研究目的
To develop a sensitive damage detection system for reinforced concrete bridge slabs using ground-penetrating radar (GPR) with a new deconvolution algorithm and higher frequency bands to detect millimeter-order thin cracks and segregations.
研究成果
The proposed time-variant deconvolution algorithm at SHF band significantly improves the detection sensitivity for millimeter-order damages in RC bridge slabs. It successfully detects 1-mm cracks, especially when filled with water, and provides accurate depth and condition assessments. The method enhances GPR's utility for noncontact, high-speed infrastructure monitoring, with potential for broader application after further validation.
研究不足
The study assumes a 1-D model and ignores 3-D effects and inhomogeneity of media. Multireflection and transmission losses are not fully considered, potentially leading to underestimation. The algorithm's performance depends on accurate noise level assumptions and initial reflection coefficient estimates. Applicability may vary for concrete with different compositions or standards.
1：Experimental Design and Method Selection:
The study uses a 1-D bridge slab model to inversely estimate reflection coefficient distributions. Algorithms include time-invariant and time-variant deconvolution, considering signal attenuation in concrete. A Wiener filter is applied for noise reduction.
2：Sample Selection and Data Sources:
Concrete specimens with varying water-cement ratios (30% and 65%) and a real bridge slab model with modeled damages (e.g., 1-mm horizontal cracks and 2-cm segregations) are used. Data is sourced from simulations and field experiments.
3：List of Experimental Equipment and Materials:
Double-ridge horn antennas, a network analyzer, data logger, measurement cart, coaxial probe system, free-space method setup with concrete specimens of thicknesses 3-30 cm, and modeled damages with spacers and sand/water fillings.
4：Experimental Procedures and Operational Workflow:
For electromagnetic property measurements, coaxial probe and free-space methods are used. Field experiments involve measuring radar signals at points on the bridge slab model, applying deconvolution algorithms, and analyzing results. Simulations add noise to synthetic signals to validate the Wiener filter.
5：Data Analysis Methods:
Data is analyzed using deconvolution in time and frequency domains, singular value decomposition for pseudoinverse matrices, and statistical methods for evaluating permittivity and attenuation coefficients. Results are compared to theoretical values.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
