Use of electromagnetic two-layer wave-guided propagation in the GPR frequency range to characterize water transfer in concrete
Introduction
For decades, a considerable body of research has focused on studying the durability of civil engineering structures through the use of non-destructive testing (NDT) methods, whose features vary in degree of reliability. Among these methods, electromagnetic techniques are primarily based on an analysis of the temporal content of electromagnetic (EM) waves propagated in concrete [1], [2], [3], [4], [5], [6], [7]. Besides the efforts involved in interpreting EM wave dispersion relative to the dielectric nature of materials [8], certain applications derived from seismic techniques have revealed the existence of a second dispersion phenomenon due to the geometry and boundary conditions of the propagation medium [9], [10], [11], [12]. Within the GPR field of application, the notion of geometric dispersion has been introduced as a multi-modal propagation of EM waves through waveguides [13], [14], [15]: in passing through the waveguide layer, EM energy undergoes multiple reflections, leading to a series of constructive interferences among the various propagation modes. Under these conditions, the use of temporal methods to estimate permittivities and/or thicknesses through time peak picking, from WARR (Wide Angle Reflection and Refraction) or CMP (Common Mid-Point) configurations [16], [17], [18] becomes totally inappropriate, mainly for two-layer media with very thin layers or without enough dielectric contrasts at the interfaces.
In the presence of a strong dielectric contrast between the various layers composing the investigated medium, the EM energy is "trapped" in the upper layer of the structure, a situation that necessitates a modal approach to studying EM wave propagation [19], [20]. Arcone [21] is recognized as the precursor of this approach for the EM characterization of a thin ice layer bounded by air on its upper surface and liquid water below. Since the air/ice and ice/water dielectric contrasts are extremely high (i.e. strong reflectors), the ice layer therefore behaves like an EM waveguide. In assuming that the dominant frequencies of dispersive waves are sufficient to calculate the waveguide cutoff frequency, Arcone [21] was able to determine the dielectric permittivity of various ice layers.
More recently, van der Kruk et al. [22], [23] demonstrated the possibility of deriving the physical properties of each layer constituting the dielectric medium from the dispersion of GPR waves propagated in waveguides with both strong [23] and weak [22] permittivities. The results of these studies have shown that the propagation of Transverse Electric (TE) modes is less sensitive to EM leakage phenomena, which implies the use of "broadside" antenna orientations in order to generate such modes [23], [24]. As a follow-up to this work, a new inversion method has been implemented to estimate the dielectric permittivity and geometry of waveguides.
This paper will develop an adapted innovative procedure, from these last works [19], [20], [21], [22], [23], [24], for applying EM waveguide dispersion within the spectral domain in order to characterize two layers present in concrete structures, whose formation is due to water penetration into dry concrete and which decompose the investigated medium into two parts, i.e. dry and wet layers presenting a gradient between the two parts. An underlying question linked to this study is about the robustness of this proposed approach on a realistic two-layer medium with gradient.
This paper will thus begin with an introduction to the Parallel model and modal propagation theory, as adapted to the case of multilayer waveguide media delimited by both air (upper surface) and a strong reflector (lower surface). The next part will develop an analytical EM model, called the Waveguide Model (WGM), to be associated with an inversion procedure in a way that allows estimating the thicknesses of the studied medium by inverting the phase velocity dispersion curves. This model will be validated using both synthetic data generated from a FDTD (Finite Difference Time Domain) numerical model and experimental data stemming from the EM characterization of standard materials (simple cases with planar layers) using multi-offset GPR antennas. The experimental study conducted during the final step will associate the inversion procedure applied to GPR measurements with both the physical and hydric characteristics of two-tested concrete mix designs presenting realistic imbibition gradient fronts.
Section snippets
Fundamental equations of the EM waveguide model
The aim of this section is to introduce a basis with respect to the Parallel model and to EM modal theory that subsequently allows setting up the two-layer WGM.
Inversion and validation process on synthetic data
In this section, the two-layer waveguide inversion process has been validated using synthetic data covering several WARR configurations.
Validation on standard materials
The second step in the WGM validation process requires introducing GPR wave fields measured on standard materials. We chose two homogeneous materials, namely PVC and limestone, in order to form a two-layer waveguide and proceeded with GPR measurements using 2.6-GHz antennas. The waveguide has an equivalent permittivity similar to concrete samples in the dry state, despite the effects of concrete heterogeneities. This validation was performed to ensure the method yields accurate measurements.
Experimental program on concrete
The objective of this section is to compare water ingress depths of one concrete and one mortar obtained by GPR measurements, in combination with WGM (waveguide model) results, whereby water content gradients are evaluated using gammadensimetry measurements conducted several times during an absorption process [33]. The gammadensimetry results are thus considered as a reference.
Results and discussion
The concrete C and mortar M used in this study have displayed high porosities as high water permeabilities, which has significantly shortened the duration of the water ingress protocol until reaching a relative saturation of the slabs and cores. The high porosity of these concretes however offers the advantage of favoring the formation of a relatively sharp water penetration front during the ingress process, as depicted in Fig. 12, Fig. 13. Gammadensimetric measurements were performed as a
Conclusion
A recent innovative non-destructive technique using EM waveguide dispersion has been studied and applied to two layers of homogeneous materials and then to both concrete and mortar slabs during a realistic imbibition process. This technique is based on the waveguide inversion of dispersive TE-mode GPR data and the dielectric mixing model described by the Lichtenecker-Rother equation. Then, although the permittivities are defined as constants according to imbibition time, independently to
Acknowledgments
This work offers a contribution to the COST Action TU1208 entitled "Civil Engineering Applications of Ground Penetrating Radar". The authors would like to thank Odile Coffec for her serious work during the experimental campaign. The authors are also grateful to the "Pays de la Loire" Regional Council (2012-9651) for their financial support in favor of the project.
References (33)
Condition monitoring of deteriorating concrete dams using radar
Cem Concr Res
(2001)- et al.
Use of capacitive and gpr techniques for non-destructive evaluation of cover concrete
NDTE Int
(2008) - et al.
Characterization of concrete properties from dielectric properties using ground penetrating radar
Cem Concr Res
(2009) - et al.
Durability diagnosis of a concrete structure in a tidal zone by combining NDT methods: laboratory tests and case study
Constr Build Mater
(2012) - et al.
Lichtenecker's equation: applicability and limitations
Opt Commun
(2000) - et al.
Gammadensimetry: a method to determine drying and carbonation profiles in concrete
NDTE Int J
(2006) - et al.
Influence of concrete relative humidity on the amplitude of ground-penetrating radar (GPR) signal
Mater Struct
(2002) Sub-surface radar testing of concrete: a review
Constr Build Mater
(2004)- et al.
Detection of chlorides and moisture in concrete structures with ground penetrating radar
Mater Struct
(2008) - et al.
Complex permittivity frequency variations from multi-offset GPR data: hydraulic concrete characterization
IEEE Trans Instrum Meas
(2012)
In situ measurements of shear-wave velocity in sediments with higher-mode rayleigh waves
Geophys Prospect
Inversion of source-generated noise in high resolution seismic data
Lead Edge
Utilization of high frequency Rayleigh waves in near-surface geophysics
Lead Edge
Limites théoriques et expérimentales de l′interprétation de la dispersion des ondes de Rayleigh: Apport de la modélisation numérique et physique [Ph.D. dissertation]
Field observations of electromagnetic pulse propagation in dielectric slabs
Geophysics
Propagation of a ground penetrating radar (gpr) pulse in a thin-surface waveguide
Geophysics
Cited by (10)
Condition assessment of concrete-made structures using ground penetrating radar
2022, Automation in ConstructionReview of moisture measurements in civil engineering with ground penetrating radar – Applied methods and signal features
2021, Construction and Building MaterialsCitation Excerpt :Another inversion-based approach evaluates the occurring dispersion caused by thin low-velocity layers that act as waveguides [75–77]. Xiao et al. recently applied this principle to monitor the water ingress and transfer in concrete [78,79]. The inversion-based methods show promising results, however they are still considered as advanced research, that requires skilled and experienced users [80].
Application of GPR for determining electrophysical properties of structural layers and materials
2020, Journal of Applied GeophysicsDesign and validation of a multi-electrode embedded sensor to monitor resistivity profiles over depth in concrete
2019, Construction and Building MaterialsCitation Excerpt :Moreover, for these same surface techniques, various problems (surface irregularities, material variability, segregation of aggregates, etc.) generate dispersion in measurements and penalize access to deeper information. Regarding the GPR technique, the surface measurements require a complex inversion procedure to assess hydric profiles [18–20] that may not yet be suitable for operational applications. As for the gammadensimetry technique, it is restricted to test specimens and is not suitable for in situ structures.
Review on microwave nondestructive testing techniques and its applications in concrete technology
2019, Construction and Building MaterialsMonitoring water transfers in limestone building materials with water retention curve and Ground Penetrating Radar: A comparative study
2018, NDT and E InternationalCitation Excerpt :Section VI provides results and discussions, while Section VII closes by drawing a number of conclusions. An innovative two-layer waveguide model [21,25] is applied to the CMP (Common Mid-Point) measurements of the GPR to monitor the water transfer in limestone block for the imbibition experiment in the laboratory. To simplify the study, the limestone block is approximated as a two-layer waveguide medium, formed by the dry layer and the saturated layer.