Anda di halaman 1dari 9




F. EL-Madaani, M. Zaid, P. Gaydecki, H. Hussin and G. Miller

School of Electrical and Electronic Engineering, the University of Manchester, PO Box
88, Manchester M60 1QD, United Kingdom

ABSTRACT. Corrosion processes with time may cause cracking, delamination and finally catastrophic
failure of concrete structures. From a safety aspect as well as economic reasons, it is necessary to monitor
concrete structures to detect corrosion at its early stages utilizing non-destructive testing techniques. In
this paper, a new monitoring technique using inductive scanning technology is presented. Results
obtained using this technique are compared with the potential mapping data. The advantages and
limitations of both techniques are discussed.
Keywords: Corrosion, Non-destructive testing, Inductive scanning sensor
PACS: 82. 45. Bb, 81. 70. -q, 85.75.Ss

Concrete structures normally create highly alkaline environments due to the reaction of
cement with water (the hydration process). The high alkalinity provides excellent corrosion
protection for reinforcing steel due to the formation of a thin passive oxide film on the
surface of the steel. Under high alkalinity (pH>13.5) steel remains passivated [1]. In the
presence of carbonated concrete or chloride ions at the steel / concrete interface, the passive
film breaks down and corrosion initiates, progresses at a steady rate on the steel reinforcing
bars. As a result, corrosion products will build up with time causing cracking, delaminating
and spalling of concrete structures. It is necessary to monitor concrete structures to detect
corrosion at its early stages utilizing non-destructive testing techniques [2]. In this paper, a
new monitoring technique using inductive scanning technology is presented. The method
generates a vector of voltage values of the reinforced concrete samples scanned in the
laboratory at a spatial interval of 1.1mm. These voltage vectors are then compared to the
potential mapping (measuring half-cell potentials at the concrete surface) data for the same
concrete samples. Results obtained using the inductive scanning technique provide a more
detailed view of the corrosion generated on the concrete samples.

CP820, Review of Quantitative Nondestructive Evaluation Vol. 25, ed. by D. O. Thompson and D. E. Chimenti
2006 American Institute of Physics 0-7354-0312-0/06/$23.00


FIGURE 1. Set-up of half-cell potential measurements.


The method of half-cell potential measurements normally involves measuring the
potential of an embedded reinforcing bar relative to a reference half-cell placed on the
concrete surface. The basic set up of this method is shown in Figure 1. The half-cell is
usually a copper/copper sulphate or silver/silver chloride cell. The concrete functions as an
electrolyte and the risk of corrosion of the reinforcement in the immediate region of the test
location may be related empirically to the measured potential difference. In some
circumstances, useful measurements can be obtained between two half-cells on the concrete
surface [3]. The procedure for measuring half-cell potentials is simple and straight-forward.
First, electrical continuity of reinforcement has to be checked prior to testing. Normally a
measured resistance of less than 0.3 ohm [2, 4] is used to indicate continuous electrical
connection of reinforcement. A proper electrical contact with the reference electrode has to
be provided by pre-wetting of concrete surface. A good electrical connection (no clip) is
made to the reinforcement, an external reference electrode (e.g. copper/ copper sulphate,
Cu/CuSO4 ,CSE) is placed on the concrete surface and potential readings are taken on a
regular grid on the free concrete surface. Potentials are measured with a high impedance
millivoltmeter (>10 Mega ohm) to ensure low current conditions during testing. This
technique has already been introduced by Stratfull [5, 6] in 1973. Since then, different
microprocessor-controlled single or multiple electrode devices have been developed and are
commercially available, e.g. the eight-wheel electrode measuring system allowing a survey
of 300 m2/h with grid dimensions of 15 cm using computer-assisted data acquisition and
processing [5,7-10] . The use of multi electrode instruments greatly improves potential
surveys of large bridge decks, sidewalls or parking decks. The best way of representing the
data has found to be a color map of the potential field, where every individual potential
reading can be identified as a small square. Alternatively, a contour plot can be used as
well. A detailed measurement procedure is described in the ASTM C876-91 standard [11],
the basis of which is that the corrosion potential of the rebar will shift in the negative
direction if the surface changes from the passive to the active state (corrosion). Potentials
from zero to -200mV against Cu/CuSO4 indicate passive steel (non-corroded steel).
Positive readings are sometimes recorded, especially in high resistivity soils. Potentials
more negative than -350mV indicate active steel (corroded steel). There is a special
difficulty in interpreting the results, which lie between -200 and -350 mV vs. copper/
copper sulphate reference electrode, (CSE). Comparing neighboring surface potentials, i.e.
potentials gradients, is a more appropriate procedure [12]. The simplified interpretation of
the potential readings is presented in Table 1.


TABLE 1. Probability of corrosion according to ASTM standard C876-91.

ECorr (V vs CSE)

Probability of Corrosion

More negative than -0.350


More positive than -0.200


Between -0.200 to -0.350



Over a period of several years, the SISP (sensor imaging and signal processing) group at
the University of Manchester has developed and reported on a number of inductive sensors
used in conjunction with motorized x-y scanning systems as shown in Figure 2 to perform
parallel, orthogonal, and hybrid line scans and imaging steel reinforcing bars embedded in
concrete [13-17]. The group recently reported on a new phase-sensitive heterodyne
inductive sensor that is capable of imaging corrosion on the surface of the steel bars in
concrete [16,17]. The major components of the inductive scanning system are x-y scanner
and an inductive sensor connected to a personnel computer. The scanner can be
programmed to scan in one or two-dimensional modes. The scan resolution of the system is
determined by the mechanical characteristics of the stepper motors for both the x and yaxes. The sensor is mounted to the vertical arm (z-axis) which is hand operated and allows
the sensor to be positioned at different heights above the surface of a sample. At steps of
1.1 mm, the signal from the sensor was read by a 16-bit ADC, filtered in real-time by a
purpose-built DSP system [18,19] and fed to the same computer for line-scan or image
generation. When performing a single line scan, the sensor produces a stream of voltage
values at each of the scanning positions of the scanned object. The PC is connected to both
the scanner and the DSP board and contains all software that controls the scanner and the
DSP board. The PC control, data acquisition and data storage software was coded entirely
in Delphi [15]. The DSP board is a dedicated real time DSP system designed to acquire and
process the voltage data obtained from the sensor. A real-time averaging algorithm
averages the voltage stream and the averaged results are stored in the local memory of the
DSP system. At the end of each line scan, the results are transferred out of the DSP system
and written onto the local hard drive of the host PC via RS232 serial communication.

FIGURE 2. Inductive scanning system.


Sensor Coil

Tuned Sensor


to Voltage

To Computer




FIGURE 3. Major components of heterodyning Sensor.

Heterodyning Sensor
The heterodyne sensor converts the change in the excitation frequency of the tuned
oscillator into a voltage value. It is designed for imaging corrosion product on the surface
of the steel bars embedded within concrete. Its major components are shown in Figure 3.
This sensor contains two oscillating circuits. The first is a standard Colpitts-type LC tuned
oscillator, which is comprised of a sensing coil, connected across the input/output terminals
of a 74HC04 inverter, with the two tuning capacitors connected across each terminal to
ground. The coil acts as both a transmitter for propagating a time-varying magnetic field
towards a metal target and a receiver for receiving the reflected secondary field. The coil
was constructed using 30 turns of 1 mm diameter enamel-insulated copper wire wound
around a 25 mm diameter hollow ferrite, giving an inductance of 60H [17]. The frequency
of the oscillator is approximately 46.5kHz. This sensor oscillator is connected to a
comparator, which converts the sinusoidal output into a square wave. The second is a
crystal oscillator, which acts as a reference oscillator and generates a square wave at fixed
frequency close to 46.5kHz. These two waves are fed to a phase sensitive detector (PSD)
and the output of the PSD is further integrated using an integrator. The integration yields a
signal which is proportional to the phase error between the two waves. This signal is then
fed to a frequency-to-voltage converter and finally to an ADC system within the DSP.
Finally, the digital signal is transmitted out of the DSP unit to a PC for storage and display.
Two reinforcing rebars of 20mm and 12mm in diameter and 450mm in length were
prepared and cast in a cement mortar at 20mm cover- thickness. The two specimens are
illustrated in Figure 4.



FIGURE 4. (a) two-block specimen. (b) continuous specimen (c) corrosion cell.



The first sample comprising two mortar blocks each is 100mm in length and separated
from each other by a distance of 100mm. The ends and middle of the bare steel were
insulated by wrapping a PVC insulation tape. The two-block specimen was made of mortar
mixed with tap water. The second sample was made as a one-block of mortar. The oneblock specimen was made of mortar mixed with de-ionized water. Mixture proportion of
mortar was three parts sand to one part cement. The water to cement ratio was 0.6. In
addition, the 28 days compressive strength was 150KN. The half of the two-block specimen
was immersed in 0.2% NaCl solution (wet conditions) whereas the one-block specimen was
placed on a cloths towel socked with 0.2% salt water (dry conditions). The two-block
specimen was immersed in 0.2% NaCl solution for 24 hours before any testing or corrosion
acceleration take place. To accelerate the corrosion process, two corrosion cells were
prepared Figure 4c. The specimens were connected to a metal sheet through a constant
power supply. The applied corrosion current was 10mA. The half-cell potential
measurement was conducted every 24 hours by using a portable half-cell of copper/ copper
sulphate reference electrode as shown in Figure 1 above. After the measurement, the
potentials were estimated as the probability of corrosion based on ASTM C876-91 standard
[11]. The specimens were scanned using the inductive scanning system every 24 hours. The
sensor was mounted on the scanner z-axis and positioned in such away that the centre of the
sensing head was aligned with the centre of the rebar. The distance between the sensor head
and the top surface of the rebar was set to 21mm. The sensor was calibrated to remove the
DC offset. The parallel line scan was conducted along the axis of the rebar at steps of
1.1mm and averaging of the signal by 600 times at each step using the line scan program on
the PC computer.
Figure 5a shows the results obtained from using the half-cell method on the two-block
sample, after 0, 3 and 7 days. In contrast, Figure 5b shows the same sample scanned after 3,
5 and 7 days using the inductive sensor. Because of the higher spatial resolution of the
latter, it is possible to distinguish between corroded and non-corroded sections. This is
especially evident from Figure 6, which shows a scan of a bar corroded in two discrete

Two-Block Sample Inductive Scanning Data

Tow- Block Sample Half- cell Potential data


Distance (cm)


Distance (mm)



100 150 200 250 300 350 400 450 500


Voltage (mV)

Potential (mV) vs
Cu / CuSO4








FIGURE 5. Scans of a partially corroded bar encased in two concrete blocks (as in Figure 4 a).
(a) half-cell potential measurement. (b) inductive sensor parallel line scans.


FIGURE 6. On-screen result of inductive scan of steel bar, corroded in two sections.

Since the two-block specimen was immersed in the solution for 24 hours before any
corrosion acceleration performed, high negative potential readings was recorded at day-0 as
shown in Figure 5a. Black and green liquid was seen on the concrete surface of the twoblock specimen as concrete begins to crack due to corrosion. Although corrosion
acceleration test was conducted for seven days, low negative potential shifts was still
recorded on the dry specimen ( One-block specimen) as shown in Figure 7a. The inductive
scanning sensor is unaffected by the presence of oxygen or moisture in concrete as can be
seen from Figures 5b and 7b. The potential measurements could not be conducted on the
steel surface of the two-block specimen at the gap between the two mortar blocks as the
steel was insulated with PVC tape and not in contact with the solution. This result is
indicated by the gap in Figure 5a. The inductive scanning is unaffected by the presence of
PVC tape barrier coating as can be seen in Figure 5b. It was observed during the testing that
a good electric contact is required between the digital millivoltmeter and the specimen
under test to obtain stable readings. Further, the concrete surface has to be pre-wetted
before taking potential measurements otherwise, readings will be fluctuating. Furthermore,
it is noticed that successful potential measurement is best performed only by an experienced
corrosion engineer.

One Block Sample Inductive Scanning Data

One- Block Sample Half- cell Potential data

Distance (cm)





Distance (mm)




Voltage (mV)

Potential (mV) VS.













FIGURE 7. Scans of a corroded bar encased in single concrete block (as in Figure 4 b).
(a) half-cell potential measurement. (b) inductive sensor parallel line scans.





Although the half cell potential mapping method is simple, fast and inexpensive, it
suffers from several disadvantages. The disadvantages are that the results are only
qualitative, without the establishment of an actual rebar corrosion rate. It only yields
information on the present state of corrosion activity. Further, structures with impermeable
membranes and epoxy coated reinforcement are non-testable (Figure 5a) because of the
barrier between the steel and the electrode. As a corollary, any reinforcement that is
electrically discontinuous is not testable. Structures under cathodic protection systems (CP)
are not testable due to the presence of current from CP system. Although structures with
galvanized reinforcement are testable, the results must be interpreted differently due to the
presence of the zinc (mixed potential). Since potentials are not measured on the rebar
surface but on concrete, compensation is unavoidable to get more reliable results. Testing
on wet concrete surfaces may result in false readings. Testing of wet concrete results in
high negative potential shift due to lack of oxygen, so passive steel may show negative
potentials similar to those of corroding steel as shown in Figure 5a. Measurements on dry
concrete show low negative potential shifts on both passive and corroding rebars as shown
in Figure 7a. It should also be stated that the half cell method requires a physical contact
point with the steel. In addition, the method is affected by the presence of carbonated
concrete and chloride ions in the concrete.
In contrast to the half-cell method, the inductive sensor requires no contact at all with
the steel (i.e. it is totally non-invasive), is unaffected by issues of electrical continuity, the
presence or absence of a coating and concrete moisture content. Furthermore, because of its
high spatial resolution, it can locate localized areas of corrosion with considerable precision
as illustrated in Figure 6. Interpretations of results are straightforward. It is simple, fast and
inexpensive. However, the major problems that have held back the development of the
inductive scanning technology is the depth penetration and drift of the sensing coil.
It is promising to use the inductive scanning as a practical non destructive technique to
visualise and monitor corrosion of reinforcing steel in concrete. It can be used for
monitoring the increase in corrosion activity on steel surfaces with time. In addition, the
inductive sensor technique is unaffected by concrete conditions, i.e. moisture content,
carbonation, chloride contamination, lack of oxygen,...etc. It is believed that to make this
technique of a high quality standard, a set of criteria has to be established between
corrosion rate, sensor voltage response, penetration depth and bar diameter.
The authors gratefully acknowledge the support of the Great Man-Made River Project
Authority in Libya for financially supporting this work.


1. Shamsad Ahmad, Cement & Concrete Composites, 25 459-471(2003).
2. Jiirgen Mietz and Bernd lsecke, Construction and Building Materials, 10- 5, 361-313
3. Guide book on non-destructive testing of concrete structures, IAEA-TCS-17, Vienna,
2002., pp.56-60.
4. CSNI Technical Note, Electrochemical techniques to detect corrosion in concrete
structures in nuclear installations, NEA/CSNI/R (2002)21.
5. Elsener B., construction and Building Materials, 15, 133-139 (2001).
6. Stratfull JR., Corrosion NACE 1957; 13:173t.
7. Elsener B, Bohni H., Potential mapping and corrosion of steel in concrete-ASTM
STP 1065. In: Berke NS, Chaker V, Whiting D, editors. Corrosion rate of steel in
concrete. Philadelphia: American Society for Testing and Materials, (1990) 143-156.
8. Elsener B, Bohni H., Electrochemical methods for the inspection of reinforcement
corrosion, Concrete structures-field experience. Material Science Forum, (1992)
635- 646.
9. Hunkeler F. Bauwerksinspektion mittels Potentialmessung. Schweizer Ingenieur
Architekt (1991) 109-272.
10. Cox RN, Cigna R, Vennesland O, Valente T, editors. COST 509 Corrosion and
protection of metals in contact with concrete, Final Report, European Commission,
Brussels, 1997 EUR 17608 EN.
11. American Society for Testing Materials, Standard test method for half-cell potentials
of uncoated reinforcing steel in concrete, ASTM Standard C876-91(1999).
12. Hansson,C.M., Cement and Concrete Research, 14, 574-584 (1984).
13. Gaydecki PA, Burdekin FM., Meas. Sci. Technol., 5, 127280 ( 1994).
14. Fernandes B, Gaydecki P, Sung Q, Miller G and Burdekin FM A Multi-Sensor Array
Inductive Scanner for High-Speed Imaging of Reinforcing and Pre-Stressing Steel in
Concrete, Proc. Conf. BInst NDT, 2000, pp.217-222.
15. Gaydecki P, Quek S, Miller G, Fernandes B, Zaid M., Meas.Sci. Technol, 13, 13271335 (2002).
16. Fernandes B., Miller G., Gaydecki P., Quek S., Damaged and corrosion visualization
of reinforcing bars embedded in concrete using a new solid- state inductive scanning
sensor, in review of progress in QNDE 21B, edited by D. O. Thompson and D. E.
Chimenti, AIP Conference Proceedings, Vol. 615, Brunswick, Maine, 2001, pp. 12331238.
17. Miller G., Gaydecki P., Quek S., Fernandes B., Zaid M., NDT&E International, 36, 19
26 (2003).
18. Quek S., Gaydecki P., Zaid M., Miller G., Fernandes B., NDT&E International, 36, 7
18 (2003).
19. Miller G., Gaydecki P., Quek S., Fernandes B., Zaid M., Sensors and Actuators A 121
339-346 (2005).