question

The idle time after a lack of materials will be simulated by the flow of items, you don't need to incorporate that into your MTBF table. I'm a bit confused by what you mean by "full speed," do you want to decrease the processing time for each station? And if so, how long are the processors at "full speed" after a break down before they go back to "normal speed?" And your 5 second set up time, I assume you want this as a single time setup, just to get the processors at "full speed?"

I think your best bet is to see if you can control the logic of your model using a process flow, I think this might give you a little more direct control over the model. Incidentally, we actually have a bottling station that uses Process Flow to simulate break-downs and maintenance. Perhaps it can give you some ideas on how to approach this.

bottlestationexample.fsm

As a side note, I noticed that you had a few things in your MTBF tables that you should be aware of. First I noticed that the MTBF tables are duplicates of each other, instead of creating a table for each processor, you can just attach each processor to the same table, and have them utilize the same logic. Also I noticed that in your down and up functions you are using a percentage, but you aren't using a full 100 percent. In your up time you say that for 90 percent of the time you want the uptime to be 1, but how long do you want the other 10 percent of the time to be? Same with the down time. I think you are confusing the two, and you are saying that you want the up time to be 90 percent of the time and down to be 10 percent of the time, but instead, these two values should be actual times or random distributions. (IE how long are the machines up for, and then how long are they down for when the break down.)