About recycling flowitem problems

Hi,
In the attachment, I try to model such a system characterized by recycling flowitems.
The procedure of my model is as follows:
1. On the beginning of simulation, two empty tanks are put on the TransporterLc1 and TransporterLc2, and the two tanks are not shown in my model.
2. The Source will generate molten steel continuously according to the arrival schedule table, and subsequently, molten steel will be infused into the empty tanks.
3. Once the tank is filled with molten steel, TransporterLc1(or TransporterLc2) will bring it to the next station by coordinating with other TEs.
4. If the OnProcessFinish event of ProcessorKr1(or ProcessorKr2) triggers, the tank will be transferred to the last station, named as ProcessorLdp1 and ProcessorLdp2.
5. In the setup procedure of the last station, molten steel will spill out to the processor and the tank will be emptied gradually.
6. Simultaneous with processing of molten steel, the empty tank will be carried to TransporterLc1(or TransporterLc2) and be placed on original position.

Now my question is: which is the best way to deal with the recycling flowitems problem? I don’t want to use the fluid objects.

In my model, the flowitem is composed of a empty tank and molten steel in it, and I am not sure whether it’s suitable to define the empty tank and molten steel as different type of item, and use Combiner/Separator to combine or separate them?

Does anybody has similar experience to deal with such flowitem that is composed of container and fluid?

My colleague and I have modelled our steel making facility and the rail system used to transport hot metal. Within the actual facility, we use rail cars (called torpedo cars) to transport molten iron from the blast furnaces to the steel shop.

We did not model the molten iron. We modeled the torpedo cars using a flowitem. We had a label on the torpedo car that tracked the quantity of molten metal in the car in addition to other characteristics of the metal such as temperature and chemistry of that hot metal load. The time taken to empty or fill the torpedo car was based on the flow rate and the quantity of hot metal being added to or removed from the car. Whenever molten iron was added to or removed from the torpedo car, the quantity label on the flow item was updated at the end of the calculated fill or pour time. A corresponding hot metal quantity label was also updated on the faciltity providing the hot metal or consuming the hot metal. There were multiple locomotives in the system so the flowitem (hot metal car) was handled by a number of TE's in the complete cycle back to the blast furnace.

The advantage of tracking hot metal in the system numerically is that you speed up your model run time significantly as the number of flowitems in the system is greatly reduced.

I hope this helps. If you have any additional questions about how we modeled the movement of molten metal, please let me know.

Danny

hi danny，i do same project ,But not very successful, I also use label to track the quantity of molten metal in the car .but in my project the hardest part is the train dispatching and collision. I don't know how you deal with this problem. do you use RailAPI library?
If you have free time,i want to exchange something with you.

Danny, thanks for your help.
I totally agree with you that having a label on the torpedo car is a good idea, but in my system, we assume that the quantity of molten metal in the car(I named it tank in the first thread) is a fixed value, 300 ton, for example. And this fixed quantity of molten metal is the basic unit in the whole processing time.

For the aboving reason, we place great emphasis on the transfering of molten metal in the tank by different TEs, not on the quantity and calculating of it.

As you have mentioned in your thread, I am interested in your modeling methods of the molten metal's movement (or multiple locomotives), and I eagerly appreciate your advice.

I am assuming that the focus of the model is on moving hot metal and not on the hot metal itself.

The quantity is not of primary importance nor is the transfer rate. It is the combination of the two, that is, the time to pour a quantity - the transfer time, that is of primary importance.

The car cannot move or undergo further processing until the pouring is complete. The time required to model the pour time can be done many ways in Flexsim (eg. delayed message with a delay of quantity/rate or a processor with a process time of quantity/rate, etc.). Either of the examples mentioned is more efficient than moving flowitems that represent some smaller quantity of molten metal. The overall result is that the quanity at the source of the hot metal is reduced by the poured quantity and the quantity at the destination of the hot metal is increased by the same amount.

I've attached a pdf of a presentation that my colleague gave at the Instititue of Industrial Engineering Conference in 2007. Most of the presentation is not relevant to this discussion as it does not go into the details of how we modeled the system but it gives an idea of the scope of the system. We did it all without hot metal flowitems. We tracked the hot metal in the system quantitatively with labels on different objects in the model. The torpedo cars in the system were flowitems and the locomotives were task executers. The torpedo cars were filled, weighed, drained, and cleaned in their cycle around the system. The hot metal in the cars was also "processed" in car to adjust its chemisry

I hope this helps.

Danny