PSI - Issue 17

C P Okeke et al. / Procedia Structural Integrity 17 (2019) 596–601

599

C P Okeke et al / Structural Integrity Procedia 00 (2019) 000 – 000

4

frequency. These response characteristics are in line with a study on dynamic behaviour of the rubber isolator, Sadhana et al (2017). The level of variation in transmissibility relates to the level of non-proportionality between the drive and the response accelerations. The drive response acceleration curve in fig 3 (b) clearly shows non-linear behavior (hyperelastic). This is a known phenomenon characterising the behavior of polymers. Okeke et al (2017), stated that care must be taken when modelling the response of polymers due to their hyperelastic behaviour. Accordingly, this level of nonlinearity in the dynamic behaviour of vacuum cast polymers, must be taken into account when designing with this particular material. In resonance based fatigue life measurement, the response acceleration of the material is theoretically calculated at every load level. This theoretically calculated acceleration must match or be very close to the experimentally measured resonance response acceleration at every fatigue load level. This is where the drive-response acceleration curve is very important as it will give understanding of the drive acceleration required to produce the desired response acceleration at a given load level.

80.0 77.5 75.0 72.5 70.0 67.5 65.0

(a)

Peak Frequency (Hz)

0 10

20

30

40

50

60

Drive acc (m/s²)

Figure 3: Dynamic response - (a): Peak frequency vs input acceleration, (b): Transmissibility / response acceleration vs drive acceleration. Table 1 shows the theoretical and experimental response accelerations at three fatigue load levels. The experimentally measured three fatigue load level response accelerations were based on the first resonance frequency. The values are averaged over five specimens tested per load level. It can be seen that the values obtained for both theoretical and experimental are closely matched. At 80% fatigue load level, the difference in both response values is 3.4% while for 60% and 40% load levels, the difference in response are 0.6% and 3.2% respectively.

Table 1: Average calculated / measured response acceleration

Average response acceleration (m/s²)

Fatigue load level

80%

60%

40%

Theoretical

440

330

220

Experimental

425

328

213

4.2. Fatigue properties

Fig 4 shows the fatigue life curves of vacuum cast polyurethane polymer obtained using resonance bending fatigue test system. Fig 4(a) is acceleration – fatigue life curve (a-N curve) which was converted to stress-fatigue life (S-N curve) shown in Fig 4(b). The curve shows scatter in the fatigue life which increases as the stress amplitude decreased. This phenomenon is known to be an intrinsic characteristic of material fatigue properties, however, the level of scatter seen at 40% load level is significantly large. The crack initiation and propagation are the two main stages of fatigue life. The fatigue mechanisms in both stages are said to be different, this is explained by Schijve (2001). The scatter arising from the first stage (crack initiation) can be substantial as the failure in this stage can be influenced by different factors. Generally, the crack initiation is attributed to specimen surface imperfection while