Non-Destructive Testing of joints in precast element structures

Claus Germann Petersen, M.Sc., MBA, Germann Instruments A/S, Denmark

([email protected])
Hugo Dolan Orozco Recillas, M.Sc., MBA, Germann Instruments A/S, Denmark
([email protected])

Keywords: Precast elements, keyed joints, grouting, quality control, impact-echo, ultrasonic-echo.

Abstract

Strength and durability of precast concrete structures are determined to a great extent by the strength and
durability of the connections between the precast elements. Most common problems are improper
grouting and packing of the filling material, e.g. by using shotcrete or stiff or improper aggregate size
mortars, as well as excess of shrinkage or inadequate bond to the boundaries.

Therefore, ensuring the proper filling of joints is a must to avoid undesirable effects such as cracks, steel
corrosion, lack of stiffness or most importantly loss of load carrying capacity. However, it is not easy to be
sure a correct workmanship has been done unless destructive techniques are performed, which means a lot
of coring or drilling of holes.

This paper presents a case study of NDT testing, carried out on vertical joints of precast concrete wall
elements improperly grouted to show an easy and totally non-destructive way to carry out quality control
or inspection survey by means the impact-echo method or, alternatively, by using an ultrasonic-echo
tomographer. Unlike other traditional NDT methods like UPV (ultrasonic pulse velocity), these techniques
require access to only one face of the element and are able to determine the depth of anomalies fast and
reliable.

Introduction

In order to form a monolithic structure using precast concrete wall elements, vertical shear connections are
usually made as keyed joints reinforced with overlapping U-bars. They form a cylindrical core in which a
locking bar is positioned, and all the joints are grouted with mortar. Another alternative is replacing the U-
bars with high strength looped wire ropes since their flexibility makes the installation of wall elements
easier. The wire ropes are usually pre-installed in “wire boxes” embedded in the precast walls with a rebar
in the middle [1].

While these kinds of joints have proved to be structurally appropriate, sometimes the grouting procedure
at the job site is not properly done [2]. A vertical shear connection joint with defects in the filling mortar can
lead to stress concentrations that may induce undesirable plastic deformations in the U-bars or a brittle
failure in the looped wire ropes due to their lack of yield plateau in the stress-strain relationship [1].
Furthermore, even if the load carrying capacity is not compromised, the structure’s service life might be
because cracks and voids make concrete susceptible to water intrusion and consequently, steel corrosion

1
may become a problem or ice can be formed during freezing temperatures, expanding and cracking the
surrounding concrete.
A demonstration project of quality inspection of joints with looped wire ropes was performed in a multi-
story building where precast concrete walls were used extensively. Knowing that the grouting process is
very susceptible to human error, looking for possible defects is critical. Two instruments were used, the
DOCter, which is based on the Impact-Echo method, and MIRA, an ultrasonic-echo tomographer. The stress
wave NDT assessment methods are totally sensitive to this type of defects and fully accomplished the
investigation. Results are discussed and compared. Cores were extracted for verification by courtesy of the
contractor Mr. Jack Pedersen, MT Hoejgaard, Copenhagen, Denmark.

The DOCter Impact-Echo principle [3],

A short mechanical impact is produced on the concrete surface with a small steel sphere. This impact
generates stress waves that propagate into the concrete. When the P-wave (the longitudinal wave) reaches
a material (e.g. air, steel, soil, etc.) with a different acoustic impedance (wave speed x density), it is
reflected and travels back to the surface where the impact was generated. The reflective surface is
normally the back side of the concrete member but if this concrete has embedded objects or defects such
as voids, honeycombs, cracks, etc., they also act as reflective surfaces. A sensitive displacement transducer
next to the impact point picks up the displacement of the surface due to the arrival of the P-wave. The P-
wave is then reflected back into the member and the cycle begins again. Thus the P-wave undergoes
multiple reflections between the two surfaces and the recorded waveform of surface displacement has a
periodic pattern that is related to the depth of the reflective surface for a given wave speed. This
displacement waveform is transformed into a frequency domain to produce an amplitude spectrum that
shows the predominant frequencies in the waveform. The analysis of these frequencies is the basis of the
Impact-Echo method which is governed by the equation shown in Figure 1. The DOCter instrument is
illustrated in operation in Figure 2. Each measurement takes only 1 second.

𝐶𝑝
𝑓=
2𝑇

Fig. 1. Impact-Echo principle and main equation.

f = Frequency (kHz)
Cp = P-Wave speed (m/s)
T = Travel path (mm)
Fig. 2. DOCter Impact-Echo in progress.

2
The MIRA ultrasonic-echo tomographer principle [4].

MIRA is a portable instrument based on the ultrasonic pulse-echo method using transmitting and receiving
transducers in a "pitch-catch” configuration. One set of transducer sends out a Shear Wave pulse and all
other sets of transducer receive the reflected pulse. The time from the start of the pulse until the arrival of
the echo is measured. For a known Shear Wave speed, Cs, the depth of the reflecting interface can be
calculated. The instrument itself measures automatically the Shear Wave speed.

The 4 x 12 transducer array antenna produces many transit time measurements during each test that are
later analyzed using the synthetic aperture focusing technique (SAFT) in real time to reconstruct a 2D image
of the cross section below the instrument. It takes about 3 to 4 seconds to complete data acquisition and
data processing at each location.

The data captured can later be transferred to a computer with a dedicated software to create a 3D image
that allows views of different slices of the reconstructed internal structure to identify the locations of the
reflecting interfaces, which can be the opposite side of the concrete member (back wall reflection),
reinforcing bars or concrete-air interfaces such as voids, delaminations, honeycombings or other defects
alike. Figures 3 and 4 show the instrument.

Fig. 3. The MIRA ultrasonic-echo tomographer. Front face at the left with the display showing
a cross section scan. To the right, the 4 x 12 transducer array antenna.

Fig. 4. The MIRA ultrasonic-echo tomographer testing in progress.

Testing approach of the joints

Both methods described above were used in a demonstration project that consisted of testing two wire
loop joints of precast wall elements during the construction of the Life Science & Bio Engineering Building
at the Technical University of Denmark, Campus Lyngby: Joint # 2 in the Ground Level (2GL) and Joint # 2 in

3
the First Level (2FL). The joints were grouted with a fluid mortar, but there were suspicions that a bad
procedure of grouting could have been done. The focus of the testing was to find the position and size of
possible defects to be repaired.

A vertical column of 17 testing points, with 100 mm spacing in between, were selected for the joint in the
ground floor and 18 points, same spacing, for the joint in the first floor. Tested concrete walls were 200 mm
thick (see Fig. 5).

Fig. 5. Lateral view of the wall and testing area

Testing with DOCter

In order to obtain the P-wave wave speed (Cp) of the mortar, the instrument was calibrated in an area of
the wall with known good quality filling (see Figure 6) and the wave speed was estimated to be 3,800 m/s
(a reasonable P-wave speed for the mortar) from the equation shown in Figure 1:

𝐶𝑝
𝑓= 𝐶𝑝 = 2 𝑇 𝑓 = 2 (200)(9.5) = 3,800 𝑚/𝑠 with f = 9.5 kHz, measured by DOCter (see Fig. 7)
2𝑇
Test core
Cross section of the joint
Steel rebar
Steel wire rope
200 mm

Steel profile or
“wire box”
Concrete wall,
front face
Filling material

Fig. 6. A solid concrete core extracted from the joint for verification together with a
drawing of the joint’s cross section.

4
 a. Solid joint

Figure 7 shows the frequency spectrum of the verification test in joint 2FL which was the reference for
comparing the rest of the testing points. It can be seen that the main peak is indicating the 9.5 kHz value
used for the calculation of the wave speed velocity and that there are no other major peaks indicating
other reflective surfaces than the one from the back side of the wall.

Fig. 7. Recorded spectrum of a filled-out joint section. One single main peak is
displayed, 9.5 kHz in this case (red line), corresponding to a thickness of 200 mm for
a P-wave speed of 3,800 m/s. A 150 mm diameter core was taken from the section
for verification, shown adjacent.

b. Joint with a limited defect

For a limited size of defect, reflections occur mainly from the back side of the member but partly also from
the internal small voids. Because the P-wave runs around the defect(s), it will travel longer and the main or
“solid” frequency will drop in the spectrum to a lower frequency value. Yet, the other reflections coming
from the defect itself will produce a high amplitude peak on the right side of the frequency spectrum. If this
peak is big enough to identify it, the depth of the defect can be estimated using the equation in Fig. 1.

22

Fig. 8. The frequency of the main peak has dropped to 7.5 kHz (blue line). The red line shows the
position at 9.5 kHz where the main peak should have been according to Fig. 7. The amplitude in
this point is still high, meaning that part of the P-wave still runs through solid material. Notice
also the 22 kHz frequency peak coming up, related to a larger possible defect. A 150 mm core
was taken from the section for verification shown adjacent.

5
Figure 8 shows this case with data obtained from joint 2FL. A mixed frequency shows up, partly the solid
frequency at 9.5 kHz and partly a frequency at 7.5 kHz related to a longer travel path than the 200 mm
thickness of the wall. The interpretation is a partly injected joint with small voids, which was confirmed by
the core drilled out shown adjacent to the spectrum. A smaller frequency peak at 22 kHz is also showing up
at the right side of the spectrum, however it is not big enough to identify it clearly in this case.

It can be also noticed that there are some other peaks on the left side of the spectrum (low frequency
region), most certainly caused by e.g. the Raleigh Wave or flexural displacement. They are not important
keeping a focus only on the main peaks of the solid frequency during the analysis, a drop in this frequency
and higher frequency peaks coming up.

c. Joint with large defect.

This time the drop of the main peak of the spectrum is followed by a large second main peak in the higher
frequency region, meaning that the defect is big enough to clearly transmit a reflection by itself for the
impactor selected. Using this high frequency with the same equation, the depth of the defect can be
calculated as follows:
𝐶𝑝 3,800
𝑇= = = 86 𝑚𝑚
2𝑓 2 (22)

Figure 9 shows this case with data obtained from joint 2FL.

86 mm

Fig. 9. The blue dotted line at 7.5 kHz indicate a longer travel path than the 200 mm related to
a frequency of 9.5 kHz. The P-wave runs around a defect. A second main peak in the spectrum
shows reflection from a large defect at 22 kHz related to the depth of the defect. A 150 mm
core was taken from the section for verification and it can be seen that the depth of the defect
was predicted accurately.

As a summary, the testing points were classified as shown in Figure 12, page 11, with 3 colors: blue for solid
or filled-out region of the joint, yellow for small voids or defects and red for large voids or defects.

A dedicated software has been developed recently to facilitate the interpretation. The application creates
2D and 3D visual representations of the results obtained from a grid of closely-spaced test points. The
technique consist on using the amplitude spectrum to construct a volume model of the test region from
which locations of reflecting interfaces can be presented as images where a color coding is used to

6
represent the amplitude of each frequency. See Fig. 10. This software was however not used in this case as
only 1 line of testing points was implemented.

Fig. 10. Screenshots of impact-echo representations made by the MIRA d’Or software.
On the left, a cross section parallel to the surface at 150 mm depth in a slab. The purple
area shows a honeycombing and the red one shows a void. On the right, a cross section
perpendicular to the surface showing the backwall represented by the bright green line.

Testing with MIRA

Regarding the use of the MIRA ultrasonic-echo tomographer, one measurement was done at each point
that had been previously tested with the Impact-Echo method. However, with MIRA, the coverage of each
measurement is about 250 mm width and the image of the plane below the instrument is displayed in 3-4
seconds (the so-called “B-scans”, see Figure 11).

Fig.11. Identification of the 3 orthogonal planes of MIRA

according to its position on the surface.

The instrument was firstly calibrated in an area of the wall with known good quality filling. This was done
only once at the beginning where the parameters such as color gain and testing emitting frequency are set.

This calibration should be made carefully before every specific testing job, otherwise the color contours in
the images could easily be interpreted erroneously. Furthermore, because reflections could be an

7
indication of other embedded objects, interpretation is recommended to be followed by some testing cores
for verification at the beginning before continuing making ample testing.

Figure 12 shows examples of B-scans from 3 different points corresponding, from left to right, to filled-out
joint, joint with small defects and joint with big defects. Since the MIRA was placed transversally to the
longitudinal direction of the vertical joint as shown in Figure 4, the plane of a B-scan corresponds to the
transversal cross section of the joint.

(A) (B) (C)

Front side of the wall Front side of the wall Front side of the wall

Reflections of the back side of the wall

Visual width 250 mm Visual width 250 mm Visual width 250 mm

Fig. 12. B-scans from the MIRA tomographer. Red zones indicate reflections from voids. (A) fill-out section of
the joint. (B) section of the joint with small defects. (C) section of the joint with big defects. Below the
verification cores are shown for each case.

All 2D images obtained from each point were afterwards processed in a laptop with a dedicated software to
reconstruct a 3D view of a joint. Figure 13 shows the final results of both instruments, MIRA and DOCter in
joint 2FL for comparison.

The table with the Impact-Echo data from the testing points has been positioned to match the height of the
images of MIRA and compare the color classification given as explained before. As it will be seen, there is
good agreement between both methods.

Some minor differences may be expected because Impact-Echo are point readings below the position of
the single transducer whereas MIRA makes a reconstructed map over the whole length of the joint with a
width of about 250 mm. However, the general interpretation of both tools for the practical purpose of
Quality Control or Quality Assurance is equivalent.

8
 Fig. 13. Final results from joint 2FL and comparison. (A) 3D view of the whole testing area of the joint.
(B) D-scan view representing the longitudinal cross section of the joint at its center (x = 125 mm) and
the position where the transversal cross sections shown in Figure 11 were taken. (C) Impact-Echo
method results and chosen color classification.

Conclusions

Testing from one side only is feasible for checking the quality of injection of joints locking precast elements
together. Two such systems are the impact-echo (DOCter) and the ultrasonic-echo method (MIRA). They
can help owners or supervisors for Quality Assurance purposes and contractors by implementing a quick
Quality Control procedure that allows them to change or improve the working conditions or methods to
avoid further expenses on repairs and delays. The quality of the repair itself can also be tested.

The impact-echo (DOCter) is simple to perform and gives immediate information of the condition of the
joint right below transducer. Once the fundamental frequencies are understood, the testing can take place
very quickly, within 1-2 seconds for each test. A newly developed impact-echo software, the MIRA d’OR,
allows 2D and 3D images of the joint quality to be produced.

9
The ultrasonic-echo method performed by means of the MIRA needs a careful calibration when starting a
testing set, takes a little longer time and it is difficult to perform within confined areas because of the size
of the instrument. The advantage is that the 2D color illustration is displayed within 3-4 seconds as a cross
section (B-Scan). The 3D reconstruction of the joint condition requires additional time to transfer the 2D
data to a laptop with the dedicated 3D software but it is a useful tool to generate friendlier testing reports.

To obtain the full benefits of both systems the theoretical background should be understood. For this
purpose, NDT Professional Workshops are offered on a regular basis.

Training in operation of the systems is equally important. For this purpose Germann Instruments offers
training of the technicians on-site or at Germann Instruments facilities in Copenhagen, Denmark or Chicago,
USA, where a number of large specimens have been constructed with build-in defects.

References

1. Joergensen, H.B. (University of Southern Denmark) and Hoang, L.C. (Technical University of
Denmark): “Load carrying capacity of keyed joints reinforced with high strength wire rope loops”
FIB Symposium Proceedings 2015, Copenhagen, Denmark.

2. Lund, Mia. S.M., Arvidsson, M. and Hansen, K.K.: “Homogeneity and strength of mortar joints in
Pearl-Chain bridges”, DTU Civil Engineering. FIB Symposium Proceedings 2015, Copenhagen,
Denmark.

3. Carino, N.J.: “Impact Echo, The Fundamentals”, International Symposium on Non-Destructive

Testing in Civil Engineering, September 15-17, 2015, Berlin, Germany.

4. Hoegh K., Khazanovich L., Yu H.T. “Ultrasonic Tomography Technique for Evaluation of Concrete
Pavements.” Transportation Research Record: Journal of the Transportation Research Board, No.
2232, pp. 85–94. 2011, Washington, USA.

10