【正文】 d is generally used, which is regarded as a faster and a more user friendly method. This cost estimation approach agreed quite closely with Menges and Mohren [2] that the respective mould costs are “often not puted at all” but “estimated based on experience or in parison with moulds made in the past”. Moreover, the fact that the number of order is only 5% of the number of quotations is also affecting the amount of time allowed for a proper quotation. However, the empirical cost estimation method requires mould makers to memorize the detailed mould construction steps and tool materials. Occasionally when they forget some mould construction details, underestimates may occur which will affect the mould cost significantly. Another problem is the imprecise estimation of material cost, which could be a main cost element for mould making in Mainland China because the labor cost will be paratively lower. It is hard to remember all the prices for standard ponents without reference to handbooks. Without calculating the ponents’ volume and density, it is difficult to know the weight of raw materials. Finally, it is hard to estimate the cost for a certain mould fabrication process. Hence, the estimated cost by the empirical method is quite inconsistent. Theoretically, cost estimators who calculate mould cost manually to a very subtle detail may be able to obtain a precise and consistent result. However, it would be extremely time consuming and impractical. A software package can quickly pute the material cost and machining time directly from its database with known data of the product.A good software package can also improve the manufacturer’s management. By providing machining data, time frame and actual fabrication data for reference, two things can be improved. Firstly, the efficiency of the shop floor or design department can be known and hence can be improved. Secondly, the accuracy of quotation can be improved when it does not match with the actual production model. This paper will discuss upon a prototype system developed to provide a more reliable and consistent method for quotation and better documentation, and will discuss the performance of the prototype system.2. Design consideration for puterized quotation system for injection mould manufacture (CQSIM)User friendly, low cost and accuracy are the basic requirement for a good CQSIM. Ease of updating for basic cost elements, such as labor cost, material cost and cutting data, should also be considered. This is important for maintaining the accuracy of the calculated results. Further, for a plicated mould, many data will need to be input for cost estimation. Too much data for input will be an obstacle to the users due to time consumption and the plicated procedure. It is best that data input can be limited to the actual size of inserts, machining method required, work size, quantity, special requirement (. subcontract) and other necessary information (. name of the client) for the mould. Suitable sizes of standard raw materials for various given insert sizes should also be considered. The final size should be large enough for preparation works such as roughing. From the volume and density of the part, and the unit price in the predefined database, its cost can be calculated. The choice of machining methods should be as free as possible. This is because different mould making processes can achieve the same product shape and finish. The choice of methods and cutting tools may be constrained by the facility of a manufacturer, and the skill of the mould maker. All mon machining methods, such as milling and CNC milling, must be provided for the user. The machining details set for a chosen process should match with actual machining conditions as far as possible. Basic activities can be found as ‘primitives’ to calculate the leadtime for a ponent in such plex conditions. Activities that can be classified as ‘primitives’ include: setup and detaching time, drawing study time before machining, roughing time, semifinishing time, finishing time, machine cleaning time (to reduce error), measuring time, deburring time. Once the time of these primitives for individual machines is defined, they can be bined to form a synthetic data table [3]. In particular, the capacity of a cutting tool would vary from different types of cutting and materials. For example, rough cutting or hard materials yield a shorter depthofcut than do finish cutting or a soft material. Moreover, the depth to be cut for an insert may not be the same as the maximum depth ofcut capacity of the cutter. That is, the needed depthofcut may be shorter than the maximum capacity of the cutter or much deeper than the maximum capacity of the cutter. For the same type of machines, by adding up these setup times to the machining time to finish the task, the time to finish a ponent can be found. Some machining tools are required for machining processes in mould making. For example, electrodes for electrical discharge machining (EDM), and press tools and/or casting models for beryllium–copper (BeCu) mould. The raw material size and cost for each electrode or model should be calculated automatically. Of course, it would be best for the machining method and quantity to be decided automatically to reduce the amount of data input. As regards mould structure, some structures should be considered and predefined. For example, a suitable quantity with a suitable size of screws can be assigned to an insert of a certain size. Another example, a wedge and its fixing screws, angle pins, stopping mechanism and springs should be assigned to a slide. In this way addition material cost and the machining time for these assumed ponents can be calculated. Generally, a mould can be made, roughly, through the procedures is shown in Fig. 1.NoMill and Grind Raw Material to Final Size (Block)Is All Inserts Done？YesPocketing For The BlockAssembly All Components To The Mould