Bundesanstalt für Materialforschung und -prüfung

International Symposium (NDT-CE 2003)

Non-Destructive Testing in Civil Engineering 2003
Start > Contributions >Posters > Material: Print

Study on Measurement of Chloride Content using Electromagnetic Wave in Reinforced Concrete Structures

Jun-ichi ARAI, Retec. Inc, Tokyo, JAPAN
Toshiaki MIZOBUCHI, HOSEI University, Tokyo, JAPAN
Kumiko SUDA, Kajima Corporation, Tokyo, JAPAN

Abstract

The feasibility of using amplitude values of reflected electromagnetic waves as a non-destructive method for estimating the chloride ion concentration in reinforced concrete structures, was studied. The amplitude of the electromagnetic wave varies owing to the effects of relative dielectric constant and conductivity of concrete. Thus, the amount of moisture and the type of cement, which are influence factors, were varied, and the accuracy in the estimation of chloride ion concentration was verified.

Keywords:
Electromagnetic wave, chloride ion concentration, relative dielectric constant, conductivity, amplitude

1 Introduction

One of the methods of investigating salt damage in reinforced concrete structures is to sample the concrete core at an arbitrary location and to analyze the chloride ion concentration. Predicting the start of corrosion of reinforcement in concrete from the distribution of chloride ion concentration [1] is considered to be important for the maintenance and management of reinforced concrete structures.

It is however, difficult to judge whether or not the sampled core is the location in the entire structure where the maximum chloride ion concentration occurs. Thus, a method that can superficially measure the distribution of chloride ion concentration needs to be developed.

The feasibility of using electromagnetic waves (used for non-destructive inspections), for superficial measurement of chloride ion concentration was studied in this research.

2 Characteristics of electromagnetic wave propagation in concrete

The speed of propagation and damping characteristics of electromagnetic waves in concrete are governed by the electrical properties of concrete (dielectric constant e, conductivity s, magnetic permeability m). The speed of propagation , can be expressed by the relative dielectric constant er (dielectric constant of medium/ dielectric constant of air) and the speed of light in vacuum, C. [2] However, the relative dielectric constant of concrete in the dry condition is in the range of 4 to 10, and in the wet condition, in the range of 10 to 20 [3]. Thus, the dielectric constant of concrete varies depending on the percentage of water content of concrete.

On the other hand, the damping coefficient a is expressed by equation (2) [4] and is governed by the electrical properties of concrete. However, the magnetic permeability of concrete is considered to remain constant since the electric field is constant. Consequently, the damping characteristics vary depending on the dielectric constant and the conductivity.

(1)

(2)

The relative dielectric constant is the same for both fresh water and sea water (about 81)[1], and no change in the chloride ion concentration is anticipated, but the conductivity is considered to vary with the amount of ions in concrete. For instance, the conductivity is considered to be 10-4 to 10-2 (S/m = 1/Wm) for fresh water, 4 (S/m) for sea water, 10-3 to 10-2 (S/m) for concrete in the dry condition, and 10-2 to 10-1 for concrete in the wet condition.[3]

In view of the above, it is concluded that the conductivity varies when chloride ions exist in the concrete, and it affects the damping characteristics of electromagnetic waves. That is, by studying the amplitude of the electromagnetic waves, the difference in the chloride ion concentration in concrete can be predicted. However, the relative dielectric constant and conductivity vary depending on the percentage of water content in concrete. Thus, the effect of percentage of water content in concrete also needs to be studied.

3 Method of investigation

Two cases were set for studying the effects of relative dielectric constant and conductivity on the damping characteristics of electromagnetic waves in concrete. Firstly, to study the effects of relative dielectric constant, tests were conducted using concrete specimens with percentage of water content that varied with the passage of time. Secondly, to study the effects of conductivity, tests were conducted using concrete specimens in which the amount of ions (other than chloride ions) was varied.

Fig 1: Conceptual sketch of electromagnetic wave measurement.

As shown in Fig. 1, specimens of size 100 x 100 x 400 mm were used for both cases. Two kinds of specimens were used - those in which no reinforcements were arranged (called "Specimen C" hereafter), and those in which D16 reinforcements were arranged (called "Specimen RC" hereafter). For measurement of electromagnetic waves, an antenna of about 1.0 GHz with specifications as given in Table 1, was installed on the specimen, and a steel plate was laid below the under-surface of the specimen for intensifying the electromagnetic waves. As shown in Fig. 3, the amount of amplitude of the reflective wave of an electromagnetic wave made 100% the maximum gain of the machine used for the experiment, and showed it by the ratio. Moreover, since the reflected wave from the reinforcement and the reflected wave from the steel plate are likely to interfere with the reflected wave of the RC specimen, the reflected waveform of the C specimen was subtracted from the measured waveform of the RC specimen, so that the amplitude values of the reflected waves of the reinforcement only were used in the study.

Fig 3: Conceptual sketch of reflected wave amplitude.
Fig 2: Conceptual sketch of electromagnetic wave measurement.

Item Specifications
Rader Frequencies1.0GHz
MethodImpulse Method
Transmission voltage17V p-p(at load 50)W
Horizontal resolution80mm
Table 1: Specifications of electromagnetic wave measuring equipment.

3.1 Study of the effects of relative dielectric constant
As shown in Table 2, six patterns of specimens were used to study the effects of relative dielectric constant, that is, the chloride ion concentration in the specimens was varied from 0 kg/m3 to 6 kg/m3 in steps of 1 kg/m3. Two kinds of specimens were prepared - Specimen C and Specimen RC. Table 3 and Table 4 show the materials used and the mix proportions of the specimens.One day after placement, the concrete specimens were removed and sealed in plastic bags so that the chloride ions in the specimens did not leach. The specimens were cured underwater for seven days, the moisture on the surface of the concrete specimens was wiped off thoroughly and subsequently, the specimens were used. To study the change in the percentage of water content, the mass of the specimen, and the temperature and humidity in the laboratory were measured during the electromagnetic wave measurement. After all experiments finished, the specimens were dried for three days in the drying oven and in the dried condition, the percentage of water content was taken as 0%. The change in the amount of moisture in concrete with the passage of time was studied in terms of the percentage of water content, the percentage reduction in mass and the percentage of dissipated water.

With reinforcement No reinforcement Chloride ion concenteration (Kg/m3)
RC-1C-10.0
RC-2C-21.0
RC-3C-32.0
RC-4C-43.0
RC-5C-54.0
RC-6C-65.0
RC-7C-76.0
Table 2: Patterns of test specimen.

Material Summary
CementOrdinary Portland cement, density 3.16Kg/m3
Fine aggregateRiver sand NII GATA,saturated surface-dry density 2.63Kg/m3,fineness modulus 2.48
Coarse aggregrateHard sandstone NII GATA,saturated surface-dry density 2.66Kg/m3,fineness modulus 6.89
AdmixtureAE water reducing agentLignin sulfonic acid compound,density 1.25Kg/m3
AE agentDenatuted rosinate-based anionic surface active agent
Table 3: Materials used.

W/C(%) s/a(%) Unit quantity Kg/m3 AE water reducing agent AE agent
W C S G
604215525879311080.811.5A
Table 4: Concrete mix proportion.
  • Percentage of moisture content, m
    The percentage of moisture content is the ratio between the amount of moisture in the concrete specimen at the time of measurement and the absolute dry mass of the specimen, with reference to the absolute dry mass of the specimen (after inserting it in the drying oven after completion of the measurement period) .

  • Mass reduction ratio Ww
    The mass reduction ratio is the ratio between the mass reduction of the specimen from the start of measurement and the mass of the specimen at the start of the measurement, with reference to the mass of the specimen at the start of measurement.

  • Moisture dissipation ratio RW
    The percentage of dissipated water is the ratio between reduction in the amount of moisture at the start of the measurement and the amount of moisture in concrete, with reference to the amount of moisture from the start of the measurement to the absolute dry condition of the specimen.

3.2 Study of the effects of conductivity
The conductivity in concrete is governed by the amount of ions in the concrete, therefore, the effects of ions other than chloride ions need to be studied. The chloride ion concentration in concrete was increased in four steps of 1 kg/m3 each from 0 kg/m3 to 3 kg/m3, the type of cement used was varied and the effects of the ions were studied. Four kinds of Specimen C were manufactured in ordinary Portland cement (N), blast furnace cement type B (BB), fly ash cement type B (FB) and low-heat Portland cement (L). Table 5 and Table 6 show the materials used and the mix proportions of the specimens.

Material Summary
Cement Ordinary Portland cement, density 3.15Kg/m3
Blast furnace slag cement typeB, density 3.04Kg/m3
Fly ash cement typeB,density 2.95Kg/m3
Lowheat Portland cement,density 3.24Kg/m3
Fine aggregateRiversand from KINU,saturated surface-dry density 2.52Kg/m3,fineness modulus 2.73
Coarse aggregrateCrushed stone from IWASE,saturated surface-dry density 2.64Kg/m3,fineness modulus 6.66
AdmixtureAE water reducing agentLignin sulfonic acid compound,density 1.25Kg/m3
AE agentDenatuted rosinate-based anionic surface active agent
Table 5: Materials used.

Cement W/C(%) s/a(%) Unit quantity Kg/m3
W C S G AE water reducing agent AE agent
N50441743487509991.088-
BB4316833673810251.0504A
FB4317234472810101.0754A
L461663328029861.0380.5A
Table 6: Concrete mix proportion.

Measurement of electromagnetic waves was started after wet curing the concrete specimens while they aged to seven days after placement. Since a correlation exists between the amount of ions in concrete and the percentage of water content, the mass of the specimen, the room temperature and the temperature in the antenna were also measured during the measurement of electromagnetic waves. The absolute dry mass could not be measured, therefore, the amount of moisture in the concrete was estimated by the mass reduction ratio (Ww), which is the ratio between the mass of the specimen at the time of measurement and the mass of the specimen at the start of the measurement.

4 Test results and discussions

4.1 Estimation of chloride ion concentration considering the effects of relative dielectric constant
4.1.1 Estimation of amount of moisture in concrete
In order to presume the concentration of the chloride-ion which exists in concrete, it is thought that it is necessary to also take into consideration the quantity of the water contained in concrete. [5 6 7].For this reason, it was decided to estimate the percentage of moisture content, mass reduction ratio and moisture dissipation ratio from the relative dielectric constant and the amplitude values. Table 7 shows the results of multiple regression analysis taking the various percentages that express the amount of moisture as target variables, and the relative dielectric constant and amplitude values as explanatory variables.

Reinforcement Target Variable Partial Regression Coefficient Constant Multiple correlation coefficient
Relative dielectric Amplitud a value
Nonem-1.56E-011.75E-030.1050.25
Ww1.07E-01-7.47E-03-2.690.89
Rw3.16E+009.74E-02810.73
Yesm-9.22E-021.57E-026.220.53
Ww-2.73E-031.05E-02-5.260.83
Rw4.67E-012.66E-010.030.79
Table 7: Coefficients and constants in multiple regression analysis.

The multiple correlation coefficient of percentage of moisture content was 0.25 for Specimen C and was 0.53 for Specimen RC. The estimation of the percentage of moisuture content in concrete from the amplitude of the reflected wave and the dielectric constant of concrete is considered to be difficult.

The multiple regression coefficient of mass reduction ratio was 0.89 for Specimen C and 0.83 for Specimen RC, indicating the best accuracy among the various tests. From the partial regression coefficient, it was concluded that effect of relative dielectric constant on Specimen C and the effect of amplitude value on Specimen RC, were significant.

The multiple regression coefficient of moisture dissipation ratio was 0.73 for Specimen C and 0.79 for Specimen RC. Although it was inferior compared with the mass reduction percentage, it was comparatively good accuracy. From the partial regression coefficient, it was concluded that the effect of relative dielectric constant was more significant than the effect of amplitude values.

Based on the findings above, it was concluded that the mass reduction ratio and the dissipation ratio can be estimated from the reflected electromagnetic waves. However, during the estimation of the mass reduction ratio, the influence of relative dielectric constant and amplitude value varies depending on whether the radiated electromagnetic wave has been reflected from the reinforcement or not. For this reason, it needs to be verified whether the measured reflected wave has been reflected from the reinforcement or not.

4.1.2 Estimation of chloride ion concentration considering the effects of moisture
The concentration of chloride ions was estimated by multiple regression analysis taking the chloride ion concentration as the target variable. The percentage of moisture content, mass reduction ratio and moisture dissipation ratio were each verified in this research, and moisture dissipation ratio was found to be most effective, therefore the study was made using the moisture dissipation ratio. The chloride ion concentration was estimated for two cases: considering relative dielectric constant and amplitude values only (referred to as "not considering moisture"), and considering the above as well as the moisture dissipation ratio referred to as "considering moisture"). Fig. 4 and Fig. 5 show the results of estimation of chloride ion concentration of Specimen C considering moisture and not considering moisture respectively. Fig. 6 show the range of variation of estimated values for Specimen C and Specimen RC considering moisture and not considering moisture respectively.

Fig 4: Results of estimated chloride ion concentration of Specimen C (not considering moisture). Fig 5: Results of estimated chloride ion concentration of Specimen C (considering moisture).

The multiple regression coefficient when moisture is not considered is 0.69 for Specimen C and 0.62 for Specimen RC.

Fig 6: Variation in the estimated results of chloride ion concentration(Specimen C).

This indicates that the estimation accuracy is not very high. The multiple regression coefficient (when moisture is considered) is 0.87 for Specimen C and 0.83 for Specimen RC, indicating a comparatively high estimation accuracy.

When variation range of estimated values was considered moisture, the variation range became small. When the moisture was considered, the average estimated values also tended to approach closer to the set values of chloride ion concentration.

4.1.3 Estimation of chloride ion concentration considering moisture and air temperature
When estimating the chloride ion concentration from the amplitude values of reflected electromagnetic waves, even if the amount of moisture in concrete was considered, the estimated value varied according to the day of the measurement. To perform the estimate considering this factor, multiple regression analysis was carried out taking the air temperature as an additional explanatory variable. As shown in Fig.7, the regression coefficient of 0.90 and Specimen C (not considering moisture) of the regression coefficient of Specimen C (considering moisture) was 0.94. Thus, estimation with extremely good accuracy could be obtained.

From the findings above, it can be concluded that the chloride ion concentration can be estimated with good accuracy from the amplitude values of reflected electromagnetic waves, taking into account the effects of temperature and moisture in concrete.

4.2 Estimation of chloride ion concentration considering the effects of conductivity
To confirm that the amplitude value varies depending on the type of cement used, Fig. 8 has been plotted showing the curve of amplitude values with the material age for a chloride ion concentration of 2.0 kg/m3. The amplitude values of fly-ash cement type B (FB) clearly tended to be smaller compared to those of ordinary Portland cement (N) and low-heat Portland cement (L).

Fig 7: estimated results of chloride ion concentration of Specimen C considering air temperature and moisture. Fig 8: Amplitude values for various types of cement.

Since it was evident that the amplitude values vary according to the type of cement used, the chloride ion concentration was estimated by performing multiple regression analysis for each type of cement. The chloride ion concentration was taken as the target variable, and the relative dielectric constant, amplitude value, mass reduction percentage, internal antenna temperature and air temperature were taken as explanatory variables. Fig. 9 shows the results of estimation of chloride ion concentration at an age of 90 days. The estimated value of chloride ion concentration for fly ash cement type B (FB) was comparatively accurate, but the estimated value of 2.0 kg/m3 as the chloride ion concentration for ordinary Portland cement (N), low-heat Portland cement (L) and blast furnace slag cement type B (BB) tended to be lower than the actual value.

Fig 9: Difference in the estimated results of chloride ion concentration for various types of cement.

5 Conclusions

The present research examined the method of estimating chloride ion concentration in reinforced concrete from the amplitude of reflected electromagnetic waves. The knowledge gained from this study was outlined as follows.

  • The amount of moisture in concrete can be estimated with comparatively good accuracy using the moisture dissipation ratio from the amplitude values obtained from measurement of electromagnetic waves. The chloride ion concentration can be estimated with fairly good accuracy from the amplitude values obtained during measurement of electromagnetic waves. The relative dielectric constant and conductivity of concrete, and the air temperature should be considered when estimating the chloride ion concentration.
  • The relative dielectric constant of concrete is correlated with the amount of moisture in concrete. Thus, estimation can be performed by substituting the relative dielectric constant of concrete by the amount of moisture in concrete. The moisture dissipation ratio indicates best the value of the amount of moisture.
  • The conductivity in concrete varies depending on the type of cement used, and it affects the amplitude value. Consequently, highly accurate results can be obtained by performing estimation appropriate for the type of cement used.

Whether a two-dimensional chloride ion distribution can be estimated or not for the same type of cement used, is a topic that needs to be investigated in the future.

References

  1. Corrosion and Protection Committee: Test methods and standards (provisional) related to corrosion and protection of concrete structures, the Japan Concrete Institute, pp. 57-58, 1987.
  2. Akihiko Yoshimura et al: Non-destructive tests for diagnosis of concrete structures - electromagnetic wave method, and non-destructive inspection, Vol. 47, No. 10, pp. 712-716, 1998.
  3. Society of Exploration Geophysicists of Japan: Handbook of Geophysical Exploration, principle edition, 1998.
  4. Toshio Hosono: Fundamentals of Electromagnetism, Shokodo Co., Ltd.
  5. Toshiaki Mizobuchi et al: Considerations on the measurement of chlorides in reinforced concrete by electromagnetic waves, Annual Proceedings of Concrete Engineering, Vol. 24, No. 1, pp. 1509-1514, 2002.
  6. Jun-ichi Arai et al: Research on the measurement of chlorides in reinforced concrete by non-destructive inspection, Annual Proceedings of Concrete Engineering, Vol. 24, No. 1, pp. 1515-1520, 2002.
  7. Kazumasa Morihama et al: Relationship between water content and relative dielectric constant in concrete, Japan Society of Non-Destructive Inspection, Proceedings of the 1999 Spring Symposia, pp. 91-94, 1999.
STARTPublisher: DGfZPPrograming: NDT.net