PSI - Issue 57

J. Torggler et al. / Procedia Structural Integrity 57 (2024) 152–160 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

159

8

The so calculated strain amplitude is plotted over the number of load-cycles in Figure 7. In the left diagram it is shown that a clean separation of the fibre angles is no longer possible as before (see Figure 5). This means, the strain amplitude ε log can be considered as a comparative parameter for different layups, as the results are generally unified leading to a sound design approach.

0.04 0.0 0.0 0.09 0.1 tra n am l t de log 0.05 0.0

0.04 0.0 0.0 0.09 0.1 tra n am l t de log 0.05 0.0

tests 15 deg P s 50 tests 25 deg P s 50 tests 35 deg P s 50

all tests P s 50 P s 9 .5

0.039

0.03

0.03

10 2

10 3

10 4

10 5

10 2

10 3

10 4

10 5

10

10

N m ero load y lesN

N m ero load y lesN

Figure 7: S/N-curve for local approach for different fibre angles (left) and master S/N-curve covering all testes in one evaluation (right)

In Figure 7 (right) the statistically evaluated master S/N-curve covering all test results utilizing the strain amplitude ε log is shown. The design curve is defined at N d = 50 000 load-cycles and a probability of survival of P s = 97,5 %. The S/N-curve parameters result to a slope of k = 7.35 and an endurable fibre strain amplitude of ε log = 3.9 % at N d = 50 000 load-cycles. Further, the statistically evaluated scatter band is adequate indicating a value of T s = 1.26. This design curve can be further utilized in the pre-design-state and is applicable for a fibre strain amplitude ε log from about 4 to 8.5 percent (caused by lateral displacement U ) and longitudinal preload of about 3.5 to 6 percent strain (caused by longitudinal preload F y ). 5. Summary Based on existing component tests, a test methodology was presented on a specially developed representative small-scale flat specimen. Different confections of this sample were tested with fibre angles φ f of ±15, ±25 and ±35 degrees. Various longitudinal preloads F y and cyclically lateral displacement amplitudes U were applied to the specimens. As failure criterion, the increase of the mean longitudinal displacement V was consulted. With the help of the µCT analysis, it can be determined whether and which damage is present in the tested sample and how pronounced it is. A global and a local approach to evaluate and interpret the measurement results obtained was presented. The global evaluation shows a distinctive influence of the fibre angle on the resulting service life. At a N d = 50 000 cycles acting as common design lifetime, a decrease in the tolerable lateral displacement amplitude of 2 mm with an increase in the fibre angle φ f of 10 degree was observed. The local evaluation leads to a master S/N curve which can be used as elaborated design method of future components with the investigated composite material. The S/N-curve parameters are statistically evaluated with a slope of k = 7.35 and an endurable fibre strain amplitude of ε log = 3.9 % at N d = 50 000 cycles for a survival probability of P s = 97.5 %. The presented evaluation is analytically applicable on the flat sample, but not on the three-dimensional component air spring bellow. Therefore, a suitable finite element modelling technique will be further on developed, which is applicable to both the specimen and the component. Furthermore, additional fatigue tests are planned focusing on lower strain amplitudes up to higher load-cycles to ensure a holistic fatigue assessment.