If Number Of Inserts=1 Then
Mold Width = (Insert Width + 2)
Mold Length = (Insert Length + 2)
Mold Thickness = Insert Thickness
End If
The rules are arranged in modular fashion using a standard programming language for the sake of convenience and clarity. Each module generates a set of outputs, which would be inputs for other modules.
2.6 Testing the application
The intelligent mold design application is validated using various test cases. For each case the part information, mold information and the machine information are varied and a human expert validates the results of feeding this info into the application. Table 2 shows one such test case where the part requires two cavities and there are no inserts present.
The application gives the approximate mold dimensions, runner dimension, sprue dimension and runner length based on the cavity image dimensions and other information.
Input
Number of insets 0
Insert Length 0
Insert Width 0
Insert Thickness 0
Cavity image Length 2.02
Cavity image Width 3.28
Cavity image Depth 0.5
Waterline Di 0.25
Number of parts to be produced 1000
Time Available 6
Cycle time 26
Reject Rate 0.1
Shots per minute 2.3
Material ABS
Output Program Output
Number of cavities 2
Mold Length 10.06
Mold Width 4.02
Mold Thickness 1.125
Runner Diameter 0.109
Runner Length 1.5
Big end sprue bush diameter 0.218
Table 2. Typical test case showing program input and output.
The mold dimensions obtained are very close to a typical human expert design for the test case but do not suggest explicitly the use of a standard mold base, like a specific mold from the D-M-E mold base catalog. The mold dimensions are however useful in selecting appropriate mold base from the mold catalogs. The runner dimensions are based on the material being used and therefore are limited to a specific range of shot size.
3 Summary
This paper presents the approach adopted towards developing an intelligent mold design application that performs mold base selection based on user input. The knowledge acquisition process is done by first designing a mold base in close consultation with an industry expert and also by collecting deterministic information from hand books and data sheets. The collected information, which can be both qualitative and quantitative knowledge about the mold selection process, is represented in the form of rules arranged in different modules. Decision tables are used to reduce the size of rule base and make the rule base comprehensive in the problem domain. The application developed using the rules in different modules is then tested for its validity when it comes to selecting appropriate mold bases for plastic parts manufactured in the industry摘要:注塑成型是一个生产热塑性塑料制品最流行的制造工艺，而模具设计是这个过程的一个重要方面。模具设计需要专业的知识、技能，最重要的是拥有该领域的经验。三者缺一不可。生产塑料组件需要选择恰当的模具，如果缺乏其中之一，这种选择就得在反复试验的基础上进行。这会增加生产成本，并造成设计上的不一致。
本文介绍了智能模具设计工具的发展。该工具捕获模具设计过程的知识，并且以符合逻辑的方式将这些知识反映出来。所获得的知识将是确定性的，但模具设计过程中的信息是非确定的。一旦开发了模具设计工具，它将指导使用者根据不同客户的要求，为其塑料零件选择合适的模具 塑料注射成型智能模具设计英文文献和中文翻译(4):http://www.751com.cn/fanyi/lunwen_5339.html