In common structural design, random variables, such as material strength or loads, are represented by fixed
numbers defined in design codes. This is also referred to as deterministic design. Addressing the random
character of these variables directly, the probabilistic design procedure allows the determination of the
probability of exceeding a defined limit state. This probability is referred to as failure probability. From there,
the structural reliability, representing the survival probability, can be determined. Structural reliability thus
is a property of a structure or structural member, depending on the relevant limit states, failure modes and
basic variables. This is the basis for the determination of partial safety factors which are, for sake of a simpler
design, applied within deterministic design procedures. In addition to the basic variables in terms of material
and loads, further basic variables representing the structural model have to be considered. These depend
strongly on the experience of the design engineer and the level of detailing of the model.
However, in the clear majority of cases [1] failure does not occur due to unexpectedly high or low values of
loads or material strength. The most common reasons for failure are human errors in design and execution.
This paper will provide practical examples of original designs affected by human error and will assess the
impact on structural reliability.
A Bayesian method for using simulator data to enhance human error probabilities assigned by existing HRA methods
