PSI - Issue 47

Bartłomiej Walczak et al. / Procedia Structural Integrity 47 (2023) 723 – 731 Bart ł omiej Walczak et al./ Structural Integrity Procedia 00 (2019) 000 – 000

729

7

3. Experimental results Among the tests of the first phase, the most demanding for the superstructure were emergency braking, slaloms, and driving over a speed bump. The first two tests caused the maximum values of forces in the suspension elements, which are visible in Fig. 9, and thus caused the highest values of stresses in the chassis. The bump test was causing the highest response of roof elements and side walls.

Fig. 9. Example of the registered signals; (a) for emergency breaking; (b) for slalom.

The data from an accelerometer sensor placed on the floor near the center of gravity will serve as input for numerical calculations. Based on this data, loading scenarios simulating braking, turning, and vertical forces will be created, and the recorded occurrences will be used for fatigue assessment. Since the original signal was sampled at a frequency of 200Hz, the power spectral density of the signal was determined and used to select filtering bands. A Butterworth filter was used for processing, and the frequency levels for the longitudinal (X) and lateral (Y) axes were reduced to 10 Hz, while for the vertical axis (Z) was reduced to 20 Hz. Similar frequency levels for this type of application can be found in literature, for instance Pérez et al. (2022). In the next step, the widely accepted rainflow counting technique, which is included in standards such as ASTM E 1049-85 (2017), was used to count occurrences, Fig. 10. For the purpose of fatigue simulations, this data are extrapolated for the equivalent of 1 million kilometers. The analysis of urban driving data involved processing the signals from the strain gauges by counting the fatigue cycles. The previously mentioned rainflow algorithm was also used for these sensors. While the majority of measuring points were located near welded connections, an appropriate S-N curve was assigned to each of them based on the type of joint. This process is described, for example, by Hobbacher (2009) or EN 1993-1-9 (2005). Based on this, the accumulated damage was calculated using the Palmgren-Miner linear damage summation law. Fig. 11 shows the total damage accumulated by several representative strain gauges in different parts of the structure during a drive through specific route sections. Taking into account this data, the surfaces where bus drives caused the greatest stresses of superstructure were scanned using a high-precision kinematic LIDAR. The examples of received 3D surfaces are visible in Fig.12.