TO STUDY OF CAD CAM INTEGRATION

1. Introduction

Industrial world has witnessed significant improvements in

product design and manufacturing since the advent of computer-aided

design (CAD) and computer-aided manufacturing (CAM)

technologies. Although CAD and CAM have been significantly

developed over the last three decades, they have traditionally been

treated as separate activities. Many designers use CAD with little

understanding of CAM. This sometimes results in design of nonmachinable

components or use of expensive tools and difficult

operations to machine non-crucial geometries. In many cases, design

must be modified several times, resulting in increased machining lead

times and cost. Therefore, great savings in machining times and costs

can be achieved if designers can solve machining problems of the

products at the design stage. This can only be achieved through the

use of fully integrated CAD/CAM systems.

The need to integrate CAD and CAM has long been recognised

and many systems have been developed. Kalta and Davies [1]

developed an IGES pre-processor to integrate CAD and CAPP for

turned components. In a different work, a rule-based system was

developed for converting engineering drawings into numerically

controlled machine programs for two dimensional punch objects

[2]. Lye and Yeo [3] and Thakar et al. [4] developed integrated

CAD/CAM systems for design and manufacture of turned

components. Bidanda and Billo [5] reported the development of

a system for generation of NC programs for producing countersink

cutting tools.

Feature recognition and feature based design have been used as

by most researchers to bridge the gap between CAD and CAM.

Many used feature extraction method. Examples are rule based [6]

and syntactic pattern based [7] feature recognition. Development of

domain specific feature modeling systems has also been reported.

Examples include OMEGA [8], PRISPLAN [9], CFACA [10],

VITool [11], IFPP [12]. OMEGA uses production rules to define

operation sequences that are then subsequently grouped into set-ups

based on the nature of the tolerances. PRISPLAN, CFADA and

IFPP focus on generating NC programs for machining centers.

Developments of such systems have also been reported by

others [13-19]. However, a fully CAD/CAM integration is not yet

achieved. The process of integration of CAD and CAM has been

relatively slow in comparison with the developments made in

each of these technologies. Researchers believe that the slowness

of the integration over the recent decades is essentially

attributable to the incompatibility of database formats and the lack

of common languages. The developed systems can only be

considered as islands of integration and still there is a missing

link fully integrating these two technologies.

In most of the systems developed, user still must determine

crucial manufacturing parameters such as cutting tools, cutting

speeds, feed rates, cutting depths, etc., requiring expertise and

considerable amount of time. In addition, contributions made to

integrate CAD and CAM systems for milling operation are very

limited, while this operation forms a considerable amount of

machining operations. This paper describes development of an

integrated CAD/CAM system for milling operations. It has been

proven that the use of this system would help reduce machining

leadtimes and cost, and it is believed that this system goes one step

closer towards achieving fully integrated CAD/CAM systems.

2. System developed

An integrated CAD/CAM system for milling operations has been

developed which helps designers to solve machining problems at the

design stage. A methodology has been employed which provides all

necessary information for machining products automatically. Use of

these system results in reduced machining leadtimes and cost through

(1) designing machinable components; (2) using available cutting

tools; (3) improving machining efficiency. The system is menu driven

with a user friendly interface. As shown in Figure 1, the system is

composed of following components: design module, expert system,

machining sequence planning module, optimisation module,

manufacturing module, executive program.

3. Design module

The design module is developed based on feature-based

design approach. It is implemented with AutoLISP programming

language and uses AutoCAD’s solid modeller to create product

models. This module has a feature library that includes two

groups of features, pre-defined and user-defined. Pre-defined

features are divided into protrusions and depressions. Protrusions

include cubic, rectangular, circular, polygonal, elliptical, semicircular,

semi-polygonal and semi-elliptical features. Depressions

include pockets, holes, slots and steps. There could be different

numbers and forms of islands inside pockets. Holes are composed

of normal, counterbore, countersink and flat bottom ones. Slots

have different shapes such as normal, round bottom, U, T and V

shapes. Steps are divided into normal, round bottom and notches.

User-defined features are complex features of common use for a

manufacturing firm that can be created by any combination of

pre-defined features, restored in the feature library, and used

whenever required.

The relationship between different features is represented by

use of a feature tree. This tree represents parent-child relationship

between features where a feature forms its root and other features

form its branches and leaves. A feature can be considered as the

child of another when at least one of its imaginary faces is

coincident with a real face of the parent feature. Feature A becomes

a parent for feature B when there is no tool access to machine

feature B before machining feature A. Each parent feature can have

an unlimited number of child features and all features must have a

parent feature except the root feature. The feature tree can be

organised only when the product design is completed.

The design module uses different co-ordinate frames and

datum points for placement of features in the design model. These

can be defined as:

�� Global co-ordinate frame (GCF) is the original or world coordinate

frame of the component. There is only one GCF for

each component.

�� Local co-ordinate frame (LCF) is the user defined co-ordinate

frame to design different features easier. There may exist no

or several LCFs in the design and each LCF can be defined

with respect to the GCF or another LCF.

�� Datum point (DP) is the reference point of the feature and

there is one DP for each feature. DP determines the position

of the feature with respect to the current co-ordinate frame

that can be either GCF or LCF. In other words, DP can be

considered as a sub-LCF that determines the location of a

feature with respect to the current co-ordinate frame having

the same orientation.

When different co-ordinate frames are employed, the system

must determine the position of each feature with respect to GCF when

it is defined in a LCF and vice versa. It is known that one co-ordinate

frame must be defined with respect to another. A new co-ordinate

frame can be described with a matrix of vectors. In addition, the

sequence of rotations and translations required to relocate from one

co-ordinate frame to another must be described; this can be done with

a transformation matrix. A matrix can be used to represent translation

and rotation because a sequence of translations and rotations can be

combined to produce a complex relocation more easily with matrix

multiplication than with vector addition. For example, if p1 represents

the DP of a feature with respect to the LCF, it can be represented by

p with respect to GCF where:

p = A p1 (when:

p1 = A-1 p (3)

where A is the general transformation matrix and A-1 represents

the inverse of A.

It is important to note that transformation matrix is composed

of four vectors; the first three vectors represent the orientation of

the three axes and the last vector represents the translation of the

origin of the new co-ordinate frame, all with respect to the old coordinate

frame.

In order to allow inversion of the 3x4 transformation matrix,

an additional dummy row is introduced to it and additional fourth

numbers are introduced to the point vectors, which have no effect

on the matrix manipulations. Therefore they become:

When nested LCFs are defined, there will be more than one

transformation matrix. In this case, to transform a point from a nested

LCF to GCF, more than one transformation matrix should be

employed. For example, in the component shown in Figure 2, LCF1 is

defined with respect to GCF, LCF11 is defined with respect to LCF1

and the DP of counterbore hole is defined with respect to LCF11. If p1

represents the DP of counterbore hole with respect to LCF11, it can be

represented by p with respect to GCF where:

p = A1 A2 p1 (5)

where A1 and A2 represent transformation matrices to transform

a point from LCF1 to GCF and from LCF11 to LCF1, respectively.

Similarly, DP of the counterbore hole can be transformed from

GCF to LCF11where:

p1 = A1

-1 A2

-1 p (6)

4. Expert system

In most machining activities, the selection of cutting tools and

determining machining parameters are still carried out manually by

expert operators using extensive searching through catalogues and

manuals, which requires expertise and time. Therefore, automation

of these functions can be considered as essential for developing

fully integrated CAD/CAM systems. In this work, an expert system

has been developed which is capable of performing the abovementioned

functions. It assists user in designing machinable

components, selection of cutting tools considering available tools

resources, and determining initial machining parameters at the

design stage. To accomplish this task, the system uses information

restored in a number of databases and a knowledge base. Totally

seventeen databases have been developed. Of these, nine databases

are devoted to restoring the information of different types of cutting

tools (face mills, end mills, drills, etc.). The other eight databases

restore appropriate machining parameters (cutting speeds, feed

rates, depths of cut, cutting fluids, etc.) recommended by machining

handbooks and based on the type of operation, quality of the cutting

tool and quality of workpiece material. Use of these databases in the

expert system has many benefits such as:

�� A limited number of rules have been developed for the knowledge

base. A large number of rules should have been developed to

determine required cutting tools and machining parameters for the

operations if there were no databases linked to the system.

�� Decreased required time for developing the knowledge base.

�� Decreased running time for the expert system.

�� Readability and easy accessibility to the information restored

in the databases for updating purposes. This is specially useful

for updating the information of available cutting tools since

some of those may not be available occasionally.

The input to the expert system includes geometric

characteristics of the operation and mechanical characteristics of the

workpiece material. Its output includes appropriate cutting tools and

machining parameters for the operation such as cutting speed, feed

rate, depth of cut, and appropriate type of cutting fluid if required.

These parameters will be determined for each step of machining

operation. When a non-machinable feature is placed on the design,

the expert system uses technological expertise restored in the

knowledge-base, and issues a warning message to the user. It also

assists user to redesign the feature in order to make it machinable.

An example of non-machinable features is a pocket with sharp

corners. In this case the system warns the user and requests him/her

to define a fillet radius for the pocket no smaller than the radius of

the smallest available end mill that can machine the feature.

As shown in Figure 3, the inference engine plays a central role in

the expert system developed. It uses backward chaining where it

actively integrates expertise rules restored in the knowledge base as

if-then rules, with tooling and machining information restored in the

databases. In order to reach its goal, the inference engine

systematically searches for new values to assign to appropriate

variables that are present in the knowledge base. Therefore, it has the

capability of adding to the known store of knowledge. Inference

engine fires appropriate rules from the knowledge base and extracts

necessary information from the databases to select appropriate cutting

tools and determine machining parameters for the operations.

5. Optimisation module

For improving machining efficiency in this system, an optimisation

module has been developed which determines optimum machining

parameters for each step of machining operation. This module is

activated only if user wishes to use optimum machining parameters

rather than those recommended by handbooks. Cutting speed, feed rate

and depth of cut have the greatest effect on the success of a machining

operation. Accordingly, these parameters have been considered as

variables in developing the optimisation models. The maximum profit

rate has been selected as the default objective function in this work.

Maximum machine power, required surface finish and maximum

cutting forces have been considered as the constraints. These models

and constraints have been programmed in FORTRAN, and the

optimisation method of feasible directions has been used to solve the

problems. It is noteworthy that details of the optimisation module

developed and its models have been published elsewhere [20].

6.Machining sequence planning

module

Machining sequence planning is one of the most important

functions in process planning activities. An algorithm has been

developed for automatic machining sequence planning of the

components designed by this system. The algorithm is based on the

bilateral precedence between machining operations and generates

feasible and optimal machining sequences, reducing machining cost

and time. It puts machining operations in a definite order based on

technological and geometric considerations such that the number of

necessary tool changes becomes minimal. This algorithm has been

described elsewhere [21]

7. Manufacturing module

The system restores technological data determined by

different components in the manufacturing data file (MDF) for

use by the manufacturing module. MDF provides all necessary

data for each step of machining operations that are required for

NC program generation. An IGES file generated by the CAD

system provides the geometric data. It is noteworthy that IGES is

the most common method for data exchange in current CAD

systems. Using these data, the manufacturing module generates

required tool paths for each step of machining operation and

determines all cutter locations. User can either accept generated

tool paths or modify them. Upon confirmation of generated tool

paths, the required NC program is generated using an existing

post-processor. The NC machine to produce the product can then

use the generated program.

8. Executive program

The executive program plays a central role in the system

developed, as shown in Figure 1. The user is in touch with all

components of the system only through this program. Upon running

the system, user gets in touch with the executive program where

he/she should determine the desired job from displayed menus. The

executive program communicates with user and manages the

operations of all components of the system, activates the design

module for designing the product, and at the same time, helps the

user in designing machinable features using the expertise rules

restored in the knowledge base of the expert system. It also

communicates with the expert system for determining cutting tools

and initial machining parameters. Communication with the

machining sequence planning module and optimisation module can

be mentioned as other functions of the executive program. It

collects data generated by all components of the system and restores

them in a file using a specific format called ‘design representation

scheme.’ This scheme is developed based on a group of LISP codes

that represent mechanical components as a related set of features.

Each component is represented using a group of pre-defined feature

codes, together with mechanical information of the workpiece

material, and geometric and topological information of its features

in the following form:

( (mechanical_data)

(geometric_data)

(co-ordinate_frames)

(feature_tree) )

This scheme gives a complete and unambiguous representation

of the product that can be used for different purposes such as

verification of generated process plans, NC program generation,

fixture design and so on. Details of this representation scheme can

be found elsewhere [19]. Based on the information restored in this

file, the executive program generates the required MDF file to be

used by manufacturing module for NC part program generation.

9. Case study

A number of test components have been designed and

produced using the system developed. Results showed significant

improvements in machining times and costs in design and

machining of these components. For example, in producing the

sample part shown in Figures 2 and 4 machining time and cost

were reduced by 38% and 42% respectively. This resulted in

a significant improvement in the total profit rate of about 350%,

which increasing it from $0.71/min to $2.49/min.

Fig. 4. The sample part dimentioned

10. Conclusions

In this work several goals have been achieved in the line of

developing fully integrated CAD/CAM systems. Different

components required for developing such systems have been

identified and various problems that arose in the development of

these systems have been dealt with, leading to an adequate basis for

complete integration of CAD and CAM technologies. The system

developed allows simultaneous generation of all information

required to satisfy machining requirements of the design such as its

machinability and availability of the required tooling resources. It

thus integrates different areas of design and manufacturing, giving

each area an appreciation of its role in the design process. It has

been proven that use of this system results in considerable

improvements in machining efficiency, time and cost.

Although much of the work described here goes beyond the

scope of published literature, however, it should be noted that the

system developed couldn’t be considered as a complete solution to

the CAD/CAM integration problem. Further work requires

including other manufacturing activities that are considered in

concurrent engineering concept. In this direction, further integration

of the system developed with systems such as MRP, MRP II and

assembly sequence planning packages are highly desirable.