Concrete Moisture Content Measurement Using ... - Semantic Scholar
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
1243
Concrete Moisture Content Measurement Using Interdigitated Near-Field Sensors Md. Nazmul Alam, Student Member, IEEE, Rashed H. Bhuiyan, Student Member, IEEE, Roger A. Dougal, Senior Member, IEEE, and Mohammod Ali, Senior Member, IEEE
Abstract—The efficacy of a meander and a circular interdigitated sensor in detecting and measuring the moisture present in wet concrete samples is demonstrated. Analytical, simulation, and measurement results of interelectrode capacitance for samples with different moisture contents show good agreement. As moisture content increases from 0% to 6%, the interelectrode capacitance that predicts the moisture content increases by 126.4% for the meander sensor and 187% for the circular sensor. Regression analysis of the measured data demonstrates that for moisture content less than 6% the relationship between the measured capacitance and the percent moisture content is predominantly linear. Index Terms—Concrete, interdigitated sensor, moisture sensor.
I. INTRODUCTION N ORDER TO ensure public safety, routine monitoring, and inspection of the health of civil infrastructures such as bridges, overpasses, and buildings is very important. According to the Bureau of Transportation Statistics (BTS), in 2005 about 26.2% bridges were structurally deficient in the U.S. [1]. The main causes of bridge failure were floods and collisions [2]. For rehabilitation and maintenance, field data are needed as mandated by the National Bridge Inspection Program. In most cases, bridge health monitoring is done by using ground penetrating radar (GPR) [3] which is time consuming and costly. Innovative sensor technology can provide a cheap and a more reliable solution. Wireless sensors can be used to detect the presence of moisture and crack in concrete or the corrosion in the steel reinforcement bar, etc. For active wireless sensors, their batteries can be charged by sending wireless power from outside without damaging the structure [4]. Moisture detection in civil infrastructure has been investigated by researchers in the past. Xing et al. [5] developed a 2.4-GHz integrated parallel-plate soil moisture sensor system. The sensor detects the phase shift caused by the changes in the dielectric constant of soil which is related to the soil moisture content. This sensor needs a number of components, such as a phase locked loop (PLL), voltage controlled oscillator (VCO), phase shifter, phase detector, microcontroller, and
I
Manuscript received September 09, 2009; revised December 10, 2009; accepted December 23, 2009. Date of current version May 21, 2010. This work was supported in part by the U.S. Office of Naval Research under Grant N00014-02-1-0623. The associate editor coordinating the review of this paper and approving it for publication was Prof. Okyay Kaynak. The authors are with the Department of Electrical Engineering, University of South Carolina, Columbia, SC 29208 USA (e-mail: [email protected]; [email protected]; [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2010.2040175
power management circuitry on board to be fully functional. In [6], Saxena and Tayal described a parallel-plate capacitor as a moisture sensor. The sensor consists of a Wein bridge oscillator, a capacitance bridge, a differential amplifier, a band-pass filter, a rectifier circuit, and an analog meter. Ong et al. [7] proposed a 24 MHz embedded LC resonant sensor for moisture content measurement, which requires a large external loop and an impedance analyzer. To maintain the integrity of concrete structures and to allow minimum disruption, an embedded sensor should occupy as small space as possible. The sensor should also only have very few and small components. To address these issues, we propose a planar interdigitated sensor fabricated on a very thin substrate. Because of its thin planar construction the interdigitated sensor can be easily embedded in concrete for moisture content measurement. Our proposed sensor occupies a very small space. In recent years, interdigitated sensors have drawn the attention of researchers due to their low cost and simplicity. They have found application in humidity sensing [8], chemical sensing [9], [10], gas detection [11], resin curing [12], etc. In our research group Bhuiyan et al. [13] have designed and developed interdigitated sensors which can detect insulation faults in unshielded power cables. This was achieved by applying a low frequency signal and then measuring the resulting interelectrode capacitance of the sensor. It was shown that a half circular sensor was less sensitive to its position with respect to the sample under test than a meander sensor. In this paper, the efficacy of two interdigitated sensors in moisture sensing in concrete samples is studied. First, the meander sensor introduced in our earlier work [13] was used. Then, a full circular interdigitated sensor with larger field penetration depth was designed, built, and tested. Experimental measurements were performed to determine the sensor interelectrode capacitance by varying the moisture content of concrete samples with both types of sensors. Measurement results are compared with results obtained from analytical equations and simulations from Maxwell 3-D solver. This paper is organized as follows. In Section II, the basic interdigitated sensor geometry, operating principles and its analytical model are discussed. In Section III, the details of sensor fabrication and the experimental setup and measurement procedure are described. All results are presented and discussed in Section IV followed by the conclusion in Section V. II. SENSOR GEOMETRY, CIRCUIT, AND ANALYTICAL MODEL An interdigitated sensor is a coplanar structure consisting of multiple parallel fingers. The sensor can measure material dielectric constant by applying fringing electric fields into the
1530-437X/$26.00 © 2010 IEEE Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1244
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Fig. 3. Unit cell representation of interdigitated sensor.
The substrate has zero conductivity, so the current is due to the capacitive effect only. Also, since the opamp is operating in the inverting mode, it is obvious that Fig. 1. Typical geometry of an interdigitated sensor.
(1) Fig. 3 shows a unit cell of an interdigitated sensor without the conducting backplane. Using the conformal mapping method Scapple [16] derived the total per unit capacitance for an interdigitated meander sensor as
(2)
where
Fig. 2. (a) Block diagram of sensor circuit and (b) its equivalent circuit.
material. The electrodes of the sensor must be in contact with the material under test. Fig. 1 shows the geometry of an interdigitated sensor [13]. To learn more about the sensor structure and its operation, the reader may review [12]. Three types of electrodes are present, namely the driving electrode, the sensing is electrode, and the guard electrode. A sinusoidal voltage applied to the driving electrode and the output voltage is measured from the sensing electrode. Guard electrodes and a conducting backplane are used to shield the sensor from the influence of external fields [14]. Another important term is the penetration depth of the field which is directly related to the spatial wavelength, . The penetration depth is approximately equal to 1/3 of [15]. A popular method to find the dielectric constant of a material is the short-circuit current method, which is shown in Fig. 2. is applied to the driving electrode A low-frequency voltage while the sensing electrode is connected to a precision opamp (AD708). The opamp acts as an inverting amplifier where a is used for feedback. known capacitor We used a simplified circuit model in Fig. 2(b) to calculate the capacitance between the driving and the sensing electrodes.
and
are the elliptic integrals
and , respectively [16]. The of modulus first part on the right-hand side of the equation resembles the capacitance of the material under test and the substrate and the second part resembles the capacitance of the trapped material in between the interelectrode gap. In (2), it is assumed that there is no conducting backplane. If a conducting backplane is present and the substrate is very thin, then the field lines that penetrate the substrate end at the backplane and not at the sensing elec. trodes. So the substrate capacitance has no contribution to A more accurate expression for per unit capacitance is found in [13]
(3)
where is the corrected per unit capacitance. is the correction factor. For the meander sensor, the correction factor is (4) where is a constant. For the typical meander sensor that was used in our experiment the value of is 0.02. If the electrode length is and the total number of unit cells is then the total capacitance
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
(5)
ALAM et al.: CONCRETE MOISTURE CONTENT MEASUREMENT USING INTERDIGITATED NEAR-FIELD SENSORS
1245
Fig. 5. Ansoft Maxwell 3-D model of full circular sensor.
Concrete block samples were prepared using concrete mix. The age of the samples was more than one year at the time of the measurement. The dimensions of the samples were 15 cm 15 cm 2 cm and 15 cm 15 cm 4 cm. The wet basis for the moisture content in a material is defined as (6) Fig. 4. (a) Experimental setup to measure (b) placement of the sensor in the samples.
C
for concrete samples and
is the mass of the specimen with water and is the where mass of the specimen in dry condition. The moisture content of can be calculated as a material by volume
III. EXPERIMENTAL SETUP To detect the percentage of moisture in concrete, two interdigitated sensors were designed and fabricated. One was a meander sensor and the other was a circular sensor. Two sensor configurations were chosen to create different field penetration depths. The meander sensor is simpler and has smaller field penetration depth than the circular sensor. For the same field penetration depth, the circular sensor covers a larger surface area than the meander sensor. The electrode width, height, gap, and spa, tial wavelength of the meander sensor were 1.125 mm, 17 1.125 mm, and 4.5 mm, respectively. Thus, the field penetration depth for the meander sensor is about 1.5 mm. For the circular , 2 mm, and 7 mm, sensor these dimensions were 1.5 mm, 17 respectively. Hence, the field penetration depth for the circular sensor is about 7/3 mm. The meander sensor was fabricated on ) substrate and the circular a 10 mil thick Duroid 5880 ( sensor was fabricated on a 31 mil thick Duroid 5880 substrate. Referring to Fig. 1, the meander sensor had seven driving electrodes, four sensing electrodes, and two guard electrodes. The length of each electrode was 20 mm. For the circular sensor (see Fig. 5), the inner diameter of the innermost sensing electrode was 6 mm. The circular sensor had 12 driving electrodes, ten sensing electrodes, and two guard electrodes. Both sensors were coated with very thin perylene coating to avoid short circuit with the concrete, especially in the wet condition. A 1-kHz, . The ampli10-V (peak) sinusoidal signal was chosen for tude of the signal was small enough to prevent the opamp from was 100 pF. saturating. The feedback capacitor
(7) is the density of concrete in dry condition and is where the density of water. For measurement the sensor was placed between two concrete was measured for samples. First, the sensor output voltage was known, the capacthe samples in dry condition. Since was found using (1). After completion of the meaitance surement in dry condition, the concrete samples were weighed . Then, the samples were placed in in a scale to determine a 13.5-l water bucket and were completely submerged in water for two days. Wet samples were taken out from the water bucket. Since excess water present on the concrete surface had the potentials for and high ) due to the high pererroneous results (high mittivity of water the outer surfaces of the samples were wiped off using a dry cloth. A detailed experimental measurement setup is shown in Fig. 4. The sensor was placed between the wet concrete samwas recorded and the ples. At regular periodic intervals, was measured. For each weight of the concrete samples was found using (1) while the moisture content reading, was found using (6) and (7). Thus, for each case, a new and its corresponding was found. Since the evaporation rate depended on the environment, the ambient temperature and the air flow, an external fan was used to make the evaporation process faster.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1246
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Fig. 7. Comparison between the measured, analytical and simulated data for the meander sensor.
The sensitivity of the circular sensor is 5.454 pF/ percent change in moisture content. B. Comparison Between Experimental and Simulation Results
Fig. 6. Measured capacitance, C versus moisture content, meander sensor and (b) the full circular sensor.
M
for (a) the
IV. RESULTS A. Experimental Results versus data for both the meander and the Measured circular sensor are shown in Fig. 6. A linear regression analysis was also performed on the measured data. From Fig. 6(a), in while for , dry condition . The presence of water in the wet concrete clearly re(126.4% increase). The sensitivity sulted in the increase in of the meander sensor is 0.519 pF/percent change in moisture content. Fig. 6(a) also shows that the coefficient of determinawhen . Measured versus tion results for the circular sensor are shown in Fig. 6(b). Since the field penetration depth of the circular sensor is larger than that of the meander sensor the resulting linear regression curve in Fig. 6(b) is steeper than the curve in Fig. 6(a). The coefficient of . In dry condition determination while with , (187% increase).
Ansoft Maxwell 3-D was used to simulate the sensor responses for both the meander and the circular sensors. Fig. 5 is an example of the simulation model which was used for the full circular sensor. In the simulation, we were not able to account for the moisture content directly. In [17], experimental data of relative permittivity and conductivity versus frequency for four different levels of moisture content in concrete (from Fig. 2 of [17]) are available. These data are valid for the frequency range of 10 MHz to 1 GHz. We extended their curves down to 1 kHz using curve fitting in Matlab. Then, we developed linear relations between the relativity permittivity and conductivity with percentage of moisture content using curve fitting. The resulting relative permittivity and conductivity values were used in our Maxwell simulations. The relative permittivity ( ) and conductivity ( ) of the concrete samples that were used in our simulation and analytical studies (3)–(4) were: 4.4323, 0.000224; 6.0333, 0.0009; 7.177, 0.00215; 8.549, 0.00444; and 9.693, 0.007 for moisture contents of 0%, 1.816%, 3.1135%, 4.67%, and 5.968%, respectively. In Maxwell 3-D, the solution type was electrostatic. The driving electrodes were set to 10 V and the other electrodes including the backplane were set to 0 V. The default boundary condition, the Neumann homogeneous condition was used. The capacitances between the driving electrodes and the sensing electrodes were computed for the four different cases. Since equations (2) to (5) can only correctly describe the capacitance for the meander sensor, analytical results were obtained only for the meander sensor. Comparison between the experimental, simulation, and analytical data for the meander sensor is shown in Fig. 7. As seen, in general the agreement between all three is quite good. Some deviation is observed at higher moisture content levels. Fig. 8
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: CONCRETE MOISTURE CONTENT MEASUREMENT USING INTERDIGITATED NEAR-FIELD SENSORS
1247
REFERENCES
Fig. 8. Comparison of measured and simulated capacitance for the full circular sensor.
illustrates the comparison between the experimental and simulation results for the circular sensor. Again, agreement is excellent for low moisture contents. The measured results are 10% to 20% higher than the simulated results for moisture contents larger than 4%. V. CONCLUSION The prospect of detecting and measuring moisture in concrete samples was investigated using two interdigitated sensors, e.g., a meander sensor and a full circular sensor. The measurement data clearly indicate that both types of sensors can detect and measure moisture in concrete. There is a distinctly linear relationship between the moisture content in the concrete and the measured interelectrode capacitance. Hence, the amount of moisture present can be easily predicted from the measured capacitance. The disagreement between the simulation and measurement data can be due to the following. First, as mentioned, in our Maxwell simulations the exact values of the relative permittivity and conductivity of concrete at 10 kHz and as function of moisture content was not available. We used curve fitting to extrapolate data from the measured permittivity and conductivity data of concrete that was valid from 10 MHz to 1 GHz. Since Maxwell does not use moisture content values directly accurate values of relative permittivity and conductivity are essential to obtain accurate simulation results. Second, perhaps during the wiping off of the concrete surface there was some standing water on the concrete surface. Although the outside concrete surfaces were wiped off with a dry cloth but the uneven surface might have allowed for some water to still remain on the surface. In the future one may envision that such sensors in miniature form will be integrated with wireless modules and embedded inside concrete samples for moisture content measurement in order to prevent premature corrosion of the steel reinforcements.
[1] [Online]. Available: http://www.bts.gov/publications/state_transporta tion_statistics/state_transportation_statistics_2006/html/table_01_07. html [2] K. Wardhana and F. Hadipriono, “Analysis of recent bridge failures in the United States,” J. Perform. Construct. Fac., vol. 17, no. 3, pp. 144–150, Aug. 2003. [3] C. Maierhofer, “Nondestructive evaluation of concrete infrastructure with ground penetrating radar,” J. Mater. Civil Eng., vol. 15, no. 3, pp. 287–297, May/Jun. 2003. [4] K. M. Z. Shams and M. Ali, “Wireless power transmission to a buried sensor in concrete,” IEEE Sensors J., vol. 7, no. 12, pp. 1573–1577, Dec. 2007. [5] H. Xing, J. Li, R. Liu, E. Oshinski, and R. Rogers, “2.4 GHz on-board parallel plate soil moisture sensor system,” in Proc. Sensors Industry Conf., Houston, TX, Feb. 2005, pp. 35–38. [6] S. C. Saxena and G. M. Tayal, “Capacitive moisture meter,” IEEE Trans. Ind. Electron. Contr. Instrum., vol. IECI-28, no. 1, pp. 37–39, Feb. 1981. [7] J. B. Ong, Z. You, J. Mills-Beale, E. L. Tan, B. D. Pereles, and K. G. Ong, “A wireless, passive embedded sensor for real-time monitoring of water content,” IEEE Sensors J., vol. 8, pp. 2053–2058, Dec. 2008. [8] R. S. Jachowicz and S. D. Senturia, “A thin-film capacitance humidity sensor,” Sens. Actuators, vol. 2, pp. 171–186, Dec. 1982. [9] H. E. Endres and S. Drost, “Optimization of the geometry of gas-sensitive interdigital capacitors,” Sens. Actuators B, vol. 4, pp. 95–98, May 1991. [10] R. Zhou, A. Hiermann, K. D. Schierbaum, K. E. Geckeler, and W. Gopel, “Detection of organic solvents with reliable chemical sensors based on cellulose derivatives,” Sens. Actuators B, vol. 24–25, pp. 443–447, 1995. [11] A. Leidl, R. Hartinger, M. Roth, and H. E. Endres, “A new SO sensor system with SAW and IDC elements,” Sens. Actuators B, vol. B34, no. 2, pp. 339–342, 1996. [12] A. V. Mamishev, “Interdigital Dielectrometry Sensor Design and Parameter Estimation Algorithms for Non-Destructive Materials Evaluation,” Ph.D. dissertation, Dept. Elect. Eng. Comput. Sci., Mass. Inst. Technol., Cambridge, 1999. [13] R. H. Bhuiyan, R. A. Dougal, and M. Ali, “Proximity coupled interdigitated sensors to detect insulation damage in power system cables,” IEEE Sensors J., vol. 7, no. 12, pp. 1589–1596, Dec. 2007. [14] A. V. Mamishev, K. Sundara-Rajan, F. Yang, Y. Du, and M. Zahn, “Interdigital sensors and transducers,” Proc. IEEE, vol. 92, pp. 808–845, May 2004. [15] Y. Du, “Measurements and Modeling of Moisture Diffusion Processes in Transformer Insulation Using Interdigital Dielectrometry Sensors,” Ph.D. dissertation, Dept. Elect. Eng. Comput. Sci., Mass. Inst. Technol., Cambridge, 1999. [16] R. Y. Scapple, “A trimmable planar capacitor for hybrid applications,” in Proc. 24th Electron. Component Conf., 1974, pp. 203–207. [17] M. N. Soutsos, J. H. Bungey, S. G. Millard, M. R. Shaw, and A. Patterson, “Dielectric properties of concrete and their influence on radar testing,” in Proc. NDT & E Int., 2001, vol. 34, pp. 419–425.
Md. Nazmul Alam (S’09) received the B.Sc. and M.Sc. degrees in electrical engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 2006 and 2008, respectively. Since 2009, he has been a Graduate Research Assistant with the Microwave Engineering Laboratory, Department of Electrical Engineering, University of South Carolina, Columbia. His research interests include sensor design and integration, RFID, and designing of microwave antennas.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1248
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Rashed H Bhuiyan (S’09) received the B.Sc. and M.Sc. degrees in electrical engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 2003 and 2005, respectively. He is currently working towards the Ph.D. degree in the Department of Electrical Engineering, University of South Carolina, Columbia. He is the author of multiple international conference and journal papers. He also served as a faculty member with the Bangladesh University of Engineering and Technology, Dhaka, from 2003 to 2005. His current research interests include miniature electromagnetic sensor design, antennas, and RF circuits.
Roger A. Dougal (M’82–SM’94) received the Ph.D. degree from Texas Tech University, Lubbock, in 1983. He leads the Power and Energy Group, Electrical Engineering Department, University of South Carolina (USC), Columbia, which has a principal research focus in the general area of power electronics. As a member of the Board of Directors of the Electric Ship Research and Development Consortium, he oversees USC’s activities related to new power generation, processing, and distribution technologies for ships. He is also the Site Director for a new NSF-sponsored Industry/University Cooperative Research Center for Grid-connected Advanced Power Electronics, which is a joint project between USC and the University of Arkansas. Also, since 1996, under sponsorship of the Office of Naval Research, he has overseen development of the Virtual Test Bed software, which is a comprehensive simulation and virtual prototyping environment for multidisciplinary dynamic systems. This environment has been applied in studies of electric systems for navy ships, electrochemical power sources, hybrid power sources, power electronics, and controls.
Mohammod Ali (M’93–SM’03) received the B.Sc. degree in electrical and electronic engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 1987, and the M.A.Sc. and Ph.D. degrees, both in electrical engineering, from the University of Victoria, Victoria, BC, Canada, in 1994 and 1997, respectively. He was with the Bangladesh Institute of Technology, Chittagong, from 1988 to 1992. From January 1998 to August 2001, he was with Ericsson Inc., Research Triangle Park, NC. Since August 2001, he has been with the Department of Electrical Engineering, University of South Carolina, Columbia, where currently he is an Associate Professor. He had also held appointments as a Visiting Research Scientist with the Motorola Corporate EME Research Laboratory, Plantation, FL, from June to August 2004. He established the Microwave Engineering Laboratory at the University of South Carolina in 2001. He is the author/coauthor of over 100 publications and five granted U.S. patents. His research interests include miniaturized packaged (embedded) antennas, meta-materials and their antenna applications, distributed wireless sensors and rectennas, and portable/wearable antennas and their interactions with humans (SAR). Dr. Ali is the recipient of the 2003 National Science Foundation Faculty Career Award. He is also the recipient of the College of Engineering and Information Technology Young Investigator Award and the Research Progress Award from the University of South Carolina in 2006 and in 2008, respectively. He was the Technical Program Co-Chair of the IEEE Antennas and Propagation Society’s International Symposium in Charleston, SC, in 2009. He has also served as a member of the Technical Program Committee for the IEEE Antennas and Propagation Society’s International Symposium for a number of years. He has served as a reviewer and panelist for grant proposals for a number of federal and local funding agencies. He is an Associate Editor for the IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1243
Concrete Moisture Content Measurement Using Interdigitated Near-Field Sensors Md. Nazmul Alam, Student Member, IEEE, Rashed H. Bhuiyan, Student Member, IEEE, Roger A. Dougal, Senior Member, IEEE, and Mohammod Ali, Senior Member, IEEE
Abstract—The efficacy of a meander and a circular interdigitated sensor in detecting and measuring the moisture present in wet concrete samples is demonstrated. Analytical, simulation, and measurement results of interelectrode capacitance for samples with different moisture contents show good agreement. As moisture content increases from 0% to 6%, the interelectrode capacitance that predicts the moisture content increases by 126.4% for the meander sensor and 187% for the circular sensor. Regression analysis of the measured data demonstrates that for moisture content less than 6% the relationship between the measured capacitance and the percent moisture content is predominantly linear. Index Terms—Concrete, interdigitated sensor, moisture sensor.
I. INTRODUCTION N ORDER TO ensure public safety, routine monitoring, and inspection of the health of civil infrastructures such as bridges, overpasses, and buildings is very important. According to the Bureau of Transportation Statistics (BTS), in 2005 about 26.2% bridges were structurally deficient in the U.S. [1]. The main causes of bridge failure were floods and collisions [2]. For rehabilitation and maintenance, field data are needed as mandated by the National Bridge Inspection Program. In most cases, bridge health monitoring is done by using ground penetrating radar (GPR) [3] which is time consuming and costly. Innovative sensor technology can provide a cheap and a more reliable solution. Wireless sensors can be used to detect the presence of moisture and crack in concrete or the corrosion in the steel reinforcement bar, etc. For active wireless sensors, their batteries can be charged by sending wireless power from outside without damaging the structure [4]. Moisture detection in civil infrastructure has been investigated by researchers in the past. Xing et al. [5] developed a 2.4-GHz integrated parallel-plate soil moisture sensor system. The sensor detects the phase shift caused by the changes in the dielectric constant of soil which is related to the soil moisture content. This sensor needs a number of components, such as a phase locked loop (PLL), voltage controlled oscillator (VCO), phase shifter, phase detector, microcontroller, and
I
Manuscript received September 09, 2009; revised December 10, 2009; accepted December 23, 2009. Date of current version May 21, 2010. This work was supported in part by the U.S. Office of Naval Research under Grant N00014-02-1-0623. The associate editor coordinating the review of this paper and approving it for publication was Prof. Okyay Kaynak. The authors are with the Department of Electrical Engineering, University of South Carolina, Columbia, SC 29208 USA (e-mail: [email protected]; [email protected]; [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2010.2040175
power management circuitry on board to be fully functional. In [6], Saxena and Tayal described a parallel-plate capacitor as a moisture sensor. The sensor consists of a Wein bridge oscillator, a capacitance bridge, a differential amplifier, a band-pass filter, a rectifier circuit, and an analog meter. Ong et al. [7] proposed a 24 MHz embedded LC resonant sensor for moisture content measurement, which requires a large external loop and an impedance analyzer. To maintain the integrity of concrete structures and to allow minimum disruption, an embedded sensor should occupy as small space as possible. The sensor should also only have very few and small components. To address these issues, we propose a planar interdigitated sensor fabricated on a very thin substrate. Because of its thin planar construction the interdigitated sensor can be easily embedded in concrete for moisture content measurement. Our proposed sensor occupies a very small space. In recent years, interdigitated sensors have drawn the attention of researchers due to their low cost and simplicity. They have found application in humidity sensing [8], chemical sensing [9], [10], gas detection [11], resin curing [12], etc. In our research group Bhuiyan et al. [13] have designed and developed interdigitated sensors which can detect insulation faults in unshielded power cables. This was achieved by applying a low frequency signal and then measuring the resulting interelectrode capacitance of the sensor. It was shown that a half circular sensor was less sensitive to its position with respect to the sample under test than a meander sensor. In this paper, the efficacy of two interdigitated sensors in moisture sensing in concrete samples is studied. First, the meander sensor introduced in our earlier work [13] was used. Then, a full circular interdigitated sensor with larger field penetration depth was designed, built, and tested. Experimental measurements were performed to determine the sensor interelectrode capacitance by varying the moisture content of concrete samples with both types of sensors. Measurement results are compared with results obtained from analytical equations and simulations from Maxwell 3-D solver. This paper is organized as follows. In Section II, the basic interdigitated sensor geometry, operating principles and its analytical model are discussed. In Section III, the details of sensor fabrication and the experimental setup and measurement procedure are described. All results are presented and discussed in Section IV followed by the conclusion in Section V. II. SENSOR GEOMETRY, CIRCUIT, AND ANALYTICAL MODEL An interdigitated sensor is a coplanar structure consisting of multiple parallel fingers. The sensor can measure material dielectric constant by applying fringing electric fields into the
1530-437X/$26.00 © 2010 IEEE Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1244
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Fig. 3. Unit cell representation of interdigitated sensor.
The substrate has zero conductivity, so the current is due to the capacitive effect only. Also, since the opamp is operating in the inverting mode, it is obvious that Fig. 1. Typical geometry of an interdigitated sensor.
(1) Fig. 3 shows a unit cell of an interdigitated sensor without the conducting backplane. Using the conformal mapping method Scapple [16] derived the total per unit capacitance for an interdigitated meander sensor as
(2)
where
Fig. 2. (a) Block diagram of sensor circuit and (b) its equivalent circuit.
material. The electrodes of the sensor must be in contact with the material under test. Fig. 1 shows the geometry of an interdigitated sensor [13]. To learn more about the sensor structure and its operation, the reader may review [12]. Three types of electrodes are present, namely the driving electrode, the sensing is electrode, and the guard electrode. A sinusoidal voltage applied to the driving electrode and the output voltage is measured from the sensing electrode. Guard electrodes and a conducting backplane are used to shield the sensor from the influence of external fields [14]. Another important term is the penetration depth of the field which is directly related to the spatial wavelength, . The penetration depth is approximately equal to 1/3 of [15]. A popular method to find the dielectric constant of a material is the short-circuit current method, which is shown in Fig. 2. is applied to the driving electrode A low-frequency voltage while the sensing electrode is connected to a precision opamp (AD708). The opamp acts as an inverting amplifier where a is used for feedback. known capacitor We used a simplified circuit model in Fig. 2(b) to calculate the capacitance between the driving and the sensing electrodes.
and
are the elliptic integrals
and , respectively [16]. The of modulus first part on the right-hand side of the equation resembles the capacitance of the material under test and the substrate and the second part resembles the capacitance of the trapped material in between the interelectrode gap. In (2), it is assumed that there is no conducting backplane. If a conducting backplane is present and the substrate is very thin, then the field lines that penetrate the substrate end at the backplane and not at the sensing elec. trodes. So the substrate capacitance has no contribution to A more accurate expression for per unit capacitance is found in [13]
(3)
where is the corrected per unit capacitance. is the correction factor. For the meander sensor, the correction factor is (4) where is a constant. For the typical meander sensor that was used in our experiment the value of is 0.02. If the electrode length is and the total number of unit cells is then the total capacitance
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
(5)
ALAM et al.: CONCRETE MOISTURE CONTENT MEASUREMENT USING INTERDIGITATED NEAR-FIELD SENSORS
1245
Fig. 5. Ansoft Maxwell 3-D model of full circular sensor.
Concrete block samples were prepared using concrete mix. The age of the samples was more than one year at the time of the measurement. The dimensions of the samples were 15 cm 15 cm 2 cm and 15 cm 15 cm 4 cm. The wet basis for the moisture content in a material is defined as (6) Fig. 4. (a) Experimental setup to measure (b) placement of the sensor in the samples.
C
for concrete samples and
is the mass of the specimen with water and is the where mass of the specimen in dry condition. The moisture content of can be calculated as a material by volume
III. EXPERIMENTAL SETUP To detect the percentage of moisture in concrete, two interdigitated sensors were designed and fabricated. One was a meander sensor and the other was a circular sensor. Two sensor configurations were chosen to create different field penetration depths. The meander sensor is simpler and has smaller field penetration depth than the circular sensor. For the same field penetration depth, the circular sensor covers a larger surface area than the meander sensor. The electrode width, height, gap, and spa, tial wavelength of the meander sensor were 1.125 mm, 17 1.125 mm, and 4.5 mm, respectively. Thus, the field penetration depth for the meander sensor is about 1.5 mm. For the circular , 2 mm, and 7 mm, sensor these dimensions were 1.5 mm, 17 respectively. Hence, the field penetration depth for the circular sensor is about 7/3 mm. The meander sensor was fabricated on ) substrate and the circular a 10 mil thick Duroid 5880 ( sensor was fabricated on a 31 mil thick Duroid 5880 substrate. Referring to Fig. 1, the meander sensor had seven driving electrodes, four sensing electrodes, and two guard electrodes. The length of each electrode was 20 mm. For the circular sensor (see Fig. 5), the inner diameter of the innermost sensing electrode was 6 mm. The circular sensor had 12 driving electrodes, ten sensing electrodes, and two guard electrodes. Both sensors were coated with very thin perylene coating to avoid short circuit with the concrete, especially in the wet condition. A 1-kHz, . The ampli10-V (peak) sinusoidal signal was chosen for tude of the signal was small enough to prevent the opamp from was 100 pF. saturating. The feedback capacitor
(7) is the density of concrete in dry condition and is where the density of water. For measurement the sensor was placed between two concrete was measured for samples. First, the sensor output voltage was known, the capacthe samples in dry condition. Since was found using (1). After completion of the meaitance surement in dry condition, the concrete samples were weighed . Then, the samples were placed in in a scale to determine a 13.5-l water bucket and were completely submerged in water for two days. Wet samples were taken out from the water bucket. Since excess water present on the concrete surface had the potentials for and high ) due to the high pererroneous results (high mittivity of water the outer surfaces of the samples were wiped off using a dry cloth. A detailed experimental measurement setup is shown in Fig. 4. The sensor was placed between the wet concrete samwas recorded and the ples. At regular periodic intervals, was measured. For each weight of the concrete samples was found using (1) while the moisture content reading, was found using (6) and (7). Thus, for each case, a new and its corresponding was found. Since the evaporation rate depended on the environment, the ambient temperature and the air flow, an external fan was used to make the evaporation process faster.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1246
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Fig. 7. Comparison between the measured, analytical and simulated data for the meander sensor.
The sensitivity of the circular sensor is 5.454 pF/ percent change in moisture content. B. Comparison Between Experimental and Simulation Results
Fig. 6. Measured capacitance, C versus moisture content, meander sensor and (b) the full circular sensor.
M
for (a) the
IV. RESULTS A. Experimental Results versus data for both the meander and the Measured circular sensor are shown in Fig. 6. A linear regression analysis was also performed on the measured data. From Fig. 6(a), in while for , dry condition . The presence of water in the wet concrete clearly re(126.4% increase). The sensitivity sulted in the increase in of the meander sensor is 0.519 pF/percent change in moisture content. Fig. 6(a) also shows that the coefficient of determinawhen . Measured versus tion results for the circular sensor are shown in Fig. 6(b). Since the field penetration depth of the circular sensor is larger than that of the meander sensor the resulting linear regression curve in Fig. 6(b) is steeper than the curve in Fig. 6(a). The coefficient of . In dry condition determination while with , (187% increase).
Ansoft Maxwell 3-D was used to simulate the sensor responses for both the meander and the circular sensors. Fig. 5 is an example of the simulation model which was used for the full circular sensor. In the simulation, we were not able to account for the moisture content directly. In [17], experimental data of relative permittivity and conductivity versus frequency for four different levels of moisture content in concrete (from Fig. 2 of [17]) are available. These data are valid for the frequency range of 10 MHz to 1 GHz. We extended their curves down to 1 kHz using curve fitting in Matlab. Then, we developed linear relations between the relativity permittivity and conductivity with percentage of moisture content using curve fitting. The resulting relative permittivity and conductivity values were used in our Maxwell simulations. The relative permittivity ( ) and conductivity ( ) of the concrete samples that were used in our simulation and analytical studies (3)–(4) were: 4.4323, 0.000224; 6.0333, 0.0009; 7.177, 0.00215; 8.549, 0.00444; and 9.693, 0.007 for moisture contents of 0%, 1.816%, 3.1135%, 4.67%, and 5.968%, respectively. In Maxwell 3-D, the solution type was electrostatic. The driving electrodes were set to 10 V and the other electrodes including the backplane were set to 0 V. The default boundary condition, the Neumann homogeneous condition was used. The capacitances between the driving electrodes and the sensing electrodes were computed for the four different cases. Since equations (2) to (5) can only correctly describe the capacitance for the meander sensor, analytical results were obtained only for the meander sensor. Comparison between the experimental, simulation, and analytical data for the meander sensor is shown in Fig. 7. As seen, in general the agreement between all three is quite good. Some deviation is observed at higher moisture content levels. Fig. 8
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: CONCRETE MOISTURE CONTENT MEASUREMENT USING INTERDIGITATED NEAR-FIELD SENSORS
1247
REFERENCES
Fig. 8. Comparison of measured and simulated capacitance for the full circular sensor.
illustrates the comparison between the experimental and simulation results for the circular sensor. Again, agreement is excellent for low moisture contents. The measured results are 10% to 20% higher than the simulated results for moisture contents larger than 4%. V. CONCLUSION The prospect of detecting and measuring moisture in concrete samples was investigated using two interdigitated sensors, e.g., a meander sensor and a full circular sensor. The measurement data clearly indicate that both types of sensors can detect and measure moisture in concrete. There is a distinctly linear relationship between the moisture content in the concrete and the measured interelectrode capacitance. Hence, the amount of moisture present can be easily predicted from the measured capacitance. The disagreement between the simulation and measurement data can be due to the following. First, as mentioned, in our Maxwell simulations the exact values of the relative permittivity and conductivity of concrete at 10 kHz and as function of moisture content was not available. We used curve fitting to extrapolate data from the measured permittivity and conductivity data of concrete that was valid from 10 MHz to 1 GHz. Since Maxwell does not use moisture content values directly accurate values of relative permittivity and conductivity are essential to obtain accurate simulation results. Second, perhaps during the wiping off of the concrete surface there was some standing water on the concrete surface. Although the outside concrete surfaces were wiped off with a dry cloth but the uneven surface might have allowed for some water to still remain on the surface. In the future one may envision that such sensors in miniature form will be integrated with wireless modules and embedded inside concrete samples for moisture content measurement in order to prevent premature corrosion of the steel reinforcements.
[1] [Online]. Available: http://www.bts.gov/publications/state_transporta tion_statistics/state_transportation_statistics_2006/html/table_01_07. html [2] K. Wardhana and F. Hadipriono, “Analysis of recent bridge failures in the United States,” J. Perform. Construct. Fac., vol. 17, no. 3, pp. 144–150, Aug. 2003. [3] C. Maierhofer, “Nondestructive evaluation of concrete infrastructure with ground penetrating radar,” J. Mater. Civil Eng., vol. 15, no. 3, pp. 287–297, May/Jun. 2003. [4] K. M. Z. Shams and M. Ali, “Wireless power transmission to a buried sensor in concrete,” IEEE Sensors J., vol. 7, no. 12, pp. 1573–1577, Dec. 2007. [5] H. Xing, J. Li, R. Liu, E. Oshinski, and R. Rogers, “2.4 GHz on-board parallel plate soil moisture sensor system,” in Proc. Sensors Industry Conf., Houston, TX, Feb. 2005, pp. 35–38. [6] S. C. Saxena and G. M. Tayal, “Capacitive moisture meter,” IEEE Trans. Ind. Electron. Contr. Instrum., vol. IECI-28, no. 1, pp. 37–39, Feb. 1981. [7] J. B. Ong, Z. You, J. Mills-Beale, E. L. Tan, B. D. Pereles, and K. G. Ong, “A wireless, passive embedded sensor for real-time monitoring of water content,” IEEE Sensors J., vol. 8, pp. 2053–2058, Dec. 2008. [8] R. S. Jachowicz and S. D. Senturia, “A thin-film capacitance humidity sensor,” Sens. Actuators, vol. 2, pp. 171–186, Dec. 1982. [9] H. E. Endres and S. Drost, “Optimization of the geometry of gas-sensitive interdigital capacitors,” Sens. Actuators B, vol. 4, pp. 95–98, May 1991. [10] R. Zhou, A. Hiermann, K. D. Schierbaum, K. E. Geckeler, and W. Gopel, “Detection of organic solvents with reliable chemical sensors based on cellulose derivatives,” Sens. Actuators B, vol. 24–25, pp. 443–447, 1995. [11] A. Leidl, R. Hartinger, M. Roth, and H. E. Endres, “A new SO sensor system with SAW and IDC elements,” Sens. Actuators B, vol. B34, no. 2, pp. 339–342, 1996. [12] A. V. Mamishev, “Interdigital Dielectrometry Sensor Design and Parameter Estimation Algorithms for Non-Destructive Materials Evaluation,” Ph.D. dissertation, Dept. Elect. Eng. Comput. Sci., Mass. Inst. Technol., Cambridge, 1999. [13] R. H. Bhuiyan, R. A. Dougal, and M. Ali, “Proximity coupled interdigitated sensors to detect insulation damage in power system cables,” IEEE Sensors J., vol. 7, no. 12, pp. 1589–1596, Dec. 2007. [14] A. V. Mamishev, K. Sundara-Rajan, F. Yang, Y. Du, and M. Zahn, “Interdigital sensors and transducers,” Proc. IEEE, vol. 92, pp. 808–845, May 2004. [15] Y. Du, “Measurements and Modeling of Moisture Diffusion Processes in Transformer Insulation Using Interdigital Dielectrometry Sensors,” Ph.D. dissertation, Dept. Elect. Eng. Comput. Sci., Mass. Inst. Technol., Cambridge, 1999. [16] R. Y. Scapple, “A trimmable planar capacitor for hybrid applications,” in Proc. 24th Electron. Component Conf., 1974, pp. 203–207. [17] M. N. Soutsos, J. H. Bungey, S. G. Millard, M. R. Shaw, and A. Patterson, “Dielectric properties of concrete and their influence on radar testing,” in Proc. NDT & E Int., 2001, vol. 34, pp. 419–425.
Md. Nazmul Alam (S’09) received the B.Sc. and M.Sc. degrees in electrical engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 2006 and 2008, respectively. Since 2009, he has been a Graduate Research Assistant with the Microwave Engineering Laboratory, Department of Electrical Engineering, University of South Carolina, Columbia. His research interests include sensor design and integration, RFID, and designing of microwave antennas.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
1248
IEEE SENSORS JOURNAL, VOL. 10, NO. 7, JULY 2010
Rashed H Bhuiyan (S’09) received the B.Sc. and M.Sc. degrees in electrical engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 2003 and 2005, respectively. He is currently working towards the Ph.D. degree in the Department of Electrical Engineering, University of South Carolina, Columbia. He is the author of multiple international conference and journal papers. He also served as a faculty member with the Bangladesh University of Engineering and Technology, Dhaka, from 2003 to 2005. His current research interests include miniature electromagnetic sensor design, antennas, and RF circuits.
Roger A. Dougal (M’82–SM’94) received the Ph.D. degree from Texas Tech University, Lubbock, in 1983. He leads the Power and Energy Group, Electrical Engineering Department, University of South Carolina (USC), Columbia, which has a principal research focus in the general area of power electronics. As a member of the Board of Directors of the Electric Ship Research and Development Consortium, he oversees USC’s activities related to new power generation, processing, and distribution technologies for ships. He is also the Site Director for a new NSF-sponsored Industry/University Cooperative Research Center for Grid-connected Advanced Power Electronics, which is a joint project between USC and the University of Arkansas. Also, since 1996, under sponsorship of the Office of Naval Research, he has overseen development of the Virtual Test Bed software, which is a comprehensive simulation and virtual prototyping environment for multidisciplinary dynamic systems. This environment has been applied in studies of electric systems for navy ships, electrochemical power sources, hybrid power sources, power electronics, and controls.
Mohammod Ali (M’93–SM’03) received the B.Sc. degree in electrical and electronic engineering from the Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, in 1987, and the M.A.Sc. and Ph.D. degrees, both in electrical engineering, from the University of Victoria, Victoria, BC, Canada, in 1994 and 1997, respectively. He was with the Bangladesh Institute of Technology, Chittagong, from 1988 to 1992. From January 1998 to August 2001, he was with Ericsson Inc., Research Triangle Park, NC. Since August 2001, he has been with the Department of Electrical Engineering, University of South Carolina, Columbia, where currently he is an Associate Professor. He had also held appointments as a Visiting Research Scientist with the Motorola Corporate EME Research Laboratory, Plantation, FL, from June to August 2004. He established the Microwave Engineering Laboratory at the University of South Carolina in 2001. He is the author/coauthor of over 100 publications and five granted U.S. patents. His research interests include miniaturized packaged (embedded) antennas, meta-materials and their antenna applications, distributed wireless sensors and rectennas, and portable/wearable antennas and their interactions with humans (SAR). Dr. Ali is the recipient of the 2003 National Science Foundation Faculty Career Award. He is also the recipient of the College of Engineering and Information Technology Young Investigator Award and the Research Progress Award from the University of South Carolina in 2006 and in 2008, respectively. He was the Technical Program Co-Chair of the IEEE Antennas and Propagation Society’s International Symposium in Charleston, SC, in 2009. He has also served as a member of the Technical Program Committee for the IEEE Antennas and Propagation Society’s International Symposium for a number of years. He has served as a reviewer and panelist for grant proposals for a number of federal and local funding agencies. He is an Associate Editor for the IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS.
Authorized licensed use limited to: University of South Carolina. Downloaded on June 09,2010 at 19:24:54 UTC from IEEE Xplore. Restrictions apply.
