8.16.2 The `SEQUENCES' element
There are many occasions on which the route to be followed is not in the order in which the stations have been defined or alternatively the routing order may have to change during the simulation (perhaps because of a breakdown or failed inspection). The `SEQUENCES' element has been provided to accomplish this task. The element is essentially a list of stations to be visited by the entity in the order specified by the list. Additionally the facility is provided to change any of the entities attributes prior to the transfer taking place. The `SEQUENCES' element can specify many individual sequences each of which is identified by a name or number
Example 8.12
In the following example jobs of type A are sent in turn to the miller, the lathe, inspection and packing whilst jobs of type B follow the sequence drill, lathe, miller, inspection and packing. In each case before the transfer takes place the attribute process time is set to the appropriate value.
SEQUENCES : 1, JOBA,As is usual with ARENA you enter values into text boxes and can ignore the syntax which is generated automatically.
The `ROUTE' block knows that it is supposed to route an entity according to a sequence if instead of the transfer station name the `ROUTE' block uses instead the reserve word `SEQ'.
WORKSHEET 9 THE ASSEMBLY STATION
Parts arrive to a three station serial assembly system according to a uniform distribution with a minimum of 5 minutes and a maximum of 15 minutes. Parts are processed at the three stations and inspected at the last two. Combined times for these operations follow a triangular distribution with the following parameters.
Station |
Minimum |
Mode |
Maximum |
Station 1 |
6 |
9 |
12 |
Station 2 |
5 |
8 |
11 |
Station 3 |
5 |
7 |
10 |
If a part fails inspection it is sent back for reworking. Failure rates and the station to which the part is sent is given by the following table.
From end of |
To beginning of |
% failure |
2 |
2 |
5 |
3 |
3 |
5 |
2 |
1 |
5 |
3 |
1 |
5 |
Rework times follow the same distributions as the original work times.
Model this system for 30,000 minutes in order to determine number of products produced, product flow time, queue sizes, and machine utilisations. Repeat the model with no failures.
New model statements to include :
DISCRETE (distribution)
New experimental statements to include :
SEQUENCES
Additional Work : Write the model with and without using a station submodel.
Solution
Your model should give close to the following results :
Total number of jobs processed = 298
Machine utilisations : M1 = 89.5%; M2 = 79.8%; M3 = 72.9%
WORKSHEET 10 `TOO HOT TO HANDLE !'
A three machine shop processes heated metal parts. Three different parts are produced by this operation with the following routings shown in the table :
Type |
Percentage |
Route |
Process time |
1 |
30 |
1 |
6,1 |
2 |
10,2 |
||
3 |
10,2 |
||
2 |
45 |
2 |
9,2 |
3 |
5,1 |
||
1 |
4,1 |
||
3 |
25 |
1 |
12,2 |
3 |
7,1 |
||
2 |
5,2 |
||
3 |
4,1 |
Parts arrive according to the exponential distribution with a mean of seven minutes and all process times follow a normal distribution with the mean and standard deviations as indicated in the table (mean,std). All transfer times are exponentially distributed with a mean of two minutes. Because parts must remain heated during the entire production process, the buffer space in front of each machine is limited to three parts. If a part arrives at a full buffer, it is placed in an oven located next to each buffer. Each oven can hold five parts and, if the oven is full, parts are balked to a disposal area, counted by machine centre and then disposed of. If a part is placed in an oven, it remains there for one hour and then tries again to enter the buffer. There is no limit to the number of times a part can be placed in an oven.
Run the simulation for 5,000 time units and keep statistics on the number in each buffer and oven and the utilisation of each machine. Also record the part flow time and the number of parts balked from each oven.