Neural networks technologies application in problems of critical places status monitoring of transport aircraft structure

Aeronautical and Space-Rocket Engineering

Strength and thermal conditions of flying vehicles


DOI: 10.34759/vst-2020-4-81-91

Аuthors

Bautin A. A.*, Svirskiy Y. A.**

Central Aerohydrodynamic Institute named after N.E. Zhukovsky (TsAGI), 1, Zhukovsky str., Zhukovsky, Moscow Region, 140180, Russia

*e-mail: andrey.bautin@tsagi.ru
**e-mail: yury.svirsky@tsagi.ru

Abstract

Air fleet developing prospects all over the world are closely associated with creation of highly efficient methods for maintaining the aircraft airworthiness. One of the tasks, being solved while such methods developing, is cost reduction during the aircraft operation. A reliable and rather effective periodic inspections system can be replaced by the structure status monitoring, which consists in continuous data collection and analysis of airframe integrity throughout the aircraft entire life span.

Status monitoring is performed by the onboard system, which basic elements are recording and analyzing unit, and sensors. The sensors are fixing the structure response at its integrity violation during operation. The damages detection effectiveness and possibility of reliable determination of the operation conditions depends in many ways on the algorithms realization, in which accordance the analyzing unit operates.

Currently, a large number of sensors types, based on various physical principles, have been developed. Strain gauges, which change of readings may indicate the presence of the structure damage, were widely employed while the experiment and approbation of the onboard monitoring systems.

The article proposes a method for determining the sensors installation scheme while fatigue damage detecting in the fuselage joints with account for the local nature of changes in the stress-strain state near the cracks and the allowable size of cracks that can be considered safe under certain conditions. The multi-site damage parameters, at which the residual strength of the joints does not decrease below the permissible level, were selected by studying the fractures of the joint samples by fractography. The optimal sensors installation scheme determining was performed based on the analysis of relation between of the measurement system readings and damages. This relation is presented herewith in the form of the neural network approximation.

The neural network training to obtain the necessary relation was performed based on the results of local deformations determining by the finite element method for various options of the of cracks location in the critical section of the joint. Various factors affecting strain measurements were accounted for while determining the places of sensors installation.

The article presents the result of the developed methodology application for the optimal sensors installation scheme determining in one of the types of longitudinal fuselage joints when detecting multi-point fatigue cracks during fatigue tests.

Keywords:

status monitoring, lap joint, multi-site fatigue damages, neural network approximation, finite element model of bolted joint

References

  1. Daletskii S.V., Derkach O.Ya., Petrov A.N. Effektivnost’ tekhnicheskoi ekspluatatsii samoletov grazhdanskoi aviatsii (Technical operation efficiency of civil aviation aircraft), Moscow, Vozdushnyi transport, 2002, 205 p.

  2. Maximov N.А., Maluta E.V., Sharonov A.V. Automated system for aircraft failures recorded during preflight inspection recordkeeping. Aerospace MAI Journal, 2015, vol. 22, no. 4, pp. 85-90.

  3. Chukhlebov R.V., Loshkarev A.N., Sidorenko A.S., Dmitriev V.G. Experimental research of an aircraft product’s structure vibrations under flight loads action. Aerospace MAI Journal, 2017, vol. 24, no. 3, pp. 51-59.

  4. Patrikeev S.A. Capabilities of onboard innovation measuring systems while ground and flight tests. Aerospace MAI Journal, 2018, vol. 25, no. 1, pp. 76-83.

  5. Doebling S.W., Farrar C.R., Prime M.B., Shevitz D.W. Damage Identification and Health Monitoring of Structural and Mechanical Systems from Changes in their Vibration Characteristics: A Literature Review. The Shock and Vibration Digest. Los Alamos National Laboratory Technical Report LA-13070-MS, New Mexico, 1996, 127 p. DOI: 10.2172/249299

  6. Pisupati P., Dewangan S.K., Kumar R. Structural Health Monitoring (SHM). Enabling Technology for paradigm shift in next generation Aircraft Design and Maintenance, Infosys, 2011, 8 p.

  7. Kamath G.M., Sundaram R., Gupta N., Rao S.M. Damage Studies in Composite Structures for Structural Health Monitoring using Strain Sensors. Structural Health Monitoring, 2010, vol. 9, no. 6, pp. 497-512. DOI: 10.1177/1475921710365391

  8. Kahandawa G.C. Monitoring Damage in Advanced Composite Structures Using Embedded Fibre Optic Sensors. PhD thesis. University of Southern Queensland, Toowoomba, Australia, 2012, 258 p. URL: http://eprints.usq.edu.au/id/eprint/24005

  9. Busarova M.V., Zhelonkin S.V., Kulesh V.P., Kuruliuk K.A. Application of optical videogrammetry technique for normal deformation fields of aircraft fuselage panel meausring. Aerospace MAI Journal, 2020, vol. 27,no. 2, pp. 52-60. DOI: 10.34759/vst-2020-2-52-60

  10. Wada D., Igawa H., Tamayama M. et al. Flight demonstration of aircraft fuselage and bulkhead monitoring using optical fiber distributed sensing system. Smart Materials and Structures, 2018, vol. 27, no. 2. DOI: 10.1088/1361-665X/aaa588

  11. Yari T., Nagai K., Hotate K. et al. Monitoring Aircraft Structural Health Using Optical Fiber Sensors. Mitsubishi Heavy Industries, Ltd. Technical Review, 2008, vol. 45, no. 4, pp. 5-8. URL: https://www.mhi.co.jp/technology/review/pdf/e454/e454005.pdf

  12. Jin B., Zhang W., Liang X. et al. Statistic model for calculating the Hole-Edge crack length using Fiber Bragg grating sensors. 9th European Workshop on Structural Health Monitoring EWSHM (10-13 July 2018, Manchester, United Kingdom). URL: https://www.ndt.net/article/ewshm2018/papers/0269-Jin.pdf

  13. Katsikeros Ch.E., Labeas G.N. Development and validation of a strain-based Structural Health Monitoring system. Mechanical Systems and Signal Processing, 2009, vol. 23, no. 2, pp. 372-383. DOI: 10.1016/j.ymssp.2008.03.006

  14. Guemes А., Fernandez-Lopez A. Damage Detection in Bolted Joints by Fibre Optics Distributed Sensing. 2nd Int. Symposium on NDT in Aerospace (22-24 November 2010; Hamburg, Germany), 6 p. URL: https://www.ndt.net/article/aero2010/papers/mo2a2.pdf

  15. Aviatsionnye Pravila. Ch. 25. Normy letnoi godnosti samoletov transportnoi kategorii (Civil Aviation Safety Regulations. Part 25. Norms of airworthiness of transport category aircraft), Moscow, Aviaizdat, 2009, 267 p.

  16. AC 25.571-1D «Damage Tolerance and Fatigue Evaluation of Structure». Department of transport. Federal Aviation Administration, 13.01.2011. URL: https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_25_571-1D_.pdf

  17. Svirskiy Yu.A., Bautin A.A., Papic L., Gadolina I.V. Methods of Modern Aircraft Structural Health Monitoring and Diagnostics of Technical State. In: Ram M., Davim J. (eds.) Diagnostic Techniques in Industrial Engineering. Management and Industrial Engineering. Springer, Cham. DOI: 10.1007/978-3-319-65497-3_1

  18. Swift T. Damage tolerance capability. International Journal of Fatigue, 1994, vol. 16, no. 1, pp. 75-94. DOI: 10.1016/0142-1123(94)90446-4

  19. Nesterenko B.G. Nauchnyi vestnik MGTU GA. Seriya «Aeromekhanika i prochnost», 2007, no. 119, pp. 70-81.

  20. Gromov V.F., Dunaev V.V., Eryomin M.V., Makarov A.F. Improvement of aviation structure quality and life in mechanical joint. Aerospace MAI Journal, 2010, vol. 17, no. 1, pp. 18-24.

  21. Haykin S. Neural Networks: A Comprehensive Foundation (2nd Edition). Prentice Hall, 1998, 842 p.

  22. Svirskii Yu.A., Bautin A.A., Luk’yanchuk A.A., Basov V.N. Approximate method for local elastic-plastic problems solving. Aerospace MAI Journal, 2020, vol. 27, no. 2, pp. 61-70. DOI: 10.34759/vst-2020-2-61-70

mai.ru — informational site of MAI

Copyright © 1994-2024 by MAI