Page 1 of 1

Crashing of model after about 1500 hours of simulating

Posted: Tuesday 04 October, 2022 - 08:58
by Rico de Bokx
Hi everyone,

The problem that I am experiencing is as follows. I made the attached model in ED after a few iterations. The process that I am modeling is pretty slow so I wanted to model it so I could test different ways of setting up the process and see what might change in output and throughput. The model is based on the fact that it is the power pack of a train that comes in to be refurbished. Of this particular customer, there are only ever 9 power packs on site in total. I tried to regulate this by attaching a card at the beginning and detaching the card at the end again, by setting the number of cards released I can regulate the number of power packs in the system.
I am now running into the problem that after running the model for 1000 hours it starts to stutter and stop, and after around 1500 hours it just crashes. I can see that the model slows down at the splitter because it is very slow to send the different parts one by one to their stations. Do you guys know a way to maybe fix this or relieve this problem so I can run my model for a longer time?

Looking forward to your guy's thoughts!

Re: Crashing of model after about 1500 hours of simulating

Posted: Tuesday 04 October, 2022 - 14:10
by Gabe-Hein Dijkstra
Hi Rico,

The problem lies in the fact that the number of atoms in your model grows exponentially. Instead of placing atoms in atoms you could maybe store the relevant information in tables or labels. Each atom has its own table (see the Help for more information) but there also exists a table atom.

Hopefully this helps!

Gr. Gabe-Hein

Re: Crashing of model after about 1500 hours of simulating

Posted: Tuesday 04 October, 2022 - 18:11
by Rico de Bokx
Thank you Gabe-Hein,

I am already using labels to store and get some information from the model however I am just using the atoms to dictate when they can be released. I don't prefer to do it this way... maybe there is a way to let a server tell the source when it can release a product? Then I don't have to let the 'cards' circle around and the exponential increase in atoms would be stopped right?

Gr. Rico

Re: Crashing of model after about 1500 hours of simulating

Posted: Wednesday 05 October, 2022 - 10:12
by Rico de Bokx
I fixed the problem!

I didn't know about the lock and unlock flow control atoms. I used those to produce the same function as the 'cards' looping around. Thank you for your help!

Gr Rico

Re: Crashing of model after about 1500 hours of simulating

Posted: Wednesday 05 October, 2022 - 10:19
by Gabe-Hein Dijkstra
Hi Rico,

Great to hear! I was just typing a response to your previous message. ;)

Gr. Gabe-Hein