Compaction Evaluation by MASW Surveys (CEMS) for Road Construction

The main and common purpose of the compaction process applied at various stages of road construction is to achieve the level of stiffness necessary
to sustain expected stress over the entire construction area.  In this sense, the compaction evaluation can be regarded identical to in-situ stiffness
measurement of road materials.  

Stiffness of a material is defined as a measure of resistance to deformation (Sheriff, 2002) and ultimately related to material's elastic moduli that
describe the material's behavior under stress.  Among the three primary types of moduli—Young's (E), shear (µ), and bulk (B)—the first two (E and µ)
are most commonly used because of what they represent.  Young's modulus (E) simply dictates the deformation tendency along the axis of stress,
whereas the shear modulus (µ) indicates the tendency of shape deformation (i.e., shearing).  In reality, deformation always accompanies both
transverse and longitudinal changes only at a different ratio.  In this sense, the most comprehensive and accurate definition of stiffness should include
both moduli of E and µ.  According to the theory of elasticity (
Sheriff and Geldart, 1982), these two moduli can be defined by a material's density (r)
and the two seismic velocities (or by Poisson's ratio, P) of Vp (P-wave) and Vs (S-wave):

    E = 2rVs**2(1+P)                        (1)        
    µ = rVs**2                                   (2)

The two defining equations indicate the heaviest dependency of both moduli on Vs.  This is why seismic shear-wave velocity (Vs), the final product
from an MASW survey (
Park et al., 1999), is often used as a direct indicator of a material's stiffness.

Minnesota Department of Transportation (MnDOT) recently recognized the potential utility of MASW surveys in compaction evaluation during road
construction and launched a feasibility field study to tap into its effectiveness and move toward the goal of making it a routine production method.  The
pilot study consisted of a series of multiple (5) MASW surveys performed in the same area during a full-depth-reclamation (FDR) road construction; a
500-ft long segment on TH56 South approximately 5 miles north of Kenyon, MN (Figure1).
Full-Depth Reclamation (FDR) and MASW Surveys

The full-depth reclamation (FDR) rebuilds old worn-out asphalt
pavements by
recycling the existing roadway.  The old asphalt and
base materials are pulverized, mixed with stabilizing agent (e.g.,
cement, emulsified asphalt, water, etc.), and compacted to produce a
new road base.  The new pavement layer of asphalt (or concrete) is
then laid on top of the new base.  The overall procedure can be
divided into
three stages; (1) the pre-grind (PG) stage in which
existing old pavement and base layers are pulverized to be reclaimed
at a nearby place, (2) stabilized FDR stage (
SFDR) in which another
phase of reclamation is made by adding stabilizing agent, and (3) the
final stage of hot-mix-asphalt (
HMA) lift.  During and between each
stages, it is important to ensure the necessary level of stiffness is
achieved over the entire area of construction.  Standard penetration
test (SPT) is traditionally conducted for this purpose at selected
locations, while more recently
Intelligent Compaction (IC) techniques
are used.  MASW surveys are proposed as a more advanced QA/QC
means that can provide distribution of stiffness in more technically
appropriate and spatially continuous form than any other approaches
can do.
(Below) Figure 1.  Quadruple land streamers used to acquire MASW data at the CEMS
test site near Kenyon, MN.
Data Acquisition

stage (SFDR and SFDRb), and two during the final HMA stage (HMA1stLift and HMA2ndLift).
The surveys took place during July and August, 2013, by using the existing seismic acquisition system built and used by MnDOT Materials with a
minimal modification, a seismic system that employed low-frequency geophones (4.5-Hz) for receivers and was originally built for subsurface
investigation at deeper depths of soil and bedrock (e.g., 1-30 m) than the current depth of interest (e.g., 0-2 m).  The system consisted of 48-channel
acquisition with quadruple land streamers (12channels/streamer) equipped with 4.5-Hz geophones and placed parallel and separated by 4 ft (Figure
1).  A weight-drop source (WD/SASS) generated surface waves 2-m ahead of the closest geophones from the transverse center of the streamers.  
One impact was delivered to generate one 48-channel field record at one location, and this source-receiver (SR) configuration moved by 1 m each
time to produce a total of 154 field records per survey ensuring the coverage of 500-ft (153 m) long segment of the test site.

Resolution Analysis

Considering the acquisition geometry of receiver array length (L=11 m) and
source offset (X1=2 m), the maximum investigation depth (Zmax) is expected
to be about 5 m.  However, most stiffness changes that will occur during the
FDR construction is confined within top layer of about 0.3-m thickness.  
Therefore, it is worth conducting an analysis of how the stiffness variations in
this relatively thin layer, as well as other underlying layers, influence on the
dispersion curve measurement.  This analysis is performed by modeling an
apparent-mode (AM0) dispersion curve (
Gucunski and Woods, 1992) for a
velocity (Vs) model that can represent a typical road base (Vs1=300 m/sec)
of 0.3-m thickness overlying a subgrade (Vs2=150 m/sec) followed by
weathered bedrock (Vs3=500 m/sec) at an arbitrary depth of 2.0 m (Figure
2a).  The modeled AM0 curve is displayed (in black color) in Figure 2b in
comparison to three other AM0 curves that are modeled after changing
(increasing) the velocity (Vs) by 30% in each of the three layers; i.e., Vs1,
Vs2, and Vs3, respectively.  The upper limit of modeled frequency is 1000 Hz
where the AM0 curves approach to the asymptotic surface-wave velocity of
top layer by more than 95%.  The curve comparison shown in Figure 2b
indicates that velocity (Vs) changes in the three layers result in phase
velocity changes mostly at those frequencies higher than 100 Hz (Vs1),
30-200 Hz (Vs2), and lower than 20 Hz (Vs3), respectively.  It is also obvious
that the greatest overall change occurs when the top layer changes its
velocity, proving the highest sensitivity due to the shallowest (depth despite
the smallest thickness).  These results therefore indicate that, as far as the
highest measured frequency (fmax) exceeds 100 Hz (and the lowest
frequency is lower than 20 Hz), velocity change in any of these three layers
will be detected.  However, velocity for top layer (Vs1) will be underestimated
while fmax remains lower than 1000 Hz.  Figure 2b indicates that, as fmax
increases and approaches to 1000 Hz, the degree of underestimation will
decrease, while the sensitivity in velocity (Vs1) change will increase.
(Below) Figure 2.  (a) Layer model used to generate (b) model apparent-mode
(AM0) dispersion curves.  The original AM0 curve is displayed in (b) in
comparison to other AM0 curves generated by increasing velocity (Vs) by
30% for top (Vs1), subgrade (Vs2), and bedrock (Vs3) layers.  
Data Analysis and Results

individual lines (1-4) (Figure 1) corresponding to each land streamer of 12-
channel acquisition during pre-processing.  Each subset then went through
the normal MASW data analysis sequence.  Different offline offsets of source
locations for each line were accounted for during the stage of dispersion
analysis.  Then, dispersion images were generated for a frequency range of
1-1000 Hz (0.5-Hz increment) and a phase-velocity range of 10-1500 m/sec
(5 m/sec increment).  Dispersion curves were next extracted from these
images in an approximate common frequency range of 15-200 Hz.  
Dispersion images for this frequency range are displayed in Figure 3 for the
field record obtained at the same surface location (in line 1) during the five
(5) MASW surveys.  

A set of dispersion curves for each line was then used for inversion analysis
to produce a 2-D velocity (Vs) map of 2-m depth, which was set intentionally
smaller than the optimum depth (e.g., 5 m) to increase the resolution at
shallower depths (e.g., ≤ 1 m).  A 15-layer earth model of varying thicknesses
was used during the inversion.  In this way, four (4) cross section Vs maps
were produced from each survey for the four parallel lines (1-4).  The Vs
cross sections for line 1 are displayed in Figure 4 for all five (5) surveys.

Considering the four (4) lines of 2-D Vs maps being located on the surveyed
area side by side with an even spacing (4-ft) between them, it is possible to
construct depth-slice (DS) maps by combining Vs data sets from all four
lines.  In this way, a DS map for 0.0-0.30 m depth range was created for each
survey from four (4) lines of 2-D Vs maps.  These velocity (Vs) DS maps are
displayed in Figure 5 for all five (5) surveys.  

Then, these DS maps are converted to Young's (E) and shear (µ) moduli
values by using the two equations of (1) and (2) above.  Corresponding DS
maps are displayed in Figures 6a and 6b, respectively.  A constant density (r)
of 2000 kg/m3 and also a constant Poisson's ratio (P) of 0.4 were used
during the conversion.


From a series of multiple (5) field surveys executed in the same area at five
different stages of FDR road construction, a feasibility study was executed to
tap into the possibility of using a seismic method (MASW) to evaluate the
compaction degree of road beds (subgrade, base, and pavement layer) by
using a pre-existing seismic acquisition system with minimal modification—a
system that employed low-frequency geophones (4.5-Hz) for receivers and
was originally built for subsurface investigation at deeper depths of soil and
bedrock (e.g., 1-30 m) than the current depth of interest (e.g., 0-2 m).

Although geophones normally have the most sensitive frequency response
within a few to a few hundred hertz (e.g., 5-200 Hz) that are most effective
with MASW in sampling materials at depths of a few to a few tens of meters (e.
g., 1-30 m), they were used in this study mainly to investigate the top few
meters of subsurface by shortening the geophone-array length to focus into
relatively shallower depths than normally used.

The acquired data showed surface waves clearly identifiable in approximately
15-200 Hz that were used to evaluate stiffness distribution within the top 2.0-
m of road beds.  Despite the reduced resolution in measurement due to
limitations in data acquisition (i.e., geophones), it seems that results
successfully showed relative variations in stiffness for the top 2-m range that
were expected between different stages of road construction as well as
between different surface locations.  This is an unprecedented approach that
deals with the most important property of road materials (i.e., stiffness)
through one of the most fundamental scientific approaches (i.e., seismic-
wave propagation).

Considering the possible range of shear-wave velocities (Vs's) for base
materials (e.g., 200 m/sec ≤ Vs ≤  500 m/sec) and bituminous pavement (e.g.,
1,000 m/sec ≤ Vs ≤  2,000 m/sec), and also possible thickness (H's) ranges
(e.g., 0.1 m ≤ H ≤ 0.5 m for base, and 0.05 m ≤ H ≤ 0.30 m for pavement), the
optimum frequency ranges necessary for absolute evaluation of each layer's
velocity are calculated as 500 Hz-5,000 Hz for the base and 5,000 Hz -
30,000 Hz for the pavement, respectively, based on the range of velocities
and thicknesses considered.  These are the conditions necessary so that
surface waves can have as short wavelengths as possible to focus into each
layer itself without being influenced by other materials existing below.  
Therefore, the evaluation results from this series of five (5) field surveys
represent underestimated velocities for base and pavement layers, especially
for the pavement, due to the lack of sufficiently-high frequencies during


The current acquisition setup using geophones is essential to investigate the
subgrade.  The current seismic source has been excellent in producing
surface waves with the necessary spectral characteristics for this purpose (i.
e., strong surface waves in 10-150 Hz).  A possible improvement may be
made by using a closer geophone spacing (dx) (e.g., dx = 0.5 m) than the
current 1-m dx.  If the current configuration of 12channels per streamer is to
be maintained, then the total length of receiver arrays per streamer will be
reduced and maximum investigation depth (Zmax) will be accordingly reduced
to Zmax ~2.5 m from the current value of Zmax ~ 5 m.  This reduced receiver
spacing (dx) by itself will increase the maximum frequency of measured
dispersion curves by, for example, two times if dx is shortened to its half.  In
addition, switching to higher-frequency geophones (e.g., 40-Hz phones or
100-Hz phones) will increase overall recording sensitivity at higher
frequencies (e.g., 100-1,000 Hz) that can ultimately improve the resolution,
especially for top about 1.0 m depth range.  

Modifying the geophone configuration with possibly higher-frequency
geophones will require a series of field tests at a typical site for road
construction.  It is recommended at this moment, however, that the next field
tests be performed using current geophones (4.5-Hz) at much tighter spacing
(e.g., dx = 0.1 m versus 1.0 m).  This will examine the highest frequency
measurable without other harmful interference such as spatial aliasing
effects.  It will then be obvious whether the higher-frequency geophones are
critically needed or not.

The current acquisition setup using geophones is not suitable for accurate
evaluation of base and pavement layers (i.e., absolute evaluation of Vs).  
Accelerometers have to be used that can record surface waves up to 50,000
Hz (50 KHz).  Their sensitivity decreases rapidly at low frequencies (e.g., ≤
100 Hz) in comparison to geophones, and therefore they are not suitable for
subgrade investigation.

The current data analysis sequence requires an operator's continuous
involvement at each stage of processing, which in turn requires successive
decision-making steps for optimum parameters before moving on to next step
of processing.  Most of these steps should eventually be fully automated
within the analysis software, minimizing the need for operator intervention.  
For this to happen, configuration of an optimum acquisition system should
ensure recording of surface waves with a high signal-to-noise ratio (SN) in
the signal frequency bands.  This fully-automated software will lead to a
complete system in the field that will produce cross sections and depth-slice
maps of stiffness in real-time mode as field survey proceeds.
(Below) Figure 3.  Dispersion images from field records obtained at the same
location (in line 1) during the five (5) MASW surveys.  They show typical
changes in dispersion trend over different stages of construction for the
frequency range used to extracted measured dispersion curves (AM0's).
(Below) Figure 4.  Velocity (Vs) cross sections of line 1 from the five (5)
MASW surveys.  
(Right) Figure 5.  Velocity (Vs) depth-slice (DS) maps from the five (5) MASW surveys
for 0.0-0.3 m depth.  Each DS map is constructed from four (4) parallel velocity (Vs)
cross sections.  

(Below) Figure 6.  Velocity (Vs) DS maps in Figure 5 are converted to (a) Young's (E)
and (b) shear (µ) moduli maps by using constant values of density (r) and Poisson's
ratio (P).