Content uploaded by Hung Manh La
Author content
All content in this area was uploaded by Hung Manh La on Sep 19, 2015
Content may be subject to copyright.
5
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of
CONCRETE BRIDGE DECKS USING MULTIPLE
NDE TECHNOLOGIES
N. Gucunski 1, A. Maher 2, B. Basily1, H. La2, R. Lim2, H. Parvardeh2 and S.-H. Kee1
1 Department of Civil and Environmental Engineering,
Rutgers University, Piscataway, New Jersey, U.S.A. Email: gucunski@rci.rutgers.edu
2 Center for Advanced Infrastructure and Transportation,
Rutgers University, Piscataway, New Jersey, U.S.A
ABSTRACT - Current assessment of concrete bridge decks relies on visual inspection and
use of simple nondestructive and destructive evaluations. More advanced, but still manual
nondestructive evaluation (NDE) technologies provide more comprehensive assessment.
Still, due to a lower speed of data collection and still not automated data analysis and inter-
pretation, they are not used on a regular basis. The development and implementation of a
fully autonomous robotic system for condition assessment of concrete bridge decks using
multiple nondestructive evaluation (NDE) technologies is described. The system named RA-
BIT (Robotics Assisted Bridge Inspection Tool) resolves issues related to the speed of data
collection and analysis. The system concentrates on the characterization of internal deterio-
ration and damage, in particular three most common deterioration types in concrete bridge
decks: rebar corrosion, delamination, and concrete degradation. For those purposes, RABIT
implements four NDE technologies: electrical resistivity (ER), impact echo (IE), ultrasonic
surface waves (USW) and ground-penetrating radar (GPR). Because the system utilizes
multiple probes or large sensor arrays for the four NDE technologies, the spatial resolution
of the results is signicantly improved. The technologies are used in a complementary way
to enhance the overall condition assessment and certainty regarding the detected deteriora-
tion. In addition, the system utilizes three high resolution cameras to image the surface of the
deck for crack mapping and documentation of previous repairs, and to image larger areas of
the bridge for inventory purposes. Finally, the robot’s data visualization platform facilitates an
intuitive 3-dimensional presentation of the main three deterioration types and deck surface
features.
Keywords: Concrete, bridge decks, corrosion, delamination NDE, automation, GPR, electri-
cal resistivity, acoustics.
INTRODUCTION
Upkeep of concrete bridge decks is one of
the biggest challenges for transportation
agencies. The Federal Highway Administra-
tion’s (FHWA’s) Long Term Bridge Perfor-
mance (LTBP) Program team interviewed a
number of state Departments of Transporta-
tion (DOTs) regarding the expenditure lev-
els for maintenance, rehabilitation, and re-
placement of bridges. The conclusion of the
interviews was that bridge decks constitute
between 50 and 80 percent of the overall
expenditures for bridges. This high expense
stems from three primary reasons. The rst
reason is that bridge decks, due to their di-
rect exposure to trafc and environmental
loads, deteriorate faster than other bridge
components. The second reason is the in-
spection practices that detect problems only
once those have reached their last stage of
progression. For example, the predominant
practice of condition assessment of con-
crete bridge decks in the United States is
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
6
by visual inspection and use of simple NDE
tools like chain drag and hammer sounding.
While such approaches have its merits, they
also have limitations in terms of the early
problem detection and characterization of
deterioration or defects with respect to their
state of progression. The third reason is that
rehabilitation practices rarely address early
problem mitigation. For all these reasons,
the performance of concrete bridge decks
was identied as the most important bridge
performance issue that needs to be ad-
dressed.
The LTBP Program initiated periodical data
collection on concrete bridge decks using
multiple NDE technologies (Gucunski et al.
2012 and 2013). It was demonstrated dur-
ing the initial phase of the program that:
1) NDE technologies can provide accurate
condition assessment, 2) condition indices
obtained from NDE survey results provide
more objective condition assessment, and
3) NDE enables monitoring of deteriora-
tion progression through periodical surveys.
However, it was also recognized that such
surveys require signicant effort and time,
and ultimately represent a signicant ex-
penditure. For example, a typical compre-
hensive survey within the LTBP Program
would require a team of ve to six special-
ists and technicians. To address the need
for evaluation of hundreds of bridges in
the next phase of the Program, the FHWA
initiated in 2011 the development of a ro-
botic system for the NDE of concrete bridge
decks. The main goal of the development
was to improve both the data collection and
data analysis components. On the data col-
lection side the concentration was on an in-
crease of speed of data collection and its
automation. On the data analysis side, the
concentration was on its automation and
the enhancement of the current data inter-
pretation and presentation. During the rst
two years of the development, many of the
stated objectives were achieved and RABIT
is being deployed on a regular basis.
The paper provides an overview of current
NDE of concrete bridge decks and their
evaluation using RABIT system. The rst
part of the paper concentrates on the de-
scription of typical deterioration in concrete
bridge decks. In the second part, a descrip-
tion of the current practice of NDE of bridge
decks, with the concentration on the NDE
methods implemented in RABIT. Finally, the
third part provides a description of the phys-
ical components of the robotic system and
their operation. The data collection process,
and data analysis and presentation/visuali-
zation are presented by sample results.
CONCRETE DECK DETERIORATION
Deterioration in concrete bridge decks can
be caused by a number of causes of chemi-
cal, physical and even biological nature.
Because of it, different NDE technologies
will be more effective in their detection and
characterization. The most common cause
of deterioration is corrosion that will typically
lead to concrete delamination and spalling,
as shown in Figure 1.
Figure 1 Typical concrete bridge deck deterio-
ration and damage: rebar corrosion (left), de-
lamination (middle), and deck spalling (right).
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
7
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
It should be also mentioned that delamina-
tion can be induced by repeated overload-
ing and fatigue of concrete (Gucunski et
al. 2013). Corrosion and delamination are
also the deterioration types of the highest
interest to bridge owners, since the most of
the repairs are related to those. However,
some other deteriorations (e.g. alkali-silica
reaction, delayed ettringite formation, car-
bonation), will primarily cause
material alterations, in terms of
a reduced elastic modulus or
strength, or changed electrical
and chemical properties. Dete-
rioration of bridge decks is often
accelerated by the lack of main-
tenance, or use of improper pro-
cedures during their construc-
tion, especially during concrete
curing. Therefore, information
obtained from the data collected
using multiple NDE technologies
will be necessary to identify the
primary causes of deterioration.
CURRENT PRACTICE OF NDE OF
BRIDGE DECKS
Today, NDE technologies are most common-
ly used to assess whether a deck requires
and what type rehabilitation, or to identify
areas that should be rehabilitated/repaired.
However, for effective bridge management,
bridge owners should develop strategies
regarding the selection of NDE technolo-
gies. Such strategies should enable captur-
ing deterioration in concrete bridge decks
at all stages of their development. For ex-
ample, in a case where deterioration is pri-
marily caused by corrosion, the process can
be described as the one initiated by the de-
velopment of a corrosive environment. One
of the ways to detect and characterize cor-
rosive environment is by using a electrical
resistivity (ER) measurement. As the corro-
sive environment becomes more severe, it
will initiate corrosion activity in rebars. Fur-
thermore, rebar corrosion will induce micro
and macro cracking of concrete. This will
be manifested in delamination of the deck,
which can be detected and
characterized using impact echo (IE). It will
also be reected in the reduction of concrete
elastic properties, which can be measured
using the ultrasonic surface waves (USW)
method. Implementation of the mentioned
and additional NDE technologies is illus-
trated in Figure 2, where measurements by
all are being taken on a 0.6 m by 0.6 m test
grid.
Figure 2 Most commonly used NDE technologies
in detection of corrosion induced deterioration.
Figure 3 Samples of probes and devices of
NDE technologies implemented in RABIT.
8
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
Manual equivalents of the probes and de-
vices of NDE technologies implemented
in RABIT are shown in Figure 3. Electrical
resistivity is a descriptor of corrosive envi-
ronment. Dry concrete will pose a high re-
sistance to the passage of current, and thus
will be enable to support ionic ow. On the
other hand, presence of water and chlorides
in concrete, and increased porosity due to
damage and cracks, will increase ion ow,
and thus reduce resistivity. Resistivity is
typically measured using a four electrode
Wenner probe, shown in the gure. The two
outer probes are used to induce the current
into concrete, while the inner two to meas-
ure the potential of the generated electrical
eld. From the two the electrical resistivity
of concrete is calculated (Brown 1980). An
impact echo (IE) probe consists of a me-
chanical impactor and a receiver. When an
impact is applied, bridge deck resonances
will be induced. The resonances represent
“reections” from the bottom of the deck or
delamination, or exural oscillations of the
delaminated part of the deck (Sansalone,
1993).
Concrete modulus is measured using the
USW method by devices similar to the one
in the gure, called portable seismic prop-
erty analyzer (PSPA) (Nazarian et al. 1993).
The device has a single impact source
and at least two receivers that measure
the velocity of surface waves (phase ve-
locity) generated by an impact. The phase
velocity prole is used to assess the aver-
age concrete modulus or modulus prole.
Qualitative assessment of concrete deck
can be made using ground penetrating ra-
dar (GPR). Electromagnetic waves gener-
ated by an emitting antenna are in part be-
ing reected from the objects and interfaces
of materials of different dielectric properties
and detected by a receiving antenna. The
strength of the reection from the top rebar,
which is typically described as the attenu-
ation of the signal, is used to characterize
corrosive environment and possible delami-
nation (Barnes and Trottier 2000). The at-
tenuation of the GPR signal is primarily af-
fected by the changes in concrete
conductivity and dielectric value. Concrete
lled with moisture and chlorides is highly
conductive and causes strong wave attenu-
ation. Therefore, the GPR assessment of-
ten provides a good description of corrosive
environment. In addition, GPR surveys en-
able rebar mapping and the measurement
of concrete cover, which in some cases may
point insufcient and variable cover as a
contributing factor to accelerated deteriora-
tion.
BRIDGE DECK INSPECTION USING
RABIT
Description of Physical Components
and Data Collection
The robotic system with its main NDE and
navigation components marked is shown in
Figure 4. On the RABIT’s front end there are
two acoustic arrays of the total width of 1.8
m, which matches the scanning width of the
system. Each acoustic array contains four
impact sources and seven receivers. They
are used in different combinations to enable
multiple impact echo and USW measure-
ments. In particular, RABIT’s acoustic ar-
rays can be considered to be equivalent to
fourteen IE and eight or more USW devices.
This large number of sources and receivers
facilitates IE data collection at about 15 cm
spatial resolution, and USW concrete mod-
ulus measurements at a 25 cm resolution
in the robot’s transverse direction. This is a
much higher spatial resolution than a previ-
ously described 60 cm resolution commonly
used in deck testing, and identied in the
current LTBP Program protocols for data
collection. The resolution in the direction of
the robot movement can be controlled by
the robot movement and sampling. Four
Proceq Resipod electrical resistivity (Wen-
ner) probes are attached on the front side
of the acoustic arrays. To establish electri-
cal contacts between the deck surface and
probes, the probe electrodes are being con-
tinuously moistened using a spraying sys-
tem. There are two high resolution cameras
that are being used to capture the
9
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
deck surface for mapping of cracks, spalls,
previous repairs and other surface anoma-
lies. Each of the cameras, once the images
are stitched, covers approximately a 60 by
90 cm area.
Figure 4 Front end of RABIT with NDE and nav-
igation components.
Two IDS (Italy) Hi-Bright GPR arrays are at-
tached on the rear side of the deployment
mechanism (Figure 4). Each of the arrays
has sixteen antennas, or two sets of eight
antennas with dual polarization. The third
camera (not visible in the gure) is placed
on a pneumatic mast in the middle of the
robot that can lift the camera up to a 4.5 m
height. The camera has a 360 degree mirror
that enables panoramic images of the sur-
rounding of the tested area.
The robot’s movement can be controlled
using a keyboard, joystick, Android type de-
vice, or even Iphone. For a fully autono-
mous movement, the robot uses three sys-
tems or devices. The primary navigation
system is a differential GPS, for which the
robot uses two Novatel antennas mounted
on the robot, and the third one on a tripod,
the base station. In addition, RABIT has on
board inertial measurement unit (IMU) and
a wheel encoder. The information from the
three systems is fused using a Kalman l-
ter to facilitate movement with an accuracy
of about 5 cm. High agility of the robotic
platform is enabled by four omni-direction-
al wheels, which allow the robot to move
laterally and to turn at a zero ra-
dius. These wheels also allow fast
movement from one test location
to the next one in any direction.
With all the NDE sensors fully de-
ployed, the robot is about 2.1 m
long and 1.8 m wide.
The survey is conducted by mul-
tiple sweeps of the robot in the
longitudinal bridge direction. Each
sweep covers a 1.8 m wide strip,
equivalent to one half width of a
typical trafc lane 3.6 m wide. At
the end of a strip, RABIT trans-
lates to the next strip and rotates
180 degrees before proceeding
with another sweep. The survey
starts with taking of the GPS co-
ordinates of the GPS base station. This
needs to be done only once for a particular
bridge. Afterwards, the data collection path
can be fully dened by taking GPS coordi-
nates at three arbitrarily selected points on
the bridge deck.
The data collection is fully autonomous. It
can be done in either the full data collection
mode, or the scanning mode. In the full data
collection mode, the robot moves and stops
at prescribed increments, typically 30 to 60
cm, and deploys the sensor arrays to collect
the data. In the scanning mode, the system
moves continuously and collects data using
only the GPR arrays and digital surface im-
aging.
The data from the sensor arrays and probes,
and digital cameras are wirelessly transmit-
ted to the “command van” shown in Figure
5. RABIT can collect data on approximately
300 m2 of a bridge deck area per hour. In
the continuous mode, the production rate is
more than 1,000 m2 per hour.
10
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
Figure 5 Command van and data collection
and robot monitoring displays.
The data collection process can be moni-
tored in real time in the command van, as
illustrated in Figure 5. Four main displays
are used for that purpose, as well as for the
display of real time, or near real time, con-
struction of condition maps for some NDE
technologies and stitched deck surface im-
ages. The summary of all the functions that
can be displayed, or will be available in the
near future, on the four monitors are listed
in Table 1. In addition, two smaller displays
enable monitoring of the robot movement
and survey progression.
Table 1 Display Functions in the Command Van
Data Analysis, Interpretation and
Visualization
The most important results of RABIT sur-
veys are condition maps. An example of
those are a delamination map from impact
echo and concrete modulus map from USW
(Figure 6).
Figure 6 Delamination map (top) and concrete
modulus map (bottom).
11
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
In addition, collected deck surface im-
ages are stitched into a single or multiple
large high resolution images of the bridge
deck. The images are imported into an im-
age viewer that allows review at different
zoom levels, and identication of position
and dimensions of identied features. Two
high resolution images of two sections of
the bridge are shown in Figure 7, the left
one showing a joint, the second transverse
cracking of the deck. Finally, a 3D visualiza-
tion platform enables integration and
Figure 7 High resolution images of the deck
surface.
visualization of the NDE results and images
in an intuitive way. The main internal dete-
rioration types: corrosion, delamination, and
concrete degradation (low quality concrete),
and the deck surface defects are presented
in a common 3D space. This data presenta-
tion is illustrated in Figure 8. Zones of low
concrete modulus concrete are described as
clouds of different translucencies and color
intensity. On the other hand, delaminations
are presented as predominantly horizon-
tal thin clouds at the depth and position as
detected by the IE test. The severity of de-
lamination is presented through the varia-
tion of translucency and color of the image.
Similarly, the corrosive environment is dis-
played through coloring of the rebars. Hot
colors (reds and yellows) are an indication
of highly corrosive environment and, thus,
expected high corrosion rates, while cold
colors (blues and greens) are an indication
of low corrosive environment and, thus, low
corrosion rates. Finally, the surface of this
3D deck volume, not shown in the gure, is
overlaid by a high-resolution image of the
deck surface, as those shown in Figure 7.
CONCLUSIONS
Implementation of NDE in condition assess-
ment of bridge deck will be essential for ef-
fective management of bridges. RABIT with
its integrated multiple NDE technologies
and vision, fully autonomous and rapid data
collection, and near real time data analysis
and interpretation, overcomes the past ob-
stacles related to slow data collection and
interpretation. In addition, rapid and fully
autonomous data collection will signicantly
reduce the required workforce and expo-
sure of the bridge inspection crews to the
passing trafc. It will also in long term re-
duce costs of comprehensive bridge decks
inspections, and make the assessment of a
large population of bridges feasible.
12
ROBOTIC PLATFORM RABIT for CONDITION ASSESSMENT of CONCRETE BRIDGE DECKS USING MULTIPLE NDE TECHNOLOGIES
ACKNOWLEDGMENTS
RABIT was developed under DTFH61-
08-C-00005 contract from the U.S. Depart-
ment of Transportation – Federal Highway
Administration (USDOT-FHWA). The au-
thors gratefully acknowledge contributions
of: Professors Jingang Yi and Kristin Dana,
and former graduate students Parneet Kaur
and Prateek Prasanna.
REFERENCES
[1] Barnes, L. and Trottier, J.-F. (2000).
“Ground penetrating radar for network level
concrete deck repair management”, Jour-
nal of Transportation Engineering, ASCE,
126(3), 257-262.
[2] Brown, R. D. (1980). “Mechanisms of
corrosion of steel in concrete in relation to
design, inspection, and repair of offshore
and coastal structures”, ACI SP-65: Perfor-
mance of Concrete in Marine Environments,
169-204.
[3] G ucunski, N., Romero, F., Imani, A.
and Fetrat, F.A. (2013). “Nondestructive
evaluation-based assessment of deteriora-
tion progression in concrete bridge decks”,
Proceedings of 92nd Annual Transportation
Board Meeting, Washington, D.C., January
13-17, on CD.
[4] Gucunski, N., Maher, A., Ghasemi, H.
and Ibrahim, F. (2012). “Segmentation and
condition rating of concrete bridge decks us-
ing NDE for more objective inspection and
rehabilitation planning”, Proceedings of 6th
Intl. Conference on Bridge Maintenance,
Safety and Management-IABMAS 2012.
Stresa, Lake Maggiore, Italy, July 8-12, on
CD.
[5] Nazarian, S., Baker, M.R. and Crain,
K. (1993). Development and Testing of a
Seismic Pavement Analyzer. Report SHRP-
H-375, Strategic Highway Research Pro-
gram, NRC, Washington, D.C.
[6] Sansalone, M. J. (1993). “Detecting de-
laminations in concrete bridge decks with
and without asphalt overlays using an au-
tomated impact-echo eld system”, NDT in
Civil Engineering, Proceedings of Intl. Con-
ference of British Institute of Non-Destruc-
tive Testing, Liverpool, U.K., April 14-16,
807-820.
Figure 8 3D visualization of the detected
deterioration and defects.
