Bundesanstalt für Materialforschung und -prüfung

International Symposium (NDT-CE 2003)

Non-Destructive Testing in Civil Engineering 2003
Start > Contributions >Lectures > Tendon Ducts: Print

Impact-Echo testing of steel cable ducts for injection grouting quality

Claus Germann Petersen, Germann Instruments A/S, Emdrupvej 102, DK-2400 Copenhagen, Denmark, E-mail: Claus.Petersen@NDT-Titans.com
Allen Davis, Construction Technology Laboratories, Inc. (CTL), 5420 Old Orchard Road, Skokie, Illinois 60077, USA, E-mail: Allen.Davis@NDT-Titans.com
Aldo Delahaza, WJE, 330 Pfingsten Road, Northbrook, Illinois 60062, USA, E-mail: Aldo.Delahaza@NDT-Titans.com

1. Purpose

Impact-echo testing during construction of a project for quality control of the injection quality of steel cable ducts is illustrated, where the reflection of the P-wave from the duct itself is used as the only criterion for evaluating whether the duct is solid or voided.

2. The structure and the ducts

The ducts were positioned in a 40 meter wide, newly constructed tunnel cover slab as illustrated in fig.1.

The ducts, 80 mm in diameter, were made of corrugated steel tubes, 1 mm thick. Each duct contained 8 steel tendons, each ~ 12 mm in diameter. After tensioning of the tendons the ducts were injected with an expansive cement grout, either from tube A in fig.1, or though tube B.

Fig 1: The 40 metre wide tunnel cover slab with an embedded cable duct.

Impact-echo testing was performed in the vicinity of tube B, the highest point of the cable ducts, fig.2.

Fig 2: The corrugated steel cable duct illustrated in the vicinity of tube B with the depth of the duct indicated in the testing area.

Testing was performed after the grout had hardened, two days after injection.

3. Impact-echo

The impact-echo test system used was the DOCter, ref. (1). The general theory for testing of the injection quality of cable ducts with impact-echo is outlined in ref. (2).

For cable ducts positioned in a concrete member with a constant thickness, the primary evaluation criterion is to compare the frequency obtained from testing above the duct to the solid frequency of the member. Should the two frequencies be the same the travel path of the P-wave passing through the duct is identical to the travel path related to the solid frequency. The duct is then solid, that is the duct is fully injected.

For a drop in the frequency, the travel path of the P-wave has to be longer than that related to the thickness of the member, and, hence, the wave has to run around an air interface, which could be an air void in the grout in the duct.

The second criterion is to evaluate the frequency stemming from the reflection from the tendons inside the duct and/or the void in the duct. For a solid steel duct, the reflection of the P-wave will occur from the tendons inside the duct at a frequency of


Cp being the wave speed of the concrete and Tt the depth to the steel tendons. The 1 mm thick cable duct steel is transparent to impact-echo.

For a void in the duct the frequency measured will be


where Tv is the depth to the void.

To have reflection from the void the P-wave has to have a sufficiently high frequency content, controlled by the contact time tc of the impactor (f max = 1250/ tc).

In addition, the lateral size of the void L has to be larger than the minimum wavelength lmin of the P-wave, where lmin = Cp / fmax, and the void has to be placed at least at a depth of 1/2 lmin and a maximum depth of 4l min.

As an example, if the lateral void size L is 50 mm, l min has to be smaller than 50 mm, which in turn means that the maximum useful frequency fmax has to be higher than Cp/l min @ 4000 m/s / 50 mm = 80 kHz. In addition, the void has to be placed at a minimum 25 mm and maximum 200 mm depth for reflection of the P-wave to occur.

4. The position of the cable ducts and the surrounding reinforcement

The cross-section of the tunnel slab, the cable ducts and their position at the highest point close to tube B are illustrated in fig. 2. Fig.3 show the reinforcement positioned close to the ducts.

Fig 3: Dimensions and the depth of the reinforcement positioned close to the cable ducts.

As will be seen, the thickness of the slab in the testing area is not constant, and is more than 1.1 meters thick. Consequently, a drop in the solid frequency cannot be used as a reliable measure for the presence of a void in the injected duct. Only the second set of criteria can be used, reflection from the steel tendons inside the duct or reflection from a void in the duct.

5. Calculation of the expected impact-echo frequencies

The P-wave speed of the concrete was measured by two displacement transducers placed on the surface 300 mm apart and a 5 mm impactor positioned 150 mm from the first transducer on the line of the transducers. The equipment used called the Longship is illustrated in ref. (1). The average plate impact-echo P-wave speed found from three measurements was 0.96 . 4410 m/s = 4234 m/s.

5.1 Reflection from the tendons (solid frequency):
For a depth of the steel tendons of 120 mm the expected frequency will be:

f = Cp / 4Tt = 4234 m/s / (4 . 120 mm) = 8.82 kHz

5.2 Reflection from a void in the cable duct:
For a depth of 120 mm to a void in the duct the expected frequency will be:

f = Cp / 2Tv = 4234 m/s / (2 . 120 mm) = 17.64 kHz

5.3 Possible reflection from reinforcement, (ref.2):
Top reinforcement 16 mm diameter, 45 mm deep: f=(4234 m/s/(4 . 45 mm) . 1.30=31 kHz
Top reinforcement 16 mm diameter, 60 mm deep: f=(4234 m/s/(4 . 60 mm) . 1.33=23 kHz

The 20 mm diameter reinforcement placed at a depth of 180 mm is positioned to deep for detection with impact-echo.

As will be seen, the possible reinforcement frequencies are clearly distinguishable from the solid and void frequencies.

6. Selection of the impactor

A maximum useful frequency of fmax = 35 kHz was selected. This relates in the actual case to an impactor diameter of 8 mm used on the concrete surface in question.

l min is then Cp / fmax = 4234 m/s / 35 kHz = 120 mm, indicating that the lateral flaw must be minimum 120 mm in diameter for detection, located at a minimum depth of 60 mm and a maximum depth of 480 mm. However, for cable duct testing where the void is not a lateral sized flaw, but rather elongated in one direction, experience shows that even a 20 mm wide void can be detected with an impactor generating a maximum useful frequency of 35 kHz, if the longitudinal dimension of the flaw is 4 to 6 times the 20 mm wide void.

7. Pilot testing program

Two cables were selected, one where the injection encountered no apparent problems and the other believed to be voided due to difficult injection caused by a combination of a new, inexperienced injection crew and hot weather conditions.

7.1 Fully injected duct
The frequency spectrum is illustrated in fig.4.

Fig 4: Impact-echo frequency spectrum at test location B-1 of cable duct AA-30. A steel tendon frequency is present at 8.79 kHz related to a depth of the tendons of Tt = Cp / 4f = 4234 m/s / (4 × 8.79 kHz) = 120 mm. A reinforcement signal is present at 22.9 kHz with its typical "crown-shaped" frequency signal. No void frequency around 17-18 kHz is present in the frequency spectrum. The cable duct is interpreted as solid.

7.2 Voided duct

Fig 5: Impact-echo frequency spectrum at test location B-4 of cable duct AA-14. A distinct frequency peak is present at 17.33 kHz, related to a void in the duct at a depth of Tv = Cp / 2f = 4234 m/s / (2 × 17.3 kHz) = 122 mm. The interpretation of the signal is a voided duct.

7.3 Partially voided duct

Fig 6: Impact-echo frequency spectrum at test location B-3 of cable duct AA-31. Two distinct frequency peaks are present in the frequency spectrum, one at 8.10 kHz and another at 18.31 kHz. The 8.10 kHz relates to reflection from steel tendons at a depth of Tt = Cp / 4f = 4234 m/s / (4 × 8.10 kHz) = 131 mm and the 18.31 kHz to reflection from an air void at a depth of Tv = Cp / 2f = 4234 m/s / (2 × 18.31 kHz) = 116 mm. The interpretation of the signal is a partially voided duct.

In all three frequency spectra illustrated the frequency peak at 1 kHz relates to the frequency of the displacement transducer itself and should be disregarded in the analysis.


The actual injection quality of the cable ducts was inspected by an endoscope inserted through the opening of tube B, after any grout in the inlet tube to the duct had been removed by drilling.

The duct tested in fig.4 was confirmed to be solid. No injection grout was found in the duct tested in fig. 5. Partially grouting was detected in the cable duct, fig. 6.

9. Testing

Following the verification of the pilot test results, the testing was conducted on 45 cable ducts in a test grid of 16 test points on all cables, 8 test points on each side of the extending tube B.

Of the cable ducts tested, 8 were found to be voided and 4 partially voided. Re-injection though the opening of tube B, was followed by impact-echo re-testing. All cable ducts were finally accepted as fully injected in the top point in the vicinity of tube B.

10. Additional findings

To test out the sensitivity of positioning the transducer-impactor on the centerline of a voided cable duct the centerline of the transducer-impactor unit was 3 cm (b), 6 cm (c) and 9 cm (d) parallel to the centerline of the duct, fig. 7, and compared to the CL signal.

Fig 7: Voided steel duct, 80 mm dia., 115 mm deep, with position of the centerline of the impact-echo transducer-impactor unit (a) on CL of duct, (b) 3 cm, (c) 6 cm and (d) 9 cm from the centerline CL of the duct.

The comparative frequency signals obtained are shown in fig. 8.

Fig 8: Comparative frequency spectra obtained when displacing the centerline of the impact-echo transducer/impactor unit from the centerline of a voided 80 mm in diameter steel duct positioned at a depth of 110 mm.

11. Conclusions

The quality of the injection with a cement-based, cement slurry in a steel cable duct, 80 mm in diameter, containing 8 tendons and positioned at a depth between 105 mm to 190 mm in a slab with varying thickness was tested successfully with impact-echo.

Properly applied, the impact-echo system is a powerful diagnostic tool for detection of the injection quality of steel cable ducts.

The term "properly applied" means that the basic limitations of the system have to be taken into account, the proper homework has to be done prior to the testing, the proper hardware selected and the test results have to be verified in a pilot study prior to conducting the main testing program.

For testing of injection of cable ducts it is important to conduct the impact-echo testing above the centerline of the duct. Should the centerline not be easily and reliably identified on the project, the only solution is to locate the duct and its depth by means of radar (GPR) prior to testing with impact-echo.

12. References

  1. Catalog NDT-2003: "NDT Systems for durability of new structures, service life estimation, fast track construction, structural integrity, repair quality and monitoring", Germann Instruments A/S, Emdrupvej 102, DK-2400 Copenhagen, Denmark

  2. Sansalone, M.J. & Street, W.B.: "Impact-Echo, Nondestructive Evaluation of Concrete and Masonry", Bullbrier Press, Ithaca, N.Y., USA, 1997.

STARTPublisher: DGfZPPrograming: NDT.net