et 3. POWELL W. D., POTTER J. F., MAYHEW H. C. and NUNN M. E.
al.3 An equation suggested by AASHTO 36 tends to overestimate The Structural Design of Bituminous Roads. Transportation
the pavement modulus for relatively high CBRs. The modulus and Road Research Laboratory, Crowthorne, 1984, TRRL
from the SSG is also correlated well with the axial stress at 1% Laboratory Report 1132.
strain in the conventional unconfined compression test. 4. BROWN S. F. The relationship between California bearing
ratio and elastic stiffness for compacted clays. Ground
The modulus and CBR are affected in a similar manner by Engineering, 1990, 23, No. 8, 27–31.
changes of pavement condition, and the relationship between 5. LEE W., BOHRA N. C., ALTSCHAEFFL A. G. and WHITE T. D.
modulus and CBR may be similar for a given soil under any Resilient modulus of cohesive soils and the effect of freeze-
condition if the soil was initially compacted by the same thaw. Canadian Geotechnical Journal, 1995, 32, No. 4,
method. Therefore the modulus for the as-compacted or the 559–568.
post-construction states may be estimated from Fig. 7, if CBR is 6. SAWANGSURIYA A., EDIL T. B. and BOSSCHER P. J.
measured or estimated from DCP on the subbase and subgrade Relationship between soil stiffness gauge modulus and
materials being subjected to the same conditions. other test moduli for granular soils. Transportation

228 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil
 Research Record 1849, TRB, National Research Council, 21. LIVNEH M., ISHAO I. and LIVNEH N. A. Effect of vertical
Washington, DC, 2003, pp. 3–10. confinement on dynamic cone penetrometer strength
7. LIVNEH M. Validation of correlations between a number of values in pavement and subgrade evaluations.
penetration tests and in situ California bearing ratio tests. Transportation Research Record 1473, Transportation
Transportation Research Record 1219, TRB, National Research Board, National Research Council, Washington,
Research Council, Washington, DC, 1989, pp. 56–67. DC, 1995, pp. 1–8.
8. SAWANGSURIYA A., BOSSCHER P. J. and EDIL T. B. Laboratory 22. KLEYN, E. G. The Use of the Dynamic Cone Penetrometer
evaluation of the soil stiffness gauge. Transportation (DCP). Transvaal Roads Department, South Africa, 1975,
Research Record 1808, TRB, National Research Council, Report No. 2/74.
Washington, DC, 2002, pp. 30–37. 23. HARISON J. A. Correlation between California bearing ratio
9. EGOROV K. E. Calculation of bed for foundation with ring and dynamic cone penetrometer strength measurement of
footing. Proceedings of the 6th International Conference on soils. Proceedings of the Institution of Civil Engineers, Part
Soil Mechanics and Foundation Engineering, Montreal, 2, 1987, 83, Technical Note No. 463, 833–844.
1965, 2, 41–45. 24. LIVNEH M. Validation of correlations between a number of
10. WU W., ARELLANO M., CHEN D.-H., BILYEU J. and HE R. penetration tests and in situ California bearing ratio tests.
Using a Stiffness Gauge as an Alternative Quality Control Transportation Research Record 1219, Transportation
Device in Pavement Construction. Texas Department of Research Board, National Research Council, Washington,
Transportation, Austin, TX, 1998. DC, 1987, pp. 56–67.
11. CHEN D.-H., WU W., HE R., BILYEU J. and ARRELANO M. 25. MCELVANEY J. and DJATNIKA B. I. Strength evaluation of
Evaluation of in situ resilient modulus testing techniques. lime-stabilised pavement foundations using the dynamic
In Recent Advances in the Characterization of cone penetrometer. Australian Road Research, 1991, 21,
Transportation Geo-Materials. GSP No. 86, ASCE, 1999, No. 1, 40–52.
pp. 1–11. 26. WEBSTER S. L., GRAU R. H. and WILLIAMS T. P. Description
12. SCALA A. J. Simple methods of flexible pavement design and Application of Dual Mass Dynamic Cone Penetrometer.
using cone penetrometers. New Zealand Engineering, 1956, US Army Engineers Waterways Experimental Station,
11, No. 2, 34–44. Vicksburg, MS, 1992, Instruction Report GL-92-3.
13. KLEYN E. G., MAREE J. H. and SAVAGE P. F. Application of a 27. LIVNEH M. and LIVNEH N. A. Subgrade strength evaluation
portable pavement dynamic cone penetrometer to with the extended dynamic cone penetrometer. Proceedings
determine in situ bearing properties of road pavement of the 7th International Association of Engineering Geology
layers and subgrades in South Africa. Proceedings of the Congress, Lisbon, Portugal, 1994, Part 1, p. 219.
2nd European Symposium on Penetration Testing, 28. SIEKMEIER J. A., YOUNG D. and BEBERG D. Comparison of the
Amsterdam, The Netherlands, 1982, 277–282. dynamic cone penetrometer with other tests during
14. CHUA K. M. Determination of CBR and elastic modulus of subgrade and granular base characterization in Minnesota.
soils using a portable pavement dynamic cone In Nondestructive Testing of Pavements and Back
penetrometer. Proceedings of the 1st International Calculation of Moduli. American Society for Testing and
Symposium on Penetration, Orlando, 1988, 407–414. Materials, West Conshohocken, PA, 1999, STP 1375,
15. NEWCOMB D. E., CHADBOURN B. A., VAN DEUSEN D. A. and 175–188.
BURNHAM T. R. Initial Characterization of Subgrade Soils 29. CHEN D.-H., WANG J.-N. and BILYEU J. Application of the
and Granular Base Materials at the Minnesota Road dynamic cone penetrometer in evaluation of base and
Research Project. Minnesota Department of Transportation, subgrade layers. Transportation Research Record 1764,
St Paul, MN, 1996, MN/RC 96-19. Transportation Research Board, National Research Council,
16. SYED I. and SCULLION T. In-place engineering properties of Washington, DC, 2001, pp. 1–10.
recycled and stabilized pavement layers. Proceedings of the 30. ASTM. Standard Test Method for Measuring Stiffness and
5th International Conference on the Bearing Capacity of Apparent Modulus of Soil and Soil-Aggregate In-Place by
Roads and Airfields, Trondheim, 1998, 3, 1619–1630. an Electro-Mechanical Method. American Society for
17. SAARENKETO T., SCULLION T. and KOLISOJA P. Moisture Testing and Materials, West Conshohocken, PA, 2002,
susceptibility and electrical properties of base course D6758.
aggregates. Proceedings of the 5th International Conference 31. ASTM. Standard Test Method for Use of the Dynamic Cone
on the Bearing Capacity of Roads and Airfields, Trondheim, Penetrometer in Shallow Pavement Applications. American
1998, 3, 1401–1410. Society for Testing and Materials, West Conshohocken, PA,
18. EDIL T. B. and SAWANGSURIYA A. Investigation of the DCP 2003, D6951.
and SSG as Alternative Methods to Determine Subgrade 32. EDIL T. B., BENSON C. H., BIN-SHAFIQUE S., TANYU B. F., KIM
Stability. Wisconsin Department of Transportation, W.-H. and SENOL A. Field evaluation of construction
Madison, WI, 2004, Report No. 0092-45-18. alternatives for roadway over soft subgrade. Transportation
19. ALLBRIGHT R. L. Evaluation of the Dynamic Cone Research Record 1786, Transportation Research Board,
Penetrometer and its Correlations with Other Field National Research Council, Washington, DC, 2002, pp.
Instruments. MS thesis, Department of Civil and 36–48.
Environmental Engineering, University of Wisconsin- 33. ASTM. Standard Test Methods for Laboratory Compaction
Madison, 2002. Characteristics of Soil Using Standard Effort (12,400 ft-lbf/
20. KONRAD J.-M. and LACHANCE D. Use of in situ penetration ft3 (600 kN-m/m3 )). American Society for Testing and
tests in pavement evaluation. Canadian Geotechnical Materials, West Conshohocken, PA, 2000, D698.
Journal, 2001, 38, No. 5, 924–935. 34. JONES R. In situ measurements of the dynamic properties of

Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil 229
 soil by vibration methods. Géotechnique, 1958, 8, No. 1, formation and foundation construction: use of falling-
1–21. weight deflectometer and light drop weight. Transportation
35. HEUKELOM W. and FOSTER C. R. Dynamic testing of Research Record 1755, Transportation Research Board,
pavements. Journal of the Soil Mechanics and Foundations National Research Council, Washington, DC, 2001, pp.
Division, ASCE, 1960, 86, No. SM1, 1–28. 69–77.
36. AASHTO. Guide for Design of Pavement Structures. 39. HUANG Y. H. Pavement Analysis and Design. Prentice Hall,
American Association of State Highway and Transportation Englewood Cliffs, NJ, 1993.
Officials, Washington, DC, 1993. 40. AASHTO. Standard Method of Test for Resilient Modulus of
37. CHEN J., HOSSAIN M., and LATORELLA T. M. Use of falling Subgrade Soils. American Association of State Highway
weight deflectometer and dynamic cone penetrometer in and Transportation Officials, Washington, DC, 1986,
pavement evaluation. Transportation Research Record T 274-82.
1655, Transportation Research Board, National Research 41. ASTM. Standard Test Method for Unconfined Compressive
Council, Washington, DC, 1999, pp. 145–151. Strength of Cohesive Soil. American Society for Testing
38. LIVNEH M. and GOLDBERG Y. Quality assessment during road and Materials, West Conshohocken, PA, 2000, D2166.

What do you think?

To comment on this paper, please email up to 500 words to the editor at journals@ice.org.uk
Proceedings journals rely entirely on contributions sent in by civil engineers and related professionals, academics and students. Papers
should be 2–5000 words long, with adequate illustrations and references. Please visit www.thomastelford.com/journals for author
guidelines and further details.

230 Geotechnical Engineering 158 Issue GE4 Evaluating stiffness and strength of pavement materials Sawangsuriya • Edil

View publication stats