|Title:||Case based reasoning system for semiconductor encapsulation mold design|
Plastics in packaging
Expert systems (Computer science) -- Industrial applications
Hong Kong Polytechnic University -- Dissertations
Department of Manufacturing Engineering
|Pages:||xvi, 168,  leaves : ill. ; 30 cm|
|Abstract:||The microelectronics industry continues to grow rapidly in size and importance. The industry has already reached the size of other major industries with sales of product and equipment totaling billions of dollars a year. Among all the options available for semiconductor assembly, plastic packaging by using epoxy based encapsulation process is less expensive and account for approximately 80% of the worldwide packaging share and this percentage is kept increasing. Plastic packaging utilizes a metal leadframe to mount the silicon chip and to provide a fanout pattern of leads to the pins of the package. During the encapsulation process, the chip bonded leadframe are loaded into a heated multi-cavity molding tool and thermoset material is then transferred into it. The molded plastic is then cured to protect the chip and leadframe from physical damage and contamination. This process involves a lot of elements working together in order to produce a quality product. Among all these elements, the molding tool and operating parameters plays a vital role. They vary from product to product. Hence, whenever a new IC package is needed, a new mold has to be built and a new set of molding parameters has to be generated. Owing to the increasingly high turnover cycle of this industry (as can be seen from the rate that new mobile phones and computers were introduced), design/building and commissioning of a new tool has to be done quickly. Although it is a very fast growing industry, it has not attracted the level of research and development effort commensurate with its value. Hence, a lot of problems associate with the encapsulation process have not yet got commercially available solution. It is a general wish of this industry that a Computer Aided Design (CAD) tool is ready for improving the design quality of semiconductor transfer molding and reduce the risk as well as time involved during its setup and commissioning. It is the goal of this project to develop such a system to aid IC transfer mold design, analysis and setup. Key features of the tool include: (1) Automating the design of the key elements of the mold, like runner and gate, based on the design requirements and constraints; (2) Proposing parameters for optimizing the performance of the mold, based on the suggested design data and standard practice; and (3) Improving its own design skill through a learning process. The development of the software tool is based on the Cased-Based Reasoning (CBR) paradigm. The underlying principle of CBR is the idea to remember solutions to already known problems for their reuse during novel problem solving. The main characteristics that attract an application in this project is that CBR does not need an explicit domain model and hence elicitation becomes a task of gathering case histories only. This is very important as there is limited research data and information about the physical model of the melt, flow and cross-linking behavior of thermoset epoxy material inside a heated metallic enclosure. On the other hand, a lot of good and bad cases do available for examination and reference. In this project, the CBR system was developed by following its operation steps. (1) Collect successful initial cases; (2) Build an indexing system for identifying cases; (3) Retrieval a case from the case base that is similar to the problem on hand; (4) Modify the retrieved case to match the exact characteristics of the problem; (5) Retain the new solution as part of the case base. Various techniques had been used and considerations had been given to turn these conceptual steps into real and operational system. Cases were collected and studied in order to determine the most appropriate case structure and indexing system. Fuzzy indexing and weighted similarity analysis had been used for case retrieval. It is also supplemented by the use of enhancement and weakening factors for final case selection. Different models had been studied and used for case adaptation in order to arrive in the most reliable results. After completion of the system, it is subjected to several tests. It had been observed that by retrieving a case which is most similar to the current problem from a case base and modifying it, the developed CBR system can provide design solution similar to that from a multi-disciplined approach.|
|Rights:||All rights reserved|
Files in This Item:
|b14853942.pdf||For All Users (off-campus access for PolyU Staff & Students only)||5.35 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: