An investigation of parts flow in the flexible manufacturing system

A certain Flexible Manufacturing System was investigated. The line of the system consists of four machines, a queueing area, and a material handling system represented by a cart. The main objective of this study was to investigate the effects of different levels of cart utilization, local storages, and the size and design of the queueing area on the FMS performance. Additionally, the examination of different scheduling rules was conducted in order to see their impact on the FMS performance.

Simulation was the chosen methodology for investigating the above stated issues. The GPSS/H simulation language was used to simulate the model of the examined FMS and to investigate the stated problems. Cart utilization was defined as the ratio of the total time that the cart was in use to the total time of the system operation. Different levels of cart utilization were obtained by changing the processing times of the part mixes manufactured in the system. The main criteria for the FMS examined were the number of parts and operations completed, as well as the machine utilization.

The experimental results showed that the FMS was strongly affected by increasing cart, or generally speaking, material handling utilization. When a high level of cart utilization was experienced, this caused machine utilization and therefore, system productivity to decline. Cart utilization is a dependent variable in any FMS, an effect of several factors in the FMS design. The major components of material handling utilization are travel times, number of machines in the system, and part processing times. A ratio factor R, taking into account the above factors was proposed and the curve between the ratio R and cart utilization was developed. The FMS proposed by the General Dynamics in Fort Worth, Texas was used to test the ability of the developed curve to predict the material handling utilization for a different system. The theoretical cart utilization based on the ratio R and the developed curve, was very close to the material handling utilization obtained through the simulation.

Adding local storages to the system configuration reduced the machine idleness due to the lack of cart service. The highest improvement in the system productivity after installing local storages was observed for the FMS operating with a cart utilization of 75% - 97%.

The experimental results showed that the system was very sensitive to any changes in the configuration which caused cart utilization to increase. Providing more queueing stations by expanding the line length, and therefore increasing cart utilization, resulted in poorer performance of the FMS.

The effects of the number of parts in process on the system productivity was examined by changing the size of the queueing area. The experimental results indicated that a given FMS provided the best performance with a certain number of parts in process. Expanding the queueing area above this level did not produce the better system output; on the other hand too few parts in process led to lower machine utilization.

The experimental results showed that for the FMS examined, such rules as SPT, SPT*TOT or FCFS to system, provided the most balanced demands for all machines and therefore the highest system output. It was noted that the scheduling problem in the FMS should be carefully investigated for any system developed in terms of the production and configuration characteristics.