serazzi

no problem Zahra, we are happy to help people use JMT! Best regards, giuseppe

Hi, you may find the JSIMg model attached. I used for the demo r1=3j/s. I have used two classes. You may solve analytically the model. The max r1 is 3.5 (the U of Q1 will be 1!). You have to solve a simple system with two equations (r1+0.5 X2=X1) and (r2+0.3 X1=X2) where X1 and X2 are the unknown throughputs of Q1 and Q2, respectively. Leave r1 as a parameter. Than to find the max value of r1 use the Utilization equation X1 S1 <=1 and find the r1. In the model attached i used r1=3 and you may see...

Hi, the four queue stations have negligible uses and therefore their response times correspond to their service times (0.04). So the response time of Queue1 will also be 0.04. It is important to change the unit of time (e.g., ms instead of sec) to have a more meaningful model.

Hi Marcello, sorry for the delay in answering. Yes, you are right. When between two queues (the second use BAS blocking) there is a logger there is a problem. The jobs are blocked at the logger but are not decreased when are released. I may suggest you to use simply the CSV files obtained automatically flagging the StatRes in the performance indices window of the "throughput" of Queue1. You have the same data of the logger 2 without using the logger. Ciao

and the last is the layout of model 2

here the layout of model 1

here the other attachments.

Hi Rahee, attached you will find two models. I adopted all your assumptions except for the routing strategy. It is known that the one you suggested tend to generate a very unbalanced load on the three stations (station three is overloaded and station 1 is underloaded). I modeled a routing startegy "Round Robin". You may change according to your needs. The service times of the servers are exponentially distributed with mean 1/3 s. The two models provide the same results but are at diferent level of...