PSI - Issue 75

Marco Bonato et al. / Procedia Structural Integrity 75 (2025) 719–729 Author name / Structural Integrity Procedia (2025)

725

7

nature of the signal that best represents the vibration stress. For example, if a component is mounted on the body of the vehicle, the vibration type would be mostly random transient, therefore an ALT provided as a power spectrum signal (PSD) would be representative of the real environment more that a sing-off signal provided as pure sinusoidal (either broad band sine sweep or fixed sine). Similarly, a component mounted on the cylinder head of a combustion engine power vehicle will suffer vibration loads that are mainly sinusoidal in nature. In this case, a design validation test would be more appropriate if the signal is a sine sweep rather than a random PSD. Comparison of Vibration Simulations As a worldwide tier 1 supplier, the authors company has had the opportunity to analyze and compare many different carmakers specifications. Moreover, thanks to collaborative works with our customers, on several occasions the authors have developed tailored vibration tests by exploiting the data provided by the carmaker (vibration acceleration data at the component mounting points, plus its user profile). A tailored accelerated vibration durability test[3] was then generated by adopting the methodology explained in the previous section. This approach has resulted in a collection of numerous vibration tests, generated on different components and for various carmakers. A comparative analysis based on such diverse vibration tests allows to capitalize on the definition of an internal standard test that should be used at the beginning of each project (and already at RFQ phase) to assess the representativeness of the vibration signal requested by the customer. In order to solve the issues encountered during the final vehicle testing, the first step is to compare the vibration signal used for the validation of the eDC system, and to compare it to the internal standard for components mounted on the body of the car. The first significant difference between the two is the nature of the signal: broadband random PSD for the Valeo standard versus a fixed sine sinusoidal signal for the customer. Being of different nature, these signals cannot be directly compared. The FDS [5] of each signal is calculated by considering that the failure affects the electronic sub components of the eDC (Basquin’s fatigue exponent b = 4, Quality factor, Q = 25), as shown in Figure 4. Alternatively, the fixed sine can be converted to a FDS equivalent PSD, and plotted versus the Valeo reference PSD only for comparative purposes (Figure 5). The conclusion of the comparison is that: ● The fixed sine fails to reproduce the broadband vibration load to which the eDC undergoes during the vehicle driving conditions. ● The severity of the customer spec is, at the resonance frequency of the component, much lower than the standard signal used within the company. In particular, the importance gap in severity of the customer specification is noticeable in the frequency range of the resonance mode of the eDC, between 350 Hz and 550 Hz. The internal reference PSD shows a relative fatigue damage which is orders of magnitude higher than the vibration damage provoked by the fixed sine. It comes as no surprise that the design flaw affecting the flyback has been revealed only during the proving ground test. Had the eDC been validated according to the internal standard, the design issues would have appeared already during the post-treatment of the FEA simulations. 3.2.