Dynamic Gates
Dynamic gates are relatively new in fault tree analysis. A dynamic gate considers the temporal order of the occurrence of input events. This means that the order in which input events occur is important to determining the output.
You can insert dynamic gates only in time dependent trees. A time dependent tree becomes a dynamic tree when a dynamic gate is present. Time dependent trees support four dynamic gates, which are described in the following topics:
When you run FTA calculations, the FTA module determines if the tree is static or dynamic. For solving dynamic fault trees, either the Exact method or Monte Carlo simulation method is recommended.
 |
You typically do not want to use any of the three cut set-based calculation methods for solving dynamic trees. They ignore the sequence-dependent analysis and produce approximate results.
|
The FTA module must then break the tree into corresponding modules, and analyze each module individually as either a static or dynamic entity.
• When you use the Exact method, the tree is broken into modules. Each module is then analyze as either a static or dynamic entity.
|
|
During the Exact calculations, only AND, OR, and dynamic gates can be part of a dynamically calculated module. When Exact calculations are performed, if an unsupported gate is found, a window opens, indicating this. Usually the “problem” gate is not directly connected to the top gate of this module, but rather it is somewhere under it. This problem gate might also be part of a module created using repeated events. For more information, see Dynamic Gate Calculations.
|
◦ For static entities, standard combinatorial techniques are used to compute exact calculated results. These techniques consider conditional event and gate probabilities in the same manner as binary decisions diagrams.
◦ For dynamic entities, equivalent Markov models are generated. These Markov models are then computed using the internal Markov calculation engine for the Markov module. For more information, see Windchill Markov.
Techniques employed for analyzing generalized static trees are then used to bring the results for the various entities back together to perform exact calculations.
• When the Monte Carlo simulation method is selected, varying sets of inputs are used to observe the dynamic behavior of the tree. Modularization techniques are not used.
For descriptions of all available calculation methods, see Advanced Page.