simulation stops before completing the operation

Hi everybody,

I have got two processors which produce boxes some of which need two transportes to transport and some which need only one.
Using coordinated task sequences I manage to plan the transports but simulation stops at some point in time and does not proceed further. I have attached the model here.

the major codes are under useTransporters of the processors and the sources produce input based on a table.


I am wondering what could be the reason for such problem.
Any comment is appreciated.

best regards,

Hello,

I haven't found anything so far. But there exists in flexscript some very short commands eg i, a or c. It is not recommended to use these command names as variablenames eg in loops "for (int i = ..."

Jörg

maybe check the eventlist (treeview -> main/events ) to find out at which event flexsim hangs. That would give you an idea where to look for a solution

Hello Shahin,

build a much more simplier model to test the coordinated tasksequence.
Insert an frunload-task.
Unload to a queue and not to a BasicFR.
Test each step in a simple model.
Only if one step is running perfectly, than put this solution in a next complex model. in contrast to that do not build maxium advanced models with various not solved problems.

Jörg

Hi Joerg,


Thanks for the comment.
In fact, that is what surprises me. I usually follow the same process as you said. Every thing is fine just before this integrated model.

Steven,

I am not sure how that could help? could you elaborate on it?

Hello Shahin,

I was curious about how you synchronize all your ressources (your other thread) so I take a look at your model and I found this:

1. In one of your coordinated tasksequence (cts) you deallocated a resource that was not previously allocated (NULL pointer). This threw an exception.
2. Crane object cannot do TASKTYPE_TRAVEL task. It can however do traveltoloc or travelrelative.
3. The dispatcher that will receive the cts do not have to be a taskexecuter allocated within the cts. It can be any dispatcher.
4. Be aware that cts aren't queued like normal tasksequence. Once dispatched, the system immediately start executing it. This may result in many cts executed simultaneously with deadlock situations where transporters and cranes are allocated in different cts and waiting for each other.
5. I suggest that you give a meaningful name to the cts you create. Its easier to debug later.

I've uploaded your model with the modifications.


Regards,




Normand

Hi Normand

Thank you for your suggestions and modifications.
Well two things
1) what did you change that it ask to compile before running?
2) I still see that simulation terminates before completion of the work. I am I wrong? I see still boxes at processor and transporters being idle close to the cranes and one of which was totally unused.
3) I don't know why I cannot find the redundant deallocate command.
how did you learn that the cranes do not have travel ? I cannot find it in documentation. is that by experience or I am missing something here?
best regards,

-------------

Shahin, see comments below


1) what did you change that it ask to compile before running?
I have inserted a c++ node in your model. I removed it.
2) I still see that simulation terminates before completion of the work. I am I wrong? I see still boxes at processor and transporters being idle close to the cranes and one of which was totally unused.
You're right. This is another problem with your model. Your processors' output ports are connected to some stacks but in your custom box transport logic you choose to send to different stacks. After a certain number of box deliveries, the internal handshaking between processors and stacks connected to their outpout ports just block and the useTransporter trigger do not fire. I disconnected all output ports of the processors and moved the useTransporter logic in the processfinish trigger and everything works fine. If you prefer to work with the output port then you have to send the box to the selected output port.
3) I don't know why I cannot find the redundant deallocate command.
When itemtype of a box is 1, you allocate only one transporter but you deallocate two

I too have a question: why do you use BasicFR objects as stacking areas ? Have you consider Rack objects (setup as floor storage) ?

Good luck,


Normand

Normand,

I very much appreciate your comment. It was very very helpful and I learned new things.
Well, I actually used one of examples in flexsim and tried to modifiy it towards my logic which apparently was not successful.
But, what you did was very nice and clean.
Well, I used BasicFr because I was not sure how to work with Racks having Crane as transporter as the transporter this area must be a crane and not a top loader.

thanks again.
best regards
Shahin