Sensitivity and Robustness of Neural Networks for Defect-Depth Estimation in CFRP Composites

  • Numan Saeed
  • Houda Al Zarkani
  • Mohammed A. OmarEmail author


Carbon fiber reinforced composites are lightweight materials that possess a desirable mechanical performance manifested in high strength to weight ratio. However, due to their anisotropic and low thermal conductive nature; the detection of delamination and impact damages using thermography is still a challenging task. This research paper aims at conducting a comprehensive analysis of the neural network when used as a post-processor to quantify defects depths from thermograms of CFRP; a pulsed thermography inspection routine was used to generate the thermograms. A thorough study is conducted to compare and analyze the effect of the different parameters on the accuracy and the robustness of the neural network outcome. The optimized neural network architecture and hyper-parameters are realized based on the accuracy yields (defect depths estimate). This research explores the (i) network related parameters and the (ii) thermography-data specific factors such as frame rate, number of considered points per defect location, etc.; showing that the robustness of the neural network is found more sensitive to: the network architecture, noise, training hyperparameters and the type, amount and diversity of the training data.


Sensitivity Pulsed thermography Non-destructive evaluation Finite elements method Neural network Defect depth detection 



  1. 1.
    Ley, O., Godinez-Azcuaga, V.: Line scanning thermography and its application inspecting aerospace composites. In: 5th International Symposium on NDT in Aerospace, pp. 13–15, 2013Google Scholar
  2. 2.
    Bonavolonta, C., Valentino, M., Peluso, G., Barone, A.: Non destructive evaluation of advanced composite materials for aerospace application using HTS SQUIDs. IEEE Trans. Appl. Supercond. 17(2), 772–775 (2007)CrossRefGoogle Scholar
  3. 3.
    Matthews, F.L., Rawlings, R.D.: Composite Materials: Engineering and Science. Elsevier, New York (1999)Google Scholar
  4. 4.
    Soutis, C.: Fibre reinforced composites in aircraft construction. Prog. Aerosp. Sci. 41(2), 143–151 (2005)CrossRefGoogle Scholar
  5. 5.
    Smith, R.A.: Composite defects and their detection. Mater. Sci. Eng. 3, 103–143 (2009)Google Scholar
  6. 6.
    Avdelidis, N.P., Hawtin, B.C., Almond, D.P.: Transient thermography in the assessment of defects of aircraft composites. NDT E Int. 36(6), 433–439 (2003)CrossRefGoogle Scholar
  7. 7.
    Theodorakeas, P., Avdelidis, N.P., Hrissagis, K., Ibarra-Castanedo, C., Koui, M., Maldague, X.: Automated transient thermography for the inspection of CFRP structures: experimental results and developed procedures. In: Thermosense: Thermal Infrared Applications XXXIII, vol. 8013, p. 80130W, 2011Google Scholar
  8. 8.
    Jasinien, E., Raiutis, R., Voleiis, A., Vladiauskas, A., Mitchard, D., Amos, M.: NDT of wind turbine blades using adapted ultrasonic and radiographic techniques. Insight-Non-Destr. Test. Cond. Monit. 51(9), 477–483 (2009)CrossRefGoogle Scholar
  9. 9.
    Ghoshal, A., Martin, W.N., Schulz, M.J., Chattopadhyay, A., Prosser, W.H., Kim, H.S.: Health monitoring of composite plates using acoustic wave propagation, continuous sensors and wavelet analysis. J. Reinf. Plast. Compos. 26(1), 95–112 (2007)CrossRefGoogle Scholar
  10. 10.
    Giordano, M., Calabro, A., Esposito, C., D’Amore, A., Nicolais, L.: An acoustic-emission characterization of the failure modes in polymer-composite materials. Compos. Sci. Technol. 58(12), 1923–1928 (1998)CrossRefGoogle Scholar
  11. 11.
    Bowler, J., Johnson, M.: Pulsed eddy-current response to a conducting half-space. IEEE Trans. Magn. 33(3), 2258–2264 (1997)CrossRefGoogle Scholar
  12. 12.
    Kharkovsky, S., Case, J.T., Abou-Khousa, M.A., Zoughi, R., Hepburn, F.L.: Millimeter-wave detection of localized anomalies in the space shuttle external fuel tank insulating foam. IEEE Trans. Instrum. Meas. 55(4), 1250–1257 (2006)CrossRefGoogle Scholar
  13. 13.
    Ganchev, S.I., Qaddoumi, N., Ranu, E., Zoughi, R.: Microwave detection optimization of disbond in layered dielectrics with varying thickness. IEEE Trans. Instrum. Meas. 44(2), 326–328 (1995)CrossRefGoogle Scholar
  14. 14.
    Ibarra-Castanedo, C., Genest, M., Piau, J.-M., Guibert, S., Bendada, A., Maldague, X.P.: Active infrared thermography techniques for the nondestructive testing of materials. Ultrasonic and advanced methods for nondestructive testing and material characterization, pp. 325–348. World Scientific, Singapore (2007)CrossRefGoogle Scholar
  15. 15.
    Giorleo, G., Meola, C.: Comparison between pulsed and modulated thermography in glass-epoxy laminates. NDT E Int. 35(5), 287–292 (2002)CrossRefGoogle Scholar
  16. 16.
    Peeters, J., Arroud, G., Ribbens, B., Dirckx, J.J.J., Steenackers, G.: Updating a finite element model to the real experimental setup by thermographic measurements and adaptive regression optimization. Mech. Syst. Signal Process. 64, 428–440 (2015)CrossRefGoogle Scholar
  17. 17.
    Sun, J.G.: Analysis of pulsed thermography methods for defect depth prediction. J. Heat Transf. 128(4), 329–338 (2006)CrossRefGoogle Scholar
  18. 18.
    Han, X., Favro, L.D., Kuo, P.K., Thomas, R.L.: Early-time pulse-echo thermal wave imaging. Review of Progress in Quantitative Nondestructive Evaluation, pp. 519–524. Springer, New York (1996)CrossRefGoogle Scholar
  19. 19.
    Omar, M., Hassan, M.I., Saito, K., Alloo, R.: IR self-referencing thermography for detection of in-depth defects. Infrared Phys. Technol. 46(4), 283–289 (2005)CrossRefGoogle Scholar
  20. 20.
    Maldague, X., Largouet, Y., Couturier, J.-P.: A study of defect depth using neural networks in pulsed phase thermography: modelling, noise, experiments. Rev. Gén. Therm. 37(8), 704–717 (1998)CrossRefGoogle Scholar
  21. 21.
    Darabi, A., Maldague, X.: Neural network based defect detection and depth estimation in TNDE. Ndt E Int. 35(3), 165–175 (2002)CrossRefGoogle Scholar
  22. 22.
    Saeed, N., Omar, M.A., Abdulrahman, Y.: A neural network approach for quantifying defects depth, for nondestructive testing thermograms. Infrared Phys. Technol. 94, 55–64 (2018)CrossRefGoogle Scholar
  23. 23.
    Book, F.A.B.: The Handbook of Brain Theory and Neural Networks. MIT Press, Cambridge (2002)Google Scholar
  24. 24.
    Rosenblatt, F.: The perceptron: a probabilistic model for information storage and organization in the brain. Psychol. Rev. 65(6), 386 (1958)CrossRefGoogle Scholar
  25. 25.
    Lowe, D., Broomhead, D.: Multivariable functional interpolation and adaptive networks. Complex Syst. 2(3), 321–355 (1988)MathSciNetzbMATHGoogle Scholar
  26. 26.
    Ibarra-Castanedo, C., Benítez, H., Maldague, X., Bendada, A.: Review of thermal-contrast-based signal processing techniques for the nondestructive testing and evaluation of materials by infrared thermography. In: Proceedings of the International Workshop on Imaging NDE, Kalpakkam, India, 25–28 April, 2007, pp. 1–6Google Scholar
  27. 27.
    Saeed, N., Omar, M.A., Abdulrahman, Y., Salem, S., Mayyas, A.: IR thermographic analysis of 3D printed CFRP reference samples with back-drilled and embedded defects. J. Nondestruct. Eval. 37(3), 59 (2018)CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  • Numan Saeed
    • 1
  • Houda Al Zarkani
    • 1
  • Mohammed A. Omar
    • 1
    Email author
  1. 1.Industrial and Systems Engineering DepartmentKhalifa University of Science and TechnologyAbu DhabiUnited Arab Emirates

Personalised recommendations