Skip to main content
SpringerLink
Log in

Modelling the scatter of E–N curves using a serial hybrid neural network

  • Original Article
  • Published:
Neural Computing and Applications Aims and scope Submit manuscript

Abstract

If structural reliability is estimated by following a strain-based approach, a material’s strength should be represented by the scatter of the ε–N (E–N) curves that link the strain amplitude with the corresponding statistical distribution of the number of cycles-to-failure. The basic shape of the ε–N curve is usually modelled by the Coffin–Manson relationship. If a loading mean level also needs to be considered, the original Coffin–Manson relationship is modified to account for the non-zero mean level of the loading, which can be achieved by using a Smith–Watson–Topper modification of the original Coffin–Manson relationship. In this paper, a methodology for estimating the dependence of the statistical distribution of the number of cycles-to-failure on the Smith–Watson–Topper modification is presented. The statistical distribution of the number of cycles-to-failure was modelled with a two-parametric Weibull probability density function. The core of the presented methodology is represented by a multilayer perceptron neural network combined with the Weibull probability density function using a size parameter that follows the Smith–Watson–Topper analytical model. The article presents the theoretical background of the methodology and its application in the case of experimental fatigue data. The results show that it is possible to model ε–N curves and their scatter for different influential parameters, such as the specimen’s diameter and the testing temperature.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16

Similar content being viewed by others

Notes

  1. The S–N curve represents the dependence between the stress (usually it is an amplitude stress) and the number of cycles-to-failure at different stress levels.

  2. The ε–N curve represents the dependence between the amplitude strain and the number of cycles-to-failure at different stress levels.

  3. The P SWT(N f ) durability curve is in its essence merely a modified ε–N durability curve.

References

  1. Orbanic P, Fajdiga M (1999) R&D process located at automotive industry suppliers. In: Fajdiga M, Jurejevcic T, Trenc F (eds) Proceedings of 4th conference and exhibition innovative automotive technology—IAT ‘99, Nova Gorica, Slovenia, 8th–9th April 1999, pp 233–240

  2. Bucar T, Nagode M, Fajdiga M (2006) Predicting the scatter of S–N curves by hybrid multilayer perceptron for structural reliability evaluation. In: Soares CG, Zio E (eds) Safety and reliability for managing risk: proceedings of the European and reliability conference 2006 (ESREL 2006), Estoril, Portugal, 18–22 September 2006. Taylor & Francis, pp 873–880

  3. Nagode M, Fajdiga M (1998) On a new method for prediction of the scatter of loading spectra. Int J Fatigue 20(4):271–277

    Article  Google Scholar 

  4. Nagode M, Klemenc J, Fajdiga M (2001) Parametric modelling and scatter prediction of rainflow matrices. Int J Fatigue 23(6):525–532

    Article  Google Scholar 

  5. Shen H, Lin J, Mu E (2000) Probabilistic model on stochastic fatigue damage. Int J Fatigue 22(7):569–572

    Article  Google Scholar 

  6. Tovo R (2001) On the fatigue reliability evaluation of structural components under service loading. Int J Fatigue 23(7):587–598

    Article  Google Scholar 

  7. Zhao YX, Yang B, Zhai ZY (2008) The framework for a strain-based fatigue reliability analysis. Int J Fatigue 30(3):493–501

    Article  Google Scholar 

  8. Dowling NE (1999) Mechanical behaviour of materials: engineering methods for deformation, fracture and fatigue. Prentice Hall, Upper Saddle River

    Google Scholar 

  9. Bucar T, Nagode M, Fajdiga M (2006) A neural network approach to describing the scatter of s–n curves. Int J Fatigue 28(4):311–323

    Article  MATH  Google Scholar 

  10. Bucar T, Nagode M, Fajdiga M (2007) An improved neural computing method for describing the scatter of s–n curves. Int J Fatigue 29(12):2125–2137

    Article  MATH  Google Scholar 

  11. Coffin LF (1954) A study of the effects of cyclic thermal stresses on a ductile material. Trans ASME 76(6):931–950

    Google Scholar 

  12. Manson SS (1965) Fatigue: a complex subject—some simple approximations. Exp Mech 5(7):193–226

    Article  Google Scholar 

  13. Buxbaum O (1992) Betriebsfestigkeit: sichere und wirtschafliche Bemessung schwingbruchgefaehredeter Bauteile. Strahleisen, Duesseldorf

    Google Scholar 

  14. Haibach E (1989) Betriebsfestigkeit: Verfahren und Daten zur Bauteilerechnung. VDI Verlag, Duesseldorf

    Google Scholar 

  15. Graham JA, Millan JF, Franklin J (1968) Fatigue design handbook. SAE, Warrendale

    Google Scholar 

  16. Manson SS, Halford GR (1981) Practical implementation of the double linear damage rule and damage curve approach for treating cumulative fatigue damage. Int J Fract 17(2):169–180

    Article  Google Scholar 

  17. Smith KN, Watson P, Topper TH (1970) A stress strain function for the fatigue of metals. J Mater JMLSA 5(4):767–778

    Google Scholar 

  18. Rosa U, Nagode M, Fajdiga M (2007) Strain-life approach in thermo-mechanical fatigue evaluation of complex structures. Fatigue Fract Eng Mater Struct 30(9):808–822

    Article  Google Scholar 

  19. Nagode M, Zingsheim F (2004) An online algorithm for temperature influenced fatigue-life estimation: strain-life approach. Int J Fatigue 26(2):155–161

    Article  MATH  Google Scholar 

  20. Agarwal M (1997) Combining neural and conventional paradigms for modelling, prediction and control. Int J Syst Sci 28(1):65–81

    Article  MATH  Google Scholar 

  21. Klemenc J, Fajdiga M (2002) A neural network approach to the simulation of load histories by considering the influence of a sequence of rainflow load cycles. Int J Fatigue 24(11):1109–1125

    Article  MATH  Google Scholar 

  22. Klemenc J, Fajdiga M (2004) An improvement to the methods for estimating the statistical dependencies of the parameters of random load states. Int J Fatigue 26(2):141–154

    Article  MATH  Google Scholar 

  23. Janezic M, Klemenc J, Fajdiga M (2010) A neural-network approach to describe the scatter of cyclic stress–strain curves. Mater Des 31(1):438–448

    Article  Google Scholar 

  24. Bishop CM (1995) Neural networks for pattern recognition. Clarendon Press, Oxford

    Google Scholar 

  25. Press WH (1988) Numerical recipes in C. Cambridge University Press, Cambridge

    MATH  Google Scholar 

  26. Haykin S (1994) Neural networks, a comprehensive foundation. Macmillan, New York

    MATH  Google Scholar 

  27. Akaike H (1974) A new look at the statistical model identification. IEE Transact Autom Cont 19(6):716–723

    Article  MathSciNet  MATH  Google Scholar 

  28. Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical information–theoretic approach, 2nd edn. Springer, New York

    MATH  Google Scholar 

  29. Hastie T, Tibshirani R, Friedman J (2001) The elements of statistical learning: data mining, inference and prediction. Springer, New York

    MATH  Google Scholar 

  30. Cohen AC (1965) Maximum likelihood estimation in the Weibull distribution based on complete and on censored samples. Technometrics 7(4):579–588

    Article  MathSciNet  Google Scholar 

  31. ASTM International (1996) Standard practice for conducting force controlled constant amplitude axial fatigue tests of metallic materials. ASTM Standard E466–96

  32. ASTM International (1992) Standard practice for strain-controlled fatigue testing. ASTM Standard E606–92

Download references

Acknowledgments

The authors would like to thank the company CIMOS d.d. and the Ministry of Science and Technology of the Republic of Slovenia, which supported our research within the project “Innovative development of parts and technologies for automotive industry in frame of PTC.”

Author information

Authors and Affiliations

  1. Faculty of Mechanical Engineering, University of Ljubljana, Askerceva 6, 1000, Ljubljana, Slovenia

    Jernej Klemenc, Miha Janezic & Matija Fajdiga

Authors
  1. Jernej Klemenc
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Miha Janezic
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Matija Fajdiga
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Jernej Klemenc.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Klemenc, J., Janezic, M. & Fajdiga, M. Modelling the scatter of E–N curves using a serial hybrid neural network. Neural Comput & Applic 21, 1517–1530 (2012). https://doi.org/10.1007/s00521-012-0828-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-012-0828-2

Keywords

Search

Navigation