1 shows the generalprocedure of mould design [1]. It can be seen that howinterrelated the conditions are and which boundary andsecondary conditions have to be met by the main functions.A mould design project normally starts with economicconsiderations, namely the question of how many parts canand should be produced in one mould in one shot in orderto meet the delivery date and other requirements. This isfollowed by consideration of the arrangements of thecavities in the mould frame, which might directly includethoughts on the ease of ejection and subsequently, theconnection between mouldings and runners and partquality (number, position and shape of gates). The feedsystem accommodates the molten plastic material comingfrom the injection nozzle of the moulding machine anddistributes it into each cavity. To remove the heat from themoulding, it is necessary to provide the mould with acooling system. After the moulding has solidified andcooled down, it has to be removed from the mould by theejection system. Mould is normally constructed by stackingseveral metal plates to form a rigid body. It has to housevarious mould components in correct positions for theproper functioning of the mould. Mould constructionnormally involves the selection of mould bases andstandard mould parts. For complicated plastic parts, someother mechanisms such as slides, unscrewing device, etc.,might also be involved in the whole mould structure.3. Related researchA number of research activities have been carried out onmould design and its related field over the years usingcomputer-aided techniques. These research activities rangefrom studying specific areas of mould design to investigat-ing mould design as a whole integrated system. They canbroadly be classified into three areas: the functional andinitial mould designs; the algorithms to automate mouldgeneration; and system development of mould design.Hui and Tan [2] presented a heuristic search approachbased on sweep operations to develop automated moulddesign systems for determining parting direction, partingline, side core, etc. Huang et al. [3] used solid modellingtechniques to build mould plates and library of standardmould components. Chen and Liu [4] have developed acost model, which depicts the relationships between costfactors and product development activities as well as theirrelationships with product geometry, for cost-effectivenessdesign for injection moulding. Fu et al. [5] proposed anefficient methodology which systematically presents theundercut feature definition, classification, undercut featureparameters, and the recognition criteria of all types ofundercut features for undercut feature recognition in aninjection mould design system. Chen et al. [6] presented amethod for determination of parting direction based ondexel model and fuzzy decision-making. Li JC [7] presenteda neural network approach for modelling and optimizationof injection mould gate parameters. Li CL [8] used afeature-based approach to design the cooling system ofinjection moulds.
Chung and Lee [9] proposed a frame-work of collaborative design environment for injectionmoulding in which geographically distributed, multi-disciplinary designers can collaborate with one another.Ma et al. [10] described the development of a standardcomponent library for plastic injection mould design usingan object-oriented approach. Low and Lee [11] introduceda methodology of providing the initial design in 3D solidmodel instead of 2D drawings using the standardizationmethod. Ashaab et al. [12] described a supporting plasticengineering development system to facilitate the sharing ofinjection moulding information and knowledge betweeninterested parties via the Internet for the collaborativedesign of injection mould. Li et al. [13] used graph-basedtechnique and developed a heuristic search algorithm toautomate the layout design of plastic injection mouldcooling system. Researchers have started to adopt a knowledge-basedapproach to solve the injection moulding and mould designproblems in recent years. DTMOULD-1 [14] is a KBS forof injection moulds cost estimation. EIMPPLAN-1 [15]incorporated mouldability considerations into part designsand addressed the conceptual design development ofinjection-moulded parts. CADFEED [16] was developedfor injection mould design. They are, however, limited tospecific design areas or simple parts, and are not matureand practical enough to cover general mould design.Bozdana and Eyercioglu [17] developed an expert systemfor the determination of injection moulding parameters ofthermoplastic materials. Chan et al. [18] presented the basicstructure of a CAD knowledge-based assisted injectionmould design system which covers both the mould designprocess and mould knowledge management.From the above review it can be seen that most of theprevious work consider only certain aspects of the totaldesign and some of them are too theoretical to be appliedfor practical mould design which involves a substantial practical knowledge component about functions andstructure of a mould, human heuristic knowledge andempirical type of knowledge. The KB system has demon-strated great potential to assist the designer to interact witha CAD system for conceptualized design as well as the finalengineering design of a mould by using engineering rules ofthumb with extensive analytical means. In addition, there isstill relatively few research works on Internet-basedintelligent system for injection mould design.4. Architecture of the Internet-based mould design system4.1. OverviewAt the present time, most CAD systems provide only thegeometric modelling functions which facilitate the draftingoperations of mould design, and do not provide moulddesigners with necessary knowledge to develop good moulddesigns. Conventional computer-aided engineeringpackages are usually good at data processing for informa-tion-intensive problems or at number manipulation forformulation-intensive problems. The former comprehendsthe computer-aided drafting and graphics, and datareduction and transformation; while the latter involvesnumerical (or mathematical) modelling and analysis.However, in design problem, especially in mould designwhich involves a substantial practical knowledge compo-nent about functions and structure of a mould, humanheuristic knowledge and empirical type of knowledge areneeded in addition to the information-intensive andformulation-intensive knowledge. Therefore, conventionalcomputer-aided design technology is unsuitable for proces-sing heuristic and empirical type of knowledge which iscritical in the mould design problems. In general, the majoradvantage of the KB system for mould design overconventional computer-aided design systems is the explicitrepresentation and manipulation of a body of knowledge,representing the human expertise.An Internet-based mould design system using Internettechnology and knowledge-based approach can provide aneffective and feasible tool to aid the collaborative develop-ment of injection moulds in the small- and medium-sizedenterprises to satisfy the stringent requirements of nowadayscompetitive global market. Moreover, there is a consistentneed to check and update the information to ensure that thedata is accurate and updated as mould manufacturers andstandard components suppliers continually improve andupgrade their equipments and processes. By taking fulladvantage of the fast evolving computer network andinformation technologies, an Internet-based mould designsystem could have the ability to continually update andupgrade the large amount of information related to injectionmould design in a prompt and convenient manner.4.2. System architectureBased on the practical mould design procedures and theissues to be considered in this stage, the architecture ofan Internet-based intelligent injection mould designsystem is proposed as shown in Fig. 2. 互联网的注射模具智能设计系统外文文献和中文翻译(2):http://www.youerw.com/fanyi/lunwen_34039.html