Identification of defects in a coating wedge based on ultrasonic non-destructive testing methods and convolutional neural networks

A.N. Soloviev, B.V. Sobol, P.V. Vasiliev, A.V. Senichev, A.I. Novikova

Author details

A.N. Soloviev
Don State Technical University, Rostov-on-Don, Russian Federation
B.V. Sobol
Don State Technical University, Rostov-on-Don, Russian Federation
P.V. Vasiliev
Don State Technical University, Rostov-on-Don, Russian Federation
A.V. Senichev
Don State Technical University, Rostov-on-Don, Russian Federation
A.I. Novikova
Don State Technical University, Rostov-on-Don, Russian Federation

https://doi.org/10.15593/perm.mech/2023.1.11

The paper deals with the identification of a crack-like defect in a coated wedge based on ultrasonic nondestructive testing. The authors propose an approach of defect identification followed by determination of its geometrical parameters. The approach is based on a shadowed ultrasonic nondestructive testing method combined with deep machine learning technologies. A wedge-shaped area is inspected for the presence of an internal defect. On one edge of the wedge there is a source of ultrasonic vibrations, on the opposite edge there is a receiver. Passing through the coating and body of the wedge, part of the signal is reflected from inhomogeneities and defects that may be present in it. The signal reaching the opposite edge of the wedge is read by the receiver. The received data is processed by a neural network model, which predicts the presence or absence of an internal defect and, if present, determines geometric parameters such as size and position. A finite element model of ultrasonic wave propagation inside the wedge is constructed. Special damping layers are used, due to which the influence of parasitic signal reflections and its further propagation into the wedge body is significantly reduced. Based on the built model, the shadow method of ultrasonic scanning is implemented. This method implies that on one side of the wedge are installed excitation devices, and on the opposite side – receiving devices. Several numerical experiments for various combinations of geometric parameters of the wedge and the defect have been performed using a distributed computing system. Based on the obtained data, a neural network model was built and trained, capable of identifying the defect and determining its characteristics. The input of the model is spectrograms of the readout signal, and the output is values characterizing the defect.

Keywords

ultrasonic nondestructive testingconvolutional neural networksthin coatingelastic wedgecrackflaw detectionmachine learningartificial neural networksdefect identificationfinite element modeling

Идентификация дефектов в клине с покрытием на основе методов ультразвукового неразрушающего контроля и сверточных нейронных сетей

Рассматривается идентификация трещиноподобного дефекта в клине, усиленного покрытием на основе ультразвукового неразрушающего контроля. Авторы предлагают подход идентификации дефекта с последующим определением его геометрических параметров. Подход основан на теневом методе ультразвукового неразрушающего контроля в сочетании с технологиями глубокого машинного обучения. Инспектируется клиновидная область на наличие внутреннего дефекта. На одной грани клина расположен источник ультразвуковых колебаний, на противоположной грани – приемник. Проходя через покрытие и тело клина, часть сигнала отражается от неоднородностей и дефектов, которые могут присутствовать в нем. Сигнал, достигший противоположной грани клина, считывается приемником. Полученные данные обрабатываются нейросетевой моделью, которая прогнозирует наличие или отсутствие внутреннего дефекта, и в случае его наличия определяет геометрические параметры, такие как размер и положение. Построена конечно-элементная модель распространения ультразвуковой волны внутри клина. Применяются специальные демпфирующие слои, благодаря которым существенно снижается влияние паразитных отражений сигнала и его дальнейшее распространение в тело клина. На основе построенной модели реализован теневой метод ультразвукового сканирования. Данный метод предполагает, что на одной стороне клина устанавливаются возбуждающие сигнал устройства, а на противоположной стороне – принимающие. С помощью системы распределенных вычислений проведен ряд численных экспериментов для различных комбинаций геометрических параметров клина и дефекта. На основе полученных данных построена и обучена нейросетевая модель, способная идентифицировать дефект и определять его характеристики. На вход модели поступают спектрограммы считанного сигнала, а на выходе получаются значения, характеризующие дефект.

Ключевые слова

ультразвуковой неразрушающий контрольсверточные нейронные сетитонкое покрытиеупругий клинтрещинадефектоскопиямашинное обучениеискусственные нейронные сетиидентификация дефектовконечно-элементное моделирование

Bibliography

1.S. Samanta, A. Mandal, Thingujam Jackson Singh. Application of ANN in Identifying Defects in Impacted Composite // Procedia Materials Science, Volume 6, 2014, pp. 926-930, ISSN 2211-8128. Scholar
2.R.J.G. van Sloun, R. Cohen and Y.C. Eldar, "Deep Learning in Ultrasound Imaging," in Proceedings of the IEEE, vol. 108, no. 1, pp. 11–29, Jan. 2020, doi: 10.1109/JPROC.2019.2932116. Scholar
3.A.V. Cherpakov, V.A. Akop'yan, A.N. Solov'yev. Algoritm mnogoparametricheskoy identifikatsii defektov sterzhnevykh konstruktsiy [Algorithm for multi-parameter identification of defects in bar structures] // Tekhnicheskaya akustika http://www.ejta.org. – No1, 2013. Scholar
4.H.N. Bar, M.R. Bhat, C.R.L. Murthy, Identification of failure modes in GFRP using PVDF sensors: ANN approach // Composite Structures, Volume 65, Issue 2, August 2004, Pages 231-237, ISSN 0263-8223. Scholar
5.Dong-Fang Chen, Xia-Ting Feng, Ding-Ping Xu, Quan Jiang, Cheng-Xiang Yang, Pin-Pin Yao. Use of an improved ANN model to predict collapse depth of thin and extremely thin layered rock strata during tunneling // Tunnelling and Underground Space Technology, Volume 51, January 2016, Pages 372-386, ISSN 0886-7798. Scholar
6.Vatul'yan A.O. Obratnye zadachi v mekhanike deformiruemogo tverdogo tela [Inverse problems in solid mechanics]. – M.: Fizmatlit, 2007, P. 224. Scholar
7.Ermolov I., Aleshin N., Potapov A. Non-destructive testing. Acoustic methods of control. – Moscow: High School, 1991. Scholar
8.Willcox M.A. A Brief Description of NDT Techniques / M.A. Willcox – Toronto: NDT Equipment Limited, 2003. – 54 p. Scholar
9.J.S. Hesthaven and T. Warburton, Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications, Springer, 2008. Scholar
10.Soloviev A., Sobol B., Vasiliev P. (2019) Identification of Defects in Pavement Images Using Deep Convolutional Neural Networks. In: Parinov I., Chang SH., Kim YH. (eds) Advanced Materials. Springer Proceedings in Physics, vol 224. Springer, Cham. Scholar
11.Vasil'yev P.V., Senichev A.V. Primenenie neyrosetevykh tekhnologiy v zadache kontrolya poverkhnostnykh defektov [Application of neural network technologies in the problem of surface defects control] // Scientific-educational and applied journaluniversity newsNorth-Caucasian region. Technical sciences series. 2020; 1:33–40. Scholar
12.Cao Vu Dung, Le DucAnh, Autonomous concrete crack detection using deep fully convolutional neural network, Automation in Construction, Volume 99, 2019, Pages 52-58. Scholar
13.Shengfeng Liu, Yi Wang, Xin Yang, Baiying Lei, Li Liu, Shawn Xiang Li, Dong Ni, Tianfu Wang, Deep Learning in Medical Ultrasound Analysis: A Review, Engineering, Volume 5, Issue 2,2019, Pages 261-275. Scholar
14.Gradient-based learning applied to document recognition / Y. LeCun [et al.] // Proceedings of the IEEE. – 1998. – Vol. 86, iss. 11. – pp. 2278–2324. Scholar
15.Ioffe, S. Batch normalization: Accelerating deep network training by reducing internal covariate shift / S. Ioffe, C. Szegedy // CoRR. – 2015. – 9 p. – DOI: abs/1502.03167. Scholar
16.Solov'yev A.N., Sobol' B.V., Vasil'yev P.V. Ul'trazvukovaya lokatsiya vnutrennikh treshchinopodobnykh defektov v sostavnom uprugom tsilindre s primeneniem apparata iskusstvennykh neyronnykh setey [Ultrasonic location of internal crack-like defects in a composite elastic cylinder using the apparatus of artificial neural networks] // Russian Journal of Nondestructive Testing. – 2016. – No3. – pp. 3–9. Scholar
17.Sobol' B.V., Solov'yev A.N., Rashidova E.V., Vasil'yev P.V. Identifikatsiya treshchinopodobnogo defekta i issledovanie kontsentratsii napryazheniy v polose s pokrytiem [Identification of a crack-like defect and study of stress concentration in a coated strip] // PNRPU Mechanics Bulletin. – 2019. – No 4. – pp. 165–174. DOI: 10.15593/perm.mech/2019.4.16. Scholar
18.Andreaus, U., &Baragatti, P. (2011). Cracked beam identification by numerically analysing the nonlinear behaviour of the harmonically forced response. Journal of Sound and Vibration, 330(4), 721–742. Scholar
19.Andreaus, U., &Casini, P. (2016). Identification of multiple open and fatigue cracks in beam-like structures using wavelets on deflection signals. Continuum Mechanics and Thermodynamics, 28(1–2), 361–378. Scholar
20.Placidi, L., Misra, A., &Barchiesi, E. (2019). Simulation results for damage with evolving microstructure and growing strain gradient moduli. Continuum Mechanics and Thermodynamics, 31(4), 1143–1163. Scholar
21.Timofeev, D., Barchiesi, E., Misra, A., &Placidi, L. (2021). Hemivariational continuum approach for granular solids with damage-induced anisotropy evolution. Mathematics and Mechanics of Solids, 26(5), 738–770. Scholar
22.Cuomo, M., Contrafatto, L., & Greco, L. (2014). A variational model based on isogeometric interpolation for the analysis of cracked bodies. International Journal of Engineering Science, 80, 173–188. Scholar
23.Volkov, I.A., Igumnov, L.A., dell'Isola, F., Litvinchuk, S.Y., &Eremeyev, V.A. (2020). A continual model of a damaged medium used for analyzing fatigue life of polycrystalline structural alloys under thermal–mechanical loading. Continuum Mechanics and Thermodynamics, 32(1), 229–245. Scholar
24.Spagnuolo, M., Barcz, K., Pfaff, A., dell’Isola, F., &Franciosi, P. (2017). Qualitative pivot damage analysis in aluminum printed pantographic sheets: numerics and experiments. Mechanics Research Communications, 83, 47–52. Scholar
25.Berezovski, A., Yildizdag, M. E., &Scerrato, D. (2020). On the wave dispersion in microstructured solids. Continuum Mechanics and Thermodynamics, 32(3), 569–588. Scholar
26.Turco, E. (2017). Tools for the numerical solution of inverse problems in structural mechanics: review and research perspectives. European Journal of Environmental and Civil Engineering, 21(5), 509–554. Scholar
27.Bilotta, A., Morassi, A., &Turco, E. (2017). Quasi-isospectral Sturm-Liouville operators and applications to system identification. Procedia engineering, 199, 1050–1055. Scholar
28.Abali, B. E., Klunker, A., Barchiesi, E., &Placidi, L. (2021). A novel phase‐field approach to brittle damage mechanics of gradient metamaterials combining action formalism and history variable. ZAMM‐Zeitschrift für Angewandte Mathematik und Mechanik, e202000289. Scholar
29.Glushkov S.V., SkvortsovYu.V., Perov S.N. Sravnenie rezul'tatov resheniya zadachi mekhaniki razrusheniya dlya truby s neskvoznoy treshchinoy [Comparison of the results of solving the problem of fracture mechanics for a pipe with a non-through crack] // PNRPU Mechanics Bulletin. – 2014. – No 3. – pp. 36–49. doi: 10.15593/perm.mech/2014.3.03. Scholar
30.Anoshkin A.N., Zuyko V.Yu., Osokin V.M. Modelirovanie tekhnologicheskikh defektov i otsenka ikh vliyaniya na staticheskuyu prochnost' kompozitnykh flantsev [Modeling of technological defects and evaluation of their influence on the static strength of composite flanges] // PNRPU Mechanics Bulletin. – 2016. – No 2. – pp. 5–21. doi: 10.15593/perm.mech/2016.2.01. Scholar

1.Samanta S., Mandal A., Thingujam Jackson Singh. Application of ANN in Identifying Defects in Impacted Composite // Procedia Materials Science. – 2014. – Vol. 6. – Р. 926–930. Scholar
2.van Sloun R.J.G., Cohen R., Eldar Y.C. Deep Learning in Ultrasound Imaging // Proceedings of the IEEE. – 2020. – Vol. 108, no. 1. – P. 11–29. DOI: 10.1109/JPROC.2019.2932116 Scholar
3.Черпаков А.В., Акопьян В.А., Соловьев А.Н. Алгоритм многопараметрической идентификации дефектов стержневых конструкций [Электронный ресурс] // Техническая акустика: электронный журнал. – URL: http://www.ejta.org. – 2013. – No 1. Scholar
4.Bar H.N., Bhat M.R., Murthy C.R.L. Identification of failure modes in GFRP using PVDF sensors: ANN approach // Composite Structures. – 2004. – Vol. 65, iss. 2. – P. 231–237. Scholar
5.Use of an improved ANN model to predict collapse depth of thin and extremely thin layered rock strata during tunneling / Dong-Fang Chen, Xia-Ting Feng, Ding-Ping Xu, Quan Jiang, Cheng-Xiang Yang, Pin-Pin Yao // Tunnelling and Underground Space Technology. – 2016. – Vol. 51. – P. 372–386. Scholar
6.Ватульян А.О. Обратные задачи в механике деформируемого твёрдого тела. – М.: Физматлит, 2007. – С. 224. Scholar
7.Ermolov I., Aleshin N., Potapov A. Non-destructive testing. Acoustic methods of control. – Moscow: High School, 1991. Scholar
8.Willcox M.A. A Brief Description of NDT Techniques. – Toronto: NDT Equipment Limited, 2003. – 54 p. Scholar
9.Hesthaven J.S., Warburton T. Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications. – Springer, 2008. Scholar
10.Soloviev A., Sobol B., Vasiliev P. (2019) Identification of Defects in Pavement Images Using Deep Convolutional Neural Networks // Parinov I., Chang S.H., Kim Y.H. (eds) Advanced Materials. Springer Proceedings in Physics. – 2009. – Vol 224. Scholar
11.Васильев П.В., Сеничев А.В. Применение нейросетевых технологий в задаче контроля поверхностных дефектов // Известия высших учебных заведений. Северо-Кавказский регион. Технические науки. – 2020. – No 1ю. – С. 33–40. Scholar
12.Cao Vu Dung, Le Duc Anh, Autonomous concrete crack detection using deep fully convolutional neural network // Automation in Construction. – 2019. – Vol. 99. – P. 52–58. Scholar
13.Deep Learning in Medical Ultrasound Analysis: A Review / Shengfeng Liu, Yi Wang, Xin Yang, Baiying Lei, Li Liu, Shawn Xiang Li, Dong Ni, Tianfu Wang // Engineering. – 2019. – Vol. 5, iss. 2. – P. 261–275. Scholar
14.Gradient-based learning applied to document recognition / Y. LeCun [et al.] // Proceedings of the IEEE. – 1998. – Vol. 86, iss. 11. – P. 2278–2324. Scholar
15.Ioffe S., Szegedy C. Batch normalization: Accelerating deep network training by reducing internal covariate shift // CoRR. – 2015. – 9 p. – DOI: abs/1502.03167 Scholar
16.Соловьев А.Н., Соболь Б.В., Васильев П.В. Ультразвуковая локация внутренних трещиноподобных дефектов в составном упругом цилиндре с применением аппарата искусственных нейронных сетей // Дефектоскопия. – 2016. – No 3. – C. 3–9. Scholar
17.Идентификация трещиноподобного дефекта и исследование концентрации напряжений в полосе с покрытием / Соболь Б.В., Соловьев А.Н., Рашидова Е.В., Васильев П.В. // Вестник Пермского национального исследовательского политехнического университета. Механика. – 2019. – No 4. – С. 165–174. DOI: 10.15593/perm.mech/2019.4.16 Scholar
18.Andreaus U., Baragatti P. Cracked beam identification by numerically analysing the nonlinear behaviour of the harmonically forced response // Journal of Sound and Vibration. – 2011. – Vol. 330(4). – P. 721–742. Scholar
19.Andreaus U., Casini P. Identification of multiple open and fatigue cracks in beam-like structures using wavelets on deflection signals // Continuum Mechanics and Thermodynamics. – 2016. – Vol. 28(1–2). – P. 361–378. Scholar
20.Placidi L., Misra A., Barchiesi E. Simulation results for damage with evolving microstructure and growing strain gradient moduli // Continuum Mechanics and Thermodynamics. – 2019. – Vol. 31(4). – P. 1143–1163. Scholar
21.Hemivariational continuum approach for granular solids with damage-induced anisotropy evolution / D. Timofeev, E. Barchiesi, A. Misra, L. Placidi // Mathematics and Mechanics of Solids. – 2021. – Vol. 26(5). – P. 738–770. Scholar
22.Cuomo M., Contrafatto L., Greco L. A variational model based on isogeometric interpolation for the analysis of cracked bodies // International Journal of Engineering Science. – 2014. – Vol. 80. – P. 173–188. Scholar
23.A continual model of a damaged medium used for analyzing fatigue life of polycrystalline structural alloys under thermal–mechanical loading / I.A. Volkov, L.A. Igumnov, F. dell'Isola, S.Y. Litvinchuk, V.A. Eremeyev // Continuum Mechanics and Thermodynamics. – 2020. – Vol. 32(1). – P. 229–245. Scholar
24.Qualitative pivot damage analysis in aluminum printed pantographic sheets: numerics and experiments / M. Spagnuolo, K. Barcz, A. Pfaff, F. dell’Isola, P. Franciosi // Mechanics Research Communications. – 2017. – Vol. 83. – P. 47–52. Scholar
25.Berezovski A., Yildizdag M.E., Scerrato D. On the wave dispersion in microstructured solids // Continuum Mechanics and Thermodynamics. – 2020. – Vol. 32(3). – P. 569–588. Scholar
26.Turco E. Tools for the numerical solution of inverse problems in structural mechanics: review and research perspectives // European Journal of Environmental and Civil Engineering. – 2017. – Vol. 21(5). – P. 509–554. Scholar
27.Bilotta A., Morassi A., Turco E. Quasi-isospectral Sturm-Liouville operators and applications to system identification // Procedia engineering. – 2017. – Vol. 199. – P. 1050–1055. Scholar
28.A novel phase‐field approach to brittle damage mechanics of gradient metamaterials combining action formalism and history variable / B.E. Abali, A. Klunker, E. Barchiesi, L. Placidi // ZAMM‐Zeitschrift für Angewandte Mathematik und Mechanik. – 2021. – e202000289. Scholar
29.Глушков С.В., Скворцов Ю.В., Перов С.Н. Сравнение результатов решения задачи механики разрушения для трубы с несквозной трещиной // Вестник Пермского национального исследовательского политехнического университета. Механика. – 2014. – No 3. – C. 36–49. DOI: 10.15593/perm.mech/2014.3.03 Scholar
30.Моделирование технологических дефектов и оценка их влияния на статическую прочность композитных фланцев / А.Н. Аношкин, В.Ю. Зуйко, В.М. Осокин [и др.] // Вестник Пермского национального исследовательского политехнического университета. Механика. – 2016. – No 2. – C. 5–21. DOI: 10.15593/perm.mech/2016.2.01 Scholar

Article information

Published: 29 May 2023
Issue date: 29 May 2023
Issue: No. 1 (2023)