Full-Depth Assessment of Concrete Bridge Decks in A GPR Survey: A Machine Learning Approach

The current practice for condition evaluation of bridge decks using GPR does not provide a full depth assessment of the deck and is limited only to the portion of the deck above the top reinforcement mat. Additionally, there are ambiguities and uncertainties associated with the GPR data interpretation especially when it comes to differentiating the concrete material condition. The amplitude or the shape of the hyperbola of reflection from rebars in the B-scan do not provide sufficient information regarding the material condition. A series of FDTD simulations were conducted to expand the GPR evaluation zone beyond the top reinforcement level, and to provide a full depth assessment of the deck. The slab was divided into three separate yet interconnected longitudinal layers. A parametric study was performed by changing the properties of each concrete layer. The quality of concrete was characterized by two electromagnetic properties – permittivity and conductivity. Using the electromagnetic properties as characteristic parameters, six concrete conditions from good to critical were simulated. A machine learning technique called gradient boosting was used to predict the layers’ condition. The amplitudes at the top and bottom reinforcing layer, and at the bottom of the slab, as well as the travel times were used as the model input features. The paper presents the study of the influence of each feature on the prediction of layer properties.

References

 

  • "Report Card for America’s Infrastructure, 2017 Grades," American Society of Civil Engineers (ASCE), 2017. [Online]. Available: http://www.infrastructurereportcard.org/. [Accessed May 2018].
  • N. Gucunski, A. Imani, F. Romero, S. Nazarian, D. Yuan, H. Wiggenhauser, P. Shokouhi, A. Taffe and D. Kutrubes, "Nondestructive Testing to Identify Concrete Bridge Deck Deterioration," FHWA, Washington, D.C., 2013.
  • N. P. Singh and M. J. Nene, "Buried Object Detection and Analysis of GPR Images: Using Neural Network and Curve Fitting," in International Conference on Microelectronics, Communication and Renewable Energy (ICMiCR-2013), Kerala, India, 2013.
  • M. R. Shaw, M. S. G, T. C. K. Molyneaux and M. J. Taylor, "Location of Steel Reinforcement in Concrete Using Ground Penetrating Radar and Neural Networks," NDT & E International, vol. 38, no. 3, pp. 203-212, 2004.
  • W. Al-Nuaimy, Y. Huang, M. Nakhkash, M. T. C. Fang, V. T. C. Nguyen and A. Eriksen, "Automatic detection of buried utilities and solid objects with GPR using neural networks and pattern recognition," Journal of Applied Geophysics, vol. 43, pp. 157-165, 2000.
  • K. Dinh, N. Gucunski, J. Kim and T. H. Duong, "Understanding Depth-Amplitude Effects in Assessment of GPR Data from Concrete Bridge Deck Surveys," NDT & E International, vol. 83, pp. 48-58, 2016.
  • T. C. K. Molyneaux, S. G. Millard, J. H. Bungey and J. Q. Zhou, "Radar assessment of structural concrete using neural networks," NDT & E International, vol. 28, no. 5, pp. 281-288, 1995.
  • J. B. Rodriguez, M. F. Pantoja, X. L. Travassos, D. A. G. Vieira and S. R. R, "A predictionalgorithmfordataanalysisinGPR-basedsurveys," Neurocomputing, vol. 168, pp. 464-474, 2015.
  • C. Warren, A. Giannopoulos and I. Giannakis, "gprMax: Open source software to simulate electromagnetic wave propagation for Ground Penetrating Radar," Computer Physics Communications, 2016.
  • "ANSYS High Frequency Structure Simulator," ANSYS, [Online]. Available: http://www.ansys.com/Products/Simulation+Technology/Electromagnetics/High- Performance+Electronic+Design/ANSYS+HFSS. [Accessed 15 July 2012].

 

Metrics
Usage Shares
Total Views
33 Page Views
Total Shares
0 Tweets
33
0 PDF Downloads
0
0 Facebook Shares
Total Usage
33