TECHNOLOGY AREA(S): Air Platform, Ground/Sea Vehicles, Space Platforms

ACQUISITION PROGRAM: PMA-299 (Rotary) H-60 Helicopter Program

OBJECTIVE: Develop an improved loading spectrum development methodology utilizing wavelet transformations, and other signal processing techniques, to produce a loading spectrum optimized to reproduce real-world loading while minimizing cost and schedule requirements.

DESCRIPTION: Ideally, a fatigue test should exactly reproduce the loading conditions experienced by a given component or part during the entirety of its service life. However, current methods make it prohibitive to run a fatigue test in this manner. The U.S. Navy has relied mainly on traditional truncation methods with arbitrary criteria to reduce the length of test spectra in full-scale and component testing. These conventional methods do not address spectrum compression in multi-axial loading situations and account for material memory effects during fatigue loading. Currently, spectrum editing and compression in the form of truncation is carried out to reduce the total number of cycles while maintaining fatigue damage imparted into the test article. Appropriate clipping levels are also determined.

Early experiments of accelerated fatigue testing in laboratory settings relied on techniques such as increased loading frequency, increased load level, and removal of small amplitude cycles from the time history to accomplish these goals. The techniques, some still in use today, are carried out with an empirical set of guidelines and recommendations and rely on experienced personnel and/or extensive coupon testing. The current “trial and error” empirical methodology approach has shortcomings in the high cycle range with both damage replacement and clipping. Success in fatigue signal editing is determined by the reduction in spectrum length achieved while maintaining equivalent fatigue damage and representative failure modes.

The automotive industry popularized and currently employs a different wavelet analysis technique to characterize stress time histories. Wavelet analysis decomposes a time series into a time-frequency space, showing the dominant frequency modes and how these modes change with respect to time. Compared to the traditional Fourier analysis decomposition of a given signal, the wavelet transform can address nonstationary signals, or signals whose characteristics change as a function of time. A wavelet bump extraction methodology was demonstrated on an automotive fatigue spectrum with promising results with respect to fatigue life comparisons between the original and compressed signals. However, the automation heuristics were based solely on global signal statistics and were not optimized to deliver the best performance in terms of maximum signal compression. Wavelet transforms, specifically the Morlet wavelet, have also been used to perform quicker and more accurate accelerated testing of wind turbine blades.

Wavelet approaches to spectrum editing provide a more rigorous and systematic way of determining the most appropriate fatigue spectrum for a given test article. A wavelet approach for spectrum editing is desired for broadband aerospace fatigue spectra [Refs 7, 8, 9] including buffet and sonic fatigue that include both high and low frequency and amplitude content to deliver reductions in spectrum length while maintaining fatigue damage, load interaction effects, and producing fleet-representative failure modes. This wavelet approach for spectrum editing should perform well in the high cycle range, achieving better damage replacement and reducing clipping. The solution should also be optimized to deliver the best performance in terms of maximum signal compression. The developed spectrum editing tool should enable the user to edit spectra based on user selected parameters (e.g., length/cycle count, damage equivalency, and upper/lower amplitude bounds) and provide a way to graphically inspect the data.

PHASE I: Design and demonstrate the feasibility of a spectrum editing concept on a publicly available aerospace fatigue spectrum including both high and low frequency and amplitude content. Buffet and sonic fatigue are specific areas of interest. Demonstrate that the edited spectrum maintains equivalent damage and representative failure modes. The Phase I effort will include prototype plans to be developed under Phase II.

PHASE II: Develop and demonstrate a prototype spectrum editing tool that can deliver reduced length spectra for a variety of publicly available aerospace fatigue profiles. Conduct critical tests on a contractor-fabricated or procured sample set of aerospace aluminum alloys to validate the tool. Extend the wavelet editing to multiaxial situations and demonstrate the approach for simple non-proportional load histories.

PHASE III DUAL USE APPLICATIONS: Develop and deliver a fully integrated spectrum editing tool that can deliver a variety of edited spectra based on user-selected parameters including length/cycle count, damage equivalency, and upper/lower amplitude bounds. Accelerated testing is required in many industry sectors to assess component life performance in order to determine required design enhancements and identify any service actions to meet service life goals. These industries include aerospace, automotive, ship building, oil and gas, heavy machinery, and electronic equipment manufacturing. The proposed spectrum editing/compression techniques can be used directly and thus save testing times in these industries.

REFERENCES:

1. Frost, N., Marsh, K., and Pook, L. “Metal Fatigue”. Oxford: Clarendon Press, 1974. http://www.worldcat.org/title/metal-fatigue/oclc/1258430

2. Nopiah, Z. and Osman, M. “Statistical Optimisation Techniques in Fatigue Signal Editing Problem”. AIP Conference Proceedings, 2015, 1643(776). http://aip.scitation.org/doi/pdf/10.1063/1.4907527

3. Torrence, C. and Compo, G. “A Practical Guide to Wavelet Analysis”. Bulletin of the American Meteorological Society, 1998, 79 (1). http://journals.ametsoc.org/doi/pdf/10.1175/1520-0477%281998%29079%3C0061%3AAPGTWA%3E2.0.CO%3B2

4. Newland, D.E. “An Introduction to Random Vibrations Spectral and Wavelet Analysis”. 3rd edition. Mineola, New York: Dover Publications, 2005. http://www.worldcat.org/title/introduction-to-random-vibrations-spectral-wavelet-analysis/oclc/828932080&referer=brief_results

5. Abdullah, S. (2007). “The Wavelet Transform for Fatigue History Editing: Is it Applicable for Automotive Applications?” Journal of Engineering and Applied Sciences, 2007, 2(2), pp. 342-349. http://docsdrive.com/pdfs/medwelljournals/jeasci/2007/342-349.pdf

6. Pratumnopharat, P., Leung, P., and Court, R. “Application of Morlet Wavelet in the Stress-Time History Editing of Horizontal Axis Wind Turbine Blades”. 2nd International Symposium on Environment-Friendly Energies and Applications, 2012. http://ieeexplore.ieee.org/document/6294048/?part=1

7. Edwards, P. R., and J. Darts. Standardised Fatigue Loading Sequences for Helicopter Rotors (Helix and Felix) Part 1. Background and Fatigue Evaluation. No. RAE-TR-84084. ROYAL AIRCRAFT ESTABLISHMENT FARNBOROUGH (UNITED KINGDOM), 1984. http://www.dtic.mil/dtic/tr/fulltext/u2/a156621.pdf

8. Edwards, P. R., and J. Darts. Standardised Fatigue Loading Sequences for Helicopter Rotors (Helix and Felix) Part 2. Final Definition of Helix and Felix. No. RAE-TR-84085. ROYAL AIRCRAFT ESTABLISHMENT FARNBOROUGH (UNITED KINGDOM), 1984. http://www.dtic.mil/dtic/tr/fulltext/u2/a156622.pdf

9. Heuler, P., and H. Klätschke. "Generation and use of standardised load spectra and load–time histories." International Journal of Fatigue 27, no. 8 (2005): 974-990. https://www.infona.pl/resource/bwmeta1.element.elsevier-12225d1a-70d6-35d1-b770-d3b42049a4f8

KEYWORDS: Accelerated Testing; Spectrum Editing/Compression; Clipping and Truncation; Damage Equivalence; Failure Mode Preservation; Wavelet Transforms