PSI - Issue 19

Thorsten Voigt et al. / Procedia Structural Integrity 19 (2019) 4–11 Dr.-Ing. Thorsten Voigt / Structural Integrity Procedia 00 (2019) 000 – 000

6

3

Figure 5: control arm, test setup

Figure 6: control arm, failure

Figure 4: control arm

Table 1: impact parameters for customer usage profile.

Table 2: maximum forces for customer usage profile

force

straight driving

cornering

braking

load direction

straight driving

cornering

braking

max.

min.

max.

min.

max.

min.

max.

min.

max.

min.

max.

min.

z y x

8.6 kN

0 4.4. kN

0.2 kN

2.1 kN

2.6 kN

vertical lateral brake

3.70 0.75

0

1.90 2.00

0.10

0.90

1.10

1.7 kN -1.7 kN

4.7 kN

0

-0.75

0

brake

1.9 kN -2.6 kN

0.80 -1.10

3.4 kN -2,7 kN

longitudinal

1.45

-1.16

3. Endurance tests on the control arm of the hinged rear axle

The considered suspension component combines the functions of a longitudinal and transverse control arm, and is therefore responsible for the transmission of longitudinal and lateral forces. Due to the lever arm between the body connection of the control arm and the force introduction at the wheel hub, the lateral load represents the more critical stress. It was therefore determined to test the AlSi7Mg cast control arms (Figure 4) in a uniaxial test by applying lateral loads. In order to determine the influence of corrosion on the fatigue strength of the aluminium component, tests using a salt spray also had to be carried out. The test setup is shown in Figure 5. As in the vehicle, the control arm is attached to two bearing points. The originally provided elastomeric bearings were replaced by plastic sleeves to avoid excessive wear. The load was applied by a hydraulic cylinder. The test setup is enclosed by a salt spray chamber. Seven variable amplitude tests were performed, three dry and four with salt spray. According to the standard DIN EN ISO 9227, a 5% NaCl solution was sprayed for five minutes during the tests followed by a 20-minute drying time. Three tests were previously run without salt spray under constant amplitudes to determine a component Woehler curve. The tests under variable amplitudes were based on the load channel of the lateral forces y of the synthetic STAMAS load program. Since the lateral forces are transmitted directly from the wheel to the control arm, the signal designed for the simulation of wheel loads can be used directly, i.e. without additional scaling. 3.1. Constant amplitude tests To determine the first load level, the maximum value 4.66 kN and the minimum value −1.72 kN were taken from the variable load signal and both scaled by a factor of 0.75. The oscillation of the constant amplitude experiment is characterised by the reversal points (3.5 kN, −1.27 kN) defined in this way. The load ratio = min / max is = −0.36 . At this load level, 2 ∙ 10 5 load cycles were completed without failure. Next, the tests were continued with the same specimen and the load level was proportionally increased by 20% after every 2 ∙ 10 5 load cycles without failure. The failure of the first control arm occurred at a peak load of about 15 kN (≙ 10.3 kN half cycle amplitude). In the same way (increase by 40% and 20%, respectively) the load levels of the other two Woehler tests were determined. The results of the constant amplitude experiments are shown in Figure 7 (solid line). From the illustration, it is clear that the Woehler curve with = 19 has a flat slope. First cracks appeared on the outside of the vehicle outer bearing eye in all tests (Figure 6). The crack of about 3 mm in length was detected with the help of zinc oxide paste and a camera. In all tests, crack growth from 3 mm to breaking point was less than 1% of the total number of cycles to breaking point. 3.2. Variable amplitude experiments The variable amplitude tests on the control arm were carried out to demonstrate that the component can safely withstand lateral loads that occur during hard operational use over 10 5 km. In addition, it was to be examined whether linear Palmgren-Miner damage accumulation on the control arm allows reliable lifespan estimates under variable amplitudes.