**1. Introduction**

Production paradigm has been changing since Henry Ford's ''we believe that no factory is large enough to make two kinds of products'' (Ford, H. 1926). With their Scion brand Toyota joined the race for offering customers an increasing product variety, a trend which has been characterizing the automotive industry throughout the last decades (Lee, H. et al., 2005). This development has been driven by two factors. On the outside demand, customization is driven by the improved competitive position of companies which address individual customer's needs (Kotler, P.1989). On the inside supply, customization strategies have been signicantly promoted – if not was made possible at all-by advances in product design and manufacturing as well as information technology(Da Silveira,G.et al.,2001). Based on these advances it became possible to quickly respond to the customer orders by combining standardized modules and cut down the cost.

As the design of the production line is a complex and systematic project, many scholars advance to apply the computer-aided design to each unit of the production line design. Sang Hyeok Han et al (S.H. Han et al., 2011) used Maxscript in 3D Studio Max for automation of the visualization process, which has been applied to the production line of modular buildings with the output of lean, simulation, and visualization in the form of animation, to automate the visualization process as a post-simulation tool through sharing interactive information between simulation and visualization. Thomas Volling(Thomas Volling&Thomas S. Spengler.2011) provided a model and simulation of the order-driven planning policies in build-to-order automobile production, comprising separate interlinked quantitative models for order promising and master production scheduling and evaluating both models in a dynamic setting. Yong-Sik Kim (Yong-Sik Kim et al., 2006) proposed that virtual reality module uses a commercial virtual manufacturing system instead of expensive virtual reality equipments as the viewer of the immersive virtual reality system on a cluster of PCs and adopts the modified simulation algorithm. GAO Chonghui(GAO Chonhui et al.,2010) constructed the virtual simulation for automobile panels based on analyzing the motion characteristics of automatic press line and extracting the corresponding data of motion. These models took better advantage of computer-aided design technology for

<sup>\*</sup> Corresponding Author

Virtual Design of Piston Production Line 3

provide the model and analysis tool to rapidly design the piston production and improve

Production lines involve multiple objects and actions with discrete, random, complexity, hierarchy and so on. Modelling for production lines is the foundation of virtual design. The traditional simulation model mainly focused on the design of algorithms that can be accepted by the computer, resulting in a variety of simulation algorithms and simulation software (Zhao Ji et al.,2000; S B.Yoo et al.,1994; H.T. Papadopolous& C. Heavey, J. Browne.,1993; Zhang Longxiang,2007). From the 1980s, due to high-level language for computer compiling, structured simulation modelling has been a great progress. Chan and Chan(F.T.S. Chan&H.K. Chan,2004) presented a review of discrete event simulation (DES) applications in scheduling for flexible manufacturing systems (FMS). Ashworth and Carley (M.J. Ashworth&K.M. Carley,2007) had conducted a review that addresses organizational theory and modelling using agent-based simulation (ABS) and system dynamics (SD).Shafer and Smunt(S.M. Shafer&T.L. Smunt,2004), Smith(J.S. Smith,2003), Baines and Harrison(T.S. Baines& D.K. Harrison,1999) targeted the larger domain of operations management and applied the simulation to it. However, most reviews limited themselves to either a single technique (DES or SD) or a single application area where more than one technique is used. However, because the interactivity of the structural simulation modelling is poor, it has not been widely application. With the developing object-oriented technology, object-oriented simulation modelling has been rapidly developed. Object-oriented modelling techniques (OMT) is a software environment applying classes, objects, inheritance, packages, collections, messaging, polymorphism and other concepts, which emphasizes the concept of the problem domain map directly to objects or the interface definition between objects, applies the modelling, analysis and maintenance of the realistic entity, so that the built model is easy to reflect the real objects, and makes the constructed model with re-configurability, reusability and maintainability. And it is easy to expand and upgrade and can reduce the complexity of systems analysis and development costs significantly. Many different OMT methods have been advanced, such as OMT / Rumbaugh, OSA / Embley, OOD / Booch et al (Zhang

Longxiang,2007; Par Klingstam& Per Gullander ,1999; Dirk Rantzau et al.,1999).

PPL) to analyse the static and dynamic characteristics of the piston production line.

**3. Virtual design of Piston Production Line (PPL)** 

**3.1 The frame of VD-PPL** 

But when these models are used for dynamic performance analysis on the production line, the modelling is more complex and difficult to describe the dynamic characteristics of the production line quickly and easily, which has greater limitations. QUEST is a virtual integrated development environment applied to queue simulation analysis in Deneb company, which is proper to simulate and analyse the accuracy of the technological process and productivity, in order to improve the design, reduce risk and cost, and make the planned production line meet the design requirements early in the design and implementation, before investing real facilities. Combining the advantages of the QUEST virtual simulation development environment and unified modelling language (UML), this paper presents the simulation, analysis and modelling methods of Virtual Design of Piston Production Line (VD-

Due to the characteristics such as multi-objectives optimization, strong resources correlation, large randomness etc., based on system theory and hierarchical design methodology, VD-

the design rationality in the end(Shao Li et al.,2000).

production line design, but these methods cannot model for the design process of the whole production line, and cannot complete dynamic analysis of production lines. Because of varieties and quantities of the piston are constantly changing, the above method is difficult to effectively and proactively verify the running condition of piston production lines.

As the part supplier of automobile assembly, piston companies also face the same problems, for example, Shandong Binzhou Bohai Piston Co., Ltd. has more than 70 piston production lines to manufacture those pistons such as car, motorcycle, marine, air compressors, chillers, engineering machinery and agricultural machinery pistons. Those size ranges from Ф30 mm to Ф350mm. However, due to the changing market and customized demand, the annual piston species is up to 800 kinds, some piston production line can change the product twice per month, some even more than five, and the production batch is ranging from small to mass. So it is difficult to quickly make the production planning under those demands with the traditional production line design methods. Therefore, it needs the advancing manufacturing technologies and methods to respond quickly to market changes and customized production.

As for production activities in production lines, it often faces the adjustment of design, and the well-designed production line can reduce operating and maintenance costs, improve equipment capacity factor and the efficiency of the system.
