PSI - Issue 38

Denis Chojnacki et al. / Procedia Structural Integrity 38 (2022) 362–371 Author name / Structural Integrity Procedia 00 (2021) 000 – 000

365

4

Reliability approach: This approach called Stress-Strength Interference (or SSI) (Lipson et al. (1967)) aims to limit the probability (risk) of failure in service under a predefined level. The same scalar variable called severity defines both the loading imposed by the customers (Stress C) and the maximum damage that a given structure can sustain (Strength R). The general idea is that it is unlikely that the most severe customer will encounter the weakest structure even if the zero risk does not exist. We then impose an objective value P f at the risk, generally lower than one over the number of designed structures for safety parts: ( < ) = ∬ , ( , ) < < (1) Where f R,C is the joint probability of the Stress C and the Strength R as illustrated in Fig. 3.

a

b

Fig. 3. (a) The joint distribution of Stress C and Strength R; (b) their bisector projections and the definition of a Target Level Fn

The definition of the so-called severity variable differentiate the applications of this approach. In mechanical engineering, the severity derives from RLDA signals using a general approach (Johannesson&Speckert 2014) that can be summarized as indicated Fig. 4.

Fig. 4. From RLDA signals to scalar severity, a method.

Under the assumptions of a linear relation between the structural loading and a local stress variable, a Rainflow counting algorithm, a linear Miner’s cumulative law for damage and a Basquin’s model for aWöhler curve, the severity is monotonous of the cumulated damage induced by the RLDA loading signal. It is also monotonous of an equivalent loading variable (Raoult & Delattre (2018)). Hence, the relative intuitive interpretation that the severity measures both the level of loading and the sustained damage (or its inverse the lifetime). From the RLDA, we assess the Stress C distribution. Since we impose P f , the