Showing results for 
Search instead for 
Did you mean: 

Dynamic dwell time



I'm wondering what happens when I change the buffer dwell time and many mus are already into the buffer.

An example to make your understanding easier.

I have two different parts: A and B.

50 parts A enter the buffer with dwell time 100 min. After the last part enters, one part B enter the buffer (let's say, after 5 min). Before B enters, the dwell time is changed to 50 min.

Assuming after the buffer there is a drain with no time, I imagine the part B wait for 95 minutes, then exit the buffer because it already waited its dwell time (50 min). All the other parts are not affected by the change.

Does it really work in this way? Or somehow changing the dwell time by code, the parts which are consuming the time into the buffer are affected?




Re: Dynamic dwell time

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom
For this behaviour it is easier to use a ParallelProc, so that each MU gets its own dwell time.

Gert Nomden
Senior Consultant | Tecnomatix Product Manager

Re: Dynamic dwell time


Thank you, but I don't know if is possible does what I have to do with a singleproc.

I try to explain my scenario.

I have three machining lines, each one simulated in separate model. From each model I extract the lead time (which changes depending on the type).
So I have for the first line 5 types (with 5 different lead times), for the second one 3 types and so on.

In another model I have the assembly lines with the logistic and I need to simulate the machining line without using the actual model (in that case I would have thousands of objects...).
My idea was: each line has a source which produces a batch sequence. I put all the batch in a buffer with a certain dwell time to simulate the lead time. The source has an interval time to simulate the line takt time. But I need to change the lead time when the batch changes.
At the end I store all the machined parts in a storage where they wait the assembly will take them out.

Maybe instead of the buffer I could use a line with a proper time to simulate the lead time, but still I would have the problem to change it according with the batch.

Do you think this is the right approach or do you have other suggestions?

Re: Dynamic dwell time




I would say instead to store the lead time on the MU as a attribute, then have a large buffer which is your production area. Have no dwell time on that buffer and instead on the enterance control have something like the following (untested code btw)


wait @.LeadTime

Ensure that the exit to the buffer is locked so it will only move along when the code sais to do so. This way you can change the lead time on a MU by MU basis as needed.



Re: Dynamic dwell time


Thank you James,
this solution seems simpler, I'll try to implement it.