Interpretation of BlockSim Maintenance Simulation Results

[Frank]

Can somebody help me out with the interpretation of the maintenance simulation results. First, I ran a simulation for the Complex System demo project that comes along with the evaluation kit, and got the following results (irrelevant parts deleted)

Simulations: 10000.0000
Time: 1000.0000
DownTime: 881.2138
UpTime: 118.7862
Mean Availability:0.1188
PM DownTime: 0.0000
Pt. Availability: 0.0000
Reliability: 0.0000
Expected Number of Failures: 0.7652
MTTFF: 118.7862

Because of settings for component L in the model, it is expected that the (expected) reliability is 0. However, I am not sure how to reconcile that with an expected number of failures of 0.7652. I would expect this number to be at least 1. What does the number mean?

Also, what is the meaning of 'PM downtime', 'Exp. NOF', 'Exp. NOPM'?

Thanks,

Frank

[RS Support]

Simulations: Number of simulation runs
Time: End time of simulation
DownTime: Total time the system was down
UpTime: Total time the system was up
Mean Availability: Average availability (ratio of uptime/total time)
PM DownTime: Time the system was down for preventive maint (if that option was set)
Pt. Availability: Point Availability at the end time
Reliability: Point Reliability at the end time
Expected Number of Failures: see below
MTTFF: Mean time to FIRST failure (system)

Expected number of failures is the expected number of downing events a single system will experience. It can be less than one as shown.

Now your question is to how to reconcile that with reliability. If you have redundancy, keep in mind that system reliability is without maintenance actions, i.e. as components fail they are not repaired. In using maintenance actions components that fail are restored. It is possible then for a system of very low reliability to experience no system failures if redundancy is present, and if a redundant component that failed is restored very quickly.

For example look at a system of two components in parallel, both components having a low reliability i.e. R(100)=.2. Then the system reliability for 1000 hrs is very low. Now if a failed component is restored in 1 second, then the probability of a system-downing event becomes low (e.g. if one of the components failed then the system would fail if the other component failed during the second that it took to restore the first one.)

In other words the system failures are a function of your system design (diagram), reliability characteristics of each component as well as maintenance characteristics of that component.

Hope this helps