|
|
|
TECHNICAL
REPORT STANDARD TITLE PAGE
|
|
1. Report No.
|
2.Government Accession No.
|
3. Recipient’s Catalog No.
|
|
FHWA-NJ-2002-026
|
|
|
|
4.
Title and Subtitle
|
5.
Report Date
|
|
Correlation of Surface Texture,
Segregation and Measurement of Air
Voids
|
October 2002
|
|
6. Performing Organization Code
|
|
NJIT/Abatech, Inc.
|
|
7.
Author(s)
|
8. Performing Organization Report No.
|
|
Jay N. Meegoda, Geoffrey M. Rowe,
Chamil H. Hettiarachchi, Nishantha
Bandara and Mark J. Sharrock
|
FHWA-NJ-2002-026
|
|
9.
Performing Organization Name and Address
|
10. Work Unit No.
|
|
New Jersey Department of
Transportation
PO 600
Trenton, NJ 08625
|
|
|
11.
Contract or Grant No.
|
|
Task order No. 29
|
|
12.
Sponsoring Agency Name and Address
|
13.
Type of Report and Period Covered
|
|
Federal Highway Administration
U.S. Department of Transportation
Washington, D.C.
|
Jan. 2001- Sep 2002
|
|
14.
Sponsoring Agency Code
|
|
|
|
15.
Supplementary Notes
|
|
|
|
16.
Abstract
|
|
Laser based systems were used in this
research to quantify segregation during paving of hot mix asphalt concrete
pavements. Two segregated test sections and a control test section were
evaluated to determine the applicability of laser texture method to detect
and quantify segregation. Laser texture data were gathered from all three
sites. Sand patch and nuclear density tests were also performed at 25-feet
intervals. In addition to the above, visual surveys were performed to confirm
the measurements.
The laser texture data consistently
showed texture peaks indicating the presence of segregation, which occurred at
approximately 100-feet intervals for one site suggesting end-of-truck-load
segregation. Test results from the control
section were used to establish a correlation between the sand patch test (a
quantitative segregation test) and the laser texture data. The poor
correlation between nuclear density and texture from the sand patch method
for the test sections suggested that nuclear density measurements could not
be used for quantification of segregation. However, the high nuclear density
values for the control section when compared with those for segregated
sections suggested that it could be used as a confirmation test. By combining
the concepts described above, a computer program, NJTxtr, to detect
segregation was developed. This
program uses pavement texture data and determines the acceptability of the
pavement section based on the level of segregation present within the
pavement section. Ratios of texture in segregated areas to that in
non-segregated areas were used as the basis for detection of different levels
of segregation. By combining levels
of segregation and the extent of each level of segregation, an AREA index was
developed to determine the acceptance of a pavement section. When a pavement section is acceptable, the
software determines the pay adjustment factor to be used. If segregation is
present, it suggests remedial actions for each segregated area. NJTxtr was
evaluated using the data collected from one control and two segregated test
sections and satisfactory results were obtained.
|
|
17. Key Words
|
18. Distribution Statement
|
|
Monitoring Construction, Surface
Texture, Segregation, Air Voids, Nuclear Density, Sand Patch, LASER, Mean
Profile Depth, Estimated Profile Depth, Asphalt Pavements, and Computer
Program
|
|
|
19. Security Classif (of this report)
|
20. Security Classif. (of this page)
|
21. No of Pages
|
22. Price
|
|
Unclassified
|
Unclassified
|
94
|
|
|
Form DOT F 1700.7 (8-69)
|
|
|
|
|
|
|
|
|
The New Jersey
Department of Transportation (NJDOT) sponsored this research. The program manager at the NJDOT is Mr. Anthony Chmiel and the NJDOT
customer for this project is Mr. Andris Jumikis. Authors wish to acknowledge
the efforts of the NJDOT project manager and the NJDOT customer. It would not
have been possible to complete this project without the assistance of Mr.
Nicholas Gephart of Pavement Management Unit, NJDOT. Authors also would like to
acknowledge the contributions of Mr. Cliff Wu of NJIT, Mr. Anthony Orsi, Mr.
Joseph Maloney, and Mr. Fred Kern of the Bureau of Materials, NJDOT and Mr.
Jaroslaw Hucul, Mr. Charles Isiadinso, Mr. Raymond Micharski, Jr., and Mr.
Kevin Hall of the Pavement Management Unit, NJDOT. The editorial assistance
from Ms. I. Martyn Nichols and technical assistance from Mr. Kurt Huber and Mr.
Paul Harbin of Roadware Group Inc., Canada, and Joseph Biggica and Michael
Manno of Newark Asphalt, NJ is highly appreciated.
Correlation of Surface Texture, Segregation,
and Measurement of Air Voids
-Table of Contents-
Summary 1
Problem Statement 2
Objectives 3
Introduction 3
Field Evaluations 17
Test Results 21
Data Interpretation and Application 29
Development of Numerical Procedures 37
NJTxtr Software for Segregation
Monitoring 44
Summary and Conclusions 54
References 60
Appendix 1: Visual Evaluation Data
for Route 9 Test Section 62
Appendix 2: Standard Test Method for
Surface Texture 71
Appendix 3: NJTxtr Results for the
Route 9 Data 82
Figure 1: Schematic Representation of Surface Texture Laser 8
Figure 2: Example of Laser Surface Texture Measurement Over
a 0.6m Length 8
Figure 3: The Dynatest RSP (from http://www.dynatest.com) 9
Figure 4: Greenwood Engineering System (from http://www.greenwood.dk) 10
Figure 5: Automatic Road Analyzer (ARAN) (from http://www.roadware.com) 10
Figure 6: WDM - HSTM Trailer Mounted Device (from http://www.wdm.co.uk) 12
Figure 7: ARRB Multi Laser Profiler (from http://www.arrb.org.au/index.htm) 12
Figure 8: ROSANv (http://www.surfan.com) 14
Figure 9: Variable ETD for a Section of Asphalt Pavement
(http://www.surfan.com) 15
Figure 10: Four Meter Wide Analysis Window Showing Probable
Aggregate
Segregation (http://www.surfan.com) 16
Figure 12: Sand Patch Tests at 25 ft intervals, Alternating
between Tests on
Each of Three Test Lines on Rt. 9 19
Figure 13: Sand Patch Tests Performed
on Route I-195 20
Figure 14: Definition of Mean Texture
Depth (MTD) in the ARAN Software 22
Figure 15: RMS Summary Data for Three
Test Lines 23
Figure 16: MTD Summary Data for Three Test
Lines 24
Figure 17:
Variation of Sand Patch MTD along Three Test Sections. 25
Figure 18:
Variation of Nuclear Density along Three Test Sections. 27
Figure 19:
Variation of Air Voids Percentage along the Test Section 28
Figure 20: Correlation between Textures from the ARAN and Sand Patch
Method 30
Figure 21: Frequency Distribution Curves for the ARAN and Sand
Patch (SP)
Test Results, Normalized by
Maximum Value 30
Figure 22:
Correlation between Textures from the ARAN and Sand Patch Method
for Rt. I-195
Control Section 32
Figure 23:
Variation of Sand Patch Test Results and Predicted Texture Depths
from ARAN for
Rt. I-195 Control Section 33
Figure 24:
Typical View of Pavement Surface of Route 9 34
Figure 25:
Mean Segment Depth (MSD) Plot for the Test Line 1 of the Route 9
Data 35
Figure 26: Longitudinal Paths for
Measurement for Each Lot 37
Figure 27: Determination
of Mean Profile Depth (MPD) from a 100mm
Base-length 39
Figure 28:
Variation of Mean Segment Depth with Base-length 40
Figure 29:
Variation of Texture Depth Ratios with Base-length 41
Figure 30:
Computation Flow Chart of NJTxtr Software 45
Figure 31:
Block Average of Mean Segment Depth for100mm Blocks of Control
Section of the Route I-195 Data. 47-49
Figure 32:
Block Average of Mean Segment Depth for 100mm Blocks of Test
Section of the Route I-195 Data. 50-52
Figure 33:
NJTxtr Suggested Treatments for 40-50 Meters of Route I-195 53
Figure 34:
Block Average of Mean Segment Depth for 100mm Blocks of Test
Line 1 of the
Route 9 Data 55
Figure 35:
Zoomed Mean Segment Depth (MSD) Plot for the Test Line 1 of the
Route 9 Data 56
Figure 36:
NJTxtr Suggested Treatments for 0-10 Meters of Route 9 57
List of Abbreviations and Symbols
AASHTO American
Association of State Highway and Transportation Officials
ARAN Automatic
Road Analyzer
ARRB Australian
Road Research Board
ASTM American
Society for Testing of Materials
DOT Department
of Transportation
ETD Estimated
Texture Depth
FHWA
Federal Highway Administration
Gsb Bulk
Specific Gravity
GPS Global
Positioning System
GPR Ground
Penetrating Radar
HMA Hot
Mix Asphalt
HSTS High
Speed Texture System
IRI International Roughness Index
JMF Job Mix Formula
MLP Multi-Laser Profiler
MMSD Mean of
the Sean segment Depths
MP Mile
Post
MP Materials
Procedure
MPD Mean
Profile Depth
MSD Mean
Segment Depth
MTD Mean
Texture Depth
NCHRP National
Cooperative Highway Research Program
NJDOT New
Jersey Department of Transportation
PC Personal
Computer
QC/QA Quality
Control/Quality Assurance
ROSANv Road
Surface Analyzer - vehicle-mounted
RMS Root
Mean Square
RN Ride
Number
RSP Road
Surface Profiler
SP Sand
Patch
TFHRC Turner-Fairbank
Highway Research Center
TR Texture
Ratio
This report
describes the research funded by the NJDOT to develop an automated technology
to quantify segregation of hot mix asphalt concrete pavements. Laser-based
systems were evaluated in this research. From the laser-based systems reviewed,
the ROSAN system appears to be the most advanced, with respect to algorithms
developed, to determine segregation in asphalt. Currently it has the widest
application in this area (Stroup-Gardiner and Brown, 2000). However, the
application of the algorithms developed could be applied to the data collected
from other laser devices. The level of technology available in the ARAN device
is considered acceptable for the application of texture measurement.
Consequently, since NJDOT owns this piece of equipment, the ARAN was selected
as the field texture measuring device and the development work applied to the
output from the laser sensors of the ARAN.
Two segregated
test sections and a non segregated control section were tested to evaluate the
applicability of the laser texture method to detect and quantify segregation.
Laser texture data was gathered from all three sites, and the sand patch and
nuclear density tests were performed in three sections at intervals of 25
feet. In addition to the above, visual
surveys were performed to confirm the measurements.
The poor
correlation between the nuclear density and the texture from the sand patch
method for test sections suggested that nuclear density test should not be used
as a quantitative method to predict segregation. The nuclear density measurements
can locate the areas of low density, which is a volume measurement, and may be
due to segregation. Whereas, the sand patch test can locate areas of
segregation, as indicated by surface texture. Hence a poor correlation between
the nuclear density and the texture from sand patch method is expected. Since
low nuclear density measurements may be due to segregation, the high nuclear
density values for the control section suggested that nuclear density
measurements might be used as a confirmation test.
The laser
texture data showed the presence of segregation with consistent texture peaks
that occurred at approximately 100-feet intervals. Test results from the
control section were used to establish a correlation between the sand patch
test (a quantitative test to determine segregation) and the laser texture data.
A computer program, NJTxtr, to detect and monitor segregation was developed by
combining the above concepts. This program uses the ARAN-collected pavement
texture data and determines the acceptability of the pavement section based on
the level of segregation present within the pavement section. Ratios of texture
in segregated areas to that in non-segregated areas were set for detection and
monitoring of different levels of segregation.
By combining levels of segregation and the extent of each level of
segregation, an AREA index can be developed to determine the acceptance or
non-acceptance of a pavement section. When a pavement section is acceptable,
the software determines the pay adjustment factor to be used. Remedial actions
are suggested when segregation is present. NJTxtr was evaluated using data
collected from one non segregated control section and two segregated sections
and found to be satisfactory based on predictions from NJTxtr with visual
observations.
The correlation between air voids and
pavement durability is well documented. Some projects have experienced high air
voids and segregation of the surface mixes due to poor construction practices
or equipment problems. By establishing a relationship between surface texture
measurements, surface segregation, and air voids, the NJDOT will have a
screening tool to identify variations in surface texture that are typical of
segregation and potentially locate pavement sections with high air voids.
The project evaluated the current
technology to develop a screening tool for assessing surface texture as a means
of locating segregated areas of the surface pavement and potentially high air
void locations. The proposed technology was compared to another method that
measures texture, namely the sand patch test, to establish acceptance limits.
Once segregation was located, cores from the pavement layer were taken and used
to determine a correlation of these areas compared to the average air void
content for the lot.
Objectives
The objectives
of this study were to:
1.
Develop
a vehicle-mounted screening device to measure variations in surface texture
that are typical for segregated sections of pavement.
2.
Correlate
the air voids and density of these sections compared to the average texture for
the lot.
3.
Recommend
development of the NJDOT specifications for implementation of the surface
texture measurement methods.
The technical approach to the work
included specific tasks that are identified as follows:
1. Literature search.
2.
Compare air voids and density from cores taken
from segregated areas to the remainder of tested lots.
·
Develop
laboratory correlations between surface texture, segregation, and measurement
of air voids.
·
Field
evaluation of laboratory correlations between surface texture, segregation, and
measurement of air voids.
3.
Evaluate
current technology to develop a screening device that can:
a.
Measure
surface texture from a moving vehicle,
b.
Locate
segregated sections of pavement surface,
·
Benchmark results
against standardized techniques.
·
Assess accuracy,
repeatability, and reliability of collected data.
·
Perform testing that
replicates field conditions.
4.
Develop
the standard Materials Procedure (MP) for testing.
Segregation may be defined as the lack of
homogeneity of constituents in Hot Mix Asphalt concrete (HMA) pavements that
accelerates pavement distresses. Constituents of HMA are asphalt cement,
aggregates, additives, and air voids. Segregation produces repetitive patterns
of non-uniformity. Therefore, standard quality control/quality assurance
(QC/QA) procedures that randomly define sampling locations would have a low
probability of adequately identifying this problem. Ideally, some type of
longitudinal pavement profile, using one or more nondestructive measurements at
selected transverse locations, can be identified. An alternative methodology is
needed to address random but localized areas of non-uniformity.
The most common HMA segregation has been identified as
gradation segregation. Gradation segregation is the non-uniform distribution of
coarse and fine aggregate materials in the finished HMA pavements. Gradation
segregation can occur as the result of aggregate stockpiling and handling,
production, storage, truck-loading practices, construction practices, and
equipment adjustments. Localized pavement areas rich in coarse aggregate are
typically associated with high air voids and low asphalt contents. These
conditions can lead to moisture damage as well as to durability-related
pavement distresses such as fatigue cracking, pothole formation, and raveling.
Conversely, pavement areas rich in fine aggregate are associated with low air
voids and high asphalt contents, making them susceptible to rutting.
There are several traditional and emerging methods to detect
and quantify texture, so that a quality control/quality assurance program can
be built into the design and construction of HMA pavements. Following is a
description of those methods.
Traditional
Methods for Detecting Segregation:
Visual identification: Visual identification of non-uniform
surface texture has been used to locate segregation (Brook et al., 1996). This
is a subjective approach, which can lead to disagreements between the agency
and representatives of the contractor. Usually visual detection of non-uniform
areas is used as the baseline for other quantitative approaches. Cross et al.,
1997, studied four Kansas field projects with suspected segregation problems.
They concluded that visual observations are better able to identify segregation
in mixtures with larger aggregate sizes and coarser (below the maximum density
line) gradations, but it was difficult to visually identify segregation for
mixtures with smaller sized aggregates and finer gradations.
Sand Patch Testing: The sand patch test method has been used
to quantify visual observations of differences in the surface macro-texture.
The ASTM E965 test method (ASTM 2001) indicates that the precision of the test
method is approximately one percent of the measured depth in millimeters and
the operator variation is about two percent. Good correlation was found between
visual observations of non-uniform textured areas and the sand patch test
results for measuring surface macro-texture.
Nuclear Density Gauges: These gauges can be used to identify
segregated areas by profiling the longitudinal density of the pavement mats. An
assumption is made that segregation will be seen as low density. However,
literature indicates limited success for this method. There are two reasons:
First, a common
assumption for using these gauges is that density decreases with increasingly
coarse aggregate segregation. However, this assumption does not consider the
relationship of the gradation to the maximum density line. If the Job Mix
Formula (JMF) begins above this line, separation of the coarse aggregate in
this type of mix may result in a higher density as the gradation shifts toward
the maximum density line.
Second, different
types of aggregates have different effects on gauge variability. If a mixture
is composed of a mixture of different aggregate types the change in testing
variability in coarse aggregate-rich and fine aggregate-rich areas may make it
difficult to adequately detect or measure segregation or both.
Of the three traditional methods
described above, the sand patch test seems to be most objective and accurate
method to detect texture and hence it was used in this research.
Innovative Technologies for Detecting Segregation: Three new technologies have been identified
as having potential to selectively identify HMA segregation:
Thermal Imaging: All objects emit infrared radiation in the form of heat, which can be
detected by an infrared scanner. These natural impulses are converted into
electrical pulses and then processed to create a visual image of the object's
thermal energy. The colors used to represent the thermal imaging can be
user-selected to represent surface temperature changes, such as blue for colder
regions and red for warmer regions. The thermal imaging technology will
indicate high-void regions, as thermal capacity of air is minimal compared to
that of aggregates and asphalt cements.
If one assumes that high segregation causes high void ratios, then the
technology can be easily adopted to detect segregation.
The
primary component of any thermal imaging system is an optical scanner, a unit
that is used to detect infrared radiation from an object. Other essential
components of the system are a display monitor, a video camera, and a computer
with appropriate software for data acquisition, analysis, and storage. Minimum
resolution requirements and the height of the equipment above the surface
determine the area surveyed by the camera. A full-lane width can be surveyed at
one time with an appropriately placed camera. Usually liquid-nitrogen-cooled
scanners provide improved resolution over other types of scanners. Although the
current technology is vehicle-mounted, operation at highway speeds (>80 kph
or 50 mph)) tends to blur the image. Resolution is improved substantially by
operating the equipment at slower speeds (<60 kph or 40 mph]). Some
disadvantages of this technology are as follows:
·
Only
surface or near-surface defects are evident.
·
Both
temperature and gradation segregation will appear as "cold" areas. A
secondary confirmation is needed to define the segregation
·
Use
of this technology for after-construction surface evaluations is doubtful as
infrared radiation from an old HMA pavement depends on gain from solar energy.
Ground Penetrating Radar (GPR):
The basic
theory used in GPR is a measurement of the dielectric constant or permittivity
(e). The electrical permittivity of air is different from that of
aggregate and asphalt cement. In highway applications, the travel time of an
electromagnetic pulse through the structure can then be used to compute the
layer thickness. Rmeili and Scullion, 1997, found that anomalies in the
reflected waveforms could be used to evaluate density and moisture content.
Saarenketo and Scullion, 1994, used this approach with reasonable success to
detect underlying moisture damaged (stripped) areas in several field projects.
Continuous
longitudinal surface texture profiles can be obtained quickly because the
technology can be operated at normal highway speeds. However, slower speeds are
needed for higher resolution. The equipment is portable and reasonably affordable
and can be mounted on any vehicle. Disadvantages of this technology include the
following:
·
A
dry pavement surface is needed. Wet surfaces will alter the deflection of the
laser beam.
·
Prior
calibration is needed, as final results are mix-type dependent.
Laser Surface
Texture Measurements: Segregation in asphalt materials can occur due to a variety
of reasons. The manifestation on site
is an uneven distribution of aggregate surface texture associated with uneven
distribution of aggregate, asphalt binder and/or air voids. Over the past twenty years the use of laser
technology to define surface texture has been gaining wide popularity. The basic concept of the measurement system
is illustrated in Figure 1. Using mathematical algorithms the distance to the
surface at a discrete point is obtained.
The measurements are conducted very rapidly as the vehicle drives along
the pavement enabling measurements at points that can be typically separated by
1mm (1/25-inch) defining a surface profile as illustrated in Figure 2.
The
assessment of texture with vehicle-mounted laser measuring devices is well
established and a number of commercial devices are available. Several firms market equipment, which can
be vehicle-mounted. A few of these are
briefly summarized below - with abstracts from the manufacturer's information
from web pages as cited in the text.
Figure 1: Schematic Representation of Surface Texture Laser
Figure 2: Example of Laser
Surface Texture Measurement Over a 0.6m Length
ABSTRACTS FROM
MANUFACTURES WEB SITES
Dynatest
The Dynatest Road Surface Profiler (RSP) from Denmark (http://www.dynatest.com) is designed
to provide an advanced automated high quality pavement roughness and related
measurements. The RSP performs
continuous highway-speed measurements of longitudinal and transverse profile,
including real-time roughness (IRI and RN), rut depth and texture evaluation,
GPS, and geometrics.
This product line is available in several levels of
sophistication, ranging from top-of-the-line version (with 11 laser sensors
standard or up to 21 lasers on special order) down to 1 laser, single wheel
path version for real-time longitudinal profile (and optionally texture)
evaluation.
Figure 3: The Dynatest RSP (from
http://www.dynatest.com)
Greenwood
Engineering
The Profilograph system, manufactured by Greenwood
Engineering of Denmark (http://www.greenwood.dk),
is installed in a standard vehicle. Original vehicle integral safety is
maintained. The driver may operate the system as it is controlled from the
front seat. With the current
configuration, up to 25 lasers are located on a front beam. Each laser measures
the distance from the beam to the pavement surface at a sample rate of 16 kHz.
On the beam an inertial unit keeps track of all movements of the beam with high
precision. The traveled distance is measured with a resolution of 20 pulses per
wheel revolution. All signals are stored for each sensor individually. An
industrial PC and signal controller interfaces handle all signals from the
sensors to storage on the hard disk.
Figure 4: Greenwood Engineering
System (from http://www.greenwood.dk)
ARAN
The first ARAN, manufactured by Roadware of Canada (http://www.roadware.com)
(Automatic Road Analyzer), was first delivered to Autostrade, in Italy, in
1984. Today, over 75 agencies in 15 plus countries use ARAN's state-of-the-art
technologies to increase the cost effectiveness of their data collection
activities.
Figure 5:
Automatic Road Analyzer (ARAN) (from http://www.roadware.com)
ARAN provides information to support better management
decisions by collecting consistent, accurate data quickly and cost effectively.
A wide variety of data can be collected continuously at highway speeds, as
follows:
·
Longitudinal
profile/roughness (IRI)
·
Transverse
profile/rutting
·
Grade,
cross-slope
·
Pavement texture
·
Pavement condition or
distress
·
GPS coordinates
·
Right-of-way video
(tape or disk)
·
Pavement video
feature location
WDM
– HSTM
WDM, from the United Kingdom (http://www.wdm.co.uk),
manufactures a High Speed Texture System that has been designed to provide an
economic method of routinely monitoring the macro texture of road networks.
The HSTS consists of a two-wheeled trailer designed to carry
a maximum of three laser sensors (one in each wheel-track and a third on the
lane center), a touch control flat screen monitor (acting as a control
console), a computer for calculating, and a disk for storing the data. The HSTS
is capable of measuring over the speed range of 0-100km/h. WDM has also
developed a hand-held device, which measures texture at walking speed.
The computer is able to calculate either or both of the two
most prolific methods of macro-texture measurement available in the world
today.
Most vehicles fitted with a 50 mm tow-ball can tow the HSTS
and the control panel and computer can be housed within the vehicle.
Interconnecting cables are supplied to suit this purpose. Data storage is by 3
1/2" floppy disc or Zip disc dependant on system requirements. The use of
a tow vehicle avoids need for special vehicle modifications.
Figure 6: WDM - HSTM Trailer
Mounted Device (from http://www.wdm.co.uk)
ARRB
Australian Road Research Board (http://www.arrb.org.au/index.htm)
manufactures a range of laser profilers that include a multi-laser and a more
limited portable system.
The Multi-Laser Profiler (MLP) is a vehicle-mounted system
that automatically collects integrated road condition data by recording laser
profiles of the road surface at highway speed. The MLP comes with an onboard
computer system and a range of software for data acquisition and analysis
tasks.
Figure
7: ARRB Multi-laser Profiler (from http://www.arrb.org.au/index.htm)
Advantages:
·
Simultaneously
measures roughness, rutting and macro texture.
·
Surveys up to 600 km
of road per day at intervals as close as 50mm.
·
Tailored to suit
individual needs.
·
Outputs compatible
with asset management systems.
·
Easy to operate with
a compact operator's console.
·
Available for
purchase or for contract network surveys.
ROSANv
Surfan
Engineering and Software Inc. (http://www.surfan.com) developed ROSANv
in cooperation with the FHWA (Stroup-Gardiner and Brown, 2000).
This technology uses a lightweight portable, bumper-mounted
laser system to evaluate pavement texture characteristics along a linear path.
The high speed, high quality laser in this system reduces electronic noise and
spikes as well as provides a high level of resolution capable of producing
measurements as defined in ASTM E1845-96 for the Calculation of Pavement
Macro-texture Mean profile Depth. The ROSANv software is a Windows
95 compatible program that provides a graphical user interface for both data
acquisition and analysis (FHWA, 1998).
The original and primary goal of the ROSANv
project at FHWA's TFHRC Pavement Surface Analysis lab was the development of a
portable and automated system for the measurement of pavement texture at
highway speeds along a linear path as a replacement of the manual Volumetric
Patch Method "Sand Patch Test" as outlined by ASTM-E965 and ISO
10844. Volumetric Patch Method
procedures are valid for concrete or asphalt paving whose surface has not been
treated or designed for improved drainage (such as grooving, tinning, or
open-graded porous asphalt) or milled to remove rutting. Prior to the
completion of the ROSANV research work, ASTM Committee E-17 approved
ASTM Standard E1845, "Standard Practice for Calculating Pavement
Macro-texture Mean Profile Depth" from a profile of pavement macro-texture
in November of 1996. The "v"
in ROSANv stands for "vehicle-mounted" in that ROSANv
can be quickly mounted on any vehicle using a temporary bumper hitch.
Figure 8: ROSANv (http://www.surfan.com)
The following areas of application are feasible with the aid
of ROSANV:
·
Texture measurements
for Pavement Management Systems
·
Site specific texture
measurements for safety investigations
·
Quality Control
measurements for new pavement for certifying pavement meeting contract
specifications for texture and aggregate segregation limits
·
Combining
friction-testing equipment such as a skid trailer with ROSANV for
simultaneous measurement
·
Texture and surface
detail measurements (grooving, thinning) in noise research studies
·
Faulting at joints
and cracks
·
Joint and crack
measurement summaries over a section of pavement
·
Quality Control
measurement of sawed concrete grooves
·
Groove maintenance
measurements for wear and debris
·
Memory of re-paved
surfaces
Asphalt pavement and
aggregate segregation as determined by ROSAN is illustrated in Figure 9 over a
148-meter length of asphalt pavement. The graph shows a large but expected
variability in texture as compared to concrete pavement. A possible recurring
pattern can be discerned within the graph. By using an analysis window that is
four meters long, the pattern is more clearly visible as shown in Figure 10.
Each of the waves is about 24-25 meters in length. These waves or delta patterns probably represent individual
truckloads of asphalt and can be interpreted as the interaction of the truck
with the paving machine and the processing of the load through the paving
machine.
Figure
9: Variable ETD For a Section of Asphalt
Pavement (http://www.surfan.com)
Table 1:
Comparison of Field Laser Measuring Devices
|
|
Laser Sensors
|
Structure
|
|
Dynatest
- Road surface profiler (RSP)
|
1-21 (Standard 11)
|
Vehicle mounted
|
|
Greenwood
- Profilo-graph System
|
Up to 25
|
Standard vehicle
|
|
ARAN
(Automatic road analyzer)
|
1-3
|
Special vehicle
|
|
WDM-HSTM (High
speed texture system)
|
Up to 3
|
Two wheeled trailer
|
|
ARRB-MLP (Multi
laser profiler)
|
Standard 13 (Expandable)
|
Vehicle mounted
|
|
ROSAN (Road surface
analyzer)
|
1-3
|
Vehicle mounted
|
Figure 10: Four-Meter Wide Analysis Window Showing
Probable Aggregate Segregation (http://www.surfan.com)
Summary of manufacturer's
information
Table 1 provides a comparison of all six field surface
texture measuring devices. Based on the available information it appears that
any device could measure the micro-texture of the asphalt pavements to detect
segregation.
From
the laser-based systems reviewed, the ROSAN appears to be the most advanced
with respect to algorithms developed to determine segregation in asphalt and
has the widest application in this area to date (Stroup-Gardiner and Brown,
2000). However, the application of the
algorithms developed could be applied to the data collected from other laser
devices. The level of technology
available in the ARAN device is considered to be acceptable for the application
of texture measurement. Consequently, since the NJDOT has this piece of
equipment, the ARAN was selected as the measuring device for field surface
texture.
Of the three emerging
technologies discussed above, the laser technology is quite capable of
quantifying the surface texture and hence the segregation of HMA pavements.
Therefore, laser technology is further evaluated in this research.
Table 2 Comparisons of Surface Texture Measuring
Technologies
|
Test
Method
|
Type
of Mix
|
Depth of measurement
|
|
Fine
Gradations
|
Dense
Gradations
|
SMA
|
|
Surface
Only
|
Depth of Life
|
Full AC
Mat Depth
|
|
Visual
Observe
|
Yes
|
Yes
|
Yes
|
|
Yes
|
No
|
No
|
|
Sand Patch
|
Yes
|
Yes
|
Yes
|
|
Yes
|
Yes
|
No
|
|
Nuclear
Density
|
Gradation
Dependent
|
Gradation
Dependent
|
Yes
|
|
No
|
Yes
|
No
|
|
Laser
|
Yes
|
Yes
|
Yes
|
|
Yes
|
No
|
No
|
|
GPR
|
Unknown
|
Unknown
|
Unknown
|
|
No
|
Yes
|
Yes
|
|
Infrared
|
Unknown
|
Unknown
|
Unknown
|
|
Yes
|
Thin Lift
|
Unknown
|
Table 2
summarizes all the surface texture methods considered in this research. Based
on the subjectivity of measurements if the visual observation is eliminated,
then sand patch and laser methods are the most appropriate methods to quantify
the segregation using surface texture.
During the course of the literature review, the team
became aware that significant advances in this technology for the detection of
segregation have recently been undertaken by the National Center for Asphalt
Technology and reported in NCHRP Report 441 (Stroup-Gardiner and Brown,
2000). In this work a draft
specification for the detection of a laser-based method for detecting
segregation was developed in AASHTO format.
The research conducted by the NJIT team used the above work as a
starting point for the development work.
In addition, considerable use has been made of the definitions given in
ASTM specification E 1845-96.
Field
Evaluations
Representatives of
the pavement
management group at the NJDOT identified two field test sites and a control
test section. The first test site was on Route 9 from Mile Post (MP) 111.5 to
112, northbound fast lane, which had segregated materials on the pavement
surface. The second test site was I-195 MP 9.1-9.8 east bound slow lane, which
also had segregated materials on the pavement surface. The control section was
I-195 MP 9.8-10.1 eastbound slow lane, which was a uniform pavement surface
with no visible segregation.
The Route 9 Site: The data collection program was carried out on October 22,
2001 and October 23, 2001. A half-mile section of the fast lane MP 111.5 to
112.0 of Route 9 was closed from 9.00 AM to 1.00 PM on both days to accommodate
the testing program.
The field-testing
program consisted of following components.
1. Texture profile measurement using the ARAN
2. Sand patch testing
3. Nuclear density measurement
4. Coring to produce specimens for density testing and
additional analysis
5. Visual observations
The test site was
prepared for the tests by marking the pavement at 5-feet intervals to enable
location referencing. The length of the pavement surveyed was limited to 1500
feet, since this is the approximate capacity of the ARAN’s data unit when
detailed laser data are being stored. Texture measurements were performed on
three separate lines along the subject lane of the Route 9 from MP 111.5 to
112.0. These test lines were located at 3, 5.5 and 8 feet from the edge of the
lane. The ARAN was operated at a speed of 40 mph along each test line. The ARAN
texture software recorded texture depths and the distance corresponding to each
measurement. The ability of the ARAN to maintain a constant distance from the
edge relied upon the skill of the driver.
Due to the speed of the survey (approximately 40mph), precise transverse
locations were difficult to achieve.
Sand patch tests were
performed every 25 feet alternating between the three lines tested with the
ARAN resulting in a total of 58 sand patch tests. Nuclear density tests were conducted with one test at each sand
patch location. Pavement cores were
taken from 11 locations, which were selected to cover the range in sand patch
measurements. Visual observations were
also conducted to enable a visual comparison of test results. Typical sand patch test locations are
presented in Figure 12.
The Route I-195 Site: The data collection program was carried out on July 20,
2002 and on July 21, 2002. One lane of
a one-mile section from MP 9.1 to 10.1 of Route I-195 was closed from 10 PM to
5.00 AM in the night to accommodate the testing program.
Figure 11:
General View of Rt. 9
(Note area in foreground is patched due to disintegration of
materials as a result of segregation, markings on pavement at 5-feet intervals)
Figure 12: Sand Patch Tests at
25 ft Intervals, Alternating between Tests on Each of Three Test Lines on Rt. 9
Figure 13:
Sand Patch Test Performed on Route I-195
Figure 13:
Sand Patch Tests Performed on Route I-195
The field-testing
program consisted of following components:
1. Texture profile measurement using the ARAN
2. Sand patch testing
3. Nuclear density measurement
4. Visual observations
The test site was
prepared by marking the pavement at 25-feet intervals to enable location
referencing. The length of the pavement surveyed was limited to 1500 feet for
the test section and 500 feet for control section. Three ARAN tests were
conducted in the test section as well as in the control section. These test
runs were located approximately three feet from the edge of the lane. The ARAN
was run at a speed of 40 mph to capture texture profile. The ability of the ARAN to maintain a
constant distance from the edge relied upon the skill of the driver. Due to the speed of the survey
(approximately 40 mph) the precise transverse location was difficult to
achieve. Therefore, a spray painting device was added to the ARAN to locate the
path of the laser. Due to the high pressure in the paint canister, the paint
nozzle failed after the first run in the test section. The sand patch and
nuclear density tests were performed at points on this spray-painted path of
the laser.
Sand patch tests were
performed every 25 feet on the paint lines generated from the ARAN resulting in
a total of 58 sand patch tests for the test section. Sand patch tests were also
performed at 25 feet intervals along the line tested with the ARAN resulting in
a total of 21 sand patch tests for the control section. Nuclear density tests were conducted at each
sand patch location for both test and control sections. Visual observations
were also conducted to enable a visual comparison of the test results.
All data, collected by the measurement sub-system was stored on the hard
drive at regular station intervals. A
high-pass filter was used, with a base length of 50mm (Roadware Corporation, 1995). The 50mm length was found to be acceptable by the ARAN to remove
wavelengths that are a result of truck axle dynamics and fluctuations in
pavement profile, which are not attributable to texture. When each test line was completed by the
ARAN, two separate data files were created and stored on the computer. One data file consisted of the raw texture
data (after passed through the high-pass filter) while the other file consisted
of texture summary data such as Root Mean Square (RMS) value and Mean Texture
Depth (MTD).
The RMS value is calculated as follows:
---------------------------(1)
where:
dn = a valid data
point number - n
n = number of
valid data points over measurement interval
