【正文】 f manufacturing systems and simulating manufacturing processes instead of their operations in the real world. VM pursues the informational equivalence with real manufacturing systems. Therefore, the concept of Virtual Manufacturing System is expected to provide dramatic benefits in reducing cycle times, manufacturing and production costs, and improving munications across global facilities to launch new products faster, improve productivity and reduce operations costs for existing product shop [1,2]. With an objectoriented paradigm, puterbased technologies such as virtual prototyping and virtual factory are employed as a basic concept for developing the manufacturing processes, including the layout of the optimal facility, to produce products. Virtual prototyping is a process by which advanced puter simulation enables early evaluation of new products or machines concept without actually fabricating physical machines or products. Bodner, et al.,[3] concentrated on the decision problems associated with individual machines that assemble electronic ponents onto printed circuit boards (PCBs). Virtual factory is a realistic, highly visual, 3D graphical representation of an actual factory floor with the real world plexity linked to the production controlling system and the real factory. Virtual factories are increasingly used within manufacturing industries as representations of physical plants, for example, VirtualWork system for representation of shop floor factory[4]. Despite its benefits and applicability, VM systems should deal with a number of models of various types and require a large amount of putation for simulating behavior of equipment on a shop floor. To cope with this plexity in manufacturing, it is necessary to introduce open system architecture of modeling and simulation for VM systems. In this paper, three models, which are product, device, and process models will be addressed. Especially process model for FMC will be emphasizedusing QUEST/IGRIP as an implementation issue. The open system architecture consists of wellformalized modules for modeling and simulation that have carefully deposed functions and welldefined interface with 2 other modules. 2. Concept of virtual manufacturing Virtual Manufacturing System is a puter model that represents the precise and whole structure of manufacturing systems and simulates their physical and logical behavior in operation, as well as interacting with the real manufacturing system. Its concept is specified as the model of present or future manufacturing systems with all products, processes, and control data. Before information and control data are used in the real system, their verification is performed within virtual manufacturing environment. In addition, its status and information is fed back to the virtual system from the real system. Virtual environments will provide visualization technology for virtual manufacturing. The virtual prototype is an essential ponent in the virtual product life cycle, while the virtual factory caters for operations needed for fabricating products. Therefore, the developments in the area of virtual prototyping and virtual factory will enhance the capabilities of virtual manufacturing. The major benefit of a virtual manufacturing is that physical system ponents (such as equipment and materials) as well as conceptual system pvonents (., process plans and equipment schedules) can be easily represented through the creation of virtual manufacturing entities that emulate their structure and function. These entities can be added to or removed from the virtual plant as necessary with minimal impact on other system data. The software entities of the virtual factory have a high correspondence with real system ponents, thereby lending validity to simulations carried out in the virtual system meant to aid decisionmakers in the real system. For virtual manufacturing, three major paradigms have been proposed, such as Design centered VM, Productioncentered VM, and Control centered VM. The designcentered VM provides an environment for designers to design products and to evaluate the manufacturability and affordability of products. The results of designcentered VM include the product model, cost estimate, and so forth. Thus, potential problems with the design can be identified and its merit can be estimated. In order to maintain the manufacturing proficiency without actual building products, productioncentered VM provides an environment for generating process plans and production plans, for planning resource requirements (new equipment purchase, etc.), and for evaluating these plans. This can provide more accurate cost information and schedules for product delivery. By providing the capability to simulate actual production, controlcentered VM offers the environment for engineers to evaluate new or revised product designs with respect to shop floor related activities. Controlcentered VM provides information for optimizing 3 manufacturing processes and improving manufacturing systems. The virtual manufacturing approach in this paper is close to Controlcentered VM. illustrates the viewpoint of the functional model of the virtual flexible manufacturing cell. Since the activity Execute real manufacturing systems depicts a model of real factory, it possibly replaces real factory. All manufacturing processes except physical elements of virtual manufacturing, such as design, process planning, scheduling, are included in the activity Operation of Virtual f