Reliability analysis goes a step further than robustness assessment as it provides methods to compute the probability of rare events, such as the probability of failure.

Each engineering system is subject to random influences, either in its own properties (for example, production tolerances) or external influences (loads, operation conditions). For a meaningful analysis these are modelled as random variables, which are defined by distribution type and parameters, and correlations.

The reliability task is formulated by the limit state function g(X), which separates the safe (or admissible) from the unsafe (inadmissible) domain. Therein, X is the vector of all random variables, input parameters as well as responses. g(X) is scaled such that a negative value denotes failure or any defined inadmissible state, positive g the safe state, while g=0 is the transition state. The probabilistic reliability measure is typically expressed as the probability of failure. This is defined as the probability of the event that a random parameter set falls into the unsafe domain.

In optiSLang, only parameters with stochastic properties (stochastic or mixed parameter) are considered by the reliability modules. The limit state is expressed in the Criteria tab by an inequality, which defines the safe (admissible) state. Multiple limit states are interpreted as a series system.

optiSLang offers the following solution methods for probabilistic reliability analysis.

Further information about methods of reliability analysis used in optiSLang can be found here.