Static Reliability Modeling

In some cases, you may want to run GoldSim statically (i.e., without timestepping). A static simulation consists of a single calculation of the system in its steady-state condition.  As such, it is equivalent to a conventional fault tree analysis (using Monte Carlo simulation instead of cut sets).

Static simulation cannot represent any true system dynamics, but is significantly faster than dynamic simulation (as GoldSim only needs to calculate element values once as opposed to the hundreds or thousands of updates in a dynamic simulation).  Hence, a properly constructed static model can provide useful information on the long-term availability of systems that do not have complex dynamic behavior.

When running a static reliability simulation, the following logic is used by GoldSim:

1.  The system is assumed to have operated for an indefinite period of time and reached a steady-state with regard to failures and repairs.

2.  To determine each reliability element's availability, GoldSim computes the steady state probability of being operable for each failure mode as follows:

MTTF / (MTTF + MTTR)

where MTTF is the Mean Time to Failure, and MTTR is the Mean Time to Repair.

3.  For each realization, each failure mode in each reliability element randomly realizes its state (operating or failed). If a reliability element has a failure mode that has failed, that reliability element is considered failed for that static reliability simulation. 

4.  Based on the status of each reliability element, the system is considered to either be operating or failed for a particular realization.  By specifying multiple Monte Carlo realizations, the mean availability can be computed.

   Note: After calculating whether it has failed due to a failure mode, the reliability element then checks that its Operating Requirements are met.  Within the requirements-tree, any nodes that reference other GoldSim elements use their time = 0 values.

Static reliability simulations have several important limitations:

   The automatic repair feature for all failure modes must be turned on in order to run a static reliability simulation (otherwise a fatal error will be displayed when you try to run the model).

   Only failure modes that are FMCV-based (and do not have a user-defined base variable) are allowed (otherwise a fatal error will be displayed when you try to run the model).

   Static simulation only provides mean availability metrics. (It also provides reliability, but for a steady-state system, this is by definition always zero, unless there are no failure modes at all).  No further results are available from a static simulation.

   Static models obviously do not allow the system to evolve in a natural way, and as a result, do not model system variability as accurately as a dynamic model.

Related Topics…

Learn more about: