PSI - Issue 75

Thomas Constant et al. / Procedia Structural Integrity 75 (2025) 660–676 Author name / Structural Integrity Procedia 00 (2025) 000–000

669

10

Selection of the

cut points

Generation of the initial DOE from by LHS:

Generation of a Monte-Carlo population

Start

Computation and union of projected points, yielding and (see figure 1 for illustration)

Calibration of

from

respectively

Selection of the best candidate points of in the sense of the learning function ( see Eq.16 )

True

?

False

ESC stopping criterion met ?

False

True

Evaluation of

with crude MCS using

with ( see Eq.15 )

Increase the size of

False

True

End

Fig. 2. Flowchart of the proposed extension of AK-SM.

Fig. 3. Numerical modal bracket durability chain.

variable X A c , which is the random multiplicative coe ffi cient C R z that amplifies the excitation signal. An illustration of the excitation profile is given figure 4. All random variables are assumed to be normally distributed with a coe ffi cient of variation of 5%. The mean values of C R z range from 0.5 to 2.0, with configuration c 1 corresponding to the most severe load case, in average.

Table 1. Properties of load configurations. LoadCase p c

Random Variable

Distribution

Mean

Coe ffi cient of Variation

Configuration c 1 Configuration c 2 Configuration c 3 Configuration c 4

15% 35% 35% 15%

X A c X A c X A c X A c

Normal Normal Normal Normal

2 . 0 1 . 5 1 . 0 0 . 5

5% 5% 5% 5%

1

2

3

4