Impact-Echo and Impulse Response stress wave methods:

Advantages and limitations for the evaluation ofhighway pavement concrete overlays.

Allen Davis1, Bemhardt Hertlein2, Malcolm Lim3 and Kevin Michols3

ABSTRACT

Concrete overlays with thickness ranging between 25 mm and 300 mm are frequently used to
restore and strengthen existing concrete pavements and bridge approach slabs. Differences in the
strengths and elastic moduli ofthe overlay and the substrate, as well as the cleanliness and roughness of
the interface between the two layers affect the medium and long term performance of these structures.
Debonding at the interface, excessive tensile stresses at the base of the overlay and delamination within
the upper layer are commonly occurring problems. Ifthese defects are not detected and corrected in good
time, the deterioration ofthe overlay under the action ofheavy axle loads is rapid and becomes expensive
to :fix. Nondestructive methods are required to identiIy budding problems ofthe type described above, by
surveying overlay systems quickly and economically. Stress wave methods for flaw detection in concrete
structures and foundations have shown great promise in recent years. The Impact-Echo (I-E) test has
been applied successfully to many diverse concrete material problems. The Impulse Response (IR) test is
proven in the detection of flaws in deep concrete foundations, as well as the location of poor support
conditions beneath and delaminations within concrete slabs on grade. This paper presents a case study
where both methods were used to examine a stepped concrete overlay on approach slabs to bridge decks
on a heavily trafficked Interstate highway. The two test methods are briefly described, and a comparison
is drawn emphasizing the advantages and disadvantages ofboth techniques.

Keywords: concrete pavement overlay, nondestructive testing, stress wave methods.

1. INTRODUCTION

In January 1995, the authors were able to examine the condition ofreinforced concrete approach
slabs by nondestructive stress wave methods at seven different bridge decks on a heavily trafficked
Interstate highway in the eastern USA. The original slabs were constructed in a stepped configuration
along their longitudinal direction, with slab thickness between 375 mm (15 inches) and 625 mm (25
inches), in 125 mm (5 inch) steps. The approach slabs were supported on soil fill. Each bridge approach
slab is approximately 6 m (20 feet) long and 4 lanes wide, with one slab width per lane. The stepped areas
had lirst been overlaid by asphalt, which was subsequently removed and replaced by reinforced concrete
overlay, intended to be bonded to the underlying slab. These concrete overlays now exhibit cracking,
delamination and spalling. It was believed that much ofthe deterioration was associated with debonding
ofthe overlaid stepped areas from the underlying approach slab.

Nondestructive methods using stress waves have been used to examine debonding of concrete
layers in previous studies. However, a comparison between the two principal practical methods
available (Impact-Echo (I-E) and Impulse Response (IR)) has not been reported to date. The opportunity
1
Madsen, Kneppers & Associates, 1245 Brickyard Road, Suite 520, Salt Lake City, Utah 84106
2
STS Consultants, Ltd., 1415 Lake Cook Road, Deerfield, Illinois 60015
Construction Technology Laboratories, Inc., 5420 Old Orchard Road, Skokie, Illinois 60077

88 / SPIE Vol. 2946 O-8194-2350-5/96/$6.OO

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 was afforded in this study to use both test methods in parallel, and to compare the relative merits of the
two techniques.

2. TEST METHOD DESCRIPTIONS

2. 1 Impact-Echo

The I-E technique applied to the detection oflayer debonding is fully described in references 4-7.
A short summary ofthe method and test equipment is given here.

The equipment used was the Germann Instruments DOCter system, which comprises a mechanical
impactor source and an electromechanical receiving displacement transducer located in a common unit.
The spherical impactor generates a broad band stress pulse, or P-wave, typically with a maximum
compressive stress of from 0.5 to 0. 1 MPa at the surface to a depth of 25 mm. Below that depth, the
compressive stress decreases rapidly, falling to between le-02 and le-03 MPa between 150 mm and 200
mni7 Reflected stress waves are detected by the displacement transducer, and the time-displacement
response is converted to a frequency response using a Fast Fourier Transform (FFT) algorithm.

The reflected stress wave from the base of a concrete slab supported on subgrade with a lower
elastic modulus is a tension wave, as is the reflected P-wave from any debonded surface between
concrete overlay and substrate. The time taken, dt, for the return wave to reach the receiving transducer
is a direct fimction ofthe P-wave velocity in the concrete, c , and the thickness ofthe tested element, H:

dt= 2H/c (1)

The plate thickness (or depth to a debonded interface) can be determined more accurately from
the displacement-frequency response curve, by locating the characteristic frequency, f for the particular
plate thickness on the response curve:

H=c1,/2f (2)

The P-wave velocity is either obtained by calibration on elements ofknown thickness, or by using
two displacement transducers at different distances from the impactor, and measuring the Rayleigh, or
surface wave velocity. Typical I-E records from this case study for debonded overlays are given in
Figures 1-2. Figure 2 also has a peak at very low frequency, which is equivalent to a response from the
slab in the bending mode. This feature is used in the Impulse Response test described below.

2.2 Impulse Response

The JR test (often referred to previously as the Transient Dynamic Response, or TDR test) also
uses a low strain impact to send a stress wave through the tested element. However, the impactor takes
the form of a 1kg sledge hammer with a built-in load cell in the hammer head. The maximum compressive
stress at the impact point in concrete is directly related to the elastic properties of the hammer tip. Typical
stress levels range from 5 MPa for hard rubber tips to more than 50 MPa for aluminum tips. This greater
stress input than for the I-E test means that plate structures behave differently. At relatively low

SPIE Vol. 2946 / 89

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 frequencies (0- 1 kHz) the plate responds to the 1K hammer impact in a bending mode. The response to
the input stress is nonnally measured using a velocity transducer (geophone). This receiver is preferred
because of its stability at low frequencies, and its robust performance in practice. The equipment used in
this project was developed by STS Consultants, Ltd.

Both the time records for the hammer force and the geophone response are processed using the
FFT algorithm, and the resulting velocity spectrum is divided by the force spectrum to obtain a transfer
function, referred to as the plate Mobility. The resultant test graph of Mobility plotted against frequency
over the 0- 1 kHz range contains information on the condition and integrity of the concrete in the slabs,
obtained from the following parameters:
S Dynamic Stiffness: The slope of the portion of the Mobility plot from 0- 100 Hz defines the
compliance, or flexibility of the test point for a normalized force input. The inverse of
compliance is the dynamic stiffliess ofthe slab at the test point. This can be expressed as:
Stiffness f [concrete quality, slab thickness, slab support conditions].
. Mobility and Daniping the slab's response to the impact-generated elastic wave will be
damped by the plate's inthnsic rigidity (body damping). The mean mobility value over the 0.1-
1 kHz range is directly related to the density and thickness of the plate. A reduction in plate
thickness corresponds to an increase in mean mobility. When total debonding of the upper
layer is present, the mean mobility reflects the thickness ofthe upper (debonded) layer; that is,
the slab becomes more mobile. Also, any cracking or low density concrete in that upper layer
will reduce the damping, and hence the response curve stability, ofthe mobility plot over that
frequency range.
. Peak/mean Mobility ratio: when debonding is present, the upper debonded layer controls the
IR result. In addition to the increase in mean mobility, the dynamic stifiuiess decreases greatly.
The peak mobility below 0. 1 kHz becomes appreciably higher than the mean mobility from
0. 1- 1 kHz. The ratio of this peak mobility to the mean mobility value is an indicator of the
presence and the degree of debonding.

Examples ofthese features are given in Figures 3-5, representing test results from this case study.
Figure 3 (test B4) shows a typical mobility plot for a well bonded overlay, with a relatively high stiffliess
and an average mobility of approximately 4e-07 rn/s/N.

Figure 4 (Test C4) is a mobility plot for an overlay which is showing incipient debonding with a
higher mobility peak between 0- 100 Hz, a resulting lower stiffness, but a similar average mobility to test
B4 in Figure 3.

Figure 5 shows a totally debonded overlay with very high initial mobility and low stiffliess, and
20e-07 rn/s/N average mobility. The peak/mean mobility ratio for this test point is 2.4. The plot for test
B4 is superimposed on Figure 5 to demonstrate the remarkable difference between the curves for a sound
overlay and a debonded layer.

3. TEST PROGRAM

3.1 Feasibility and correlation testing

90 / SPIE Vol. 2946

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 Initial testing with both methods was performed on a trial section offout slabs on one bridge:
a) to demonstrate the ability ofthe test methods to detect debonded areas, and
b) to correlate the test results obtained with cores taken at locations exhibiting different
features reported by the stress wave tests.
This prooftest series was carried out during one night, with traffic diverted around the test slabs.

The concrete surface was considerably roughened by wear, and was grooved for skidding
resistance. As a result, it was often difficult to guarantee a good coupling between the I-E displacement
transducer and the concrete surface, and some test points had to be retested several times. The very small
contact area ofthe I-E impactor also meant that the point ofinipact could not be located immediately at
the crest of a concrete groove, because the impactor would crush the relatively weak surface concrete at
that point. The IR test equipment with a 50 mm diameter rubber hammer tip and a 50 mm diameter
geophone base adapted to the grooved surface was not as sensitive to surface variability. As a result, the
IR testing output was much faster for a given area tested.

Six cores were taken by the local Department of Transportation at selected test points, and Table
1 summarizes the correlation ofthe core findings with the NDT observations.

Table 1. Correlation ofNDT results with cores

LOCATION I-E Results JR Results CORE SUMMARY

delamination depth Stiffness (MN/mm)
(mm) & Mean Mobility
1 140 0. 18 / 18 Delaminated at interface between two
different concretes at 133 mm
2 300 0.27 1 15 Delaminated at interface between two
different concretes at 290 mm
3 1 15 0.23 / 40 Large void in core between 75 and
125 mm from core top
4 150 0. 13 / 10 Broken at interface between two
different concretes at 150 mm depth
5 165 0.3 5 / 5 Intact core with epoxy resin layer
between 140 mm and 280 mm depth
6 140 0.07 / 30 Core separated at 140 mm, with
drilling water loss. Layer ofioose
aggregate present in hole.

A good agreement was obtained for both tests with the core observations. The dynamic stiffliess
for the IR test decreased with increasing severity of the overlay debonding, and the mean mobility
increased with decreasing effective layer thickness. The depths to layer separations were clearly recorded
by the I-E test, except for Location #5, where the depth to the intact, epoxy-bonded interface was seen.

In view of the performance of the equipment during this preliminary test series, it was decided
that all subsequent testing would be done using 1K equipment to locate and measure the area of potential
debonded areas, followed by I-E testing to quantify the depths ofthe debonded interfaces.

SPIE Vol. 2946 I 91

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 F If If If If I F I I
flMP 1

29 JIN 95 17:44:39

Peak kHz ri

THE 2.62 .61O 328

TOP 58Z 188Z

• 1824: :2: uSEC: :

..:.:
13.5 27.8 48.5 54.8 67.5 kHz

Figure 1. I-E test response from debonded overlay

IMP 1
29 JN 95 16:59:82 :
•
• •: : •
•
Peak kHz n
THE 2.62 8.618 3280

BEG
/
REFLECTION

TOP 58X

:1Q24:&:2:uSEC:
188Z

13.5 27.8 48.5 54.8 67.5 lcHz

Figure 2. I-E test response with reflection and bending peaks

92 / SPIE Vol. 2946

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 . .
Site INTERSTATE
•1
Test fl:B4 01/31/95
6 .75
. Type Slab

:
1;..
5 25
..../
'4

z
x
3 .75 ;LJ±2/ :: . :
0
Lfl

E 2 .25

. 750
.75 MN/riri
a
z 72.
80 0 240. 400. 60.
FREQUENC'Z < Hz )

Figure 3. [R test response from sound slab/overlay interface

Site: INTERSTATE
. Test $*:C4 01/31/95
6.30 ..

Type: Slab

H
— .700 . .
E .58 MNJ,inE
C
E ___ ___ ___ ___ I
80.0 240. 400. 560. 720.

FREQUENC'! ( Hz )

Figure 4. JR test response fron incipient overlay debond

SPIE Vol. 2946 1 93

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 Site INTERSTATE
I Test ft:1 01/31/95
54.0 I
LType: Slab
p 42.8
r.

4
x
z\
3.0

18 0

600
80.0 240. 400. 560. 720.
FREQUENCX! ( Hz )

Figure 5. IR test response from debonded area

Direction
of traffic

LANE 3

Surface Cracks
lnterp(eted Defamination

Figure 6. Mapping of debonded areas

94 / SPIE Vol. 2946

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 3.2 Production Testing

The remaining approach slabs to the seven bridges were tested during the following two nights. A
total of 54 slabs was tested at all bridges. The test results are summarized in Table 2, and a typical area of
debonding for one bridge approach is shown schematically in Figure 6.

The results indicated that 9 slabs were debonded over their total area, 27 slabs had separations at
50% or more of the area tested, 6 slabs were less than 50% debonded, and 12 slabs showed no
debonding. The measured depth of debonding generally corresponded to the recorded depth of the
interface between the overlay concrete and the original slab.

Table 2. Summary ofProduction Test Results

Bridge No. No. of IR No. of I-E % debonding Comments

points points
1 62 49 0 - 100 Testing on wheel path
2 1 12 127 0 - 100 Testing on wheel path & lane
center
3 85 75 50 - 100 Testing on wheel path
4 1 14 1 13 44-100 Testing on wheel path and lane
center
5 90 65 0-67 Testing onwheelpath
6 57 57 0- 100 Testing on wheel path
7 81 52 0-100 Testing on wheel path

The I-E tests were used to verify the IR test results, and to measure the depth of debonding. The
percentage debonding for each slab was calculated by interpolation ofthe JR test results.

Table 3. Comparison ofthe two test methods

Test Method Impact-Echo Impulse Response

Advantages • Apparatus very light and portable, • Apparatus is robust and portable
battery operated • Easy data storage and data recall
• Easy data storage and data recall for analysis and reporting
for analysis and reporting • Not affected by relatively rough
• Measures direct depth to reflected concrete test surfaces
features • Fast testing (each test point covers
• Single operator, 15-20 sec/test an area of2O square feet)
Disadvantages • Good sensor coupling and inipactor • Apparatus requires electricity
points are difficult to achieve on supply (generator)
rough and grooved surfaces • No depth measurement to
• each test measures the layer depth debonded surface
only immediately below the test • Requires skilled data interpretation
point

SPIE Vol. 2946 I 95

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms

 4. DISCUSSION

The complementary advantages ofthe two test methods (summarized with their disadvantages in
Table 3) were combined to give a fast, certain evaluation ofthe state ofthe interface between the original
concrete approach slabs and their concrete overlays. The IR test has the disadvantage of not giving the
depth to the debonded layer, but does identify wider areas of delamination around each test point. The
relatively rough working surface of the overlays resulted in difficult testing conditions for the I-E
equipment commercially available at the time of these tests, with uncertain coupling of the displacement
transducer and awkward positioning for the inipactor. The authors understand that new I-E equipment is
now available to overcome these problems. The high frequency, reflected signal from the I-E method
gave the thickness ofthe reflected interface immediately below the test point, but did not give information
on the state ofthe interface on either side ofthis point.

The combination ofthe two test methods provided an extremely cost-effective and rapid approach
for the evaluation of the 54 approach slabs in less than two nights' work on the Interstate. Finding the
right combination oftest techniques for a specific problem is the key to success in nondestmctive testing.
In the authors' experience, it is unusual to find a single test method that can completely resolve the issues
raised by any project.

5. REFERENCES

1. A. G. Davis and B.H. Hertlein, "Nondestmctive testing of concrete pavement slabs and floors with
the transient dynamic response method," Proc. mt. Conf Structural Faults and Repair, London, July
1987, Vol 2, pp 429-433.
2. T.P. Scull and AG. Davis, "Experiences in nondestructive testing with impulse radar and impedance
methods in the evaluation of concrete highways," 6th mt. Symp. On Concrete Roads, Madrid,
October 1990, pp 103-112.
3. A.G. Davis and B.H. Hertlein, "Assessment of bridge deck repairs by a nondestmctive technique,"

ASCE Structures Conf, 2nd USA-European Worl&iop on Bridge Evaluation, Repair and
Rehabilitation, Baltimore, April 1990, pp 229-233.
4. M. Sansalone and N.J. Carino, 'Detecting delaminations in concrete slabs with and without overlays
using the Impact-Echo method," AClMaterials Journal, VoL 86, No. 2, March 1989, pp 175-184.
5. C. Cheng and M. Sansalone, "Impact-Echo response of concrete plates containing delaminations -
numerical, experimental and field studies," Materials and Structures, RILEM, Vol. 26, No. 159, June
1993, pp 274-285.
6. J-M. Lin and M. Sansalone, "impact-Echo studies of interfacial bond quality in concrete: Part I -
effects of unbonded fraction of area," ACI Materials Journal, VoL 93, No. 3, May-June 1996, pp
223-232.
7. J-M. Lin, M. Sansalone and K Poston, "Impact-Echo studies of interfacial bond quality in concrete:
Part II - effects of bond tensile strength," ACI Materials Journal, VoL 93, No. 4, July-August 1996,
pp 3 18-326.

96 / SPIE Vol. 2946

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/21/2013 Terms of Use: http://spiedl.org/terms