Recording Average MTBF Data

Question

question

Manocher Djassemi asked Mar 28, '17 Joerg Vogel commented Jun 28, '18

Recording Average MTBF Data

I am trying to record average MTBF or Uptime for each processor. I tried Dashboard and State Stat Analysis and couldn't get such data. Is there a built-in Flexsim function for this or should I create some codes for collecting such data?

Thanks

Manocher

mtbf mttr

Answer 1 · 2017-03-29T19:55:19Z

Sam Stubbs answered Mar 29, '17 Joerg Vogel commented Jun 28, '18

In order to get the average, you can create a simple Process Flow and get the statistics using that. So essentially you're just going to create a Process Flow that generates a single token at time 0. Then you're going to have that token hit a Wait For Event activity (waiting for the break down trigger on your MTBF table) Then it will wait on another Wait For Event activity waiting for the table to go back up again.

On that waiting for break down activity, you can get the average staytime that a token is waiting in that activity. (This will measure the average of how long any objects attached to that table are "Up")

Here is a simple example to illustrate:

example.fsm

example.fsm (24.3 KiB)

· 4

Manocher Djassemi commented · Apr 20, 2017 at 09:15 PM

@Sam Stubbs

My model is constructed using objects connection approach, not with process flow approach. What do you suggest for recording the average MTBF for individual processors in an objects networked model?

Manocher

0 ·

Sam Stubbs ♦ Manocher Djassemi commented · Apr 21, 2017 at 03:42 PM

Even if the rest of your model is connected via ports, this little process flow logic could be added. It is just looking at the MTBF table's down function whenever one of its members goes down and up again. The example I attached should be able to be added to your model relatively easily.

0 ·

Joerg Vogel commented · Apr 21, 2017 at 07:44 AM

You can add a state pie chart to your dashboard. In the Utilization Analysis Tab you mark all states to yellow except the state breakdown, which you mark as green. Then the State Pie shows you the state breakdown relative to all marked states.

0 ·

Manocher Djassemi commented · Apr 21, 2017 at 10:23 PM

Sam

I added your process flow chart without any problem. But I see two new problems. Based on your example, the time between " whenever one of its members goes down and up again", is simply the repair time, not the time between failures. Time between failures is measured from the end of a repair to start of the next repair. Which parameters from MTBF table should be selected to make such measurement? The second problem is I need to generate this data for processors individually, not as grand average of all processors.

Manocher

0 ·

question