PSI - Issue 58

Radovan Minich et al. / Procedia Structural Integrity 58 (2024) 80–86 R. Minich et al./ Structural Integrity Procedia 00 (2019) 000–000

83

4

walled bodywork profiles and the results of measuring their operational stresses. The knowledge base will contain not only summary evaluations of the performed experiments, i.e., primarily S-N curves and stress spectra but all source data, including the results of detailed experimental stress analyses and supporting calculations using the finite element method. Previous measurements made it possible to investigate whether any of the usual procedures can approximate the measured stress spectra. The load amplitude spectra according to Heuler and Klätschke (2005), Fig. 3, proved to be very practical. These spectra can be described by the equation: log = � 1 − � , ⁄ � � log 0 (1) As ν values decrease, the shapes of the amplitude spectra become increasingly hollow. In the table in Fig. 3 there are the details of some ν numbers and the associated "spectral shape factor", SSF. The SSF represents the dis tance (factor, life ratio) between the (arbitrary) Wöhler S–N curve and the associated spectral fatigue life curve (sometimes referred to as the Gassner curve in Germany in honor of Ernst Gassner).

Fig. 3. Opportunity to describe / to compare aggressiveness of different stress spectrums.

Over the years, in cooperation with bus manufacturers, the basis for the creation of a database of laboratory fatigue test results for various configurations and material solutions of structural nodes of welded bodyworks has been created. To describe the S-N curves (S-N lines), a relationship using a linear mathematical model, which includes the standard deviation, proved to be suitable: log � � = 0 − ∙ − ∙ log( ) (2) The standard deviation of log can be calculated on the basis of experimental data; the value of the standard deviation can also be found in the literature, e.g., in BS 7608 (1993). It depends on the type of welded joint. The value is determined based on a large number of tests and varies around 0.2. Fatigue curves for various certainty of survival can be described by selecting standard deviation. For instance, at =0 , equation (2) describes a mean fatigue curve (failure probability of 50 %). Fatigue curves shifted by two