機(jī)械外文文獻(xiàn)翻譯-CAE分布式協(xié)同設(shè)計(jì)系統(tǒng)在有限元復(fù)雜產(chǎn)品的SOOA分析【中文14900字】【PDF+中文WORD】,中文14900字,PDF+中文WORD,機(jī)械,外文,文獻(xiàn),翻譯,CAE,分布式,協(xié)同,設(shè)計(jì),系統(tǒng),有限元,復(fù)雜,產(chǎn)品,SOOA,分析,中文,14900,PDF,WORD Contents lists available at ScienceDirect Advances in Engineering Software journal homepage: www.elsevier.com/locate/advengsoft Advances in Engineering Software 41 (2010) 590–603 A CAE-integrated distributed collaborative design system for ?nite element analysis of complex product based on SOOA Jiaqing Yu a,b,*, Jianzhong Cha a, Yiping Lu a, Wensheng Xu a, M. Sobolewski b a School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing 100044, China b Computer Science Department, Texas Tech University, Lubbock, Texas 79409, USA a r t i c l e i n f o Article history: Received 23 January 2009 Received in revised form 5 October 2009 Accepted 24 November 2009 Available online 23 December 2009 Keywords: Collaborative design Service object-oriented architecture (SOOA) Distributed CAE resources a b s t r a c t Large-scale ?nite element analysis of complex product needs a wider support from external CAE resources. This paper proposes a method of the distributed concurrent and collaborative design in the distributed intelligent resources environment to use Web-based extended manufacturing resources for product development. The CAE-integrated distributed collaborative design system architecture and enabling techniques are presented. Distributed CAE service resources on the Internet are achieved through JiniTM and service object-oriented architecture technologies. These CAE software tool resources are encapsulated as service providers in the system. A case study of railway bogie development is pre- sented, and the Pro/E, HyperMesh, Ansys and HumanExpert service providers dynamically invoked and integrated as a temporary service federated environment with no need to know the exact location of a provider beforehand, ?exibility can be well achieved, and collaborative design in the system can be implemented in the distributed dynamic environment. Compared with the traditional design system, some preliminary results indicate this architecture can shorten the cycle of service exchange, and strongly support concurrent and collaborative design of the complex product. 。 2009 Elsevier Ltd. All rights reserved. 1. Introduction Owing to the continuously increasing possibilities provided by related technologies and their wider application, today’s products are becoming more and more complex both from technical and managerial standpoints [1–3]. Products or services to be developed (the deliverable) with complexity in their formations are called complex products, such as launch vehicles, robots, ships, railway locomotives and airplanes, automobiles, weaponry systems. Com- plexity is the opposite of simplicity and is characterized by number of parts, number of part types, number of interconnections and interfaces, and ?nally by number of functions [4,5]. The main feature of complexity is structural uncertainty, viewed in terms of differentiation and interdependence [6]. Complexity is de?ned by the number of parts and variants in a module [5,7], the type of interaction between elements [8], interactions among subsys- tems [9], integrated design [10], high coupling between several different technologies and integrated architecture [11]. Generally, the development process of complex product in essence not only considers the distributed intelligent resources environment and the concurrent design process in the entire life * Corresponding author. Address: School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing 100044, China. Tel.: +86 010 51467594; fax: +86 010 51688253. E-mail address: 06116256@bjtu.edu.cn (J. Yu). cycle of the manufacturing and assembly, but also the integrated concurrent and collaborative design in the mechanical, control, dynamics, etc. Traditionally a large design team will be engaged, the communication and collaboration among members are signi?- cant to enable design task of complex product to be carried out effectively. However, the conventional product design, which is usually or- ganized within a laboratory or enterprise due to being geographi- cally limited and isolated Computer Aided Engineering (CAE) systems, cannot meet the customers’ growing diversi?ed demands of complex products. One side, much higher quality is strong de- mand on condition that these products should be developed in the same or even shorter life periods of time. While simulta- neously, for the development contains interconnection, suf?cient communication and interfaces, managerial issues arises [12]. In order to tackle managerial issues and transact complex prod- uct in possibly multi-disciplinary design and geographically dis- persed resources environments, design and analysis of product is considered to divide into separate, in most cases, concurrent design phases, and apply a distributed development approach [13–17]. Concurrent engineering, collaborative design and the optimization theory in the multi-disciplinary design are popping up in this area. Similarly, for the cross-sector, cross-region and cross-country alliance of virtual enterprises develop quickly, the design and develop environment has changed a lot, and many designs for complex product have to be collaboratively completed 0965-9978/$ - see front matter 。 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.advengsoft.2009.11.006 J. Yu et al. / Advances in Engineering Software 41 (2010) 590–603 591 by the product design staff and the other related staff in different places. Then the distributed collaborative design technology came into being and extensive research and development works have also been carried out [18–21]. Another issue is that the distribution and utilization of manu- facturing resources are relatively unbalanced in China. There are a considerable amount of underutilized manufacturing resources in the scienti?c research institutions, universities and large enter- prises [22]. A challenging issue on networked manufacturing resource services is how to discover these socialized manufactur- ing resources and provide the effective resource support services [22,23]. Now the rapid development of the Internet technology enlarges the variety and scale of manufacturing resources available [22], and still enables global networked manufacturing resource ser- vices to be unprecedented access and optimized con?guration [22–26]. Future CAE systems are moving towards supporting dis- tributed and collaborative design, in which geographically dis- persed systems can be integrated and a virtual design team can be set up within an Internet/Intranet environment [27]. Developed by Dassault Systèmes, CATIA (Computer Aided Tri-dimensional Interface Application) is designed to optimize all stages of the pro- duction life cycle and offers the most comprehensive applications portfolio available in a single system, aiming to maximize concur- rent product development practices and process re-engineering [28]. As an industrial application example, CATIA for Industrial Equipment offers a secured collaborative environment that is accessible throughout the supply chain. This enables you to share and transfer the most up-to-date information and work concur- rently, harmonizing the work environment even further. In recent years, the ‘‘software crisis” prompted the beginnings of software engineering, and the appearance of Web services developments and standards supporting automated business integration has impelled core technology advancements in the integration software space, most notably, the service-oriented architecture (SOA) [29–32]. The purpose of SOA is to address the requirements of loosely coupled, standards-based, and protocol- independent distributed computing, mapping enterprise informa- tion systems (EIS) appropriately to the overall business process ?ow [33,34]. SOA is designed to enable developers to conquer many distributed enterprise computing challenges including appli- cation integration, transaction management, security policies, while allowing multiple platforms and protocols and leveraging numerous access devices and legacy systems [35]. The driving goal of SOA is to eliminate these barriers so that applications integrate and run seamlessly. In this paper, a CAE-integrated distributed collaborative design system (CAEi-DCDS) based on SOA is proposed to automatically or semi-automatically accomplish ?nite element analysis of complex product. By CAEi-DCDS, various engineering software tools and utilities can be easily integrated as loosely coupled services that can form a temporary dynamic service federation for a speci?c design process while being requested. The CAEi-DCDS provides reusability, scalability, reliability, and ef?ciency that can be achieved by the service-oriented programming [36] and relevant dynamic service object-oriented architectures (SOOA) infrastruc- ture [37]. In particular, we concern on the underlying enabling technologies and implementation ?nite element analysis of com- plex product in the Web-based CAE resource services environment on the basis of the state-of-art of related research. Note that in this paper, the scope is CAE-oriented, though other engineering soft- ware tools, such as Computer Aided Design (CAD), Computer Aided Manufacture (CAM), Computer Aided Process Planning (CAPP), Product Data Management (PDA), Design for Assembly (DFA), De- sign for Manufacturing (DFM), system simulation, etc., can also be integrated as services into CAEi-DCDS.  The main contribution of our work is summarized as the four aspects. Firstly, this paper de?nes the concept of distributed con- current and collaborative design (DCCD) for complex product development in accordance with the theory of the optimization theory in the multidisciplinary collaborative design. Secondly, this paper provides an integration system to link various CAE tools and a means to invoke these software resources remotely at runtime based on the SORCER infrastructure [36,38]. Thirdly, CAEi-DCDS may dynamically organize the software resources and provide a temporary service federated environment for different exertions [37], so the CAE software resources can dynamically participate in different project jobs. Fourthly, an application case of the static stress analysis of railway bogies is developed and indicates the reli- ability and ef?ciency of CAE-integrated distributed collaborative design system for complex product development. Following Section 1, this paper is structured as follows. In Sec- tion 2, the related work is investigated and summarized. System requirements are taken into account as viewed from the system development in Section 3. In Section 4 discusses the solutions to achieve the above requirements on the proposed system. Section 5 outlines the system framework and functions. Section 6 demon- strates an application case of the proposed prototype system for railway bogies. Section 7 summarizes the paper. 2. Related work 2.1. Collaborative design When a product is designed through the collective and joint ef- forts of many designers, the design process may be called collabo- rative design, or co-operative design, distributed concurrent design and inter-disciplinary design [18]. The research for the distributed collaborative design started in the 1990s, and Cutkosty from the Design Institute of Stanford University ?rstly began the research in this area [39]. In 1990, National Institute of Standards and Tech- nology invested 21.5 million dollars for a develop team in a project called Federated Intelligent Product Environment (FIPER) planning to exploit a collaborative supported work environment architec- ture in ?ve years. General Electric Company successfully found good application of FIPER to develop the American key weapon equipment [37,40]. Various distributed collaborative applications have been devel- oped for different engineering domains in various system architec- tures. There are Web-based open architecture of product data management [41], the architecture of a software distributed col- laborative exploitation environment [42], distributed collaborative product customization system based on Web3D, which provides distributed collaborative product customization for product users in a virtual environment [43], Intelligence Concurrent Design Sys- tem under network aimed at the application actuality of CAD sys- tem in corporation [44]. The architectures of these systems can be generally grouped into two categories, i.e., client/server (C/S) [45,46] and peer-to-peer (P2P) [47]. In C/S architecture, the server plays as an information agent and broadcast CAD model and commands generated by one client to other clients [48–51], or the geometric kernel is in the server side and the commands are passed from client to the server to carry out the modeling activities [52,53]. A server plays as an information exchanger to broadcast CAD ?les or commands generated by a client to other clients during a collaborative design process [54]. CollaCADTM is developed [55] and a thin/strong representation in client/server has been proposed to enhance the performance of the system effectively [56], such as Alibre DesignTM [53], and other developed systems [56–58]. A network-centric feature-based 592 J. Yu et al. / Advances in Engineering Software 41 (2010) 590–603 virtual prototyping system in a distributed computing architecture was proposed [59]. In C/S architecture, the distributed collaborative design system uses Internet and Web technology as a medium to share data, information and knowledge [60,61], and with a Web-based prob- lem-solving environment distributed modeling and simulation using CAE technologies was reported [62]. A comprehensive review of some Web-based tools and systems can be found in [21,63]. In most Web-based collaborative design systems, Java and CORBA are used to develop them [64–66], some others are developed using Common Lisp (WWDL [67]), Prolog (Web CADET [68]), Acti- veX [69,70] and VRML [66,67]. The translation of terminology among disciplines, locating/ providing engineering analysis ser- vices, virtual prototyping services, and project management may be coordinated [71–73]. P2P architecture supports the sharing and manipulation of ser- vices like Inventor collaborative toolTM [74–76] and RDF [77]. P2P architecture is widely used in agent-based system [78–81], such as PACT [82], SHARE [83], SINE [84], DIDE [85], ICM [86], A-Design [87], system interoperability [88], process coordination [89], knowledge collaboration [90], and con?ict management [91]. A distributed collaborative design framework is presented with a hy- brid of grid and P2P technology [92]. However, in a distributed environment, an agent system typi- cally has some pitfalls: lack of scalability, robustness and security [93]. Hence, integration of Web and agent technologies for collab- orative design has been carried out [94–100]. A comprehensive review of the R&D literature on computer supported collaborative design (CSCD) can be found in [101], from the pre-CSCD technologies of the 1980s to today’s state-of-the-art CSCD. 2.2. Grid computing The term ‘‘Grid” was created to express a proposed distributed ‘‘cyberinfrastructure” for advanced science and engineering [102], which is now understood to refer to technologies and infrastruc- ture that enable coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations [103]. The Globus Toolkit is a community-based, open-architecture, open- source set of services and software libraries that support Grids and Grid applications [102,104], and considerable progress has since been made on the construction using the Globus Toolkit [105–110]. The term ‘‘grid computing” originated in the early 1990s as a metaphor for accessing computer power as easy as an electric power grid [111]. Grid computing stands for the new kind of sys- tems that combine heterogeneous computational resources con- nected by the Internet and make them available to a wide user community [112]. Today there are many de?nitions of grid com- puting with a varying focus on architectures, resource manage- ment, access, virtualization, provisioning, and sharing between heterogeneous computer domains. Thus, diverse computer re- sources across different administrative domains form a grid for the shared and coordinated use of resources in dynamic, distrib- uted, and virtual computing organizations [113]. Therefore, the grid requires a platform that describes some sort of framework to allow software to run utilizing virtual organizations. These orga- nizations are dynamic subsets of departmental grids, enterprise grids, and global grids, which allow programs to use shared re- sources—collaborative federations. Software users typically install and run programs on a local ma- chine in a traditional computing environment, which requires developers to create and maintain versions of their software for the different platforms and update their local installations as the software is updated [114]. But applications, systems and other computing resources are abstracted into services in a grid comput- ing environment [113], which allows users to invoke services on local or remote hosts without concerning themselves with the de- tails of how such services are implemented. Any computational grid must include the capability to invoke codes remotely, and facilitate resource sharing, which can also reduce the cost of com- puting resources. The Fusion Grid Monitor (FGM) is used to monitor the status of services as well as the status of individual code runs [115]. 2.3. SOOA A service is a well de?ned, coarse-grained, discoverable, and self-contained software entity that interacts with applications and other services through a loosely coupled, synchronous or asyn- chronous, message-based communication model [116–118]. Ser- vice-oriented architecture (SOA) is a collection of services with well-de?ned interfaces, implementation and a shared communica- tions model [31,119], and is also an emerging approach that ad- dresses the requirements of loosely coupled, standards-based, and protocol-independent distributed computing [33]. Nowadays SOA becomes the leading architectural approach to most Grid developments. Different from the client–server architec- ture, which separates a client from a server, SOA introduces a third component, a service registry, using which service providers and requestors are made available as independent service components that can be accessed without a priori knowledge of their underly- ing platform or implementation. In SOA, the service provider de- ploys a service on the network, publishes its available service to one or more registries. Then the service registries intercept these announcements and add published services. By sending queries to registries, the service requestor looks up a service, makes selec- tion from the available services, and binds and executes the service. Discovery and join protocols are used between the service pro- viders and requestors to locate registries and then publish or ac- quire services on the network. According to whether the communication protocol between service requestor and service provider is a ?xed standard protocol or not, SOA can be divided into two different major types: service protocol oriented architectures (SPOA) and SOOA [37,120]. In SPOA, the communication protocol is ?xed and generic stan- dard (e.g., SOAP in Web/Globus services, IIOP in CORBA [121]) and known beforehand by the provider and requestor. A requestor can use this protocol and a service description obtained from a service registry to create a proxy for binding to the service provider and for remote communication over the