PSI - Issue 57

Sven Maier et al. / Procedia Structural Integrity 57 (2024) 731–742

735

5

S. Maier et al. / Structural Integrity Procedia 00 (2023) 000±000

is referred to Benasciutti and Tovo (2005); Bishop (1999). For the evaluation of the spots a structural stress concept is applied. To automatise the procedure and perform a Monte-Carlo-Simulation to scatter the parameters defined in Table 1 the software Optimus is utilised.

2.4. Reference simulation

The described workflow above is used to perform a reference analysis for each of the three variants. For the nominal condition, the result of the vibration fatigue analysis is shown as a representative in Figure 2.

Hotspot 2

Hotspot 3

Hotspot 1

Hotspot 1

Fig. 2. Nominal reference simulation for variant 1. The three identified hotspots are marked (left side). Detailed view of hotspot 1 with lowest durability value (right side).

Firstly, it can be seen on the left graph that there are three hot spots allocated to the FemFat spot welds. Secondly, it can be noticed in the reference simulation that the hotspot on the lower left side has by far the lowest computed lifetime, which is referred to as hotspot 1 and focused. For the other two variants, a very similar durability distribution with the same ranking of the hotspots can be observed. Table 2 shows the durability results with regard to the target load duration of 720 hours for all three hotspots of all variants. For the material strength of the plates a 90% SN-curve with k = 11 is used. The modified miner rule with an underlying damage sum of D = 1 is applied for the damage accumulation. A calculated durability of 100% implies a failure of the component at exactly the target load time of 720 hours. Equivalently, a durability of 50% represents a failure after the half of the target load duration and 200% in turn the expected double runtime. The calculated durability can be regarded as the reciprocal value of the computed damage. Furthermore, the first three eigenfrequencies f 1 − f 3 obtained from the Nastran simulation are also illustrated for each of the variants.

Table 2. Results of a deterministic reference simulation for all three variants.

Durability [%]

Eigenfrequency f n [Hz]

Variants

Hotspot 1

Hotspot 2

Hotspot 3

Mode 1

Mode 2

Mode 3

Variant 1: nominal

410.98 197.71 117.48

2416.9 973.36 325.57

16518.7 7410.36 5452.88

40.97 38.43

76.12 70.85 62.87

81.42 77.56 76.01

Variant 2: additional mass Variant 3: thinner plates

36.1