الفهرس يوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 163 |  |  |
| | | from | 163 | | |
المستخلص
CHAPTER6
CONCLUSIONS AND RECOMMENDATIONS
Many researches have been carried out on CAPP and numerous papers have been published in the last two decades. It is difficult to apply a uniform approach to automate the process planning, since it has a diverse nature. Therefore, the process planning activity can be broken down into various phases (sub activities) like; operation selection, operation sequencing, tool selection and set-up planning [33]. The number of those sub activities determines degree of divergence and capability of a process planning system.
Two primary interfaces are required in order to integrate a computer aided process planning system with CAD and CAM is:
A CAD to CAPP interface (feature recognizer) which has the ability to accept part information from CAD systems and to translate CAD language to CAPP language
A CAPP to CAM interface (CNC code generator) which has the ability to pass the process information to CAM systems
One of the main ideas behind this thesis was to establish a step to CAD/CAM integration, and to provide a home-made and domestic. Motivation of the work was mainly given to the development of a CAD/CAPP/CAM package for rotational parts that can serve as one of the cornerstones of an intelligent and integrated manufacturing system to be developed further.
Most of the systems reported in the literature are not modular and not open to any modifications. The utilization of STEP standard which can be used to translate both technological and geometrical information between CAD and CAM is rare. This work in the first place attempts to find practical solutions to those problems encountered with a process planning system.
As everyone knows a CNC system is composed of three main units; a machine tool, a controller unit, and program of instructions. If a well constructed CNC machine tool can not provide a flexible programming interface (due to poor controller unit), expected CNC advantages can not be realized. The specification of machine controller unit (MCU) which can be used as the base. For the format of part programs to be generated is therefore also an important consideration. However, machine tools are generally equipped with their standard control unit provided by MCU producers in agreement. Therefore, the company or the user has usually no chance to select the type of MCU. The MCU coming with the machine tools is to be used.
Another problem with the MCUs is that CNC industry has so far been unable to come up with a compatible coding standard. The problem is that each machine tool/controller unit is different and machine readable codes (G and M codes) running the equipment is non-standard. Each control uses different codes and functions specific to a model. Therefore, the same G word may represent different functions on different machines. Part programs generated for a specific control type can not be used for another, even if the machine is the same kind and size, because of those incompatible codes. Besides, the number of CNC words supported by the machine tool control determines the capabilities of the machine. For instance, although there are more than two hundred G words standardized, many CNC machines may not support even a few of them. The problems concerning the non-standard CNC codes can be overcome by specifying the properties of different machine tool controllers into the system and presenting a simple selection mechanism at the beginning of the process planning session.
6.1. INCLUSION OF AI
The 1990.s produced a remarkable growth in the area of Artificial Intelligence (AI) that is accepted as to be one of the most promising key technologies for the factories of the future. In recent years, numerous programs and programming packages have been developed to solve manufacturing problems and many papers have reported the successful applications of AI to engineering design and manufacturing planning. Knowledge-based systems also known as expert systems, neural networks, genetic algorithms, fuzzy logic and case-based reasoning are some of the examples of AI tools. Unfortunately these tools, for instance, their compilers or shells are not easily found. Of course, one can also implement AI strategy or tool for different purposes by writing her/his own computer program, but this is somewhat difficult and a tedious task. For example, writing a neural network algorithm for a complex problem is a difficult process including a lot of trial and error methods for establishing the structure of the network. It is also impossible to write such an algorithm without using an object-oriented programming language.
The expert systems are the most famous AI tools. They are the first attempt to simulate human behaviors and thinking process in industry. In some old books, AI is equivalent to only expert systems. Initially, one of the principal objectives of this work was to develop a fully expert system for process planning. However, the tools for building expert systems are not sufficiently developed and are difficult to apply.
Several thousand rules and huge software systems should be contained (since the process planning is very inclusive and divergent) in order to build such a structurally expert system for process planning, which are difficult to use on conventional computer systems. A partially expert system which can take out of the responsibility of the experienced manager in the shop floor is developed in this work. It has been enriched with other AI tools to optimize the ability of some process planning modules.
The basic difference between a conventional program and expert system is actually the way in which knowledge is presented and processed. The expert system consists of a knowledge data base containing explicit knowledge of a human expert in a specialized domain, and an inference engine which can access the knowledge base to come to a decision for a described problem. In principle, any algorithm and its associated data contain expert knowledge about a limited domain [3].
6.2. CONTRIBUTIONS AND CONCLUDING REMARKS
In this study, a Computer Aided Process Planning (CAPP) package for rotational parts is developed together with its two interfaces (CAD/CAPP interface and CAPP/CAM interface) to integrate CAD/CAM facilities. The package is able to recognize a model and to generate a machining program for producing the model on the machine tool. The developed package is composed of five main sections;
Section 1: A representation scheme for the geometry and topology of component and put it as features in the parts database.
Section 2: provides a modeling platform (CAD tool) with a feature recognizer (CAD/CAPP interface) based on STEP (STandard for the Exchange of Product) standard.
Section 3: includes the process planning works such as workpiece material selection, tool selection, cutting parameters selection and set-up planning.
Section 4: involves optimization of process planning events
Section 5: is the CAPP/CAM interface that consists of the generation, verification and down-loading of CNC codes.
The main objective of this research was to develop an intelligent and integrated CAD/CAM system for shop-floor use that can be used by an average operator and produce fully (globally) optimized results (process plans and part programmes). For this purpose, in this work, an attempt has been made to include the impact and potential of Case-based Reasoning techniques in process planning applications and to optimize all events (parameters) in an integrated CAD/CAM environment.
The following functions of CAD/CAM integration are fulfilled by this work;
A CAD/CAPP interface (feature recognizer) based on STEP standard,
A CAPP system, including the applications like; .workpiece size selection, tool selection, cutting parameters selection and set-up planning,
CAPP/CAM interface, including both the automatic generation and verification of CNC part programs together with tool paths,
Most important significant features of the developed system are as follows:
Extensive use of Case-based reasoning for optimization of process parameters
Process planning tasks were individually optimized
Global optimization of process plans was accomplished
Feature recognition (CAD/CAM interface) was based on STEP standard.
A design for manufacturing utility was provided
Tool selection module is capable of selecting alternative tools for machining operations
The first generation of process planning systems were stand alone with limited links to CAD and CAM, and has employed a static approach not able to meet advances in manufacturing technology. Any CAD system capable of generating standard STEP files can serve as the front end (CAD environment) to the developed CAD/CAM package.
The experience gained from this work can easily be applied to other domains of machining, i.e., for process planning of prismatic parts and sheet metals.
Recent surveys on CAD/CAM systems have revealed an increasing trend for small and medium sized industries to employ CAPP systems for their manufacturing tasks. Lower cost of these systems will probably increase this trend. The results obtained from this study can provide an additional degree of flexibility for the planning and manufacturing resources in machining industry. With the developed package, it would be possible to reduce the times spent in the planning of machining, part programming and verification, and to contribute to the success of the manufacturing industry. This will lead to increased utilization of CNC machine tools and maximization of CNC productivity which are commonly considered as ultimate goals of CIM.
6.3. RECOMMENDATIONS FOR FURTHER STUDY
The fast moving world of software and hardware developments makes it nearly impossible to predict trends in the long-term development of integrated CAD/CAM systems. However, the work presented in this thesis can be extended to the following fields;
Future research should take real time considerations into account,
An adaptive control system for cutting parameters (feed, speed, depth of cut, etc.) might be implemented. Hence, on-line optimization of cutting parameters would be possible. This will maximize CNC productivity, extend tool life and minimize the operator supervision,
Chatter is announced as one of the most critical parameters to be taken into consideration in mass production. So, the installation of on-line chatter control systems in CAD/CAM packages will increase the product quality considerably,
Quality control is a vital element in CIM. The 21st century is expected to be a Century dedicated to Quality. Therefore, integration of Computer Aided Inspection systems with other CAD/CAM applications has a higher importance,
Thinking mechanism of experienced process planners should be investigated in order to reflect their knowledge to process planning more effectively,
Further process planning works should include Reverse Engineering. CNC code generation systems can also be incorporated with the image processing technology. from a scanner or photograph image of the workpiece, the CNC part programs can automatically be generated,
CAPP should be integrated and furnished with additional CIM functions like inspection, production planning, other than CAD and CAM
CAPP systems should be affordable to the small and medium-scale manufacturing companies,
Tolerance issues should be considered in the process planning,
Future works should be placed in the concurrent engineering context,