Modeling a process which can produce 2 different items with
Posted: Wed Sep 29, 2010 8:12 am
Hello @ all.
I got a question about modeling a process, which is able to produce 2 items.
Lets say I want to model an engine block casting production which is able to produce an usual steel engine block (1) and an magnesium engine block (2).
I want to model the production either just with a delay or with a delay and a queue (actually the production planner will only put as much material in the process as capacity is available, but as far as some production steps are processed cyclical or as a batch of 100 engines there will be still a queueing in the process).
The production of (1) has a certain delay distribution and the production of (2) has a different delay distribution (a longer production, because some steps take longer and other steps have to be done more often than for (1)).
There are 2 options to model this:
1. The first one is I calculate the overall/average delay (maybe depending on the input fraction) for all materials together and calculate the "input fraction" of (1) and (2). This fraction I will delay like the process and use it to split up the produced materials.
This option is quite inaccurate because it will not reflect the reality well, but it is simple.
2. I will split up the process in two paralell and equal processes with different delay times. Over here I have to see how I can handle capacity constraints and so on, but the result is more accurate, because all items of (2) will stay the correct time in the process.
My question now is, if someone is having some experience on the difference of both and on that issue.
Lets say you don't have 2 but 10 different materials, this will be a lot more work to do!
Maybe you have an idea, if you don't understadn what I am trying to say let me know and I will explaing better.
Thanks, Mattes_01
I got a question about modeling a process, which is able to produce 2 items.
Lets say I want to model an engine block casting production which is able to produce an usual steel engine block (1) and an magnesium engine block (2).
I want to model the production either just with a delay or with a delay and a queue (actually the production planner will only put as much material in the process as capacity is available, but as far as some production steps are processed cyclical or as a batch of 100 engines there will be still a queueing in the process).
The production of (1) has a certain delay distribution and the production of (2) has a different delay distribution (a longer production, because some steps take longer and other steps have to be done more often than for (1)).
There are 2 options to model this:
1. The first one is I calculate the overall/average delay (maybe depending on the input fraction) for all materials together and calculate the "input fraction" of (1) and (2). This fraction I will delay like the process and use it to split up the produced materials.
This option is quite inaccurate because it will not reflect the reality well, but it is simple.
2. I will split up the process in two paralell and equal processes with different delay times. Over here I have to see how I can handle capacity constraints and so on, but the result is more accurate, because all items of (2) will stay the correct time in the process.
My question now is, if someone is having some experience on the difference of both and on that issue.
Lets say you don't have 2 but 10 different materials, this will be a lot more work to do!
Maybe you have an idea, if you don't understadn what I am trying to say let me know and I will explaing better.
Thanks, Mattes_01