Cancel
Showing results for 
Search instead for 
Did you mean: 

Modeling Failure and Repair in Parallel Stations

Experimenter
Experimenter

I want to run a production line that has unreliable stations. I use Interval and Duration to model exponential Failure and Repair. My understanding is that ParallelProc will fail all servers in ParallelProc and not individual stations? In the model I want to validate, each individual station fails according to the Interval and Duration times. Is this a correct interpretation of ParallelProc?

 

To overcome the above problem I plan to use SingleProc to model each station, see attached model.

 

 

Is there an in build method of the Exit of the IOBuffer that will move Mu to an empty and not failed station? The reason I want an in build function is that I want to construct the models using SimTalk using information on Buffers, Phases (i,e., no of servers in a station), process times, failure times and repair times. If the answer to this is no, will the code in the IOBufferExit method work to model the above?

 

 

The waituntil does not allow a complex boolean, so loops I believe will not work. To overcome this is it possible to insert a reference to SingleProc.empty or SingleProc1.empty or SingleProc2.empty using the IOBuffer object and its connections to these objects (SingleProc, etc)?

 

Thanks

3 REPLIES 3

Re: Modeling Failure and Repair in Parallel Stations

Phenom
Phenom

Plant simulation does not move Mus to stations which are empty but down - which makes sense.

 

Therefore you do not have to model this kind of behaviour.

 

 

 

Re: Modeling Failure and Repair in Parallel Stations

Siemens Phenom Siemens Phenom
Siemens Phenom

you can use the exit strategy "Start at successor 1"

------------------------------------------------------------------------------------------------------
Did you like the answer? Then click the Thumbs Up button.
Did the answer solve your problem? Then accept the answer as solution.

Re: Modeling Failure and Repair in Parallel Stations

Experimenter
Experimenter

Examining this further I believe using the inbuilt  Cyclic Exit and turning off blocking will model what I require. Help states that for Cyclic Exit

When you clear Blocking, the object moves the MU to the next object in the sequence of successors, which can receive a part.

I belive this captures the modeling I require.