PSI - Issue 77

Miloslav Kepka et al. / Procedia Structural Integrity 77 (2026) 272–278

277

In total, 14 representative spectra were evaluated, corresponding to the empty and the loaded vehicles. The parameter ν , which is critical for characterising the shape of the spectrum according to Equation (1), was evaluated by regression analysis. All calculated values of this parameter are presented in Table 1.

Table 1. Evaluation of the shape exponent ν of all measured loading spectra. Vehicle City Loading state

Strain gauge

Shape exponent ν

Empty

T14 T14 T35 T35 T14 T14 T35 T35 T15 T15 T45 T45 T10 T10

0.99 1.08 0.99 0.92 1.00 0.95 0.97 0.89 0.91 0.92 0.91 0.81 0.95 0.92

SKODA 27Tr

Bratislava

Full

Empty

Full

Empty

SKODA 27Tr

Pilsen

Full

Empty

Full

Empty

SKODA 36Tr

Pilsen

Full

Empty

Full

Empty

SKODA SOLARIS 24M Pilsen

Full

Based on the results of this measurement campaign, it may be concluded that, in service, the welded joints of city bus bodies in the regions of doors and windows are subjected to load spectra whose shape exponent ν ranges between 0.81 and 1.08 when described by Equation (1). These values correspond to an almost linear form of the cumulative load spectrum. With this statistically verified knowledge, the design of city bus bodywork structures can be optimised already at the design stage to achieve the required fatigue life. 5. Follow-up research The derived stress spectra were incorporated into the systematic database, and the finding concerning their characteristic shape (predominantly linear) was taken into account in subsequent research activities. Follow-up research included the following activities: - Model test specimens were manufactured to represent a welded assembly of thin-walled bus body profiles. - The S–N curve of this structural joint was determined based on harmonic loading tests performed on 14 specimens. - For variable loading, a load block with a shape parameter of ν = 1 was designed, and 4 specimens were subjected to this loading regime. - The experimentally obtained fatigue lives under variable loading were compared with the lives predicted using fatigue damage accumulation hypotheses (Palmgren–Miner original, Palmgren–Miner elementary, Haibach). - Refined limit values of fatigue damage were established for the applied hypotheses. The analysis was conducted with consideration of the limited number of specimens available for variable loading tests, which are time-consuming. The approach proposed by Sonsino (2007) was used. A detailed presentation of the subsequent research lies beyond the scope of this paper. For this reason, only a summary presentation is provided here in Fig. 7, with the understanding that readers will soon be presented with specific findings and results in a more comprehensive scientific publication.