Printer Friendly

The application of ground-penetrating radar in highway engineering.


The demands upon our Nation's highway system continue to increase as the need for greater productivity extends the burdens it must carry. Increasing traffic as well as freight demands pose a significant challenge for improved construction quality and rehabilitation effectiveness.

This situation has generated the need for more efficient and expedient ways to identify and evaluate highway condition. One promising technology, which has been demonstrated to be effective, is ground-penetrating radar (GPR). GPR is a noninvasive, nondestructive tool that can be used in quality assurance investigations for new construction and in evaluation of structural condition prior to rehabilitation.

Radar (Radio Detection and Ranging) has been in use since the 1920's. The U.S. Army began 30 years ago to use a form of radar-ground-penetrating radar--to locate nonmetallic mines. The success of this program, as well as GPR's use in weather tracking and in mapping planetary surfaces from space probes, prompted the highway community to experiment with this technique for locating voids underneath pavements, determining pavement layer thicknesses, detecting delaminations in bridge decks, and investigating scour around bridge piers.

The Federal Highway Administration (FHWA) conducted its initial GPR research in the mid-1970's to investigate the feasibility of radar in tunneling applications. In the mid-1980's, FHWA's research focus shifted to the use of radar for the detection of subsurface distress in bridge decks. Under a 1985 research contract, a van-mounted radar system was developed for the FHWA for additional radar evaluation and testing. This van was loaned to State highway agencies and universities for use in their radar research efforts.

Highway departments have found that radar can provide useful information that was not previously accessible or available in a complete and continuous form. Because radar surveys are continuous rather than random, the radar technique can be much more objective and accurate than those methods presently used by agencies. Costs for performing radar investigations are generally considered reasonable. This article describes the theory, equipment, and applications involved in highway agencies' current use of GPR technology.


GPR operation requires an understanding of electromagnetic wave propagation and geophysical investigation concepts. Radio waves are those wavelengths on the electromagnetic spectrum between 0.001 m (0.04 in) and 10 m (33 ft). The waves travel through a vacuum at the speed of light--0.3 m (1 ft) per nanosecond.

When an electromagnetic wave encounters an interface between two materials of differing dielectric properties, one portion of the wave travels through the interface into the new material, and the rest is scattered or reflected in other directions (see figure 1). When the two materials have similar dielectric properties, most of the wave passes through the interface and little is reflected back. On the other hand, when the two materials have greatly different relative dielectric constants, a large reflection and a correspondingly small transmission occur at the interface. The relationship between the relative dielectric constants and the proportion of energy reflected at the interface is shown in equation (1): (1) [Mathematical Expression Omitted] where

p= reflection coefficient

[[Epsilon].sub.r1] = relative dielectric constant of medium 1

[[Epsilon].sub.r2] = relative dielectric constant of medium 2

In equation 1, when [[Epsilon].sub.r1] is less than [[Epsilon].sub.r2] p is negative; when [[Epsilon].sub.r1] is greater than [[Epsilon].sub.r2] p is positive. The reflected pulse may thus be in phase or out of phase with the signal emitted. This reflection phenomenon is illustrated in figure 2 and provides much of the theoretical basis for understanding specific waveform signatures encountered in actual applications.

As radar waves pass through different materials, their speeds vary because of changes in the electromagnetic properties of these materials. The velocity is a function of the material's dielectric constant, and the velocity of an electromagnetic wave as it passes through different materials varies in inverse proportion to the square root of the materials' relative dielectric constants. For example, the velocity of an electromagnetic wave through a material with [[Epsilon].sub.r] = 4 is half its velocity through air ([[Epsilon].sub.r] = 1) and twice its velocity through a material with ([Epsilon].sub.r] = 16.

Table 1 contains typical values of [[Epsilon].sub.r] for some materials often encountered in the highway environment. The extremes of these values are for air ([[Epsilon].sub.r] = 1) and water ([[Epsilon].sub.r] = 81). Ranges of values are shown for construction materials; actual values will depend on such factors as aggregate type, binder or cement source, density, and moisture content. Accurate determination of [[Epsilon].sub.r] for any given application can be found by coring a sample in the area of interest.
Table 1.--Representative dielectric constants for
construction materials
Material Dielectric Constant [Epsilon].sub.r]
Air 1
Water (fresh) 81
Water (salt) 81
Sand (dry) 4-6
Sand (wet) 30
Silt (wet) 10
Clay (wet) 8-12
Ice (fresh water) 4
Granite (dry) 5
Limestone (dry) 7-9
Portland Cement Concrete 6-11
Roller-Compacted Concrete 5-7
Asphaltic Concrete 3-5

Another concept that must be considered in radar work is the relationship between wavelength and frequency which defines the resolution capability. Since velocity through a medium remains constant, there is an inverse relationship between wavelength and frequency. Resolution capability is a function of wavelength, as shorter wavelengths can discern smaller or finer anomalies than can longer wavelengths. Longer wavelengths, however, penetrate deeper but only resolve larger discontinuities.

The depth of radar penetration depends on wave frequency. Higher frequencies can only penetrate shallow depths (within 0.6 m [2 ft] of surface at 900 to 1,000 MHz). Although the depth is limited, the wavelengths are small, permitting resolution of smaller anomalies. Conversely, lower frequencies can penetrate much deeper (typically 30 to 40 m [100 to 130 ft] at 100 to 300 MHz), but the wavelengths are longer resulting in a reduced resolution capability.

This is not a serious limitation in highway work since problems close to the surface tend to be smaller in size, thus requiring the higher frequencies. Since aberrations at greater depths tend to be larger, they can be resolved by the lower frequency, deeper penetrating waves. This dichotomy of increased depth with reduced resolution versus decreased depth with greater resolution results from electromagnetic wave theory as expressed in equation (2):

(2)[Lambda] = c. 1/f


[Lambda] = wavelength (distance)

c = velocity (speed of light)

f = frequency (1/time)

Interpreting the return signals relies on standard geophysical investigative techniques. In performing the analysis to determine whether an anomaly exists--and, if it does exist, its location and extent--the mechanics formula given in its basic form by equation (3) is used: (3) d= v.t where

d= distance

v= velocity

t= time

Since pulse velocity depends on the material's

dielectric properties, equation (3) can be rewritten


(4) d= c/[square root][[Epsilon].sub.r].t

In the GPR application of equation (4), the measured time, t, represents the time from transmission to the time of reception at the antenna or a "round trip" time. Thus the true travel time for the signal to the point of interest is half the measured time. Adding this correction to equation (4) produces equation (5), which expresses the time-distance relationship used in GPR technology: (5) d= c/[square root][[Epsilon].sub.r].t/2

With this time-distance relationship defined, area (delaminations, debonding), volume (voids), and thickness (overlay) can be determined; spacings (rebar) verified; and quantities (excavations) calculated.


The primary components of a GPR system are: * An antenna. * A transducer--this consists of a transmitter,

receiver, and timing and control electronics.

* One or more display devices--these may be

an oscilloscope, analog tape recorder, grey-level

chart recorder, or a video monitor.

Figure 3 illustrates these primary components. Typically, the antenna and transducer are located close to one another. The antenna, particularly, should be lightweight and maneuverable so that it may easily be positioned over the test area. The display devices are generally heavier and are often mounted in a van or cart for easy mobility. Alternatively, where real-time data interpretation is not a necessity, data can be stored on tape or disk to permit analysis in the office.

The GPR system also might include a mechanical or electrical unit (e.g., fifth wheel or electronic odometer) for precise distance and location information. A computer is an essential component for data storage, retrieval, and evaluation.

The overall system configuration will depend to a certain degree on the intended GPR use, but a typical schematic of a complete GPR system is shown in figure 4. The interior and exterior of a van-mounted radar system appear in figures 5 and 6. Instruments must be properly selected so surveys can be conducted with maximum accuracy.

GPR operates by generating the microwave signal and passing it from the control unit to the transmitter/receiver, through the antenna, and into the test surface. The reflected waves are received by the antenna and returned to the control unit for processing. The resultant stream of data can be further processed and displayed. There are two main types of radar, based on the modulation of transmitted waves, currently being used in highway surveys. These types are short-pulse radar and continuous wave, frequency-modulated radar.

In short-pulse radar, the signal generated by the transmitter is amplitude-modulated to produce pulses of energy. The transmitted pulses are extremely short, only about one nanosecond in duration. The spaces between the pulses of energy, however, are tens of thousands of times as long as the pulses. This length lets the reflected signal be received before another pulse is generated. Signal frequency is fixed and is dependent on the phenomenon being investigated; typically, the frequency employed by GPR systems ranges from 100 to 1,000 MHz.

In continuous wave, frequency-modulated radar, the signal is swept in frequency in sawtooth fashion over time. The frequency difference depends on the time delay between transmission of the signal and reception of the corresponding echo. Thus, this can be used for distance measurement. The speed of frequency variation must be slow enough so that frequencies of the return from the surface and subsurface are essentially the same.

Antenna design is an important factor for both

types of radar. The antenna serves to: * Provide a smooth electromagnetic transition

from the transmitter to the environment. * Direct the radiated energy into the ground in a

desirable pattern.

In general, the antenna directs energy in all directions, with most directed into the ground. The shape of the beam directed into the ground is of interest. If the beam is broad, a relatively large area--or "footprint"--can be covered on the ground. However, the return energy may be of low intensity since the signal is attenuated over the large area. Antennas are generally classified as ground-coupled (in contact with the surface) or air-coupled (suspended above the surface, typically 0.3 m [1 ft]), depending on the condition that is to be investigated. Pavement surveys are usually conducted with air-coupled or air-launched antennas to take advantage of their ability to scan the surface rapidly, thereby reducing or eliminating the need for traffic control. Geophysical explorations, on the other hand, often use the ground-coupled systems that are more transportable and generally constructed for field use.


Ground-penetrating radar has been successfully used to identify many problems associated with highway structures and is gaining acceptance as a technique to replace older, less reliable methods. These methods are subjective and often inconclusive, prompting the need for rapid, objective, and nondestructive methods for surveying structural conditions. Moreover, some of these methods (for example, the chain drag) are too slow and costly to be used on large sections of interstate highway. Geophysical, pavement, and bridge investigations can thus all be conducted more reliably and efficiently using radar.

Delamination detection

Delaminations, a major cause of bridge maintenance problems, are separations of the concrete around the rebar layer due to the forces resulting from corrosion. An early test of GPR's capabilities was an evaluation of GPR as a network-level tool that could be used to quickly assess the general condition of bridge decks with respect to delaminations. (1)[1]

Inservice bridge decks were inspected at a speed of 65 km/h (40 mi/h) without closing the decks to traffic. The results of the evaluation were very encouraging, as distressed areas with a longitudinal dimension of 0.6 m (2 ft) or more could be detected. The surveys were performed using a strip chart recorder to obtain the type of display shown in figure 7. Data interpretation, however, was subjective, being based primarily on qualitative differences in apparent wave velocity and/or the attenuation of the inspection wave.

A research program sponsored by five New England States led to the further development and verification of GPR for bridge deck evaluation. (2, 3) The program involved both network-level surveys (to assess general condition; 30-percent GPR coverage) and detailed project-level surveys (to obtain a mapping of unsound areas; 100-percent GPR coverage). The focus was on asphalt-overlaid bridge decks where the subsurface distress included freeze-thaw damaged concrete as well as delaminated concrete.

Comparisons of GPR results with the traditional coring method of analysis showed GPR predictions of deck deterioration were within [+ or -] 4.4 percent of the actual proportion of deck deterioration. Survey speed, which varied up to 80 km/h (50 mi/h), had no significant influence on predictions; thus, both network and project-level surveys can be performed at high speed, if desired. For network-level surveys, 20 or more bridge decks can easily be surveyed in 1 day, depending mainly on their relative location.

Perhaps the biggest step forward in the New England research was the use of improved automated data interpretation techniques. The researchers developed quantitative analysis techniques to predict deterioration from the variations in the concrete dielectric constant as computed directly from the radar waveforms. Besides providing a better separation of return signals, the computer processing of signals permits noise and extraneous information to be removed. Also, better methods of displaying the information have now eliminated the obscurity that images such as figure 7 once projected. Some current systems provide the capacity to delineate the problem areas using a mapping technique as illustrated in figure 8. (Note that figures 7 and 8 are of different sites, and no comparison between them is intended.)

Voids beneath pavements

Voids often develop beneath concrete pavements because of consolidation, subsidence, and erosion of the support material. Many of the voids occur beneath joints where water enters the foundation soil and, aided by the pumping action of heavy traffic, carries out fine materials. A National Cooperative Highway Research Program (NCHRP) study in 1979 was the first to demonstrate the feasibility and practicality of using GPR to locate and measure voids beneath pavements. (4) The study showed GPR to be capable of spacially locating voids to within [+ or -] 150 mm ([+ or -] 6 in) with a depth distinction of [+ or -] 13 mm ([+ or -] 0.5 in).

Numerous void surveys have been performed for State highway agencies since the NCHRP research. These experiences have been somewhat mixed, as excessive amounts of water in the subbase and subgrade tend to disrupt the radar signal and give false readings. (5, 6) However, recent improvements in equipment and data interpretation techniques have enabled the detection of voids as small as 3 mm (0.12 in).5, 6, 7) The average thickness of a void and an estimate of its area can be calculated to determine volume and the quantity of grout needed to fill the void for pavement stabilization. The estimated area can be determined more accurately when a three-antenna (or even a four-antenna) radar system is used. GPR can be very useful not only in detecting and locating voids before planning the stabilization of a concrete pavement, but also for checking on the effectiveness of complete stabilization. (8)

Pavement thickness

Determining pavement layer thickness is one of the simplest applications of GPR. The procedure, detailed in the American Society for Testing and Materials (ASTM) Standard D 4748-87, can be used to determine the thickness of newly built pavements and overlays (to ensure thickness is as specified) or of older pavements (to obtain structural values or other inventory information). (9) The procedure has some limitations. To determine the thickness of any individual pavement layer, the difference between the relative dielectric constants of adjacent layers must be large enough to reflect a sufficiently large echo from the interface. Also, when determining the thickness of reinforced concrete pavements, reflections from the bottom of the slab may sometimes be too weak to identify. (10) Further, the procedure is not recommended for wet pavements or pavements that exhibit a large variation in moisture content.

Despite these limitations, the advantages of determining thickness with GPR are considerable. With GPR, layer thickness to a depth of 0.6 m (2 ft) can be measured to an accuracy of [+ or -] 6 mm ([+ or -] 0.25 in). In contrast, the standard deviation of core thickness measurements is about 6 mm (0.25 in) for portland cement concrete and can vary from 5 to 19 mm (0.2 to 0.75 in) for bituminous concrete depending on design thickness. Therefore, a large number of core samples must be taken to provide pavement thickness information with the desired degree of confidence. Any reasonably accurate nondestructive thickness method that permits 100-percent inspection has the potential for much future application.

Other applications

Other highway applications for GPR are being investigated. These include: * Determining the degree of hydration of cement. * Determining the water content of new concrete. * Locating reinforcing bars and wire mesh. * Detecting dowel misalignment. * Detecting debonding of overlays. * Evaluating scour around bridge piers. * Back-calculating layer moduli (in conjunction

with the falling weight deflectometer).

GPR has also been used for geophysical investigations such as profiling the bottoms of lakes and rivers and locating rock formations and fractures, cavities, abandoned mines, archeology sites, pipes, sewers, cables, tanks, and ice lenses. Geophysical applications have traditionally--and very successfully--been conducted using a strip chart recorder. Geophysical examinations can be easily identified on strip chart recorders because features having surface areas of more than 0.3 m (1 ft) develop an easily identified pattern. Smaller features are frequently much more difficult to discern unless computer enhancement is used.


There are many methods for conducting surveys and determining the properties of a feature before extensive maintenance or repair is considered. The older methods such as chain drag and coring are often time consuming, costly, or destructive, with definite limits on the amount of sampling that can be performed. The objective of sampling surveys should be to determine the most information at the least cost. This need becomes especially compelling given the vast amount of infrastructure evaluation that must be conducted over the next decade.

Ground-penetrating radar affords great potential as an expedient and cost-effective evaluation tool. Initial testing and evaluation have proved successful. As a result, ASTM has developed a test method for determining the thickness of bound pavement layers, and other testing standards are currently being developed. The future use of ground-penetrating radar in highway work will depend on the extent to which State highway agencies adopt it as a viable construction and maintenance tool.


(1) Richard P. Joyce, Rapid Nondestructive Delamination Detection, Publication No. FHWA-RD-85-051, Federal Highway Administration, Washington, DC, April 1985. (2) Kenneth Maser, "Bridge Deck Condition Surveys Using Radar: Case Studies of 28 New England Decks," Transportation Research Record 1304, Transportation Research Board, Washington, DC, 1991. (3) Infrasense, Inc, Bridge Deck Evaluation Using High Speed Radar, New Hampshire Department of Transportation, Concord, NH, November 1991. (4) W.J. Steinway, et al, "Locating Voids Beneath Pavement Using Pulsed Electromagnetic Waves," National Cooperative Highway Research Program Report 237, Transportation Research Board, Washington, DC, November 1981. (5) John A. D'Angelo, Portland Cement Concrete Pavement Stabilization, Federal Highway Administration, Washington, DC, February 1986. (6) Floyd Petty, Using Radar to Detect Voids and Other Anomalies Under Concrete Pavement on a Section of 1-75 from 1-40 to 1-275 in Knox County, Experimental Project Number 9, Tennessee Department of Transportation, Nashville, TN. (7) Penetradar Corporation, Radar Inspection of Interstate 66, Project Number 6522-DS, Virginia Department of Transportation, Richmond, VA, May 1987. (8) Gerardo G. Clemena, et al, Use of Ground-Penetrating Radar for Detecting Voids Underneath a Jointed Concrete Pavement, Publication No. FHWA/VA-86-R36, Virginia Highway and Transportation Research Council, Charlottesville, VA, May 1986. (9) Standard Test Method for Determining the Thickness of Bound Pavement Layers Using Short-Pulse Radar, Standard Designation: D 4748-87, American Society for Testing and Materials, Philadelphia, PA. (10) Gerardo G. Clemena and Richard E. Steele, Measurements of the Thickness of In-Place Concrete with Microwave Reflection, Report No. FHWA/VA-88-16, Virginia Transportation Research Council, Charlottesville, VA, April 1988.

Kevin Black is a highway engineer in the Materials Branch, Construction and Maintenance Division of the Federal Highway Administration (FHWA). Formerly, he was with the Pavements Division in the Office of Research and Development.

Peter Kopac is a program manager for FHWA's Nationally Coordinated Program of Highway Research, Development, and Technology. Since 1977, he has managed, monitored, and contributed to numerous studies involving radar.
COPYRIGHT 1992 Superintendent of Documents
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Black, Kevin; Kopac, Peter
Publication:Public Roads
Date:Dec 1, 1992
Previous Article:ATMS technology - what we know and what we don't know.
Next Article:The Highway Economic Requirements System: an introduction to HERS.

Related Articles
Developing NDE Technologies for Infrastructure Assessment.
Handheld Standoff Minefield-Detection System (HSTAMIDS).
Advanced ground-penetrating radar may be used to spot tunnels dug by criminals.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters