Sensitivity and accuracy in rebar diameter measurements from dual-polarized GPR data

According to some authors, high-frequency GPR (Ground Penetrating Radar) can provide on-site non-destructive measurements of the diameter of concrete reinforcements. The procedure should be based on the analysis of radar energy scattered by reinforcements when illuminated with parallel and perpendicular polarization. The theory of Radar Cross Section (RCS) supports this assumption. However, laboratory test performed on concrete specimens indicate that many practical problems must be solved to obtain reliable and stable measurements from real data. Synthetic data were used to analyze the problem. A data processing sequence was optimized to obtain the best fit between synthetic results and theoretical expectations. Results on real data processed with the same sequence were less encouraging. The comparison with synthetic data was essential to understand that background subtraction is the most critical issue. This problem is of larger importance than the fact that commercial GPR systems generate wide band radar pulses while RCS theory is frequency dependent. Once identified, the problem was directly addressed and results improved significantly. Although delicate, the methodology is potentially showing a high sensitivity which is proportional to wavelength. An optimal selection of antenna frequency versus rebar diameter range is proposed to preserve the highest sensitivity. Accuracy is related with sensitivity but also with stability and repeatability of the measurements. On-site experiments performed to explore this issue suggest that the problem must be addressed statistically. This makes the procedure more complex but the use of dual-polarized antennas might be the solution to save efficiency and cost-effectiveness. © 2012 Elsevier B.V. All rights reserved.