TL;DR
This paper introduces a novel interferometric approach using a multistatic ground-penetrating radar (GPR) system to detect subtle variations in airport pavement thickness and wave propagation velocity, indicative of thin cracks. This method offers higher resolution and simpler analysis compared to traditional GPR techniques, enabling more effective inspection of critical infrastructure.
What Problem Does This Paper Address?
Maintaining the integrity of airport runways and taxiways is crucial for safety. Pavement deterioration, particularly the formation of thin cracks, can compromise structural integrity. Conventional inspection methods are often inefficient and lack the necessary accuracy to detect these subtle defects. While Ground-Penetrating Radar (GPR) is a promising non-destructive testing (NDT) method, its resolution is often insufficient to directly image millimeter-scale cracks. Existing GPR techniques, like Common Midpoint (CMP) analysis, can estimate subsurface properties but suffer from limitations in resolution and data acquisition complexity, making them less practical for large-scale airport inspections. This paper aims to bridge this gap by developing a more practical and higher-resolution GPR-based method for detecting these hard-to-see pavement anomalies.
Methods
The core of this paper is a novel interferometric approach that leverages the Yakumo multistatic GPR system. Instead of relying on direct reflection signals from thin cracks, which are often too weak to detect, the method focuses on identifying slight spatial variations in the pavement's dielectric permittivity. These variations are caused by the presence of thin cracks filled with water and air.
The proposed method builds upon the concept of Common Midpoint (CMP) analysis but introduces a significant enhancement:
- Subsample Time Delay Estimation: The key innovation is the use of subsample time delay estimation, achieved through an iterative cross-correlation approach. This allows for precise measurement of the time difference in reflected signals between two CMP datasets acquired at slightly different locations or under different conditions (e.g., a reference "healthy" section versus a potentially damaged section).
- Interferometric Comparison: By comparing two CMP datasets (a reference and a monitoring point), the method analytically estimates deviations in both propagation velocity ($\Delta v$) and layer thickness ($\Delta z$) from their reference values. This is achieved by solving a system of equations derived from the first-order differential arrival time, effectively treating the problem as an overdetermined system solved using least squares.
- Yakumo Multistatic GPR System: The study utilizes the Yakumo system, a multistatic GPR array with eight transmitting and eight receiving antennas. This configuration allows for rapid acquisition of multiple CMP datasets simultaneously, significantly improving data acquisition efficiency for large-scale surveys.
This approach moves away from graphical extraction of velocity from velocity spectra, offering a more direct and analytical estimation of deviations.
Datasets
The paper utilizes both simulated and real-world datasets for evaluation:
- Simulated Datasets:
- Finite-Difference Time-Domain (FDTD) simulations were used to generate CMP datasets for single-layer models with known thickness and velocity variations. These simulations helped in validating the theoretical underpinnings of the method and assessing its performance under controlled conditions.
- Ray tracing simulations were also employed to generate data for a simple model, demonstrating the initial concept of time delay estimation.
- Field Datasets:
- Experimental Model Site: Data were collected at an airport taxiway model at the Port and Airport Research Institute in Nobi, Japan. This site featured manmade fracture zones embedded within asphalt layers, allowing for controlled testing of the method's ability to detect known anomalies.
- Actual Airport Site: Data were acquired at Haneda International Airport in Tokyo, Japan, in a test area with known damaged pavement. This provided a real-world scenario to evaluate the method's practical applicability.
Key Insights and Ideas
- Indirect Detection of Thin Cracks: The most significant insight is that thin cracks, which are difficult to image directly with GPR, can be effectively detected by identifying the subtle changes they induce in the pavement's dielectric permittivity and, consequently, its wave propagation velocity and thickness.
- Time Delay as a Key Indicator: The method's reliance on precise time delay estimation between CMP datasets, rather than direct waveform analysis, makes it more robust to signal distortions and less dependent on the exact shape of the reflected signal.
- Efficiency of Multistatic GPR: The Yakumo multistatic GPR system's ability to acquire multiple CMP datasets concurrently is crucial for making this high-resolution inspection practical for large airport areas.
- Analytical vs. Graphical Estimation: Shifting from graphical interpretation of velocity spectra to analytical calculation of velocity and thickness deviations simplifies the analysis process and potentially improves accuracy.
- Crosstalk as a Challenge: The study acknowledges that in real-world scenarios, distinguishing between velocity and thickness deviations can be challenging due to "crosstalk," especially when subsurface inhomogeneities are complex.
Key Results
- Simulation Performance: In simulation tests, velocity deviations on the order of a few millimeters per nanosecond were successfully detected. The FDTD simulations demonstrated that the iterative cross-correlation method could recover time delay changes where conventional cross-correlation failed.
- Model Site Results: At the experimental model site, the method successfully identified the manmade fracture zones at approximately 4 m and 9 m along the survey line, correlating well with ground truth.
- Actual Airport Site Results: At Haneda Airport, the method identified damaged pavement regions that agreed with impact-echo method results. While the estimated deviations in velocity and thickness were larger than expected compared to ground truth (maximal values of 0.05 m/ns and 8 mm), the method effectively indicated the presence and location of anomalies.
- Robustness: The proposed method showed good performance in detecting slight deviations and was less affected by noise compared to conventional methods.
Limitations and Future Work
- Crosstalk: The paper acknowledges that the estimated velocity and thickness deviations can be correlated (crosstalk), making it difficult to perfectly distinguish between the two, especially in complex subsurface environments. This means the results should be considered as indications of inhomogeneities rather than exact numerical values.
- Overestimation in Real Data: The real-world airport data showed larger deviations than expected, attributed to factors like wavelet distortion from scatterers, the choice of reference point, and potential air gaps between the antenna and the ground.
- Numerical Solution as Additional Information: The authors suggest that numerical solutions should be considered as supplementary information, used in conjunction with known site parameters for practical applications.
- Future Work: While not explicitly detailed, the findings suggest future work could focus on improving the separation of velocity and thickness deviations, refining the reference point selection, and further investigating the impact of environmental factors on the measurements. The method's suitability for real-time inspection is highlighted, implying potential for further development in that direction.
Summary updated