Abstract
The steel industry (SI) produces parts for different industrial markets, such as aeronautics and tools. The last few years have shown that the demand from these markets can be subject to variations, which can be drastic (e.g. consequences of the events of 11 September 2001 on the aeronautical industries) or more progressive. To cope in the best way with these unexpected changes, steel industries have to evaluate their ability to react, at a strategic decision level, to these changes. In particular, they need to adapt the long-term management of their resources in the best possible way. Although issues related to short-term perturbations appear in the literature, seldom seen are methodologies to analyse the long-term behaviour of manufacturing systems. In this article, we are concerned with the evaluation, using simulation, of the effect of unexpected variations in the demand of products on the firm performance, in order to determine in what extent the system may be disturbed, and whether or not it can cope with the consequences induced by these variations. We address this problem through a system dynamics approach. Several families of parts, which correspond to different markets, are distinguished. Different types of signals are associated to these families to model the demand variations, in order to analyse the system's sensitivity. A real industrial application illustrates this approach. The results are discussed.