Deriving fatigue equivalent power spectral density spectra for the vibration testing of engine components: Ableitung schädigungsäquivalenter Leistungsdichtespektren für die Vibrationsprüfung von Motoranbauteilen. Issue 3 (27th March 2018)

• Record Type: Journal Article
• Authors: Decker, M.; Kinscherf, S.; Bauer, N.; David, P.; Serifsoy, M.
• Is Part Of: Materialwissenschaft und Werkstofftechnik. Volume 49:Issue 3(2018:Mar.)
• Page Start: 392
• Page End: 405
• Publication Date: 2018-03-27
• Subjects: Broadband vibration testing -- shock response -- fatigue response -- power spectral density -- structural durabilty; Breitbandrauschen -- Vibration -- Prüfung -- Schock-Antwort -- Ermüdungs-Antwort -- Leistungsdichtespektrum -- Betriebsfestigkeit; Materials -- Periodicals; Materials -- Testing -- Periodicals; 620.1
• DOI: 10.1002/mawe.201700193
• Languages: English
• ISSNs: 0933-5137

Abstract

Abstract: Vibration testing for the proof of structural durabilty often is performed on the basis of power spectral density spectra. At present, there are large uncertainties concerning the derivation of appropriate spectra for fatigue testing. Especially for non‐periodic, non‐stationary excitations, the commonly used procedures based on the fast Fourier transformation usually yield inaccurate results. Additional problems arise if the analyzed signals do not follow a Gaussian amplitude distribution, which is a prerequisite for describing a time history as a random process with a power spectral density. To overcome these problems, the FatiResponse method for deriving fatigue equivalent power spectral density spectra was developed and is presented here. It takes into account both the possible dynamic and fatigue responses of virtual structures. Shock response and fatigue response spectra are used to describe the impact of an arbitrary signal on virtual structures having virtual critically loaded locations. From these, test spectra are derived that yield the same shock and fatigue responses as the measured vibrations. The method is explained and demonstrated using vibration data of a passenger car engine consisting of harmonic and stochastic excitations. The procedure is validated using a dummy structure and compared to commonly used procedures.