This paper shows the application of open queueing network models to support the design (or redesign) decision making and planning of manufacturing systems. Approximate decomposition methods are used in order to evaluate the queueing network performance. From an actual metallurgical industry queueing network, it is shown that these methods are suitable and effective to estimate, for instance, the work-in-process (WIP) and the production leadtime. It is also shown that optimization models to allocate capacity at network stations, based on decomposition methods, are effective in optimizing the performance of the network under different configurations. Trade-off curves between capacity investment and WIP or leadtime are generated and are useful in supporting a manager to estimate how much capacity should be allocated to the network to achieve the performance goal. These curves are also useful in analysing the effects in terms of capacity if the variability of external arrivals of products, the product mix or the rate of production of network changes.
Manufacturing system design; Performance evaluation methods; Open queueing network; Optimization models; Trade-off curves
