You are on page 1of 178

COMPUTER-AIDED MOULD DESIGN MODIFICATION AND TOOL PATH REGENERATION

ZHANG LIPING
(B. ENG.)

A THESIS SUBMITTED FOR THE DEGREE OF PHILOSOPHY DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE


2004

ACKNOWLEDGMENTS

First and foremost I would like to wholeheartedly thank my supervisors, Professor Andrew Nee Yeh Ching and Associate Professor Jerry Fuh Ying Hsi for their support (morally and academically) and for giving me invaluable guidance, suggestions, encouragement and patience throughout the duration of my graduate study in National University of Singapore. I appreciate very much all they have done for me and I will gratefully remember all forever.

Sincere appreciation is expressed to Associate Professor Loh Han Tong and Associate Professor Zhang Yun Feng for their kind words of advice during my research.

Thanks are conveyed to National University of Singapore for providing me with the research scholarship and to Department of Mechanical Engineering and

CAD/CAM/CAE Center for the use of the facilities.

Finally, I wish to express my deepest thanks to some very special people in my life: my husband, Dr. Ding Xiaoming, thank to his love, patience, understanding and great support during my graduate study, for always encouraging me to do my best, for never losing faith in my abilities; my son, Ding Changzhao, of his smiles, laughter and understanding; my dear parents and in-laws, for their continuous concern, confidence and moral support. This thesis is specially dedicated to them.

There are many others who have indirectly contributed to my research and although I am not mentioning any name here, I am grateful to all of them.

II

TABLE OF CONTENTS

Acknowledgements.I Table of ContentsIII NomenclatureVIII List of Figures..IX Summary...XIII

Chapter 1 Introduction...1
1.1 Product development processes...3 1.2 Mould design and modification...6 1.2.1 1.2.2 1.2.3 Mould structure and subsystems..6 Mould design...8 Mould design modification10

1.3 Tool path generation and regeneration11 1.3.1 1.3.2 1.3.3 CNC machines and cutters.11 Tool path generation..12 Tool path regeneration...13

1.4 Research objectives.14 1.5 Outline of the thesis....15

Chapter 2 Literature Review17


2.1 Mould design..18 2.1.1 Parting direction and parting line design.18

III

2.1.2 2.1.3

Insert design.21 Mould system and subsystem design..23

2.2 Tool path generation.24 2.2.1 2.2.2 CC-point method.25 CL-point method.29

Chapter 3 Mathematical Background of Curves and Surfaces.34


3.1 Curve parameters..34 3.2 Surface parameters...36 3.3 Curves on surfaces.....40 3.4 Offset surface.41

Chapter 4 Mould Design Modification42


4.1 Basic concepts and principles of mould design modification...43 4.1.1 4.1.2 Concepts and assumptions of mould design modification44 Principles of the insert and pocket design.45

4.2 Architecture of the mould design modification system48 4.3 Identify the solid bodies of material to be added and removed...50 4.4 Identify the mould insert that material needs to be added to or removed from51 4.5 Remove material from the mould...53 4.6 Add material to the mould...53 4.6.1 4.6.2 Create pocket and insert.53 Detect interference.60

4.7 Illustrative examples64

IV

Chapter 5 Tool Path Regeneration with the CL-point Method70


5.1 Terms and notations.71 5.2 Basic concepts of tool path regeneration73 5.3 Methodology for detecting affected CL-points..74 5.4 Tool path regeneration algorithms for CL-point method.78 5.4.1 5.4.2 Identifying and replacing affected CL-points80 Calculating scallop height values and adding new CL-points to the modified region86 5.4.2.1 Machining scallop height and step-over size.86 5.4.2.2 Algorithm for checking the scallop height value and adding new CL-points90 5.5 Illustrative examples.93

Chapter 6 Tool Path Regeneration with the CC-point Method.102


6.1 Terms and basic concepts..103 6.2 Methodology of identifying the affected CL-points.104 6.3 Tool path regeneration algorithms for the CC-point method108 6.3.1 6.3.2 Identifying and replacing the affected CL-points109 Adding new CL-points for the modified region..116 6.3.2.1 Calculating the CL-points with the given tolerance116 6.3.2.2 Calculating scallop height and adding new CL-points for the modified region..119 6.3.2.3 Detecting and removing gouging CL-points...119 6.4 Illustrative examples...119

Chapter 7 System Implementation and Case Studies..128


7.1 The computer-aided mould design modification and tool path regeneration system..128 7.1.1 7.1.2 The platform.128 The architecture of computer-aided mould design modification and tool path regeneration system...130 7.1.3 7.1.4 7.1.5 The user interface.133 The mould design modification module...133 The tool path generation and regeneration module..134

7.2 Case studies...136 7.2.1 7.2.2 Case 1136 Case 2142

Chapter 8 Conclusions and Recommendations149


8.1 Conclusions.149 8.1.1 8.1.2 Mould design modification...150 Tool path regeneration..150

8.2 Recommendations..151 8.2.1 8.2.2 8.2.3 Mould design modification...151 Tool path generation.152 Tool path regeneration..152

List of Publications from this Study.154

VI

References...155

VII

NOMENCLATURE

2D 3D C Co CAD CAE CAM CAD/CAM CAD/CAM/CAE CC CC-point CL-point CLSF CNC EDM EC MRR UFUN UG

Two Dimensional Three Dimensional The Boundary of Mould Modification Region The Boundary of Affected CL-points Computer-Aided Design Computer-Aided Engineering Computer-Aided Manufacturing Computer-Aided Design and Manufacturing Computer-Aided Design/Manufacturing/Engineering Cutter Contact Cutter Contact point Cutter Location point Cutter Location Source File Computer Numerical Control Electrical Discharge Machining Engineering Change Metal Removal Rate User-Function Unigraphics

VIII

List of Figures

Figure 1.1 Figure 1.2 Figure 1.3 Figure 2.1 Figure 2.2 Figure 2.3 Figure 3.1 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Figure 4.8 Figure 4.9

Plastic part development processes..4 A typical injection mould structure..7 A general mould design process......9 Parting lines and parting surfaces..19 The iso-parametric method26 The inverse tool offset method...31 The Frenet frame....35 Different shapes of inserts and pockets..47 Framework of the mould design modification system...49 Material to be added to or removed from designed mould50 Cavity and core are affected by product design modification51 Designed insert and pocket....57 An ejector hole interferes with the pocket.58 Fix the insert with screw or welding process.59 Lifter and cooling holes.61 Minimum distance between pocket and other holes..63

Figure 4.10a Old product file..65 Figure 4.10b New product file....65 Figure 4.10c Old cavity...66 Figure 4.10d Modified cavity..67 Figure 4.10e Old core.....68 Figure 4.10f Modified core.69 Figure 5.1 Cutter contact (CC) and cutter location (CL) points..72

IX

Figure 5.2 Figure 5.3 Figure 5.4 Figure 5.5 Figure 5.6 Figure 5.7 Figure 5.8 Figure 5.9 Figure 5.10 Figure 5.11

Surfaces A and its offset Ao...75 Boundaries of A and TA.77 Surfaces A and A..78 Extreme points and affected points81 Difference between z and z...85 Overcut due to the wrong tool path direction86 Machining scallop height...87 Calculation of tool path interval....88 Circles interpreted from CL and CC-points..89 Changes of tool path direction due to an odd number of tool path lines91

Figure 5.12a Part surface of a workpiece before modification...93 Figure 5.12b Tool paths of the workpiece before modification..94 Figure 5.12c Part surface of work-piece after modification94 Figure 5.12d Regenerated tool paths with replaced points and added lines after modification...95 Figure 5.12e Tool paths before and after modification with replaced points and added lines....96 Figure 5.13a Part surface of bezel mould before modification...98 Figure 5.13b Tool paths of bezel mould before modification.98 Figure 5.13c Part surface of bezel mould after modification..99 Figure 5.13d Regenerated tool paths with replaced points and added lines after modification.100 Figure 5.13e Tool paths with replaced points and added lines before and after modification.101

Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Figure 6.6 Figure 6.7a Figure 6.7b Figure 6.7c Figure 6.7d

CC-point, CL-point, section plane, CC-curve and tool path row104 Curves on surface r and offset surface ro.105 Removing undercut and interference from surface A..107 Identifying affected CL-points ....110 Step length L117 Convex gouging..118 Part surface of a workpiece before design modification..120 Tool paths of the workpiece before design modification.120 Part surface of work-piece after design modification.121 Regenerated tool paths with replaced points and added lines after design modification.122

Figure 6.7e Figure 6.8a Figure 6.8b Figure 6.8c Figure 6.8d

Tool paths with added lines before and after design modification..123 Part surface of T-cover mould before design modification.124 Tool paths of T-cover mould before design modification...124 Part surface of T-cover mould after design modification125 Regenerated tool paths with replaced points after design

modification.126 Figure 6.8e Figure 7.1 Tool paths of T-cover before and after design modification127 Framework of the mould modification and tool path regeneration system...131 Figure 7.2 Figure 7.3 Figure 7.4 Figure 7.5 New product selection..133 Input mould machining status..134 Item selection..135 Input machining parameters.136

XI

Figure 7.6a Figure 7.6b Figure 7.6c Figure 7.6d Figure 7.6e Figure 7.6f Figure 7.7a Figure 7.7b

The old plastic part of a dairy clear door.137 The new plastic part of a dairy clear door137 The original cavity of a dairy clear door..138 The modified cavity of a dairy clear door...138 The original core of a dairy clear door138 The modified core of a dairy clear door...139 Tool paths of the workpiece before modification140 Regenerated tool paths with replaced points and added lines after mould modification.141

Figure 7.7c

Tool paths before and after modification with replaced points and added lines..142

Figure 7.8a Figure 7.8b Figure 7.8c Figure 7.8d Figure 7.8e Figure 7.8f Figure 7.9a Figure 7.9b

The old plastic part of a riser...143 The new plastic part of a riser..143 The original cavity of a riser144 The modified cavity of a riser..144 The original core of a riser...144 The modified core of a riser.145 Tool paths of the workpiece before modification146 Regenerated tool paths with replaced points and added lines after mould modification.147

Figure 7.9c

Tool paths before and after modification with replaced points and added lines..148

XII

SUMMARY

To shorten the product development time, a new plastic part is usually sent to the mould manufacturing company before it is finalized. The design of the plastic part may need to be changed many times during the mould manufacturing processes. It may be necessary to modify the mould design and to regenerate the affected tool paths many times. However, the existing CAD systems cannot automatically modify the mould design for non-parametric moulds. In addition, the design of the plastic part may be modified at different stages of mould manufacturing process, while the mould design modification method is related to the mould machining status. Moreover, with the existing CAM systems, no matter how small the portion of a mould is to be modified, the entire tool path that covers this region will need to be recalculated, which could take as much time as for generating a new tool path could increase the chance of NC programming errors. Consequently, the process of mould design modification and tool path regeneration is still very time-consuming. To solve these problems, this research focuses on the following aspects:

Mould Design Modification. A new mould design modification algorithm that can automatically modify the mould design for parametric and non-parametric parts according to the mould machining status has been developed in this research. The developed algorithm does not rely on the product parameters that increases with the complexity of the product and the mould. Therefore, it is not sensitive to the

XIII

complexity of the plastic part and the designed mould. Different methods have been developed to modify the mould design according to the mould machining status. When an insert needs to be designed to add extra material, the interference between the pocket to be designed and the existing holes is detected automatically. Different pockets, inserts and fasteners will be designed based on the result of interference detection.

Tool path regeneration. Four propositions have been made and proven in this research for tool path regeneration. The propositions indicate two important properties of a gouge-free tool path: 1) the affected CL-points are enclosed by the boundary of CLpoints that is corresponding to the interference-free boundary of the modified region; 2) when projected onto the XY-plane, if one CL-point is before another one, their corresponding CC-points follow the same topology. With these propositions, the affected CL-points can be efficiently identified by only noting their x- and y-values. New tool path regeneration algorithms have been developed to identify and replace the affected CL-points for 3-axis NC machining. By utilizing the unaffected CL-points, the tool path can be regenerated efficiently. Since the tool path regeneration method is related to the tool path generation method, and the tool path methods can be categorized into CC- and CL-point methods, two different tool path regeneration algorithms are developed for them respectively.

From this research, a computer-aided Mould Design Modification and Tool Path Regeneration System has been developed to modify the mould design and regenerate tool paths automatically and efficiently.

XIV

CHAPTER 1 INTRODUCTION

Thermoplastic parts, as they can provide properties of low density, superior corrosion resistance, electrical insulation and suitability for volume production, are widely used in engineering and consumer products. The most prevalent method for producing thermoplastic parts in large quantities is injection moulding, which is a highly costeffective, efficient, and precise manufacturing method. The process can be highly automated and produces almost no waste.

Facing the challenges of an increasingly more competitive market, time-to-market plays a crucial role in new product development. Technology tools such as ComputerAided Design/Manufacturing/Engineering (CAD/CAM/CAE) have been developed to help manufacturers achieve the goals of an ever-decreasing life cycle of a product from concept to market. On the other hand, as they are required to accomplish more functional and aesthetic requirements, plastic parts are becoming more complex. To meet the tighter product development schedule, a plastic part may very often be passed to the mould manufacturing company for manufacturing before it is finalized. The design of a new plastic part may be changed many times during the mould design and manufacturing processes. If the design of a plastic part has been changed, the mould design would need to be modified and the affected mould inserts re-machined.

Chapter 1

Introduction

Although some commercial CAD/CAM systems allow for automatic modification of the design for parametric parts, they cannot automatically modify the mould design for non-parametric moulds. Here, a parametric part means that geometric definitions of the design, such as dimensions, can be varied at any time in the design process by changing the parameters. In addition, it may be necessary to modify the design of the plastic part at different stages of mould manufacturing process, while the method of mould design modification is dictated by the mould machining status. However, this is not taken into consideration in the existing CAD/CAM systems. Moreover, with the existing CAM systems, no matter how small the portion of a mould is modified, the entire tool path that covers the modified region will have to be recalculated. This could take as much time as for generating a new tool path. Therefore, the process of mould design modification and tool path regeneration in NC programming is still very tedious, time-consuming and error-prone.

This research aims at developing a Computer-Aided Mould Design Modification and Tool Path Regeneration System. With this system, when the design of the plastic part is changed, the modified regions are identified and located by comparing the old and the new plastic parts. The mould design is then modified automatically based on the designed mould and the mould machining status. The affected tool path is regenerated efficiently by identifying and replacing the affected Cutter-Location (CL) points. The unaffected CL-points are reused directly to machine the modified mould, which would greatly reduce the tool path regeneration time.

This chapter introduces the background and basic concepts of the proposed research. The general process of plastic part development is summarized in the first section.

Chapter 1

Introduction

Mould design and modification processes are analyzed in the second section. This is followed by tool path generation and regeneration methods in the third section, and the objectives of the research in the fourth section.

1.1 Product development processes


As shown in Figure 1.1, the processes of a product design and manufacturing include: plastic part design, mould design, mould manufacturing, surface polishing, mould assembly, mould try-out, design and manufacturing modifications if necessary, moulding and product launch.

A plastic part is generally a component of a functional product. After the concept design of a new product is created, the plastic part is designed based on the functional and aesthetic requirements. The mould capability and manufacturability should be considered when designing the plastic part. In addition, some rules for mould design should be followed to avoid the common pitfalls, e.g., design optimal wall thickness according to the function, the plastic material and the cost; keep the wall thickness of the plastic part uniform; locate the gate so that the melt plastic enters the cavity in the thickest area and flows to the thinner areas; design radius at all corners of the part; and design optimal draft angles for removing the plastic part easily for the core and cavity.

Traditionally, plastic parts are designed with general CAD software based on the designers experience, which is time consuming and error-prone. The development of CAE systems has eliminated various trial-and-error practices and greatly streamlined the product development cycle. However, as there are many parameters that affect the quality of a moulded part, the design of the plastic part may not be optimal and the

Chapter 1 Plastic part design

Introduction

Mould design EC Design and manufacturing modification process N Mould manufacturing

Surface polishing

Mould assembly

Mould try-out

Satisfy? Y Moulding

Launch

Figure 1.1. Plastic part development processes

design may need to be modified many times before the product can be launched. For example, the design of a space mouse was changed 9 times during the mould design and manufacturing processes. Over the past decade, The Engineering Change (EC) is becoming more and more frequent. The reasons can be summarized as follows: (1) The new product design files are often passed to mould companies before it is finalized; (2) The product design engineer is less experienced; (3) There are some changes from the

Chapter 1

Introduction

related parts; (4) The feedback from the market; (5) Many new products have only six months to prove themselves in the marketplace.

After a three dimension (3D) solid model of the plastic part is designed, a prototype may be created for checking and verification. If the prototype meets the requirements, the mould can be designed and manufactured, which is generally done in another company. Otherwise, the design of the plastic part will need to be modified and verified again.

Since the plastic part and the mould are usually designed in different companies, different CAD systems may be used. In this case, the designed plastic part file will be transferred from one format to another, and the design parameters may be lost during file transformation. When the mould designer receives the plastic part, he/she will study and analyze the mould capability and then design the mould. If the design of a plastic part is found to be not suitable for moulding, it will be returned to the product designer for modification.

When the mould is designed, the material will be ordered and the mould can be machined. Computer Numerical Control (CNC) milling and Electric Discharge Machining (EDM) are the two common processes that are used to machine the mould inserts with 3D profiles. Mould inserts are most commonly machined with CNC milling machines. The EDM process is used when the part is difficult or impossible to machine with milling cutters. This research focuses on the CNC milling process.

After the mould inserts are machined, the surfaces that form the profile of the plastic

Chapter 1

Introduction

part will be polished. The mould is then assembled and tested. If the tested part satisfies the requirements, the mould can be used for mass production and the product can be launched. Otherwise, the design of the mould will be modified and the modified portion will have to be machined again. If the failure is caused by the product design, the design of the plastic part will be modified, which will in turn affect the design and manufacturing of the mould.

Although the time and cost have been reduced dramatically in most areas of the product development process, mould design and manufacturing is still the most time consuming and costly phase for plastic parts. These aspects will be discussed in the following two sections.

1.2 Mould design and modification


Mould design is an important process for a successful plastic part. Proper design of an injection mould is crucial to producing a functional plastic component. Mould design has great impact on productivity and part quality, directly affecting the efficiency and profitability of the moulding operation. About 70% of the mould manufacturing cost is decided in the mould design process.

1.2.1

Mould structure and subsystems

Injection moulding is a process that softens the plastic material with heat and forces it to flow into a closed mould. After the material cools and solidifies, a product with the specific shape is formed. A mould is what determines the shape, and in most cases, the final finish of the part. As shown in Figure 1.2, a typical injection mould system usually includes mould base, guiding and alignment, sprue and runner system, cooling

Chapter 1 Guiding & alignment Sprue & runner system

Introduction

Cavity system

Ejection system

Cooling system

Mould base

Figure 1.2 A typical injection mould structure

system, ejecting system and cavity system. Among them, the mould base is used to securely retain all the inserts in the mould. The guiding and alignment system ensures that the two mould halves (fixed half and moving half) remain in correct alignment and the mould is accurately positioned on the machine. The sprue and runner system

Chapter 1

Introduction

ensures that the impression can be filled properly and completely. It also positions and orientates the various functional parts to ensure that the whole mould assembly works properly. The cooling system makes sure that the hot material can be cooled down rapidly to a temperature, at which it solidifies sufficiently to retain the shape of the impression. When the plastic material cools down, it often shrinks onto the core insert, which makes it difficult to remove. The ejecting system provides a means to eject the moulded part from the core insert.

The cavity system forms the shape of the plastic part. It usually includes the cavity and core inserts. When there is undercut in a plastic part, sliders and lifters will be designed. Generally, a lifter is designed for internal undercut while a slider is designed for external undercut. There are usually some cooling, ejecting, lifter and screw holes in the inserts of the cavity system.

1.2.2

Mould design

Since injection moulds are of multi-functionality and have various configurations, the design of an injection mould is a complex task. In addition to satisfying the functional requirements, many other aspects, such as geometric complexities, equipment and tooling requirements, process capabilities, must also be taken into account in mould design. In order to shorten the mould design lead-time, 3D CAD systems have replaced the traditional drawing boards as a design tool in most mould manufacturing companies. Several mould design systems have been developed to automate some of the mould design processes and shorten the design time.

Figure 1.3 shows a general mould design process: Upon receiving the design of the

Chapter 1

Introduction

Designed product model

Shrinkage

Determine parting direction and parting line

Select the mould type and mould machine

Design the initial cavity layout

Design detailed layout, inserts and components

Create drawings

Order insert material and standard components

End

Figure 1.3 A general mould design process plastic part, the mould designer studies the requirements and the geometries. The plastic part is enlarged to compensate shrinkage. The shrinkage factor is the ratio of the expected reduction of the plastic part dimension as the part solidifies in the mould and cools to room temperature. The parting line and the parting surfaces are then identified and designed. Based on the geometry and dimension of the part, the ordered quantity and manufacturing cost, the injection machine and the mould type are

Chapter 1

Introduction

decided. The cavity layout and the feeding system including the runner and the gate are designed.

After the initial mould layout design, the mould design engineer proceeds to the detail design of mould inserts and components. The mould base, core, cavity, slider and lifter are designed. The ejecting and cooling systems are also designed. The accessories such as the fasteners, springs, heaters, etc. are added in last. When the mould is designed, materials and standard components are ordered and the inserts with 3D profiles are passed to the CAM department.

1.2.3

Mould design modification

When the design of a plastic part is changed, the mould design will need to be modified. Since a mould has been designed and some mould inserts may have been machined, the mould design modification method is different from the mould design method. Besides the geometrical and functional constraints, the structure of the designed mould also needs to be considered in modifying the mould design. In addition, the product and the mould design may be modified in different mould manufacturing stages. Different mould design modification methods should be applied for different mould machining status.

Some parametric-based CAD systems can automatically modify the design of simple moulds as they can preserve the design history. However, it may take very large storage space to preserve the design history. Sometimes, this makes the file very large, and the operation becomes very slow. This is a critical issue in designing complex moulds. To solve this problem, some commercial CAD/CAM systems (e.g.,

10

Chapter 1

Introduction

Unigraphics [Unigraphics 2000]) provide a function which allows the user to remove the design parameters and history. In the mould manufacturing industry, many CAD/CAM engineers do remove the design history to reduce the part file size and improve the mould design and tool path generation speeds. Therefore, design modification for complex mould parts is still time-consuming and tedious.

If the plastic part is designed with one CAD system while the mould is designed with another one, the parameters of the plastic part many be lost when the part is transferred, which makes it difficult to modify the mould design automatically.

In addition, the design of the plastic part may be changed in different mould manufacturing stages, the existing mould design modification systems cannot modify the mould design automatically according to the mould machining process status. Therefore, when a complex or non-parametric mould design needs to be changed, it is still done by the mould designer with the general CAD system based on his/her experience, which is time-consuming and error-prone.

1.3 Tool path generation and regeneration


CNC milling machines are widely used in mould manufacturing industry. Most mould manufacturing companies use CAM software to generate tool paths for NC machining. The CNC machine, the milling cutter and the tool path are three key factors that decide the mould machining accuracy and efficiency.

1.3.1

CNC machines and cutters

According to the degree of the freedom of the tool relative to the workpiece, CNC

11

Chapter 1

Introduction

milling machines can be classified into 3-, 4- and 5-axis machines. 3-axis and 5-axis CNC machines are most commonly used in mould manufacturing. A 5-axis CNC machine, with 2 additional rotating axes, can machine free-form surfaces with flat and fillet end-mills instead of ball end-mills, which can drastically reduce the machining time. However, this technology is being accepted only gradually in mould manufacturing because the programming for 5-axis CNC machines is somewhat more difficult and error-prone. Therefore, 3-axis milling machines are still most widely used in the mould manufacturing industry, which is also the emphasis of this research.

Flat, fillet and ball end-mills are most commonly used in mould machining. Since in sculpture surface machining with 3-axis milling machines, only a ball end-mill with spherical face at its end can finish the machining task, ball end-mills are widely used in the finish machining of moulds. This research will focus on ball end-mills.

1.3.2

Tool path generation

The CNC programming of a complex mould part typically consists of two general sequences: rough and finish. Rough machining removes most of the unwanted raw stock material while keeping the tool a safe distance from the parts finished surface. It is during finish machining that the cutter contacts the part surface and removes the remaining unwanted stock material.

Material Removal Rate (MRR) is the most important factor to be considered for rough machining. Machining accuracy is not the critical issue in this process. For finish machining, accuracy, surface finish and machining efficiency are the three most

12

Chapter 1

Introduction

important parameters. Minimizing the machining time with required accuracy and surface finish is the main objective in tool path generation for finish machining. During the past decades, great improvement in tool path generation has been achieved, and many tool path generation methods have been developed. Many commercial CAM systems are available. Most CAM systems (e.g., CATIA, Unigraphics and Pro/Engineer) can generate gouge-free tool paths automatically with the given cutting tool and machining parameters. However, despite of rapid increase in computer speed, the process of tool path generation is still time consuming. It may take more than two hours to generate a finish machining tool path for a complex mould insert. Since the NC codes for rough machining are generally much fewer than those for finish machining, the time for generating rough machining tool path is also usually much less than that for finish machining.

1.3.3

Tool path regeneration

When the mould design is modified, the modified mould needs to be re-machined. CNC and EDM are two common processes used to machine the modified mould. If the modified mould is to be machined with a CNC machine, the tool paths need to be regenerated.

With the existing CAM systems, no matter how small the portion of a mould insert is to be modified, the entire tool path that covers this region will need to be recalculated. As much time is needed to regenerate the tool path for the modified mould as that for generating a new tool path, which is highly unproductive and very time consuming. Sometimes, the mould needs to be modified when it is being machined. In this case,

13

Chapter 1

Introduction

the tool path should be regenerated in a very short time. Otherwise, the NC machine will be idle while waiting for the new tool path to be regenerated.

1.4 Research objectives


Many researchers have studied mould design and tool path generation in the past decades, and improvements have been achieved in automating the design and manufacturing process and increasing the efficiency. Some CAD systems can design moulds automatically and most CAM systems can generate tool paths automatically with given machining parameters. However, the following limitations are recognized in the existing work:

(1)

Some CAD systems can modify mould design automatically for parametric parts. However, it is very difficult to automatically modify mould design for complex and non-parametric parts with existing CAD systems. Moreover, the mould machining status is not considered in most CAD systems for modifying the mould design.

(2)

With the existing CAM systems, if the tool path needs to be regenerated, all CLpoints of the tool path will be recalculated. It will take as much time to regenerate an affected tool path as that for a new one.

The objective of this research is to develop a mould design modification and tool path regeneration system that can solve the above two problems. This research will focus on the following two issues:

(1)

Mould design modification. Develop a mould design modification system that

14

Chapter 1

Introduction

can automatically and efficiently modify a mould design based on the designed mould and the mould machining status. The system should be applicable to both parametric and non-parametric parts. (2) Tool path regeneration. Develop new algorithms and methodologies that can regenerate tool paths efficiently. Since the time for generating a rough machining tool path is generally short, this research will focus on finish machining. As the tool path regeneration method is related to the tool path generation method, and the tool path generation methods can be classified into CL-point method and (Cutter Contact) CC-point method, two tool path regeneration methods will be developed for them respectively.

It is assumed in this research that the product design, mould design and tool path generation are all based on 3D solid models, while an original mould has been designed and the corresponding tool paths have been generated. Since 3-axis CNC machines with ball end-mills are widely used in mould machining, this research will focus on regenerating tool paths for these configurations. In addition, the same size cutters will be used to machine both the original and the modified moulds.

1.5 Outline of the thesis


The remaining chapters of this thesis are organized as follows. Chapter 2 presents a review of the current research status in computer-aided injection mould design and tool path generation. Chapter 3 introduces the mathematical background of curves and surfaces for mould design modification and tool path regeneration. The mould design modification system is introduced in Chapter 4. The tool path regeneration algorithms for CL-point and CC-point tool path generation methods are introduced in Chapter 5

15

Chapter 1

Introduction

and Chapter 6, respectively. Chapter 7 presents the computer-aided mould design modification and tool path regeneration system. Some examples of mould design modification and tool path regeneration are also implemented in this chapter. Conclusions and future research recommendations are discussed in the last chapter.

16

CHAPTER 2 LITERATURE REVIEW

Nowadays, product designers integrate more sculptured surfaces into their product components to enhance product aesthetic and improve their designs. The increasing complexity of manufactured parts requires more advanced CAD software and effective NC programming capabilities to design moulds and generate tool paths efficiently. Mould design and the mould manufacturing are two important factors that determine the success of a mould.

During the past decades, many researchers have studied mould design and tool path generation, and numerous papers have been published in these two areas. Most reported research work on mould design has concentrated on one of the following three topics: determining the optimal ejecting direction for a plastic product, automatically generating side cores, and developing interactive CAD systems for injection mould design; while research works on tool path generation have focused on how to generate gouge-free tool paths efficiently. In this context, machining quality, efficiency and accuracy are the three key issues studied by most researchers.

Although the reported literature can hardly be used directly for mould design modification and tool path regeneration, these researchers work can provide useful inputs in these two areas. A review on mould design and tool path generation is given

17

Chapter 2 in this chapter.

Literature Review

2.1 Mould design


In general, an injection mould includes cavity system, guiding and alignment system, runner system, ejection system and cooling system. Among them, the cavity system is most important. Determining optimal parting direction and automatically generating side cores are two of the most difficult tasks in designing a cavity system. Many researchers have studied these two problems. In addition, as mould design is very tedious and time-consuming, many researchers have tried to develop systems that could design moulds automatically and efficiently. A brief literature review in these three research areas is given in the following subsections.

2.1.1

Parting direction and parting line design

Parting direction and parting line are very important for a successful mould design as they decide the number and the shape of side cores, and this decision will affect all the subsequent steps in the design of a mould. Many researchers have studied how to automatically identify the optimal parting direction and design the parting lines.

As shown in Figure 2.1, a parting direction is a direction along which a mould piece is separated from the mould assembly. One of the principles of selecting the parting direction is to minimize the undercut, where an undercut is the recess or protrusion region on a plastic part that prevents its removal from a mould along the parting direction. When the parting direction is determined, the parting line and the parting surface can be designed. A parting line is a continuous closed curve on the surface of the product part that defines the faces to be split into different mould pieces, and the

18

Chapter 2 parting surface is the contact surface of two mould pieces.

Literature Review

Parting direction Cavity

Plastic part

Parting lines

Core

Parting surface Parting direction

Figure 2.1 Parting lines and parting surfaces

Chen et al. [Chen 1993] developed a method to determine the parting direction based on the minimization of local external undercuts. Using the Gaussian and visibility maps, they presented an algorithm to obtain a set of feasible parting directions. While their approach was able to minimize the number of external undercuts, internal undercuts were not considered. Based on two levels of visibility: complete and partial, Chen et al. [Chen 1995] extended their work to parts with internal undercuts by decomposing an internal undercut feature into two portions: the separable and the internal undercuts.

19

Chapter 2

Literature Review

Nee et al. [Nee 1997] classified the undercut features into two types, namely external and internal undercuts. The external undercuts were further divided into outside external undercuts and inside external undercuts, while internal undercuts were divided into outside internal undercuts and inside internal undercuts. A group of recognition criteria for undercut features were presented in their paper. After all the potential undercuts were extracted, the optimal ejection direction was chosen based on the number of possible undercuts and their corresponding undercut volumes.

Ye et al. [Ye 2001] developed a hybrid method to recognize undercut features from moulded parts with planar, quadric and free-from surfaces. Their hybrid method took advantage of graph-based and hint-based approaches, and various undercut features, including interacting undercut features, could be recognized.

When the parting direction is determined, the parting line can be designed. Ravi and Srinivasan [Ravi 1990] introduced sectioning and silhouette methods for parting line generation. Chin and Wong [Chin 1996] presented a slicing strategy for generating the parting line. Through a recursive uneven slicing method, several parting surfaces are generated for further evaluation. Weinsten and Manoochehri [Weinsten 1997] formulated the parting line determination problem as an optimization problem. Their objective function is defined as a function of the flatness of the parting line, draw depth, number of side cores required to form the undercuts, machining complexity, etc. Majhi et al. [Majhi 1999] presented an algorithm for computing an undercut-free parting line that is as flat as possible for a convex polyhedral object.

In summary, many researchers have studied how to identify the undercut, determine

20

Chapter 2

Literature Review

the parting direction and design parting lines. Generally, if the design of a plastic part is changed, the parting direction should not be changed so as to minimize the modification of the mould structure. With the given parting direction, the possible undercut caused by the design modification can be identified with existing methods. If the design change affects the existing parting line, the affected parting line can be redesigned and the modified portion of the part can be split and united with the core and cavity inserts automatically.

2.1.2 Insert design Insert design includes the design of cavity, core, slider and lifter. These inserts form the shape of the plastic part. The cavity and core form the profile without undercut, while the slider and lifter are designed for the undercut profiles. Some researchers studied how to generate inserts automatically.

Hui and Tan [Hui 1992] proposed a four-step sweeping method to create the cavity and core: 1) Generate a solid by sweeping the plastic part in the parting direction; 2) Subtract one end of the swept solid from the first mould block; 3) Subtract the other end of the swept solid from the second mould block; and 4) Subtract the result of step 2 from that of step 3 with the mould plates in the closed position.

Shin and Lee [Shin1993] presented a procedure to recognise undercuts by checking the interference faces between a product and its core/cavity. They also discussed the generation of side cores by using the Eulers operations. For free-form surfaces represented by u, v parameters, their algorithm determines an interference face by checking the normal vectors at the points corresponding to the grid points in the

21

Chapter 2

Literature Review

parameter domain. Since it is difficult to determine the u, v parameters for trimmed free-form surfaces, this method is not suitable for designing side cores that contain trimmed surfaces.

Rosen [Rosen 1994] presented a procedure to design side cores based on the identified undercuts. For external undercuts, accessibility directions were identified and used to design side cores. For internal undercuts, form pins were constructed that accessed the undercuts through the core of an injection mould. However, this approach is only suitable for polyhedral solid models.

Zhang et al. [Zhang 2002] introduced an algorithm that could create complete lifter subassemblies. With their algorithm, the virtual core and cavity were generated first with the given parting direction without considering the undercuts. The undercuts were then identified and grouped. For each group of undercuts, the releasing direction was identified and a lifter head was designed. By attaching other standard components of the lifter to the head, a complete subassembly of the lifter was designed.

In summary, most researchers have studied how to automatically generate inserts of cavity, core, slider and lifter with the given parting direction and parting lines. However, how to modify the design of the inserts accordingly based on the modified plastic part was not considered in their research. More research is needed to automatically modify the insert design based on the modified plastic part and the mould machining process.

22

Chapter 2 2.1.3 Mould system and subsystem design

Literature Review

Even with the help of the general CAD systems, injection mould design is still a very time-consuming and tedious process, and it is highly dependent on the experience of the mould designer. To automate the mould design process and shorten the design time, some researchers studied how to build up specific CAD systems for injection mould design.

Yuan et al. [Yuan 1993] developed an integrated CAD/CAE system for injection mould design and analysis. With their system, the drawings of a plastic product were first transformed interactively into the drawings of mould impressions, the mould design was then carried out by using a group of design tools for injection moulds. The system could analyze the balance of the runner and simulate the flow process of the plastic melt.

Kruth et al. [Kruth 1997] developed a design support system (IMES/DSS) for injection mould design. The system supported the design of injection moulds through high-level functional mould objects, e.g. basic assemblies, components and features. The system managed the low-level CAD entities and allowed additional design information such as process planning information to be incorporated. The user could create or modify standard design objects and link them with a relational database.

The above two systems could only support 2D design of injection moulds. Thus, the design facilities provided by the system were actually a group of tools for editing and generating 2D drawings. To solve this problem, Lee et al [Lee 1997] developed a knowledge-based injection mould design system, which supports 3D modeling. The

23

Chapter 2

Literature Review

system contains the design libraries for mould bases and standard parts. Mould design tools such as parting line selection, parting surface generation, ejecting pin design, cooling hole design, etc., are also provided by the system. Similar knowledge-based systems were also introduced by Chan et al [Chan 2003] and Mok et al [Mok 2001].

Besides the entire mould design system, some mould design subsystems, such as the feeding system [Ong 1995, Ravi 1997], the ejection system [Wang 1996] and the cooling system [Lin 2001], have also been developed. Most of these researches focused on how to optimally and automatically design the subsystems.

In summary, some CAD/CAE systems have been developed to automate mould design processes and improve mould design quality. However, none of these systems could automatically modify the mould design according to the changed plastic profile for non-parametric parts. Research in this area is needed.

2.2 Tool path generation


After the mould is designed, the CAM engineer generates tool paths for rough and finish machining based on the designed mould parts. The purpose of rough machining is to remove excess material from a stock, while finish machining aims to accurately machine the part shape. This research focuses on tool path generation for finish machining.

According to whether the topology of CC-points or CL-points is controlled, the tool path generation can be classified into CC-point method and CL-point method. A brief review of tool path generation in these two categories will be given in the following

24

Chapter 2

Literature Review

two sub-sections. Since 3-axis CNC machines with ball end-mills are widely used in finish machining of sculptured surfaces, this review will emphasize on this configuration.

2.2.1

The CC-point method

During the past few decades, many methods have been developed for tool path generation of compound surfaces. Most of the available algorithms are based on the CC-point method. With this method, a set of CC-points are planned on the compound surface. CC-points are then offset along the surface normal vectors to compensate the effect of the cutter size. The CL-points are thus obtained, and the final tool paths are calculated by removing the interference CL-points. According to how the CC-points are planned, the CC-point tool path generation methods can be further classified into the iso-parametric method, the constant geodesic distance method, the constant scallop height method, the principal direction method, the spiral tool path method and the section curve method.

The iso-parametric method [Broomhead 1986, Loney 1987, Kuragano 1992, Yu 1996], also known as the flow-line machining method, generates tool paths along the surface constant parameter lines. With this method, the step-over size is first evaluated at each cutter contact point along the CC-path such that the scallop height is within the tolerance. The minimum step-over size within a single path will become the step-over size. By keeping one of the two parameters constant, the iso-parametric curves are formed and employed as the CC-paths. The method of iso-parametric machining takes advantage of the parametric representation of the sculptured surface. It is very easy to calculate tool paths for a parametric surface patch and can avoid the costly surface-to-

25

Chapter 2

Literature Review

surface intersection computation. Hence, it was widely used in the early CAM systems. However, since each uniform tool path interval in the parametric space between adjacent tool paths is constrained by the scallop-height requirements, the generated iso-parametric tool paths are often much denser in one surface region than others due to the non-uniform transformation between the parametric and Euclidean spaces (see Figure 2.2). This results in varying scallop-height distribution on the machined surface and non-optimal machining time [Elber 1994, Sarma 1997]. Another shortcoming of this method is the difficulty in consistently generating tool paths for the regions consisting of several trimmed surfaces on the part surface [Loney 1987, Sarma 1997].

Workpiece surface Tool path

Figure 2.2 The iso-parametric method

The constant geodesic distance method is also called the iso-distance method [Elber 1994]. With this method, the CC-points are planned on the surface curves such that the geodesic distance between every two consecutive curves is constant. The tool path is then calculated along these curves. The tool path length with this method is generally shorter than the iso-parameter method. However, the machining scallop height is not

26

Chapter 2

Literature Review

only determined by the distance between two tool paths, it is also decided by the surface curvature value. The tool path generated with this method may not be the shortest.

The tool path generation method to achieve constant scallop height was first reported by Suresh and Yang [Suresh 1994]. With this method, an initial curve, usually the edge of the surface is selected first. The CC-curves are then planned on the surface to make sure that the machining scallop height at every CC-point is constant. This method can reduce the redundant machining in the iso-parametric and iso-planar methods since the scallop height is kept constant, and the overall tool path length generated with this method is generally shorter than with other methods. The disadvantage of this method is that the selection of the initial curve has a direct impact on the total tool path length. If the initial curve is not appropriately selected, the total tool path length can even be longer than other methods in some special cases.

It was assumed in [Suresh 1994] that the corresponding swept sections on the adjacent tool paths be coplanar. This assumption had caused inaccuracy in the calculation of tool path intervals and compromised the generation of optimal tool paths. Sarma and Dutta [Sarma 1997] improved the method of calculating the offset curve on the surface by using swept sections along the tool paths to calculate the tool path intervals. Nevertheless, the derivation by Sarma and Dutta [Sarma 1997] assumed that the undetermined swept sections of the following tool path be in planes perpendicular to the tangent vectors of the common scallop curve. This is in effect equivalent to the assumption that the two corresponding swept sections on adjacent tool paths are in the same plane, which also has its limitations.

27

Chapter 2

Literature Review

With the principal direction method [Marciniak 1991, Jensen 1993, Bedi 1997], the shortest tool path can be obtained. With this method, the lines of curvature of the surface are identified first. The tool path is planned along these lines of curvature. This is because the step over size d can be approximated as in [Marciniak 1991]:

d=2

2h

(2.1)

where =

1 1 - , R is the cutter radius and is the signed surface curvature radius in R

the direction that is perpendicular to the tool path direction, > 0 for concave surface and < 0 for convex surface.

It can be found from Eq. (2.1) that if the tool path is planned along the surface principal direction that has the larger principal curvature radius, the step over size is maximum. However, this method needs to calculate the lines of curvature which are very difficult to compute. In addition, in some cases, the principal direction is not unique for some surface shapes (e.g., a planar or a spherical surface). The lines of curvatures cannot be identified for these areas.

The above two methods can only generate constant scallop height tool paths for single surface machining. As there are usually many surfaces in a mould workpiece, it is difficult to apply these methods in mould machining. To solve this problem, Lee [Lee 2003] proposed a spiral tool path generation method. With this method, a set of offset curves of the workpiece boundaries are generated first. The distance of the offset is determined to be the tool path interval such that the scallop height is maintained constant. The tool paths are then planned on these offset curves. With this method, the

28

Chapter 2

Literature Review

possible intersection and self-intersection of the offset curves need to be removed before calculating the CL-points. Another problem could be caused by the offset curves in separating the surface into several machining areas, which may result in machining marks along the boundaries of these areas.

The section curve method [Bobrow 1985, Choi 1988, Choi 1989, Suresh 1994 and Huang 1994] calculates the tool path by intersecting the part surface with a series of drive surfaces. The CC-points are then planned on the intersection curves. In general, with the section curve method, the cutter follows planar cross-sections generated by intersecting the designed surface with a set of parallel planes. It is characterized with a uniform interval between adjacent tool paths in the Euclidean space. Each interval is determined according to the scallop-height requirement. The main advantages with this method include: the resulting tool paths are non-adaptive to the local surface geometry and the CL-points of multiple adjoining surfaces are joined together into a single tool path. Since there are usually many surfaces in a complex mould, to achieve satisfactory surface quality, the tool paths for the entire workpiece should be as continuous as possible. Therefore, the section curve method is widely used in compound mould machining.

2.2.2

The CL-point method

With the CL-point method, the x- and y-values of the CL-points are planned first, the z-values are then calculated for each set of x- and y-values. According to how the CLpoints are planned, the CL-point method can be further classified into the surface offset method, the highest point method, the inverse tool offset method and the Z-map method.

29

Chapter 2

Literature Review

The surface offset method [Faux 1979, Tang 1995, Kim 1995, Maekawa 1997, Lartigue 2001] is usually used in machining sculptured surfaces with ball-end cutters. With this method, the offset surface is generated first by offsetting the original workpiece surfaces. The intersection and self-intersection of the offset surfaces are then trimmed off. The points on the trimmed offset surfaces are the gouge-free center points of the cutter. The tool paths are generated directly on the offset surfaces. One distinctive advantage of this method is that it allows flexibility in tool path planning. It is therefore possible to generate constant scallop height tool paths with the offset surface method [Lartigue 2001]. However, this method needs to calculate the offset surfaces and remove the possible intersection and self-intersection of the offset surfaces, which is computationally very intensive.

To solve this problem, Lai and Wang [Lai 1994] and Jun et al. [Jun 2002] approximated the sculptured surfaces into triangles first. These triangles were then offset. The CL-points were generated by slicing the offset triangles with a series of drive planes. If two intersection curves overlapped, the lower one was removed. The intersection curves were then sorted, trimmed and linked. Gouging is removed during the trimming process. One disadvantage of this method is the need to convert the original workpiece surfaces into triangles, which may lose the accuracy and the surface properties information.

Hwang [Hwang 1992] and Yang and Han [Yang 1999] used the highest point method to calculate the CL-points. With this method, the tool paths are planned in three steps: 1) the surfaces are converted into a set of facets; 2) the 2D tool paths are planned on the xy-plane; and 3) for a given cutter center axis at point (x, y), the highest cutter

30

Chapter 2

Literature Review

position z is determined by lowing the cutter until it touched a triangular facet under the cutter. With this method, the generated CL-points are gouge-free. This method can also be applied to tool shapes other than ball end-mills.

Takeuchi et al. [Takeuchi 1989] introduced the inverse tool offset method that can be used to calculate the tool paths. With this method, the offset surface is generated as an envelope of a virtual tool transposed upside down moving along the workpieces surface (see Figure 2.3). The gouge-free CL-points are then planned on the offset surface. Besides the ball-end-cutters, this method is also applicable to other tool shapes. When the tool path is generated with the z-map method, the inverse tool offset method is usually used to calculate the CL-points.

Offset surface

Workpiece surface

Figure 2.3 The inverse tool offset method

The Z-map method [Choi 1988, You 1995, Huang 1996, Choi 1997, Lin 1998, Maeng 2003] is a common strategy used for machining compound surfaces. It is based on a computer graphics z-buffer discretization of 3D objects which allows the approximation of 3D objects by 2D arrays. A z-buffer is a collection of z coordinate

31

Chapter 2

Literature Review

values of the CL-points, computed at the sampled grid points. With the z-map method, a set of grid points is first sampled on the xy-plane, in the domain of interest. It can be obtained by the intersection of surfaces and vertical lines passing through the grip points. Tool paths are generated from the z-map by the inverse offset method. The step-over size between tool paths is determined by the sampled grid points and the highest z value at each sampled point is accepted as the CL-point. With the z-map method, the main advantages are that the final tool path is interference free and low memory storage for the representation of complex surfaces is needed since the z coordinates of the object are the only values stored. Another merit of this method is that tools other than ball end-mills such as fillet end-mills can be used easily. However, it has the weakness of providing only global control over the scallop height and the chordal error, which may lead to low machining efficiency.

In summary, according to whether the CC-points or the CL-points are controlled, tool path generation methods can be classified into the CC-point method and the CL-point method. It is easy to control the scallop height with the CC-point method. The main drawback of CC-point method is that it is prone to gouging, while the shortcoming of CL-point method is that it is good only for preventing concave gouging [Choi 1997]. Generally, the CL-point method is better for satisfying the requirements of high-speed machining, such as cutting-load smoothing, tool path smoothness and chip-load leveling [Choi 1998].

Many researchers have studied tool path generation, and the research work focused on how to efficiently generate gouge-free tool paths that can machine moulds in minimum time while maintaining the required surface quality. There is no research work on how

32

Chapter 2

Literature Review

to regenerate tool path efficiently based on the original tool path and the modified workpiece. More research studies in this area are needed.

33

CHAPTER 3 MATHEMATICAL BACKGROUND OF CURVES AND SURFACES

In order to obtain a better understanding of the mould design modification and tool path regeneration methodologies and algorithms, the mathematical background of curves and surfaces will be introduced in this chapter. The chapter is organized as follows. The parameters of curves are discussed in the first section. The second section discusses the parameters of surfaces. Some specific curves on surfaces are studied in the third section. The offset surface is studied in the last section.

3.1 Curve parameters


A space curve can be defined as the trajectory of a point moving in three-dimensional space (3) with one degree of freedom. The Cartesian coordinates of a point on the parametric curve r can be expressed as the function of parameter t as follows:
x(t ) r(t) = y (t ) z (t )

(3.1)

where the Cartesian coordinates x, y and z are differentiable functions of parameter t. Any point on the space curve r can be obtained by specifying a value of the parameter
t. When a different parameter is used to represent the curve, the curve shape will not

34

Chapter 3 Mathematical Background of Curves and Surfaces

change. The arc length of the curve is often used as the curve parameter. The arc length s can be calculated as follows:
s(t) =

t1 t0

dr dt = dt

t1 t0

dx dy dz + + dt dt dt dt

(3.2)

The tangent (t), normal (m) and binormal (b) vectors of the curve form a local Cartesian system with original at r(t) and three axes of t, m and b (see Figure 3.1). They are given as follows:
D r t = D r m = b t DD rD b= r DD r rD

(3.3)

Osculating plane

Normal plane

Rectifying plane
P b t m

Figure 3.1 The Frenet frame [Faux 1979]

35

Chapter 3 Mathematical Background of Curves and Surfaces


C and C rC represent the first and second order derivatives of r with respect to t. where r

The frame of t, m and b is called the Frenet frame. The planes through a given point on the curve and contains the vectors t and m, m and b, and b and t are the osculating plane, the normal plane and the rectifying plane, respectively (see Figure 3.1).

Using the arc length (s) as the curve parameter, the derivatives of t, m and b with respect to arc length (s) yield the Frenet-Serret formulas: t = m m = b t b = m

(3.4)

where the prime () represents the differentiation with respect to arc length; and are the curvature and torsion of the curve at the point of evaluation:

(s) = r ' ' (s) =


1 det [r, r, r] k (s)

(3.5) (3.6)

The reciprocal of curvature is the radius of curvature of the curve: = 1/.

3.2 Surface parameters


A surface can be defined as the locus of a point moving in a three-dimensional space (3), with two degrees of freedom. It can be given in parametric form
x(u, v) r(u, v)= y (u, v) z (u, v)

(3.7)

36

Chapter 3 Mathematical Background of Curves and Surfaces

where the Cartesian coordinates x, y and z are differentiable functions of the surface parameters of u and v. The shape of the surface will not change when different parameters are used to represent it. If the surface parameter u is related to v, r(u, v) represents a curve on the surface. The unit normal vector n of a point on the surface is given by:
n= ru rv | ru rv |

(3.8)

where
ru = r u rv =
r v

ru rv 0

The partial derivative r/u represents a tangent vector of the curve v = constant, and it is denoted as ru, and so as that of the partial derivative rv.

The square of the infinitesimal distance between two points (u, v) and (u + du, v + dv) on surface r defines the first fundamental form I of the surface: I = dr dr = E du2 + 2F du dv + G dv2 (3.9)

where the fundamental magnitudes of the first order E, F and G are:


2 E = ru ,

F = ru rv ,

2 G = rv

(3.10)

The inner product of the infinitesimal distance dr and the infinitesimal variation dn defines the second fundamental form II of the surface: II = dr dn = L du2 + 2M du dv + N dv2 (3.11)

where the second order fundamental magnitudes of L, M and N are: 37

Chapter 3 Mathematical Background of Curves and Surfaces

L = n ruu ,

M = n ru ,

N = n r

(3.12)

The values of the first and second order fundamental magnitudes depend on the location of the point on the surface but they do not depend on the direction of the normal plane through that point.

At each point of a nonsingular parametric surface, a unique unit surface normal n is defined by Eq. (3.8). The family of planes containing the normal n at point P on surface r cut the surface in a family of normal section curves passing through that point. For each normal section curve C, its tangent vector t defines the normal section direction. The curvature n of the normal section curve C is defined as the surface normal curvature in direction t. The surface normal curvature value n at point P can be calculated from the first and second fundamental forms at this point:
II Ldu 2 + 2Mdudv + Ndv 2 n = = I Edu 2 + 2 Fdudv + Gdv 2

(3.13)

Although the first and second order fundamental magnitude values E, F, G, L, M and N are fixed on a point of the surface, the value of
du II depends on the ratio of . I dv

Therefore, the normal curvature n may have different values in different directions of the normal section. The extreme normal curvature values 1 and 2 are called principal curvatures, and the corresponding directions t1 and t2 are the principal directions. In the special case, where E:F:G = L:M:N, the normal curvature is independent of the normal section direction t. The point with this property is called the umbilical point. For any non-umbilical point on the surface, its principal directions are orthogonal, i.e. (t1 t2) =

38

Chapter 3 Mathematical Background of Curves and Surfaces

0. The principal curvatures 1 and 2 can be calculated from Gaussian (K) and mean (H) curvature values, and
LN M 2 K = 1 2 = EG F 2

(3.14)

H=

EN + LG 2 MF 1 (1 + 2) = 2 2( EG F 2 )

(3.15)

The Gaussian and mean curvature values determine the surface shape around a point:

If K > 0, the surface shape is either concave or convex around the point.
If H > 0, it is a concave point; otherwise, it is a convex point.

If K = 0, at least one of 1 and 2 is zero.


If H = 0, it is a planar point. If H 0, the surface is cylindrical or through-shaped around the point. If H > 0, it is a concave point; Otherwise, it is a convex point.

If K < 0, it is a saddle point.

From Eqs. (3.14) and (3.15), the principal curvatures can be calculated as:

1,2 = H

H2 K

(3.16)

In tool path generation, if the signed curvature value of the surface is greater than that of the cutter, concave interference arises.

39

Chapter 3 Mathematical Background of Curves and Surfaces

The differential dn of a unit normal vector in the principal direction in terms of a principal curvature i has the following relation: dn + i dr = 0 (3.17)

where dr is the differential of r(u, v) at that point. Eq. (3.17) is called Rodrigues formula.

3.3 Curves on surfaces


A curve r(t) on the surface can be expressed as: r(t) = r ( u(t), v(t)) (3.18)

The curve r(t) becomes an iso-parametric curve if u(t) = constant or v(t) = constant.

If the tangents of curves on a surface coincide with the principal directions there, these curves are called the lines of curvature. The lines of curvature form an orthonormal net which is given by: (GM - FN) dv2 + (GL EN)dudv + (FL - EM)du2 = 0 (3.19)

It is possible to generate the shortest tool path along the surfaces lines of curvature.

Let be an angle between the unit normal n of the surface r and the vertical direction T. The angle can be expressed as: cos = T n (3.20)

If = 90 o, the points given by Eq (3.20) form the silhouette curve with respect to direction T. When there is no undercut, the parting lines are the silhouette curves with respect to the parting direction.

40

Chapter 3 Mathematical Background of Curves and Surfaces

3.4 Offset surface


An offset surface can be defined as the locus of points, each of which has a constant distance R from a corresponding point of surface r(u,v) in the direction of its normal vector n at that point: ro(u, v) = r(u, v) + R n(u, v) (3.21)

So long as ro(u, v) does not intersect with itself, a sphere of radius R with its center on the offset surface, touches the original surface at one point. If there are regions bounded by the self-intersection curve of the offset surface, the sphere with its center in this region will touch more than one point. In tool path generation, gouging occurs.

It was proven by Willmore [Willmore 1959] that the surface normal of the offset surface ro(u, v) at point (u, v) is:
no(u, v) =

(1 R 1 )(1 R 2 ) n(u, v) (1 R 1 )(1 R 2 )

(3.22)

It can be seen from Eq. (3.22) that the normal vectors of surface r(u, v) and its offset surface ro(u, v) is parallel. If (1 - R1)(1 - R2) > 0, the normal directions are the same.

41

CHAPTER 4 MOULD DESIGN MODIFICATION

When the design of the plastic part is changed, the design of the affected mould inserts need to be modified. The existing commercial CAD systems cannot automatically modify mould design when the parameters of the mould and the plastic part are lost. In addition, when the mould has been machined and material needs to be added due to product design modifications, an insert may need to be designed to provide the extra material for the machined mould. Existing commercial CAD systems cannot realize this process automatically. In practice, when the design of a plastic product is modified and passed to the mould manufacturing department, the mould designer studies and identifies the modified region first. The mould design is then modified according to the modified product and the mould manufacturing process status. The mould design modification is based on the designers experience, which is time-consuming and error-prone.

This chapter introduces a mould design modification system based on the designed mould and the new plastic part. With this system, the old and the new plastic products are compared. The mould is then modified automatically according to the designed mould and the mould machining process. If material needs to be added to a machined mould, inserts and pockets will be designed automatically. An algorithm of detecting the interference between the pocket of the insert and the cooling, ejector, lifter and screw holes of the mould insert is also introduced. 42

Chapter 4 Mould Design Modification

This chapter is organized as follows. The basic concepts and principles of mould design modification are discussed in the first section. Based on these principles, the mould design modification system is developed, and the architecture of the system is introduced in the second section. Section 3 introduces how to identify solid bodies of material to be added or removed. Section 4 introduces how to identify the mould insert that material needs to be added to or removed from. The methods of adding and removing material from the designed mould are studied in Sections 5 and 6, respectively. The algorithm of detecting interference between the pocket and the cooling, ejector, lifter and screw holes of the mould insert is also introduced in Section 6. An example of modifying mould design with the developed algorithm is illustrated in the last section.

4.1 Basic concepts and principles of mould design modification


Since the mould has been designed, the method of mould design modification is different from that of designing a new mould. Besides the basic mould structure and functional requirements, the designed mould and the mould machining status should also be taken into account in modifying the mould design. The mould design may need to be modified before or after the mould has been machined. If the mould has not been machined, the mould design can be modified directly by uniting or subtracting metals that need to be added to or removed from the mould. If the mould has been machined, the mould needs to be modified with different methods according to whether metals need to be added to the mould. This research emphasizes on the case that the mould has been machined. Some basic concepts and principles of mould design modification are introduced in this section.

43

Chapter 4 Mould Design Modification

4.1.1 Concepts and assumptions of mould design modification

When the design of a plastic product is changed, the mould insert corresponding to the modified region should also be modified. According to how the plastic part is changed, material may need to be added to or removed from the designed mould. One or more portions of the plastic part may be changed which may affect more than one mould insert. To minimize the affected mould design and manufacturing processes, only the mould inserts corresponding to the modified profile of the plastic product will be modified. The other mould inserts and components should be kept unchanged as possible.

After the mould design is modified, it is usually machined to the designed profile with CNC or EDM machines. Since the machining process can only remove material from the mould, the modification of the mould design should make sure that there is extra material left for machining. If the mould has been machined and material needs to be added to the designed mould, different methods may be used to modify the mould design according to how the extra material will be added. In practice, depositing material with a welding process and designing an insert are the two commonly used methods to add material to a machined mould. With the welding process, the molten metal is deposited directly to the portion that needs extra material. In this case, the mould design can be modified through joining the solid body of the metal with the designed mould and marking the area to be welded. The welded region will then be machined to the design profile. However, only a limited amount of metal can be added with the welding process, as this process may lead to internal stresses due to the high welding temperature. Therefore, designing insert is widely used in practice to add extra

44

Chapter 4 Mould Design Modification

material to the machined mould. This research will emphasize on this method for adding material to the mould.

In order to shorten the mould design and manufacturing lead-time, many mould manufacturing companies are using 3D CAD/CAM software. In practice, the plastic part and mould inserts are usually managed in a way that they can easily be identified. It is assumed in this research that the mould has been designed based on the old plastic part with 3D CAD software and the same CAD system is used to modify the designed mould and unique names have been assigned to the old plastic part and the designed mould inserts. The system can thus automatically identify these parts. In addition, it is assumed that the new plastic part has been enlarged with the same scale factor as for the old one, and it has been oriented and positioned to the same orientation as the old plastic part.

4.1.2 Principles of insert and pocket design If the mould design modification needs to remove material from the machined mould, the mould design can be modified by subtracting the material from the affected mould insert directly. If material is to be added to the machined mould, an insert needs to be designed for the added material.

If an insert needs to be designed for the added material, a pocket should be created in the mould insert to hold and fix the insert. There are some principles in designing the insert and the pocket:

1) Since the insert will form the profile of the plastic part, the gap between the insert

45

Chapter 4 Mould Design Modification

and the pocket should be small enough to prevent the molten plastic from escaping. The pocket and the insert body should be designed with the same dimension.

2) The designed insert and the pocket should be easy to machine and assemble. The insert body is usually designed with a cylindrical or a rectangular shape. To make it easier to fix the insert in the mould, a through pocket should be designed as much as possible. In practice, a shoulder is usually designed at the bottom of the insert and the pocket to fix the insert (see Figure 4.1 a).

3) If a rectangular shape insert is designed, since only the z dimension of the shoulder needs to be controlled, to make it easier to machine and assemble, the size of the pockets shoulder is usually designed to be bigger than that of the insert in x- and y-directions with two semi-cylinders created at the two sides (see Figure 4.1a). The radius of the cylinder usually equals to the cutter radius that will be used to machine the block.

4) When a cylindrical shape of insert is designed, a planar face that is parallel to the ZX-plane or ZY-plane is usually designed to prevent the insert from rotating about its axis (see Figure 4.1b).

5) There are usually some cooling channels, screw and lifter holes in a mould insert. When a pocket is designed, it should not interfere with these holes. The minimum distance between the pocket and these holes should be greater than a given value, which is the minimum thickness value for enough strength and machining tolerance. This value is decided by the mould size and is usually in the range of 3

46

Chapter 4 Mould Design Modification

Rectangular insert and its shoulder Designed mould

Cylindrical insert and its shoulder

Blind insert

c Cooling system

Figure 4.1 Different shapes of inserts and pockets

47

Chapter 4 Mould Design Modification

to 8 mm. If a through pocket interferes with other holes, a blind pocket should be designed (see Figure 4.1c).

Based on these concepts and principles, a mould design modification system is developed, which will be introduced in the following sections.

4.2 Architecture of the mould design modification system


A mould design modification system has been developed in this research. The system includes four main parts (see Figure 4.2): identify solid bodies of the material to be added or removed; identify the mould insert that the solid body of the material will be added to or removed from; remove material from the designed mould; add material to the designed mould.

By comparing the old and new plastic parts, the solid bodies of the material that need to be added or removed can be identified. The corresponding mould inserts that are affected by these solid bodies are then identified. The design of these affected mould inserts are modified according to the modified profile and the status of the mould machining process.

When material needs to be removed from the designed mould, the system will subtract the solid body of the metal from the corresponding mould insert directly. If material needs to be added to the designed mould, the mould design will be modified according to the mould machining process status. If the mould has not been machined, the mould design can be modified directly by uniting the solid with the affected mould insert. Otherwise, an insert and a pocket will be designed. After the mould design is modified,

48

Chapter 4 Mould Design Modification

the affected tool path will be generated or regenerated based on the mould machining status.

New product model

Old product model

Identify the solid bodies of material to be added or removed

Identify the mould inserts that corresponding to the solid bodies

N Add material?

Y Mould insert machined? Y Create pocket & insert

Remove metal from mould N

Add metal to mould

End

Figure 4.2 Framework of the mould design modification system

49

Chapter 4 Mould Design Modification

4.3 Identify the solid bodies of the metal to be added and removed
Since unique names have been assigned to the old plastic part and the designed mould inserts, when the new plastic part is input, the solid bodies of the material to be added to or removed from the designed mould can be identified automatically. If plastic material is to be added to the modified plastic part, metal will be removed from the designed mould. Otherwise, material needs to be added to the designed mould (see Figure 4.3). The solid bodies of the material to be added or removed can be identified using the Boolean operation of subtraction: subtracting the old plastic part by the new one, and the result is the solid bodies of the material to be added to the designed mould insert (add_metal); subtracting the new plastic part by the old one, the solid bodies of the material to be removed from the mould (sub_metal) can be created (see Figure 4.3).

Cavity

Old product Metal to be added to cavity

Metal to be removed from core Core x

Figure 4.3 Material to be added to or removed from a designed mould

When a solid body of the material to be added or removed is identified, the boundary (bd) of the solid body, that is the intersection of the old and the new plastic parts, is 50

Chapter 4 Mould Design Modification

identified. The boundary of the solid body will then be used to identify the mould insert that the solid body will be added to or removed from. It will also be used for tool path regeneration.

Cavity Two inserts affected

Product modification affecting parting line z Core x

Parting line

Figure 4.4 Cavity and core are affected by product design modification

4.4 Identify the mould insert that the material needs to be added to or removed from
A solid body of metal to be added or removed may affect one or more mould inserts (see Figure 4.4). When the solid body affects more than one mould insert, it needs to be split into different solid bodies according to the parting direction and the designed mould inserts. Many researchers have studied this problem [Chen 1993, Shin 1993, Rosen 1994, Weinstein 1996, Nee 1997, Zhang 1997]. This research will not discuss

51

Chapter 4 Mould Design Modification

this aspect in detail. The research is emphasized on the case that one solid body of material to be added or removed affects only one mould insert. When a solid body of material will be added to or removed from a mould insert, the minimum distance between them is zero. However, some of the solid bodys edges may coincide with the parting line (see Figure 4.4). In this case, there may be more than one mould insert that the minimum distance with the solid body is zero. To solve this problem, a set of points on the boundary (bd) of the solid body are sampled, the minimum distance between the sampled points and the mould inserts is measured. If all the distance values are small enough, the solid body will be added to or removed from this mould insert.

Generally, changing the design of the plastic part should not lead to the formation of undercut. Otherwise, new slider or lifter needs to be designed and the structure of the mould will be changed. Therefore, when the solid body and the corresponding mould insert are identified, the possible undercut will be detected. Many researchers have studied how to identify the undercut [Chen 1995, Nee 1997, Ye 2001]. This research area will not be discussed in detail.

Product design modification may affect the location of the parting line. In this case, more than one mould insert will be affected. The affected parting line and the corresponding parting surface need to be modified based on the parting direction and the unaffected parting surfaces. Many researchers have studied the identification of parting lines and the creation of parting surface automatically with a given parting direction [Ravi 1990, Chin 1996, Weisten 1997]. This research will not emphasize on this aspect.

52

Chapter 4 Mould Design Modification

After the solid bodies and the corresponding mould inserts are identified, the mould design can be modified. This process will be discussed in the following sections.

4.5 Remove material from the mould


When material is to be removed from the designed mould, there is extra metal for machining. The mould design can be easily modified by subtracting the solid body from the designed mould directly. The modified mould can then be used to regenerate the tool path.

4.6 Add material to the mould


When material needs to be added to the mould, the mould design should be modified with different methods according to whether the mould has been machined or not. If the mould has not been machined, the mould design can be easily modified by uniting the solid body with the designed mould. If the mould has already been machined, an insert will be designed for the added material. A pocket will also be created in the mould to fix the insert, the algorithm of designing an insert and pocket will be introduced in the following subsections.

4.6.1 Create pockets and inserts An insert is designed to provide additional material for mould design modification of a machined mould. The top surfaces of the insert form the shape of the plastic part. The body of the insert is fixed in the pocket. To make it easier to modify the mould, the body of an insert is usually designed with a rectangular or cylindrical shape. The dimensions of an insert in x- and y-directions are decided by the modified profile, and they are the same as the corresponding dimensions of the pocket. The depth of the

53

Chapter 4 Mould Design Modification

pocket is decided by the interference condition. The height of the insert is decided by the pockets depth. Therefore, the insert and the pocket should be designed together. Based on these ideas, an algorithm of designing the insert and the pocket has been developed. The pseudo codes of the algorithm are as follows:

Create_pocket_and_insert { /* Design pocket and insert with the designed mould insert and the solid body (sld) of the material to be added to the designed mould */ m_z (get the extreme z-values of the mould insert); bd (get boundaries of sld); if (the projection of bd forms a circle) { D (round the diameter of the circle to integer); create a cylinder with D and m_z; } else { (x0, y0) (get the x and y values of bds range box); (x, y) (round the lengths of the range box to integers); create a block with x, y and m_z; } blk (get the common part of the cylinder/block and the mould insert); detect interference between blk and the holes in the mould; if (there is interference)

54

Chapter 4 Mould Design Modification { if (ejector hole interferes with the pocket) enlarge blk to include the ejector hole in it; if (only ejector hole interferes with the pocket) { ist (unite blk with sld); create the ejector hole in blk; create shoulder at the bottom of ist and unite it with ist; create a pocket for ist; } else { z_bottom (calculate the z value that the bottom surface of the pocket can reach); create a blind pocket in the mould insert with x, y and z_bottom; ist (unite blk with sld and trim off the portion below z_bottom); if (there is enough space) create screw holes in the mould insert and ist; else mark at the drawing for point welding; } } else { ist (unite blk with sld); /* no interference detected, a through pocket is created */

55

Chapter 4 Mould Design Modification create a shoulder at the bottom of ist and unite it with ist; create a pocket for ist; } }

To create the pocket and the insert, the extreme z values m_z of the mould insert are calculated first. The boundary bd of the solid body (sld) to be added is then projected onto the xy-plane (see Figure 4.5). The projected boundary is identified whether it belongs to the same circle: 1) Sample three points on the projected boundary; 2) Fit an arc with the three points; and 3) Check the difference between the rest of the points of the projected boundary and the arc. If the projection of boundary bd forms a circle, the diameter of the circle is rounded to an integer. A cylinder is then created with the rounded diameter D and the mould inserts extreme z-values m_z.

If the projection of boundary bd does not form a circle, a block will be created. The blocks dimensions in x- and y-directions are decided by the range box of the boundarys projection and its z-values are decided by m_z. To make it easier to machine, the lengths of the blocks edges in x- and y-directions are rounded to integers.

With the Boolean operation of intersection, the common part (blk) of the cylinder or the block and the mould insert is created (see Figure 4.5). The possible interference between the holes in the mould insert and blk is then detected, which will be introduced in the next subsection.

56

Chapter 4 Mould Design Modification

There are cooling, lifter, screw and ejector holes in mould inserts. If a pocket is designed, it may interfere with these holes. When the pocket interferes with an ejector

Modified profile

Projection of modified boundary

Insert boundary Modified boundary (a) Creation of cylinder

(b) Designed insert and pocket

Figure 4.5 Designed insert and pocket

57

Chapter 4 Mould Design Modification

hole only, the ejector hole can be created in the insert, and it does not affect the function of the mould. In practice, the insert is enlarged to enclose the hole and a through pocket will be designed (see Figure 4.6).

Old product file

Original core

Cooling holes (a) Old mould

Ejector hole

New product file

Minimum distance Ejector hole Insert and shoulder

(b) An ejector hole is enclosed in the insert

Figure 4.6 An ejector hole interferes with the pocket

58

Chapter 4 Mould Design Modification

If the pocket interferes with the various holes in the mould insert, a blind pocket should be designed. The z-value (z_bottom) of the pockets bottom surface is decided by the interfering holes. After the depth of the pocket is decided, the insert can be created. By uniting blk and sld, the insert ist is created. The portion that is below
z_bottom is then trimmed off. In practice, screw and point welding are usually used to

fix the insert firmly to the mould (see Figure 4.7). If there is enough space, one or two

Modified area Old product file

Modified core

(a) Designing screw for blind insert

Point welding Modified area Old product file

Modified core

(b) Welding blind insert to the mould

Figure 4.7 Fix the insert with screw or welding process

59

Chapter 4 Mould Design Modification

screws will be used. Otherwise, the point welding process is used to bind the insert with the mould. If a screw is to be used, the screw holes will be designed automatically for the mould and insert ist.

If there is no interference, a through pocket is designed. The insert is created by uniting blk and sld directly. To fix the insert in the mould, a shoulder is designed at the bottom of the insert and the pocket (see Figure 4.1).

4.6.2 Detect interference When a pocket is designed, it may interfere with the various holes in a mould insert. If an ejector hole interferes with the pocket, since an insert is fixed in the pocket, the ejector hole can be created inside the insert. A through pocket can be designed (see Figure 4.6). However, if the interference is caused by cooling, lifter and screw holes, the pocket should not go through them, a blind pocket needs to be designed. To detect the interference caused by these holes, they need to be identified first.

A screw hole is usually located at the bottom or side surface of the mould insert. It will not go through the top surface of the mould insert. A lifter hole will go through the bottom and the top surfaces of the mould insert, and its axis/surface is not parallel to the z-axis (see Figure 4.8). The shape of a cooling hole is usually cylindrical. A cooling hole goes through the bottom or the side surfaces of the mould insert (see Figure 4.1c). Different from the lifter and ejector holes, a cooling hole will never go through the top surface of the mould insert. The axis of a cooling hole may be horizontal, vertical or parallel to none of the coordinate systems axes. The main difference between an ejector hole and a lifter hole is that the axis/surface of the

60

Chapter 4 Mould Design Modification

ejector hole is parallel to the z-axis while the surface/axis of the lifter hole is not. Therefore, if a hole other than the ejector hole interferes with the pocket, there are some undercut surfaces in the block (blk). Product file Lifter hole

Core

Cooling holes

Figure 4.8 Lifter and cooling holes

The pocket has the same dimension as the block (blk). The block (blk) can be considered as one portion of the mould insert with the x- and y-dimensions defined by the range box of the modified profile. If there is a hole inside blk, the hole interferes with the pocket. Hence, blk can be used to detect interference. In some cases, a hole is not inside the block (blk), but the distance between the hole and the pocket is too small. This hole should also be considered as interfering with the pocket. Therefore,
blk should be enlarged before interference is detected.

With these concepts, an interference detection algorithm is developed as follows:

Detect_interference_between_pocket_and_holes {

61

Chapter 4 Mould Design Modification blk1 (create a block with the same method as blk, with the x- and y-values being bigger than blk for the given thickness value); if (there are ejector holes in blk1); enlarge blk1 to enclose the ejector holes; (fce, fce_num) (get all faces of blk1); z_bottom (the z-value of the mould inserts bottom surface); while (fce_num > 0) { (pnts, pnt_num) (sample points along the edges of fce(fce_num)); for (i=0; i<pnt_num; i++); { detect face undercut at pnts(i); if ( undercut is detected) { z (calculate the z-value of fce(fce_num)s lowest point); if (z > z_bottom) z_bottom = z; break; } } fce_num--; } }

To detect interference, the block (blk1) is created first with the same method as for creating blk, with its dimensions in x- and y-directions being enlarged for the given

62

Chapter 4 Mould Design Modification

thickness value. The thickness value is to make sure that there is enough thickness allowance for mould strength and machining tolerance. It is then detected whether there is any ejector hole inside blk1. If there is an ejector hole in blk1, blk1 is enlarged to enclose the ejector hole inside blk1 (see Figure 4.6). Interference will be detected for the enlarged blk1. In practice, the thickness value is set between 3 to 8 mm, which is determined by the size of the mould insert (see Figure 4.9).

Minimum distance

New product Minimum distance

Minimum distance

Modified core

Figure 4.9 Minimum distance between pocket and other holes

As discussed above, when a cooling, lifter or screw hole interferes with the pocket, there are some undercut points on the holes surface. Therefore, the interference detection for the holes inside blk becomes detecting its uncut surfaces. If there is undercut points on a holes surface, there is at least one undercut point on the surface edge. Therefore, the undercut can be detected for the points on the edges of the holes surfaces.

If an undercut point is detected, the z-value (z) of the corresponding surfaces lowest point is calculated. It is then compared with z_bottom, which initial value is set as the

63

Chapter 4 Mould Design Modification

z-value of the mould inserts bottom surface. If z is greater than z_bottom, its value is assigned to z_bottom. It needs to be pointed out here that, in the above statements, the lowest point and the cooperation are based on the assumption that the z-value of the mould profiles surface is greater than its bottom surface (e.g. the core insert in Figure 4.3). For the cavity insert in Figure 4.3, the z-value (z) should be the minimum z-value of the undercut surface and the comparison condition should be reversed (z<z_bottom). When an undercut point is detected for a surface, the rest of the points of the surface need not be detected. The undercut is then detected for the rest of the surfaces, until all of them are detected. Blk1 and the value of z_bottom are then used to create the pocket and the insert.

4.7 Illustrative example


The system has been tested with several industrial parts. This section shows an example of modifying a mould design with the algorithm developed in this research. If the mould has not been machined, the mould design can be modified by adding or removing material from the affected mould inserts directly, which is much easier than the case when the mould has already been machined. Therefore, it is assumed in this section that all of the affected mould inserts have been machined.

Figure 4.10a shows the original plastic part (Deckel Flash Box) while Figure 4.10b shows the modified one. It can be found that some plastic materials are removed from the old plastic part (two circular holes and two rectangular pockets) while some are added to it (two bosses and two ribs). Both core and cavity are affected by product design modification. Hence, both of them need to be modified. Material would need to be added to the cavity while material has to be removed from the core side for the two

64

Chapter 4 Mould Design Modification

Figure 4.10a Old product file

Modified areas (both sides mirror)

Figure 4.10b New product file bosses and two ribs. Figure 4.10c shows the original cavity insert. To show the cavity structure better, both shading and wire frame views are given here. Figure 4.10d shows the modified cavity insert. As material needs to be added to the cavity for the two

65

Chapter 4 Mould Design Modification

Figure 4.10c Old cavity

bosses, two pockets and inserts are designed for them. Since the projection of the modified region boundaries form circles, cylindrical inserts are designed for the two added bosses. As none of the holes interfere with the pocket, a through pocket is designed. The two shaded cylindrical inserts and their pockets (displayed with wire frame) are shown in the figure. Shoulders and the pocket are designed to fix the inserts with the cavity. Figure 4.10e shows the original core insert. Some cooling and ejector

66

Chapter 4 Mould Design Modification

Created inserts and their pockets

Figure 4.10d Modified cavity holes can be seen from the wire frame view. Figure 4.10f shows the modified core. Two inserts need to be designed to add materials for the two rectangle pockets. Since a cooling hole interferes with the through pocket for insert A, a blind pocket is designed for it. A screw is designed to fix the insert to the pocket. As none of the holes interfere with pocket B, a through pocket is therefore designed for it. The two cylindrical

67

Chapter 4 Mould Design Modification

profiles in Figure 4.10f show material removed from the designed mould for the added bosses in the plastic part.

Figure 4.10e Old core

68

Chapter 4 Mould Design Modification

A B

Figure 4.10f Modified core

69

CHAPTER 5 TOOL PATH REGENERATION FOR THE CL-POINT METHOD

In practice, if a mould design is modified, all the tool paths that cover the modified region must be regenerated, no matter how small the modified region may be. With the available tool path generation systems, if a tool path needs to be regenerated, all the cutter location (CL) points must be recalculated, and none of the generated CL-points can be reused any more. It needs as much time to regenerate the tool path for the modified mould as that for the original one. On the other hand, the region of the mould modification is usually small in comparison with the entire mould. If the generated CL-points can be utilized in the new tool path for the modified mould, the tool path regeneration time can be greatly reduced, resulting in saving of NC-data preparation cost, minimizing possible reprogramming errors and increasing production efficiency.

The tool path regeneration method is related to the tool path generation method. According to whether the topology of the CL-point or the CC-point is controlled, tool path generation methods can be classified into the CL-point method and the CC-point method. This chapter studies the regeneration of tool paths for the CL-point method. The tool path regeneration algorithm for the CC-point method will be studied in Chapter 6.

70

Chapter 5 Tool Path Regeneration for CL-Point Method

To regenerate tool paths efficiently, the affected and the unaffected CL-points should first be identified. It is proven in this chapter that the CL-points corresponding to the boundary of the mould modification region form a closed boundary. Only the CLpoints that are enclosed by this boundary are affected by mould design modification and need to be replaced. The algorithm of identifying and replacing the affected CLpoints for the CL-point method will be introduced in this chapter. The algorithm of checking the scallop height values for the replaced CL-points will also be introduced. With the developed algorithm, if the scallop height is greater than the given value, new CL-points will be added to maintain the required surface finish.

When the tool path is generated with the CL-point method for mould machining, the tool paths are usually planned to be parallel to a plane. This chapter will focus on this tool path generation method. Without losing the generality, it is assumed that the tool path lines are parallel to the X-axis, i.e., the y values of all the CL-points in the same tool path row are the same.

This chapter is organized as follows. Terms and notations are introduced in the first section. The basic concepts are introduced in the second section. Based on the basic concepts, the methodology of detecting the affected CL-points is presented in the third section. The algorithm of tool path regeneration for the CL-point topology is introduced in the fourth section. Two examples of tool path regeneration for the CLpoint method are illustrated in the last section.

5.1 Terms and notations


Some terminologies of tool path generation are introduced in the following to avoid

71

Chapter 5 Tool Path Regeneration for CL-Point Method

ambiguities in the later discussions.

Cutter axis: It is assumed that the cutter axis T is parallel to the Z-axis, and T = (0, 0,

1). If the surface normal vector n points to the -Z direction, the surface cannot be machined with a 3-axis milling machine. As such, it is assumed in this research that all the surface normal vectors point to the +Z direction, i.e. (T n) > 0.

Cutter Contact (CC) point: The cutter contact (CC) point (Pcc) of a ball end-mill is the

point on the part surface at which the ball end-mill makes a tangential contact (see Figure 5.1).
T = (0,0,1)

Cutter

Pc Pcc

Pcl

Y X

Figure 5.1 Cutter contact (CC) and cutter location (CL) points

72

Chapter 5 Tool Path Regeneration for CL-Point Method Cutter Location (CL) point: The cutter location (CL) point (Pcl) is the intersection of

the tool spherical surface and the axis of the cutter.

Cutter center point: The cutter center point (Pc) is the center point of the tool spherical

surface. It is located on the offset surface of the workpiece.

As shown in Figure 5.1, let Pcc be the CC-point, R be the radius of the ball end-mill, the cutter center point Pc is given by
Pc = Pcc + R n

(5.1)

where n is the surface normal at point Pcc.

The CL-point (Pcl) is calculated using


Pcl = Pc - R T

(5.2)

It can be found from Eq. (5.2), that a CL-point has the same x and y values as the corresponding cutter center point. Sometimes, the cutter center points are used to represent the cutter location points. To avoid any confusion, in this research, a cutter location point is represented as in Eq. (5.2).

5.2

Basic concepts of tool path regeneration

To regenerate tool path for the modified mould, one direct and commonly used method is to recalculate all the CL-points for the modified mould. However, there may be hundreds or even thousands of surfaces in a mould, whilst the modified region could be relatively small. It is time consuming and unproductive to recalculate the CL-points

73

Chapter 5 Tool Path Regeneration for CL-Point Method

for the entire mould machining area. Since most of the CL-points may not be affected by the mould design modification, it is not necessary to calculate the unaffected CLpoints and they can be reused to machine the modified mould. As such, the affected CL-points that cannot be used to machine the modified mould should be identified and replaced.

5.3

Methodology for detecting the affected CL-points

As there are a large number of CL-points in mould machining, it is necessary to develop an efficient algorithm to identify the CL-points that are affected by mould design modification. For a gouge-free tool path, there are some special properties that can be used to identify the affected CL-points efficiently, and these will be studied in this section.

Proposition 5.1. Let A be a profile with all the surface normal vectors pointing to the

+Z direction, Ao is the offset of A along the surface normal direction with the offset distance equal to the cutter radius R (see Figure 5.2). If A is interference-free (it can be machined by this cutter without over-cutting the workpiece), then when looking along the Z-axis, all points of Ao are enclosed by its boundary, and for every point on the boundary of A, there is a corresponding offset point on the boundary of Ao.

Proof: Since all the surface normal vectors of profile A point to the +Z direction,

profile A is enclosed by its boundary when looking along Z-axis. As profile A is interference-free, it can be shown that the surface normal no of profile Ao and the normal n of profile A have the following relation [Willmore 1956]:

74

Chapter 5 Tool Path Regeneration for CL-Point Method

no(u, v) =

(1 R 1 )(1 R 2 ) n(u, v) (1 R 1 )(1 R 2 )

(5.3)

where R is the cutter radius, (u, v) are the parameters of the corresponding point, 1 and 2 are the principal curvature values of profile A.

Ao

Figure 5.2 Surfaces A and its offset Ao

It can be seen from Eq. (5.3) that the normal directions of Ao and A are parallel at the point with the same (u, v) parameter values. Since profile A is interference-free, (1 R1) > 0 and (1 - R2) > 0, n and no have the same direction for the corresponding

75

Chapter 5 Tool Path Regeneration for CL-Point Method

points. Therefore, all the surface normal vectors on the offset surface point to the +Z direction. The profile of Ao is enclosed by its boundary when looking along the Z-axis, and every point on the boundary of A has a corresponding offset point on the boundary of Ao.

Since the surfaces of the mould may not be interference-free, Proposition 5.1 cannot be applied in practice directly. The following proposition shows that it is not necessary for every point of profile A to be interference-free.

Proposition 5.2. Let TA be the interference-free tool path for profile A; Co and C are

the boundaries of TA and A, respectively (see Figure 5.3). When looking along the Zaxis, Co encloses all the points of TA. If C is interference-free, then every point on C has a corresponding point on Co.

Proof: Offset A for a distance of the cutter radius, and move the offset surface along

the -Z direction for a distance of the cutter radius. If there is no interference for the entire profile A, the tool path points are on the moved offset profile. From Proposition 1, every point on C has a corresponding point on Co, and thus every tool path point of TA is enclosed by Co.

If profile A is not interference-free, a new interference-free surface A can be generated from it: rolling a ball with the cutter radius along profile A, profile A is formed by the envelope of the ball that is on the side of A (see Figure 5.4). Then the points of the interference-free tool path TA are on the moved offset surface of A.

76

Chapter 5 Tool Path Regeneration for CL-Point Method

TA A Co

Figure 5.3 Boundaries of A and TA

Since the boundary of A is interference-free, A has the same boundary of A. From Proposition 5.1, every point on the boundary of profile A has a corresponding point on the boundary of the interference-free tool path TA.

77

Chapter 5 Tool Path Regeneration for CL-Point Method

It can be found from Proposition 5.2 that if the boundary of the mould modification region is interference-free, the tool path points that are affected by the mould design modification can be identified through the boundary of these tool path points.
A

Figure 5.4 Surfaces A and A

If there are interference points on the boundary of the modified region, Proposition 5.2 can be applied by enlarging the boundary of the modified region to a new one, making all the points on the boundary interference-free. Since many researchers [Seiler 1997, Glaeser 1999, Pottmann 1999, Yang 1999, Ding 2001] have studied the problem of identifying the interference-free boundary, this research will not focus on this aspect.

5.4

Tool path regeneration algorithms for the CL-point method

As discussed in the last section, when boundary C of the modified region is interference-free, the corresponding CL-points form a closed boundary Co. All the CLpoints enclosed by Co are affected by mould design modification, and they must be removed and replaced by the new CL-points. While the CL-points outside Co are unaffected by the modification of this region. If there is no other modification, these points can be used to machine the modified mould. To identify the unaffected CLpoints, Co must be calculated first. For a given point Pcc on C, the corresponding CL-

78

Chapter 5 Tool Path Regeneration for CL-Point Method

point Pcl on Co can be calculated using Eq. (5.2). Connecting all the Pcl points, the boundary Co can be created.

When the mould design is modified, if the same intersection planes are used to generate the CL-points, the machining scallop height may be greater than the given value for the modified region. Therefore, the machining scallop height values for the modified region need to be checked. If the scallop height is greater than the given value, new CL-points should be added.

With these concepts, the tool path can be regenerated with the following four steps:
1) Identify the interference-free boundary C of the modified region; 2) Calculate the boundary Co of the affected CL-points from C; 3) Identify and replace the affected CL-points that are enclosed by Co; and 4) Check the scallop height values for the modified region. If the scallop height is greater than the given value, add new CL-points.

In practice, the boundary of the mould modification region may not be interferencefree. When the boundary is input, it will be checked whether there are interference points on the boundary. If there are interference points, a new interference-free boundary is identified by enlarging the modified region. This research uses the algorithm proposed by Ding et al. [Ding 2001] to identify the interference-free boundary. From the interference-free boundary C, the boundary Co that encloses the affected CL-points is calculated. When Co is calculated, the affected CL-points can be identified and replaced. The algorithm of replacing the affected CL-points is introduced in the following subsection.

79

Chapter 5 Tool Path Regeneration for CL-Point Method

In some cases, the surface may be changed slightly. As long as it exceeds the required calculation tolerance that is usually set to 1/10 or 1/5 of the machining tolerance value in industry, the modified region will be identified and the affected CL-points can be identified and replaced. If the change is within the calculation tolerance, the tool path does not need to be regenerated since it cannot be realized in the machining process.

5.4.1 Identifying and replacing the affected CL-points

When the tool path lines are parallel to the x-axis, the y-values of all the CL-points in the same tool path row are the same. By comparing the y-values of a tool path row with the extreme y-values (Ymin and Ymax) of Co, it can easily be identified whether this tool path row is affected by the mould design modification.

For an affected tool path row, it is likely that only some of the CL-points are affected by mould design modification (see Figure 5.5). The affected CL-points are those enclosed by the two extreme points. Since the y-values of all the CL-points in the same tool path row are the same, these two points can be identified by intersecting Co with the Y = y plane, where y is the y-value of that tool path row. Moreover, for an interference-free CL-point, there can be only one z-value for the given set of (x, y) values. As a result, the x-values (e.g. Xmin and Xmax in Figure 5.5) of the two intersection points can be used to identify whether the CL-point is affected. When all the affected CL-points in an affected tool path row are identified and removed, new CL-points will be calculated to replace the deleted ones.

The algorithm of identifying and replacing the affected CL-points is developed and its pseudo codes are listed as follows.

80

Chapter 5 Tool Path Regeneration for CL-Point Method

Tool path direction

Xmin
x

Xmax
x0

Ymax

Ymin
A

y Unaffected CL-points x Affected CL-points Extreme points

Figure 5.5 Extreme points and affected points

Identify_and_replace_affected_CL_points { /* Identify and replace the affected CL-points stored in fold, put the unaffected and the replaced CL-points to the new CL file fnew */ C (select the interference-free boundary of the modified region); Co (calculate the boundary of the affected CL-points from C); (Ymin, Ymax) (calculate extreme y values of Co);

81

Chapter 5 Tool Path Regeneration for CL-Point Method do { (x, y, z) (get x, y and z values of a CL- point from fold); if (Ymin y Ymax) { if (it is the first CL-point in this tool path row) { y-plane (create Y = y plane); (Xmin, Xmax) (calculate the x values of the intersection points of the y-plane and Co); } if ((x < Xmin) or ( x > Xmax)) { if ((the previous CL-point in this row is an affected CL-point) or (x < Xmin < Xmax < x0) or ( x0 < Xmin < Xmax < x)) { calculate CL-points for the affected region in this tool path row and add them to fnew; } add the current CL-point to fnew; } } else /* unaffected tool path row */ add the CL-point to fnew; } while not reaching the end of fold; } /* unaffected CL-point*/ /* affected tool path row*/

82

Chapter 5 Tool Path Regeneration for CL-Point Method

In the above algorithm, the boundary Co of the affected CL-points is first calculated. For a given point P on C, the corresponding point Po on Co can be calculated with Eq. (5.2). Connecting all the points of Po, the boundary Co can be created. The extreme y values (Ymin and Ymax) of Co are then calculated.

When a CL-point is read in, its y-value is compared with Ymin and Ymax to determine whether it is in an affected tool path row. If the CL-point is not in the affected tool path row, it is output to the new CL file fnew directly. Otherwise, when the first CLpoint in an affected tool path row is identified, Co intersects with a plane, which is parallel to the X-axis and with the same y-value of that CL-point. The x-values (Xmin and Xmax) of the intersection points are calculated. All the points of that row in the range of (Xmin, Xmax) are the affected CL-points and they must be replaced with new CL-points.

As shown in Figure 5.5, when a CL-point in an affected tool path row is identified as unaffected, there are two cases: the previous CL-point in that tool path row is identified as an affected or unaffected point. The first case indicates that all the affected CL-points prior to the current CL-point have been detected and removed (e.g. point A in Figure 5.5), and new CL-points need to be added before it.

The second case can be further classified into two situations by comparing the values of x, x0, Xmin and Xmax. Where x is the x-value of the current CL-point, x0 is the x-value of the previous CL-point in that row (see Figure 5.5). If (x < Xmin < Xmax < x0) or (x0 < Xmin < Xmax < x), the intersection points of the y-plane and Co is between the current and the previous CL-point, and new CL-points need to be added between them (see the

83

Chapter 5 Tool Path Regeneration for CL-Point Method

second tool path row in Figure 5.5); otherwise, no new CL-point needs to be added. The first situation occurs when the profile of the mould before the modification is relatively flat, and there is no CL-point between the two intersection points of the Y =
y plane and Co.

When replacing the affected CL-points, the y-values of the new CL-points are the same as that of other CL-points in this row. The z-value of the CL-point is calculated with the highest point method [Hwang 1992, Yang 1999], i.e., moving the cutter along the Z-axis, until the surface of the cutter is tangent to that of the workpiece. Since the profile of the modified region is different from that of the original one, the x-values of the newly generated CL-points may not be the same as the deleted ones. They must be calculated according to the machining tolerance.

As shown in Figure 5.6, when the point (x0, y, z0) is calculated and accepted as a CLpoint, the z values of z1 and z are calculated with the highest point method using x1 (=
x0 + ) and x (= x0 + /2) respectively, where is a predefined tool path step-size. The

linear center point's z-value (z) of the two points (x0, y, z0) and (x1, y, z1) are then worked out as z= (z0 + z1)/2. If the difference between z and z is within the given tolerance , the point (x1, y, z1) is accepted as the new CL-point. Here, equals to the machining tolerance value, which is specified by the user when the original tool path is generated. Otherwise, the step-size is shortened by half until the new point meets the given tolerance.

When calculating the replacing CL-points, they are calculated from the small x-value to the larger ones. After the replacing CL-points in a row are calculated, they cannot be

84

Chapter 5 Tool Path Regeneration for CL-Point Method

output to the new CL file directly. Otherwise, the mould may be over cut (see Figure 5. 7). These CL-points should be rearranged according to the machining direction, which

(x, z)

(x1, z1)

(z-z)

(x0, z0)
Z

(x, z)

Figure 5.6 Difference between z and z is determined by the x-values of two adjacent CL-points of the affected tool path row.

The CL-points in file fold that stores the CL-points before mould design modification are identified one by one, until all the CL-points are identified.

When the affected CL-points are replaced, the scallop height values for the new CLpoints need to be checked, and it is studied in the following subsection.

85

Chapter 5 Tool Path Regeneration for CL-Point Method

overcut

tool path direction

Figure 5.7 Overcut due to the wrong tool path direction

5.4.2 Calculating scallop height values and adding new CL-points to the modified region

As shown in Figure 5.8, in surface machining, there is a small distance between two adjacent tool paths, which is called the step-over size (g). The un-machined material between the two tool paths is called the cusp or scallop, and the upper limit height of the scallop is called the scallop height (h) [Suresh 1994]. Scallop height is an important parameter in tool path generation that decides the surface finish. It is decided by the machining step-over size g, the cutter radius R and the surface shape. In this research, it is assumed that the scallop height for the unaffected region is not greater than the given value. Therefore, only the scallop height values for the modified region need to be checked.

5.4.2.1 Machining scallop height and step-over size

86

Chapter 5 Tool Path Regeneration for CL-Point Method

h
y

h : Scallop height g : Step-over size : Distance between two intersection planes : Slop angle

Figure 5.8 Machining scallop height As shown in Figure 5.9, let the surface curvature radius in the direction that is perpendicular to the tool path direction be , from triangles OAC and OBD: AC = ( R)l 2

From triangle ADC and OAC: DC2 = R2 AC2, OA2 = ( - R)2 = AC2 + OC2

Consider OC = ( - h) CD, the distance l between two tangent points can be calculated with the given scallop height value h and the cutter radius R:

87

Chapter 5 Tool Path Regeneration for CL-Point Method

A B

C
R

D F

Figure 5.9 Calculation of tool path interval

4 R 2 ( h) 2 [( R) 2 R 2 ( h) 2 ] 2 l= ( R)( h)

(5.4)

where the sign of is decided by the shape of the surface, and > 0 when the surface shape is concave; < 0 when the surface shape is convex.

When the tool path is generated with the CL-point method, the distance (g) between two cutter center points are usually used as the step over size:

88

Chapter 5 Tool Path Regeneration for CL-Point Method

g=

( R)l =

4 R 2 ( h) 2 [( R ) 2 R 2 ( h) 2 ] 2 h

(5.5)

As shown in Figure 5.8, when the slope angle of the surface at the CC-point is , the distance between the two adjacent intersection planes is:

= g cos

(5.6)

Eqs. (5.5) and (5.6) are calculated based on , which is the curvature radius of the workpiece surface. When the tool path is generated with the CL-point method, cannot be calculated directly. Since the circle formed by the three corresponding cutter center points is in fact the offset of the one formed by the corresponding CC-points, and their radii have the following relation (see Figure 5.10):

0
B

Circle approximated from CL-points

Circle approximated from CC-points

Figure 5.10 Circles interpreted from CL and CC-points

89

Chapter 5 Tool Path Regeneration for CL-Point Method

= 0 + R
where 0 is the radius of the circle formed by three CL-points.

(5.7)

Given three points A, B and C (see Figure 5.10), the radius of their circumscribed circle can be calculated as in the following [Faux 1979]:

0 =
where a = AB and b = AC .

a b ab 2 a b

(5.8)

5.4.2.2 Algorithm for checking the scallop height value and adding new CL-points If the scallop height for the modified region exceeds the given value, new CL-points need to be added. Two methods can be used to add CL-points: one method is to regenerate an entire row of CL-points for that y-plane, including the modified as well as the unmodified region. The other method is to add new CL-points only for the modified region. With the second method, new CL-points only need to be added for the modified region, the tool path calculation and machining times are shorter. However, the generated tool path is not smooth and continuous with this method. In addition, the modified region is relatively small when compared with the entire region of a mould, and adding additional CL-points for the non-modified region will not affect the calculation and machining times significantly. Therefore, the first method is implemented when new CL-points need to be added, i.e., an entire row of CL-points that covers the modified and non-modified regions will be added.

90

Chapter 5 Tool Path Regeneration for CL-Point Method


If an odd number of tool path rows is to be added between two adjacent tool path rows, the sequence of the CL-points will change (see Figure 5.11), which will affect all the CL-points after this row. To avoid changing the sequence of the existing CL-points, the newly added CL-points will not be inserted between the existing CL-points. Instead, they will be added to the end of the CL file.

Original tool path

Added tool path

y x
Original tool path direction New tool path direction

Figure 5.11 Changes of tool path direction due to an odd number of tool path lines

Based on these concepts, the algorithm for checking the scallop height value and adding new CL-points has been developed and its pseudo codes are listed as follows:

Check_scallop_height_value_and_add_new_CL_points {

91

Chapter 5 Tool Path Regeneration for CL-Point Method /* Check the scallop height values for the modified region and add new CL-points if the scallop height exceeds the given value */ CL lines (connect the CL-points in the affected CL-point rows, including one row before and one row after the affected rows); Sample a set of points on the CL line of the regenerated CL-points Calculate intersection points by intersecting the CL lines with a set of planes that are parallel to the Y-axis;
for each row of the affected CL-points

{ Calculate and values from the intersection points;


if (the value is greater than the distance between the existing section planes) {

Identify the smallest value in this CL-point row; Add new CL-points according to the minimum value; } } }

To check whether the scallop height values for the replaced CL-points have exceeded the given value, the radii of the circles formed by adjacent tool path rows need to be calculated. Connecting the CL-points in the affected CL-point rows including rows before and after the affected rows, a set of CL curves are generated. The additional CL curves before and after the affected CL-point rows are used to calculate the radii and scallop heights for the CL-points in the first and last affected CL-point rows. A group of planes that are parallel to the Y-axis intersect with these CL curves, and a set of

92

Chapter 5 Tool Path Regeneration for CL-Point Method


intersection points are created. The radii of the circles in the direction that are perpendicular to the tool path direction can be approximated using Eq. (5.8) with these intersection points. The maximum distance between the two adjacent tool path planes are then calculated using Eq. (5.6). If is greater than the distance between the existing tool path planes, new CL-points will be added. The y-value of the intersection plane for the new CL-points is decided by the smallest value in that row.

5.5

Illustrative examples

Presented in the following are two examples for regenerating tool paths based on the proposed methodology. Shown in Figure 5.12a is a simple sculptured surface workpiece before mould design modification, and its tool paths that were generated with the commercial CAM software [Unigraphics 2000] are shown in Figure 5.12b. Shown in Figure 5.12c is the workpiece after removing some material from it. Figure

Figure 5.12a Part surface of a workpiece before modification

93

Chapter 5 Tool Path Regeneration for CL-Point Method

Figure 5.12b Tool paths of the workpiece before modification

Figure 5.12c Part surface of workpiece after modification

94

Chapter 5 Tool Path Regeneration for CL-Point Method


5.12d shows the new tool path generated with the tool path regeneration algorithms. The points in the figure show the replaced CL-points. Since the profile of the workpiece is changed, some new CL-points were added to maintain the required surface finish. To show the tool path clearly, points are not shown for these added CL-

Newly added tool path lines Replaced tool path points

Figure 5.12d Regenerated tool paths with replaced points and added lines after modification

95

Chapter 5 Tool Path Regeneration for CL-Point Method


points. The tool paths before and after modification with replaced points and newly added tool path lines are shown in Figure 5.12e. The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000] CAM functions and with the algorithm developed in this research are 22941.7 mm and 23726.8mm, respectively.

Original tool paths

Figure 5.12e Tool paths before and after modification with replaced points and added lines

96

Chapter 5 Tool Path Regeneration for CL-Point Method


Figure 5.13a shows another example of a front bezel industrial mould. There are 1852 trimmed surfaces in the original mould, and the surface area to be machined is 466,166 mm2. Some materials are added to the mould (Figure 5.13c). The area of the added surfaces is about 492 mm2. Figure 5.13b shows the tool path for the original mould part, which was generated with a commercial CAM software [Unigraphics 2000]. The machining parameters selected are as follows: the radius of the ball end-mill was 5 mm, the machining tolerance of the tool path was 0.01mm, and the scallop height was 0.03mm. The tool path was generated with the parallel tool path method. It took about 35 minutes to generate the tool path on a HP C240 workstation. Figure 5.13d shows the new tool paths with replaced points using the developed tool path modification algorithm. The new tool path used the same machining parameters. It took about 5 seconds to identify the affected and the unaffected CL-points. It can be found from the enlarged view in Figure 5.12d that the new tool path is interference-free and smooth. Figure 5.13e shows the new tool paths that combine the original and modified tool paths.

The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000] CAM functions and with the algorithm developed in this research are 64370.2 mm and 64882.6mm, respectively. It can be found from the two examples that the length of the tool path regenerated with the algorithm developed in this research is slightly larger than that generated by recalculating all the CL-points. This is because more tool path rows are added in the former method to maintain the scallop height requirement. However, the difference of the tool path lengths is relatively small. If the feed rate of the CNC machine is 500mm/min, the additional machining time for the second example is only about 1 minute.

97

Chapter 5 Tool Path Regeneration for CL-Point Method

Figure 5.13a Part surface of bezel mould before modification

Figure 5.13b Tool paths of bezel mold before modification

98

Chapter 5 Tool Path Regeneration for CL-Point Method

Surface modified

Figure 5.13c Part surface of bezel mold after modification

99

Chapter 5 Tool Path Regeneration for CL-Point Method

Newly added tool path lines Replaced tool path points

Figure 5.13d Regenerated tool paths with replaced points and added lines after modification

100

Chapter 5 Tool Path Regeneration for CL-Point Method

Original tool paths

Figure 5.13e Tool paths with replaced points and added line before and after modification

101

CHAPTER 6 TOOL PATH REGENERATION FOR THE CC-POINT METHOD

The tool path regeneration algorithm for the CL-point method has been studied in Chapter 5, this chapter will emphasize on regenerating tool path for the CC-point method. Since among all the CC-point methods, the section curve method is most widely used in mould machining, this chapter emphasizes on this method.

To regenerate tool path efficiently, the affected CL-points must be identified effectively. As studied in Chapter 5, when the tool path is generated with the parallel CL-point method, the affected CL-points can be easily identified by only comparing their x- and y-values. However, due to the variation of surface normal directions, when the tool path is generated with the CC-point method, the corresponding CL-points may not follow the same topology as the corresponding CC-points. This makes it difficult to identify whether a CL-point is affected or not. To solve this problem, it is proposed and proven in this chapter that, for a gouge-free tool path, the difference in signs of the two CL-points x- and y-values are the same as that of the corresponding CC-points. With this, it can be easily identified whether one CC-point is behind another one by comparing their corresponding CL-points x- and y-values. This makes it possible to efficiently identify whether a CL-point is enclosed by Co and whether a set of CLpoints belong to the same tool path row. Based on this concept, a tool path

102

Chapter 6 Tool Path Regeneration for CC-Point Method


regeneration algorithm for CC-point method has been developed and will be introduced in this chapter.

This chapter is organized as follows. The terms and the basic concepts are discussed in the first section. Based on the basic concepts, the methodology of identifying the affected CL-points is introduced in the second section. The algorithm of tool path regeneration for the CC-point method is introduced in the third section. Two examples of tool path regeneration for the CC-point method are illustrated in the last section.

6.1

Terms and basic concepts

Before studying the tool path generation algorithm, some terms and basic concepts will be introduced in this section.

CC-curve. As shown in Figure 6.1, a CC-curve is a curve on the workpiece surface,


along which CC-points are sampled. In mould machining, when the tool path is generated with the CC-point method, CC-curves are usually generated by intersecting part surfaces with a set of parallel planes. Without loss of the generality, it is assumed in this research that the intersection planes are parallel to the X-axis, i.e. the y-values of all the CC-points on the same CC-curve are the same. Since there are many surfaces in a mould, there may be more than one CC-curve with the same y-value. By comparing the y-value of a CC-curve with the extreme y-values of the modified regions boundary, it can be identified whether this CC-curve is affected by the mould design modification.

Tool path row. All the gouge-free CL-points corresponding to the CC-curves with the

103

Chapter 6 Tool Path Regeneration for CC-Point Method


same y-value are defined to be in the same tool path row (see Figure 6.1). Due to the variation of the surface normal directions, the y-values of the CL-points for the same tool path row may not be the same, which is different from that of the CL-point tool path generation method.

Tool path row

Tool path row

CC-curve

CC-point CL-point
Y X

Section plane

Figure 6.1 CC-point, CL-point, section plane, CC-curve and tool path row

6.2

Methodology of identifying the affected CL-points

To identify the CL-points inside Co efficiently, two propositions are made and proven in this section.

104

Chapter 6 Tool Path Regeneration for CC-Point Method


Proposition 6.1. Let S(t) be a curve on surface r(u, v), So(t) be the corresponding

curve on the offset surface. St and Sot are the derivatives along t direction for S(t) and
So(t) respectively (see Figure 6.2). If S(t) is interference-free, then St Sot 0.
n1

So So(t)

n2

ro

S(t)

Figure 6.2 Curves on surface r and offset surface ro

Proof: Let ro(u, v) be the offset surface of r(u, v) with the offset distance of R. ro(u, v)

can be expressed as:


ro(u, v) = r(u, v) + R n

(6.1)

where u and v are the surface parameters, n is the surface normal at point (u, v).

Let the parameters u and v be taken along the lines of curvature of r(u, v), and the principal curvatures at point P corresponding to u and v directions be 1 and 2 respectively. The derivative of ro along u direction can be expressed as:
rou = ru + R nu

(6.2)

105

Chapter 6 Tool Path Regeneration for CC-Point Method


From Rodrigues formula, nu can be expressed as:
nu = -1 ru

(6.3)

Insert Eq. (6.3) into Eq. (6.2),


rou = ru R 1 ru = (1 R 1) ru

(6.4)

Similarly,
rov = rv R 2 rv = (1 R 2) rv

(6.5)

Let S(t) = r(u(t), v(t)) be a curve on surface r(u, v). The corresponding curve on the offset surface is So(t) = ro(u(t), v(t)). The derivatives along t direction for S and So are:
St = ut ru + vt rv

(6.6)

Sot = ut rou + vt rov = ut (1 R 1) ru + vt (1 R 2) rv

(6.7)

Therefore,
St Sot = u t2 ru (1 R 1) + v t2 rv (1 R 2)
2 2

(6.8)

Since S(t) is interference-free, (1 R 1) 0 and (1 R 2) 0. Thus,


St Sot 0

(6.9)

There is one situation when (1 R i) = 0 (i = 1, 2): the concave surfaces curvature radius is the same as the cutter radius.

Proposition 6.1 indicates that the dot product of an interference-free curve and its offset tangent vectors are not less than zero. This can be further extended to the surface 106

Chapter 6 Tool Path Regeneration for CC-Point Method


points with Proposition 6.2.

Proposition 6.2. a and b are two points on surface r(u, v). O-XYZ is a coordinate

system with points a and b in the xz-plane. ao and bo are the surface offset points of a and b. Points a and b are interference-free and they are not undercut points, i.e., they can be seen from +Z direction. If Xb > Xa, then Xbo Xao, where Xp is the x coordinate value of point p.

Proof: Assume that S(t) is a curve generated by intersecting the workpiece surface

with the ZX-plane, points a and b are on S(t). Since the positions of ao and bo are decided by points a and b only, it can be assumed that there is no undercut and interference point on S(t). The possible undercut and interference portion on S(t) can be replaced with a new surface by moving a ball-end cutter along the surface without gouging (see Figure 6.3). The envelope of the cutter is the undercut-free and interference-free surface.

Figure 6.3 Removing undercut and interference from surface A

Since S(t) is a planar curve and there is no undercut point on it, the projection of St on the XY-plane is parallel to the x-axis. Without loss of the generality, it is assumed that

107

Chapter 6 Tool Path Regeneration for CC-Point Method


the projection of St on the XY-plane points to the +x direction. From Proposition 6.1, for every point on So(t), the projection of Sot on the X-axis will not point to the x direction. Therefore, Xbo Xao.

It can be found from Proposition 6.2 that for a gouge-free tool path, when the CCcurves are parallel to the ZX-plane, if the x-value of one CL-point is between the xvalues of the other two CL-points, the corresponding x-value of the CC-point is also between them or the three values are the same. With this proposition, it can be identified efficiently whether a CC-point is between or out of the two other points by comparing the x-values of their corresponding CL-points.

6.3

Tool path regeneration algorithms for the CC-point method

In tool path regeneration, to reuse the unaffected CL-points, the affected CL-points must be identified first. These points will then be replaced by new CL-points. Since the workpiece surface has been modified, the machining scallop height value with the original section curve interval may be greater than the given value. The scallop height values for the modified region need to be checked. If it is greater than the given value, new CL-points need to be added. After the CL-points are regenerated, gouging must be checked and removed. The new gouge-free CL-points can then be used to machine the modified workpiece.

With these concepts, the affected CL-points can be identified and replaced with the following four steps:

1) Identify the interference-free boundary C of the modified region and calculate the

108

Chapter 6 Tool Path Regeneration for CC-Point Method boundary Co of the affected CL-points from C; 2) Identify and replace the affected CL-points enclosed by Co; 3) Check scallop height values for the modified region. If the scallop height value is greater than the given value, add new CL-points; and 4) Detect and remove the gouging CL-points.

As in Chapter 5, when the boundary of the modified region is input, the program will check whether there are interference points on it. If the boundary is not interferencefree, identify a new interference-free boundary by enlarging the modified region. From the interference-free boundary C, the boundary Co that encloses the affected CL-points can be generated by connecting the CL-points corresponding to the CC-points on C. After C and Co are generated, the affected CL-points can be identified and replaced. The algorithm is introduced in the following subsection.

6.3.1 Identifying and replacing affected CL-points

There are a large number of CL-points in a CL file. To replace the affected CL-points, they must be identified first. However, it is not easy to identify whether a CL-point is inside the boundary of the affected CL-points (Co) directly when the tool path is generated with the CC-point method. An algorithm that can efficiently identify the affected CL-points needs to be developed.

Since the tool path is generated row-by-row, if a tool path row is not affected by mould design modification, all CL-points in this row are not affected, and they can be used to machine the modified mould directly. The affected CL-points only need to be identified for the affected tool path row. However, when the tool path is generated with

109

Chapter 6 Tool Path Regeneration for CC-Point Method


the CC-point method, it is not as easy to identify whether a CL-point belongs to an affected tool path row as for the CL-point tool path regeneration method.

Xmin x0 D

x1

x2

Xmax

Ymax H

G F E B Ymin

Tool path direction

Co

Y X Unaffected CL-points
Affected CL-points Extreme points

Figure 6.4 Identifying affected CL-points

As shown in Figure 6.4, with the section curve method, although the projection of CC curves on the XY-plane are parallel to the X-axis, the corresponding tool path rows may not be parallel to the X-axis due to the variation of surface normal directions. It cannot be identified whether two CL-points belong to the same tool path row by only

110

Chapter 6 Tool Path Regeneration for CC-Point Method


comparing their x-values. A CL-point may not be in the affected tool path row when its

x- and y-values are in the range of the extreme x- and y-values of Co (point A in Figure
6.4). On the contrary, even if the x- and y-values of the CL-point are out of the range of the extreme x- and y-values of Co, it may be in the affected tool path row (point B in Figure 6.4).

Let the extreme x- and y-values of the affected CL-points boundary Co be Xmin, Xmax, Ymin and Ymax (see Figure 6.4), x and y represent the x- and y-values of a CL-point to be studied. If a tool path row is affected by mould design modification, there is at least one CL-point which x- and y-values meet the following conditions: Xmin x Xmax and Ymin y Ymax

If the first CL-point (point D in Figure 6.4) in a tool path row with the x- and y-values meet the above condition, the corresponding CC-point can be identified, and the yvalue of the section plane for that CC-point can be identified. If the y-value of the section plane is out of the range of the modified regions boundary, the entire CL-point row is not affected. All the CL-points in this row can be used directly to machine the modified mould. Otherwise, some CL-points in this CL-point row are affected, and they must be identified and replaced by the new CL-points.

To identify all the CL-points in the same tool path row, the two end-points of this row need to be identified. Intersecting the workpiece boundary with the Y-plane, the two section points can be calculated, which are the end points of the CC-curve in this row. The corresponding CL-points are two end CL-points (points E and F in Figure 6.4) in this tool path row. From Proposition 6.2, the x-values of all the CL-points in this tool

111

Chapter 6 Tool Path Regeneration for CC-Point Method


path row are within the range of the x-values of these two CL-points. By comparing the x-values of the CL-points with that of these two CL-points, it can be identified whether the CL-point belongs to this tool path row.

When an affected tool path row is identified, the affected CL-points in this row can be identified and replaced. Since the affected CL-points are connected in sequence, the first and the last affected CL-points corresponding to the CC-points on the boundary C of the modified region in this row would need to be identified (points G and H in Figure 6.4). All the CL-points between these two points are affected CL-points. Intersecting the boundary C with the Y-plane, the two intersection points are the first and last affected CC-points in this row. The corresponding CL-points determine the first and last affected CL-points in this row.

Let the x-values of the CL-points on Co be x1 and x2 respectively. Without loss of generality, assume that x1 x2. From Proposition 6.2, if the x-value of a CL-point in this tool path row fulfils the condition of x1 x x2, it is an affected CL-point. Otherwise, it is not.

With these concepts, the pseudo-codes of an algorithm of identifying the affected CLpoints for CC-point method are developed and introduced in the following:

Identify_and_replace_affected_CL_points() { /* Regenerate tool path with the original CL-points stored in fold and save the regenerated CL-points to fnew */

112

Chapter 6 Tool Path Regeneration for CC-Point Method bd (select the workpiece boundary); C (select the interference-free boundary of the modified region); Co (Calculate the boundary of the affected CL-points from C); (Xmin, Xmax, Ymin and Ymax) (calculate extreme x and y values of Co); do { (x, y, z) (get x, y and z values of a CL-point from fold); if ((Xmin x Xmax) and (Ymin y Ymax)) { if (it is the first point in this tool path row fulfils the condition) { Calculate the y-value of the corresponding CC-point; Calculate the two CL-points corresponding to the two CC-points on bd in this row; if (it is an affected tool path row) { (x1, x2) (calculate the x values of the two CL-points on Co in this tool path row); } else add all CL-points in this row to fnew; } else { if ((x < x1) or ( x > x2)) /* in the affected tool path row */

113

Chapter 6 Tool Path Regeneration for CC-Point Method { if ((the previous CL-point in this row is an affected CL-point) or (x < x1 < x2 < x0) or (x0 < x1 < x2 < x)) { calculate CL-points for the affected region in this row and add them to fnew; } } add the CL-point to fnew; } } else /* x and y values are not in the range of the Cos extreme values */ { if ((the previous CL-point in this row is an affected CL-point) or (x < x1 < x2 < x0) or (x0 < x1 < x2 < x)) { calculate CL-points for the affected region in this row and add them to fnew; } add the CL-point to fnew; } } while not reaching the end of fold; }

In the above algorithm, when the workpiece boundary bd and the modified regions

114

Chapter 6 Tool Path Regeneration for CC-Point Method


boundary C are selected, the boundary Co is calculated from C. bd will be used to identify whether a CL-point belongs to a tool path row. The extreme x and y values of Co (Xmin, Xmax, Ymin and Ymax) are then calculated.

When a CL-point is read in from the old CL-point file fold, its x and y values are compared with the extreme values of Co (Xmin, Xmax, Ymin, Ymax). If they are not in this range, the CL-point is not affected, and it will be output to fnew to machine the modified mould. Otherwise, the CL-point will be further checked. If it is the first CLpoint of the tool path row in the range of Co, which can be identified by setting a sign to check whether the status of a point in the range is changed, the corresponding CCpoint is identified. From the y-value of the CC-point, the two CC-points on C in this row can be identified by intersecting the y-plane with C, and the corresponding CLpoints can be calculated. The two CL-points corresponding to the CC-points on the workpiece boundary bd in that tool path row can also be calculated with the same method. These two CL-points are used to identify whether a CL-point belongs to this tool path row. By comparing the y-value of the CC-point with the extreme y-values of the affected boundary C, it can be identified whether this tool path row is affected. If it is not affected, all CL-points in this tool path row can be output directly to machine the modified mould. By comparing the x-value of the CL-point with that of the CL-points corresponding to the CC-points on the boundary (bd) of the region to be machined, it can be identified whether it is in the tool path row.

If the first CL-point in an affected tool path row is identified, the x-values (x1, x2) of the CL-points on Co in this row are calculated. All the CL-points of this row in the range of (x1, x2) are the affected CL-points and they should be removed and replaced

115

Chapter 6 Tool Path Regeneration for CC-Point Method


by the new CL-points.

When a CL-point in an affected tool path row is identified as unaffected, there are two cases: the previous CL-point in that tool path row is identified as an affected or unaffected point. The first case indicates that all the affected CL-points prior to the current CL-point have been detected and removed, and new CL-points need to be added before it.

The second case can be further classified into two situations by comparing the values of x, x0, x1 and x2. Where x is the x-value of the current CL-point, x0 is the x-value of the previous CL-point in that row. If (x < x1 < x2 < x0) or (x0 < x1 < x2 < x), the intersection points of the Y-plane and C is between the current and the previous CCpoint, and new CL-points need to be added between them; otherwise, no new CL-point need to be added. The first situation occurs when the profile of the mould before the modification is relatively flat, and there is no CL-point between the two points on Co in this tool path row.

6.3.2 Adding new CL-points for the modified region

When all the affected CL-points in a tool path row have been identified and removed, new CL-points will be added to replace them. There are three key issues in adding new CL-points: machining tolerance, machining scallop height and gouging checking and removing, which will be discussed in the following subsections.

6.3.2.1 Calculate the CL-points with a given tolerance


When an affected tool path row is identified, the y-value of the corresponding CC-

116

Chapter 6 Tool Path Regeneration for CC-Point Method


curve is calculated. A Y-plane with this y-value will be created. Intersecting the workpiece with the Y-plane, a set of CC-curves is created. A mould usually consists of many surfaces, when a plane intersects with the workpiece, there may be several CCcurve segments. To obtain the correct sequence of CC-points, the CC-curves need to be connected end-by-end first. A bubble algorithm is used to reorder these curves by comparing the x-values of the section curves end points. Trimming off the curves that are out of the modified regions boundary C, the remaining curves are the affected CCcurves in this row. CC-points are then sampled along these CC-curves. The step length

L of the sampled CC-points is calculated as follows (see Figure 6.5) [Hosaka 1992]:

L = 2 (2 )

(6.10)

where is the given machining tolerance and is the radius of curvature of the CCcurve.

Figure 6.5 Step length L As shown in Figure 6.6, when the difference of two CC-points normal vectors is too large, an over cut occurs, which is called convex gouging. Convex gouging may occur in two cases: the normal vectors of two surfaces are not continuous or the distance between the two adjacent CC-points is too large, which leads to large differences

117

Chapter 6 Tool Path Regeneration for CC-Point Method


between the two normal vectors. To eliminate convex gouging, the surface normal n1 and n2 for the CC-points are calculated. If the difference between them is larger than a given value and the two CC-points belong to the same surface, additional CC-points will be added between them (see Figure 6.6a). If the two CC-points belong to two different surfaces and the distance between them is very small, an arc will be added for the tool path. The arc is located at the plane defined by n1 and n2 with the arc radius being the cutter radius and the arc center being the common vertex point of the two CC-curve (see Figure 6.6b).
n1

n2

Workpiece surface (a) Convex gouging caused by large normal difference

n1

n2

Workpiece surface

(b) Convex gouging caused by discontinuity surface normal Figure 6.6 Convex gouging

118

Chapter 6 Tool Path Regeneration for CC-Point Method 6.3.2.2 Calculating scallop height and adding new CL-points to the modified region
As the profile is modified, the scallop height for the modified region needs to be checked. With Eq. (5.6), the allowed distance value () between the two adjacent intersection planes can be calculated, where g should be replaced by l in the equation for the CC-point method. By comparing the value with the distance of the two adjacent Y-planes, it can be identified whether the interval between the two section planes is too large. If it is, new CL-points need to be added. As discussed in Chapter 5, when new CL-points are to be added, an entire row of CL-points is added, and the newly added CL-points will be included at the end of the CL file.

6.3.2.3 Detecting and removing gouging CL-points


After all the CL-points are calculated, gouging is detected. Since the boundary C is interference-free, gouging does not need to be detected for the unaffected CL-points, only the newly added CL-points need to be detected. The algorithm proposed by Choi and Jun [Choi 1989] is used to detect and remove gouging CL-points in this research.

6.4 Illustrative examples


Presented in the following are two examples for regenerating tool paths based on the proposed methodology. Figure 6.7a shows a simple sculptured surface workpiece before mould design modification. Figure 6.7b shows the tool paths for the original workpiece. Figure 6.7c is the workpiece after adding some material to it. Figure 6.7d shows the new tool path generated with the tool path regeneration algorithms. The points in the figure show the replaced CL-points. Since the profile of the workpiece is changed, two new CL-points were added to maintain the required surface finish. The tool paths before and after modification are shown in Figure 6.7e. To show the tool

119

Chapter 6 Tool Path Regeneration for CC-Point Method


path clearly, points are not shown for these added CL-points. The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000] CAM functions and with the algorithm developed in this research are 12014.1 mm and 12235.7mm, respectively.

Figure 6.7a Part surface of a workpiece before design modification

Figure 6.7b Tool paths of the workpiece before design modification

120

Chapter 6 Tool Path Regeneration for CC-Point Method

Figure 6.7c Part surface of workpiece after design modification

Figure 6.8a shows another example of an industrial mould.

The surface area to be

machined is 244,318 mm2. Some materials are added to the mould (Figure 6.8c). The area of the added surfaces is about 8693 mm2. Figure 6.8b shows the tool path for the original mould part. The machining parameters are set as follows: the radius of the ball end-mill was 5 mm, the machining tolerance of the tool path was 0.01mm, and the scallop height was 0.03mm. The tool path was generated with the CC-curve tool path method. It took about 15 minutes to generate the tool path on a HPC240 workstation. Figure 6.8d shows the new tool paths with replaced points using the developed tool path modification algorithm. The new tool path used the same machining parameters. It took about two seconds to identify the affected and the unaffected CL-points. It can be found from the enlarged view in Figure 6.8d that the new tool path is interferencefree and smooth. Figure 6.8e shows the new tool paths that combine the original and modified tool paths. The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000] CAM functions and with the algorithm developed in this research are 43762.4mm and 43787.9mm, respectively.

121

Chapter 6 Tool Path Regeneration for CC-Point Method

Newly added tool path line

Replaced tool path points

Figure 6.7d Regenerated tool paths with replaced points and added lines after design modification

122

Chapter 6 Tool Path Regeneration for CC-Point Method

Original tool paths

Figure 6.7e Tool paths with added lines before and after design modification

123

Chapter 6 Tool Path Regeneration for CC-Point Method

Figure 6.8a Part surface of T-cover mold before design modification

Figure 6.8b Tool paths of T-cover mold before design modification

124

Chapter 6 Tool Path Regeneration for CC-Point Method

Modified area

Figure 6.8c Part surface of T-cover mold after design modification

125

Chapter 6 Tool Path Regeneration for CC-Point Method

Replaced tool path

Figure 6.8d Regenerated tool paths with replaced points after design modification

126

Chapter 6 Tool Path Regeneration for CC-Point Method

Original tool paths

Figure 6.8e Tool paths of T-cover before and after design modification

127

CHAPTER 7 SYSTEM IMPLEMENTATION AND CASE STUDIES

The methodologies and algorithms of mould design modification and tool path regeneration for the CL-point and CC-point methods have been studied in the previous chapters. With these methodologies and algorithms, a computer-aided Mould Design

Modification and Tool Path Regeneration System has been developed, and this system
will be introduced with examples in this chapter.

7.1

The computer-aided mould design modification and tool path regeneration system

The computer-aided Mould Design Modification and Tool Path Regeneration System has been developed to modify mould design and regenerate tool paths automatically with little or no user interaction for 3-axis mould machining. The architecture and the modules of the system will be introduced in this section. Before introducing the system, the platform of the system is described.

7.1.1 The platform

Although the developed methodologies and algorithms can be implemented in many systems, an appropriate development platform and support software will help effectively develop the computer-aided mould design modification and tool path 128

Chapter 7 System Implementation and Case Studies


regeneration system. In this research, the HP-C240 workstation, which uses the HPUX operating system and the HP Visual User Environment (VUE), has been chosen to implement the algorithms. The operation system provides powerful graphical interfaces and application tools, which simplifies many of the involved tasks.

Since the developed system requires a lot of geometric functions for modelling and tool path generation, a commercial CAD/CAM system, Unigraphics (UG)

[Unigraphics 2000] was selected. UG is an interactive 3D Computer-Aided Design and Computer-Aided Manufacturing (CAD/CAM) system and it is widely used in the mould manufacturing industry for mould design and tool path generation. UG has the following characteristics and functions that make it suitable for developing the current system:

(1)

It provides an easy UG/Open API (User-Functions) interface between UG and third party developers. User-Functions (UFun) consist of a large set of user callable functions and subroutines that access the Unigraphics Graphics Terminal, File Manager, and Database. UFun programs can be written in Fortran, C and C++ programming languages. The developed system was written in C.

(2)

It is easy to assign certain types of information to parts and geometry, including solids, faces, and edges. Using object attributes and part attributes, the developers and the end users can assign specific information, meaningful in mould design application, to most objects and parts.

(3)

The CAM functions can automatically generate a Cutter Location Source File (CLSF) with the specified machining parameters, which can be used to drive NC

129

Chapter 7 System Implementation and Case Studies


machines. To generate the tool path, the user needs to select the geometry describing the part and set some machining parameters such as cutter diameter, feed-rate, cutting depth, scallop height/step-over size, etc.

7.1.2 The architecture As shown in Figure 7.1, the computer-aided mould design modification and tool path regeneration system includes three parts: (1) user interface, (2) mould design modification and (3) tool path generation and regeneration. The system works as follows. The old plastic part and the designed mould are stored in the part file. Unique names are assigned for each of them. The new plastic part is also stored in the part file. When the user selects the new plastic part, the system automatically identifies the differences between the old and the new plastic parts. The affected mould inserts are identified from the modified profile. Then the system highlights and prompts the affected mould inserts one by one whether the insert has been machined. The system will modify the mould design automatically according to the designed mould, the old and new plastic parts and the mould machining status: 1) If the mould has not been machined, the mould design is modified directly according to the new plastic part and the designed mould; 2) If the mould has been machined, the mould design will be modified with two different methods based on whether material needs to be added to the mould insert: if material needs to be added, an insert and a pocket will be designed; otherwise, material will be removed directly from the designed mould.

After the mould design is modified, the modified mould design and the boundary of the modification region are output. Since the mould design and tool path generation are

130

Chapter 7 System Implementation and Case Studies

Mould Design Modification

New product model

Old product model

Identify the boundary and profile of the modified product

Mould machined? N Y

Add metal?
Y

Tool path generated? N


User Interface

Create pocket & insert

Modify mould design

Modify mould design

Modify mould design

Generate tool path for this insert

Generate tool path for modified area

Generate tool path

Any repeated mould? N

Tool path regeneration

Tool Path Generation & Regeneration

End

Figure 7.1 Framework of the mould modification and tool path regeneration system

131

Chapter 7 System Implementation and Case Studies


usually done by different engineers in two different departments, the system will not generate or regenerate the tool path until the user clicks the tool path generation or regeneration buttons. Tool path generation/regeneration methods are different according to how the mould design is modified, and the tool path generation and mould machining status:

1) If the mould has not been machined, the tool path will be generated or regenerated according to whether an original tool path has been generated: a) If the tool path has not been generated, new tool paths will be generated for the modified mould directly. The commercial CAM functions will be used to generate the original tool path. b) If the tool path has been generated for the old mould, the tool path will be regenerated with the developed algorithm, and the user will be asked to input the CLSF file name and the tool path generation method for the original tool path.

2) If the mould has been machined, tool path will be generated for the added insert (when metal needs to be added to the machined mould) or the modified region (when metal is removed from the designed mould). In practice, several sets of the mould may be manufactured after the product design is confirmed. In this case, the tool path will also be regenerated for the entire mould with the developed algorithm.

The system includes three main parts: (1) the user interface, (2) the mould design modification module, and (3) the tool path generation and regeneration module. The

132

Chapter 7 System Implementation and Case Studies


inputs to the system includes the old and the new plastic parts, the designed mould for the old plastic part, the mould machining status, and the generated tool paths for the old mould if it has been generated. The outputs are the modified mould design and the generated or regenerated tool paths. These modules will be introduced in the following subsections.

7.1.3 The user interface

The user interface allows the user to input the mould machining status. If the tool path has been generated for the designed mould, the user will also input the tool path file name and the tool path generation method through the user interface. When a command is given by the user, the system performs the mould design modification or the tool path regeneration function. The user can input the parameters for mould design modification and tool path regeneration with the menus provided by the system.

7.1.4 The mould design modification module

The mould design modification module modifies the mould design according to the difference between the old and the new plastic parts and the mould machining status. When the Mould Design Modification command is given, a dialog is prompted and the user is asked to select the new plastic part (see Figure 7.2). After the new plastic part is

Figure 7.2 New product selection

133

Chapter 7 System Implementation and Case Studies


selected, the old plastic part and the designed mould inserts are identified automatically from their attribute names. The solid bodies of the material to be added to or removed from the designed mould are created automatically. The corresponding mould inserts that are affected by plastic part design modification are then identified. If material is to be removed from the designed mould, the designed mould is subtracted by the solid body of the material to be removed.

If material is to be added to the designed mould, the affected mould insert is highlighted and the user is asked to input the machining status of this mould insert (see Figure 7.3). If the mould has not been machined, the mould design is modified by uniting the solid body with the designed mould insert directly. If the mould insert has been machined, an insert and a pocket are designed. Before the insert and the pocket are designed, the possible interference between the pocket and the cooling holes will be checked, which decides whether a through or blind pocket will be designed.

Figure 7.3 Input mould machining status

7.1.5 The tool path generation and regeneration module

When the mould design is modified, the tool path will be generated or regenerated based on the mould machining status. If a new tool path is to be generated, the UGs CAM functions will be used. If the tool path is regenerated with the developed

134

Chapter 7 System Implementation and Case Studies


algorithm and the user gives the tool path regeneration command, a dialog will display and asks the user to select the boundary of the modified region (see Figure 7.4). As the original tool path is generated with commercial CAM functions, the user also needs to select the workpiece and the boundary of the profile to be machined, which is normally the boundary of the workpiece.

Figure 7.4 Item selection

The user is then asked to input the old and new tool path file names and the machining parameters, which include the cutter radius, the machining tolerance, the scallop height and the tool path generation method (CC- or CL-method) (see Figure 7.5). After all these parameters are input, the original tool path will be displayed to make sure that the correct tool path file is input. The system will then regenerate the tool path automatically and output the regenerated tool path to the new tool path file.

135

Chapter 7 System Implementation and Case Studies

Figure 7.5 Input machining parameters

7.2

Case studies

The computer aided Mould Design Modification and Tool Path Regeneration System has been tested with several industrial parts. Two examples will be illustrated in this section. Both plastic parts for the two examples were designed with another CAD system (Pro/Engineer), and they were transferred to the UG system using STEP. To implement the functions of the developed system, it is assumed that for both cases, tool paths have been generated for the cavities but the cavities have not been machined yet, while the cores have already been machined.

7.2.1 Case 1

Figures 7.6a and b show the old and new plastic parts of a dairy clear door, respectively. Five portions of the plastic part are modified (the magenta regions in the figure). The cap shape region affects both the core and the cavity. The added two

136

Chapter 7 System Implementation and Case Studies

Figure 7.6a The old plastic part of a dairy clear door

Figure 7.6b The new plastic part of a dairy clear door

rectangular pockets and two circular holes affected the core only. Figures 7.6c and d show the old and the modified cavity. Material needs to be added to the designed cavity. As the cavity has not been machined, the design is modified by uniting the solid of metal with the old cavity directly. The cyan region in the figure shows the modified region. Figure 7.6e shows the original core. To show the cooling, ejector and screw holes clearly, these holes are shaded while the other surfaces of the core are displayed in wire frames. Figure 7.6f shows the modified core. The cap shape region shows the area that material is removed from the core. As the core has been machined,

137

Chapter 7 System Implementation and Case Studies


four inserts are designed for the areas that need to add material. Two cylindrical inserts are designed for the circular shape blind holes. Two through cylindrical pockets are designed for them and shoulders are designed at the bottom of the inserts and the pockets. Two rectangular shape inserts are designed for the two rectangular pockets. A through pocket is designed for the left side insert, while a blind pocket is designed for the right side. A screw and screw hole are designed to fix the insert of the blind pocket, while a shoulder is designed for the through pocket and the insert.

Figure 7.6c The original cavity of a dairy clear door

Figure 7.6d The modified cavity of a dairy clear door

Figure 7.6e The original core of a dairy clear door

138

Chapter 7 System Implementation and Case Studies


As the core has been machined, tool paths only need to be generated for the modified region. UGs CAM function is used to generate tool paths for these regions.

Figure 7.6f The modified core of a dairy clear door

139

Chapter 7 System Implementation and Case Studies


The tool path for the cavity is regenerated with the developed algorithm. A ball endmill with radius of 5mm was used to machine the old cavity. The tool path was generated with the CC-point method, i.e., the CC-curves were parallel to the x-axis. The machining tolerance and scallop height values are 0.003mm and 0.05mm, respectively. There are 75 trimmed surfaces in the original cavity, and the surface area to be machined is 154,529.8mm2. The area of the modified region is 2,545.3 mm2.

Figure 7.7a Tool paths of the workpiece before modification Figure 7.7a shows the tool path for the old cavity. It took about 9 minutes to generate the original tool path on a HP-C240 workstation. Figure 7.7b shows the regenerated tool path with the replaced CL-points using the developed tool path regeneration algorithm. The same machining parameters were used to regenerate the tool path. It took less than 2 seconds to identify the affected and unaffected CL-points, and the tool path regeneration time is about 3 minutes. Figure 7.7 c shows the new tool paths that

140

Chapter 7 System Implementation and Case Studies

Newly added tool path lines Replaced tool path points

Figure 7.7b Regenerated tool paths with replaced points and added lines after modification

combine the original and the modified tool paths. The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000] CAM functions and with the algorithm developed in this research are 68527.5 mm and 69201.8mm, respectively.

141

Chapter 7 System Implementation and Case Studies

Original tool paths

Figure 7.7c Tool paths with replaced points and added lines before and after modification

7.2.2 Case 2

Figure 7.8 shows another example. Figures 7.8a and b show the old and the new plastic part of a riser. The magenta color region shows the three modified areas. The modified

142

Chapter 7 System Implementation and Case Studies


region A affects both the core and the cavity. Material needs to be added to the designed cavity, while the designed core needs to remove material. The modified regions B (two rectangular blind holes) affect the core only, and material needs to be added to the core for these two affected areas. Figures 7.8c and d shows the old and the modified cavities respectively. As the cavity has not been machined, the design is modified by uniting the cavity with the material to be added. The modified area is displayed with in cyan in the figure. Figure 7.8e shows the old core. The core has been machined. For clarification, the cooling, ejector and screw holes are partially shaped as

Figure 7.8a The old plastic part of a riser

B Figure 7.8b The new plastic part of a riser

143

Chapter 7 System Implementation and Case Studies


in the last example. Figure 7.8f shows the modified core. The modified regions are shown in magenta. The shaded surface of the middle area shows the profile after material is removed from the core. Two rectangular blocks are designed for the material to be added. Since through pockets will interfere with the cooling holes, blind pockets are designed for both of them. As there is not enough space, screws cannot be designed to fix the insert to the core. Both inserts will be bound to the core with the point welding process.

Figure 7.8c The original cavity of a riser Figure 7.8d The modified cavity of a riser

Figure 7.8e The original core of a riser

144

Chapter 7 System Implementation and Case Studies

Figure 7.8f The modified core of a riser

Since the core has been machined, the tool path will be generated for the modified regions only. UGs CAM functions were used to generate tool path for the three regions. The tool path for the cavity is regenerated with the developed algorithm.

145

Chapter 7 System Implementation and Case Studies

Figure 7.9a Tool paths of the workpiece before modification

Figure 7.9a shows the old tool path for the original cavity. There are 202 trimmed sculptured surfaces in the original mould, and the surface area to be machined is 180,052 mm2. The area of the modified region is 5418.5 mm2. The tool path for the old mould was generated with the CL-point method, i.e., the tool path rows are parallel to the x-axis. A ball end-mill with radius of 5mm was used to machine the old cavity. The machining tolerance and scallop height values were 0.003mm and 0.05mm, respectively. It took about 11 minutes to generate the tool path on a HP C240 workstation. Figure 7.9b shows the regenerated tool paths with the replaced CL-points using the developed tool path regeneration algorithm. The same machining parameters were used to generate the tool path. It took about 3 seconds to identify the affected and

146

Chapter 7 System Implementation and Case Studies

Newly added tool path lines Replaced tool path points

Figure 7.9b Regenerated tool paths with replaced points and added lines after modification

unaffected CL-points, and the tool path regeneration time is about 5 minutes. Figure 7.9c shows the new tool paths which combine the original and the modified tool paths. The lengths of the tool paths that are regenerated with Unigraphics [Unigraphics 2000]

147

Chapter 7 System Implementation and Case Studies


CAM functions and with the algorithm developed in this research are 61634.2 mm and 62296.1mm, respectively.

Original tool paths

Figure 7.9c Tool paths with replaced points and added lines before and after modification

148

CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS

8.1

Conclusions

To shorten the new product development time, the design of a new product may be changed many times during the mould design and manufacturing processes. With the existing CAD systems, the mould design cannot be modified automatically if the design parameters are lost when the part is transferred from one CAD system to another. When the mould design is changed, the tool path that covers the affected region needs to be regenerated. With the existing CAM systems, when the tool path is regenerated and error-prone in NC-machining, all the CL-points will need to be recalculated, which is very time-consuming. To solve these problems, this research has developed a computer-aided Mould Design Modification and Tool Path Regeneration

System. With this system, when the design of a plastic part is changed, the mould
design can be modified automatically with different methods according to the mould machining status. The tool path can then be regenerated efficiently by reusing the unaffected CL-points.

New theories, methodologies and algorithms for mould design modification and tool path regeneration have been developed in this research. The system has been tested with several industrial parts. It is proven that the system can efficiently modify mould design and regenerate tool paths according to the mould machining status. The

149

Chapter 8 Conclusions and Recommendations


contributions of this research can be summarized into two areas: mould design modification and tool path regeneration, as will be discussed in the following subsections.

8.1.1

Mould design modification

A new mould design modification module that can automatically modify the mould design for parametric and non-parametric parts according to the mould machining status has been developed in this research. With the developed module, the mould design modification does not rely on the product parameters that increase with the complexity of the product and the mould. Therefore, the developed module is not sensitive to the complexity of the plastic part and the designed mould.

The mould design may need to be modified at different mould machining stages. Different methods have been developed to modify the mould design according to the mould machining status. When an insert needs to be designed to add extra material, the system automatically detects the interference between the pocket to be designed and the existing holes. Different pockets, inserts and fasteners can be designed based on the interference detection result.

8.1.2

Tool path regeneration

New tool path regeneration algorithms have been developed in this research. With these algorithms, the affected and unaffected CL-points can be identified efficiently. The affected CL-points are then removed and replaced by new CL-points. Since the unaffected CL-points are used to machine the modified mould directly, the tool path regeneration time can be reduced greatly.

150

Chapter 8 Conclusions and Recommendations


Four propositions have been proposed and proven in the thesis for tool path regeneration. The propositions indicate two important properties of a gouge-free tool path: 1) the affected CL-points are enclosed by a boundary of CL-points that correspond to the interference-free boundary of the modified region; 2) when projected onto the XY-plane, if one CL-point is before another, their corresponding CC-points follow the same topology. With these propositions, the affected CL-points can be efficiently identified by only noting their x- and y-values.

The tool path regeneration algorithm is related to the tool path generation method. Since the tool path generation methods can be categorized into CC- and CL-point methods, two different tool path regeneration algorithms have been developed for them respectively. The tool path regeneration algorithms were implemented with the parallel tool path and CC-curve tool path generation methods. These algorithms can easily be applied to other tool path generation methods.

8.2

Recommendations

The system developed in this research has solved some of the mould design modification and tool path regeneration problems. However, it does not cover all aspects in these areas, more research is needed in mould design modification, tool path generation and regeneration, as will be discussed in the following subsection.

8.2.1 Mould design modification

This research emphasizes on modifying the mould design for the core and cavity. Sometimes, the change of the plastic part design may affect the parting line and the parting surfaces. More research is needed to develop algorithms that can efficiently

151

Chapter 8 Conclusions and Recommendations


and optimally identify the parting line for the modified profile based on the given parting direction and the unaffected parting lines.

The plastic design modification may also lead to new undercuts. More research is needed to develop algorithms to automatically and optimally design sliders and lifters based on the given parting direction and the designed mould structures.

When an insert is to be created, the optimal design of the insert which affects the quality of the moulded plastic part least is also an issue to be studied further.

8.2.2 Tool path generation

With the existing algorithms, when a tool path row is to be planned, the maximum scallop height value for this row needs to be identified point by point. A challenging problem is how to identify the point with maximum scallop height value from the surface and the CC-curve information directly. This can be applied to generating tool paths efficiently and selecting cutters more accurately without generating the tool paths.

8.2.3 Tool path regeneration

With the system developed in this research, as it is assumed that the original tool path has been developed by other systems (e.g., the commercial CAM software), the user needs to input the original machining parameters before the tool path can be regenerated. If these parameters can be stored in the original tool path file when it is generated, these parameters can be retrieved directly from the file, which will reduce the tool path regeneration errors caused by user input.

152

Chapter 8 Conclusions and Recommendations


In this research, tool path regeneration was focused on 3-axis machines with ball endmills. It will even be more time consuming and error-prone to generate tool paths for 4-axis and 5-axis CNC machines. More research is needed to develop algorithms that can efficiently identify and replace the affected CC-points from the generated CLpoints for them. Flat and fillet end-mills are also used in tool path generation for mould machining, especially for rough machining. New algorithms need to be developed to identify and replace the affected CL-points for these types of cutters.

153

RELATED PUBLICATIONS

1. Zhang, L. P., Fuh, J.Y.H. and Nee, A.Y. C. Tool path regeneration for mold modification. Computer-Aided Design, 35(9), pp.813-823, 2003.

2. Zhang, L. P., Fuh, J.Y.H. and Nee, A.Y. C. A Tool Path Re-Generation Algorithm For Die & Mold Machining. Proceedings of the 10th International Manufacturing Conference, Oct. 2002, Xiamen, China, Panel1-205.

3. Zhang, L. P., Nee, A.Y. C. and Fuh, J.Y.H. An efficient CC-curve tool path regeneration algorithm for sculptured surface machining, Journal of Engineering

Manufacture, Proceedings of IMechE, Part B218, pp. 389-402, 2004.

4. Computer-aided mould design modification for product design change, to be submitted to International Journal of Production Design.

154

REFERENCES

[Bedi 1997] Bedi, S., Gravelle, S. and Chen, Y. H., Principal curvature alignment technique for machining complex surfaces, Transactions of the ASME, 119, pp.756765, 1997.

[Bobrow 1985] Bobrow, J. E., NC machine tool path generation from CSG part representations, Computer- Aided Design, 17, pp.69-76, 1985.

[Broomhead 1986] Broomhead, P. and Edkins, M., Generating NC data at the machine tool for the manufacture of free-form surfaces, International Journal of Production Research, 24(1), pp.1-14, 1986.

[Chan 2003] Chan, W. M., Yan, L., Xiang, W. and Cheok, B. T., A 3D CAD knowledge-based assisted injection mould design system, International Journal of Advanced Manufacturing Technology, 22, pp.387395, 2003.

[Chen 1993] Chen, L.L., Chou, S., Y. and Woo, T. C., Parting directions for mold and die design, Computer-Aided Design, 25, pp.762-768, 1993.

[Chen 1995] Chen, L. L., Chou, S. Y. and Woo, T. C., Partial visibility for selecting a parting direction in mold and die design, Journal of Manufacturing Systems, 14(5), pp.31930, 1995.

155

References
[Chin 1996] Chin, K. S. and Wong, T. N., Knowledge-based evaluation for the conceptual design development of injection molding parts, Engineering Applications of Artificial Intelligence, 9(4), pp.359-376, 1996.

[Choi 1988] Choi, B. K., Lee, C. S., Hwang, J. S. and Jun, C. S., Compound surface modelling and machining, Computer-Aided Design, 20(3), pp.127-136, 1988.

[Choi 1989] Choi, B. K. and Jun, C. S. Ball-end cutter interference avoidance in NC machining of sculptured surfaces, Computer-Aided Design, 21, pp.371-378, 1989.

[Choi 1997] Choi, B. K., Kim, D. H., and Jerard, R. B., "C-space approach to tool path generation for die and mould machining", Computer-Aided Design, 29, pp.657-669, 1997.

[Choi 1998] Choi, B. K. and Jerard, R. B., Sculptured surface machining, Kluwer Academic, The Netherlands, 1998.

[Ding 2001] Ding, X. M., Fuh, J. Y. H., Lee, K. S. Interference detection for 3-axis mold machining, Computer-Aided Design, 33, pp.561-569, 2001.

[Elber 1994] Elber, G. and Cohen, E., Toolpath generation for freeform surface models, Computer-Aided Design, 26 (6), pp.490496, 1994.

[Faux 1979] Faux, I. D. and Pratt, M. J., Computational geometry for design and manufacture, Halsted Press, New York, 1979.

156

References
[Glaeser 1999] Glaeser, G., Wallner, J., Pottmann, H., Collision-free 3-axis milling and selection of cutting tools, Computer-Aided Design, 31, pp.225-232, 1999.

[Hosaka 1992] Hosaka, M., Modeling of Curves and Surfaces in CAD/CAM, Springer-Verlag, New York, 1992.

[Huang 1994] Huang, Y. and Oliver, J. H., Non-constant parameter NC tool path generation on sculptured surfaces, International Journal of Advanced Manufacturing Technology, 9, pp.281-290, 1994.

[Huang 1996] Huang, Y. S., Webster, P. D. and Dean, T. A., An image detection approach to NC rough-cut milling from solid models, International Journal of Machine Tools and Manufacturing, 36, pp.1321-1333, 1996.

[Hui 1992] Hui, K.C. and Tan, S., Mold design with sweep operation a heuristic search approach, Computer-Aided Design, 24, pp.81-91, 1992.

[Hwang 1992] Hwang, J. S., "Interference-free tool-path generation in the NC machining of parametric compound surfaces", Computer-Aided Design, 24, pp.667676, 1992.

[Jensen 1993] Jensen, C. G. and Anderson, D. C., Accurate tool placement and orientation for finish surface machining, Journal of Design and Manufacturing, 3, pp.251-261, 1993.

157

References
[Jun 2002] Jun, C. S., Kim, D. S. and Park, S., A new curve-based approach to polyhedral machining, Computer-Aided Design, 34 (5), pp.379-389, 2002.

[Kim 1995] Kim, K. I. and Kim, K., "A new machine strategy for sculptured surfaces using offset surface", International Journal of Production Research, 33, pp.1683-1697, 1995.

[Kruth 1997] Kruth, J. P., Willems, R. and Lecluse, D., A design support system using high level mould objects, CIRP International Conference and Exhibition on Design and Production of Dies and Moulds, pp.39-44, Turkey, 1921 June, 1997.

[Kuragano 1992] Kuragano, T., Fresdam system for design of aesthetically pleasing free-form objects and generation of collision-free tool path, Computer Aided Design, 24(11), pp.573-581, 1992.

[Lai 1994] Lai, J. Y. and Wang, D. J., A strategy for finish cutting path generation of compound surfaces, Computers in Industry, 25, pp.189-209, 1994.

[Lartigue 2001] Lartigue, C., Thiebaut, F. and Maekawa, T., CNC tool path in terms of B-spline curves, Computer-Aided Design, 33, pp.307-319, 2001.

[Lee 1997] Lee, K. S., Fuh, J. Y. H, Zhang, Y. F., Nee, A. Y. C. and Li, Z., IMOLD: an intelligent plastic injection mold design and assembly system, Proceedings of the 4th International Conference On Die and Mould Technology, pp.3037, Malaysia, 46 June 1997.

158

References
[Lee 2003] Lee, E. Contour offset approach to spiral toolpath generation with constant scallop height, Computer-Aided Design, 35, pp.511-518, 2003.

[Lin 1998] Lin, A. C. and Liu, H. T., Automatic generation of NC cutter path from massive data points, Computer-Aided Design, 30, pp.79-90, 1998.

[Lin 2001] Lin, J. C., From Optimum gate design of free form injection mould using the adductive network, International Journal of Advanced Manufacturing

Technology, 17, pp.297304, 2001.

[Loney 1987] Loney, G. C. and Ozsoy, T. M., NC machining of free form surface, Computer-Aided Design, 1, pp.85-90, 1987.

[Maekawa 1997] Maekawa, T., Cho, W. and Patrikalakis, N. M, Computation of selfintersection of offsets of Bezier surface patches, Journal of Mechanical Design, 119, pp.275-283, 1997.

[Maeng 2003] Maeng, S. R., Baek, N. and Choi, B. K., A Z-map update method for linearly moving tools, Computer-Aided Design, 35(11), pp. 995-1009, 2003

[Majhi 1999] Majhi, J, Gupta, P, Janardan, R., Computing a flattest, undercut-free parting line for a convex polyhedron, with application to mold design, Computer Geometry, 13(4), pp.22952, 1999.

159

References
[Marciniak 1991] Marciniak, K., Geometric modeling for numerically controlled machining, Oxford University Press, Oxford, New York, 1991.

[Mok 2001] Mok, C. K., Chin, K. S. and Ho, K. L., An interactive knowledge-based CAD system for mould design in injection moulding processes, International Journal of Advanced Manufacturing Technology, 17, pp.2738, 2001.

[Nee 1997] Nee, Y. C., Fu, M.W., Fuh, Y. H., Lee, K. S. and Zhang, Y. F., Determination of optimal parting directions in plastic injection mold design, Annals of CIRP, 46, pp.429-432, 1997.

[Ong 1995] Ong, S. K., Prombanpong, S., Lee, K. S., An object-oriented approach to computer-aided design of a plastic injection mould, Journal of Intelligent Manufacturing, 6, pp.110, 1995.

[Pottmann 1999] Pottmann, H., Wallner, J., Glaeser, G. and Ravani, B., Geometric criteria for gouge-free three-axis milling of sculptured surfaces, Transactions of the ASME, Journal of Mechanical Design, 121, pp.241-248, 1999.

[Ravi 1990] Ravi, B. and Srinivasan, M. N., Decision criteria for computer-aided parting surface design, Computer-Aided Design, 22, pp.11-18, 1990.

[Ravi 1997] Ravi, B, Intelligent design of gating channels for casting, Material Science Technology 13, pp.785790, 1997.

160

References
[Rosen 1994] Rosen, D. W., Towards automated construction of mould and die design, Proceedings ASME Computers in Engineering Conference, 1, pp.317-326, 1994.

[Seiler 1997] Seiler, A., Balendran, V. and Sivayoganathan, K., Tool interference detection and avoidance based on offset nets, International Journal of Machine Tools and Manufacturing, 37, pp.717-722, 1997.

[Shin 1993] Shin, K. H. and Lee, K., Design of side cores of injection molds from automatic detection of interference faces, Journal of Design and Manufacturing, 3, pp.225-236, 1993.

[Suresh 1994] Suresh, K. and Yang, D. C. H., Constant scallop-height machining of free-form surfaces, Journal of Engineering for Industry, 116, pp.253-258, 1994.

[Sarma 1997] Sarma, R. and Dutta, D., An integrated system for NC machining of multi-patch surface, Computer-Aided Design, 29, pp.741-750, 1997.

[Takeuchi 1989] Takeuchi, Y., Sakamoto, M., Abe, Y. and Orita, R., Development of a personal CAD/CAM system for mould manufacturing, Annals of the CIRP, 38(1), pp.429-432, 1989.

[Tang 1995] Tang, K., Cheng, C. C. and Dayan Y., Offsetting surface boundaries and 3-axis gouge-free surface machining, Computer-Aided Design, 27, pp.915-927, 1995.

161

References
[Unigraphics 2000] Unigraphics on-line documentation, Version 17.0, UG Solutions Inc. Maryland Heights, 2000.

[Wang 1996] Wang, Z., Lee, K. S. and Fuh J. Y. H., Optimum ejector system design for plastic injection moulds, International Journal of Computer Applications in Technology 9(4), pp.211218, 1996.

[Weinstein 1996] Weinstein, M. and Manoochehri, S., Geometric influence of a mould part on the draw direction range and parting line locations, Journal of Mechanical Design, 118 (1), pp.2939, 1996.

[Weinstein 1997] Weinstein, M., Manoochehri, S., Optimum parting line design of moulded and cast parts for manufacturability, Journal of Manufacturing Systems, 16(1), pp.112, 1997.

[Willmore 1959] Willmore, T. J. An Introduction to Differential Geometry, Clarendon Press, Oxford, 1959.

[Yang 1999] Yang, D. C. H. and Han, Z., Interference detection and optimal tool selection in 3-axis NC machining of free-form surfaces, Computer-Aided Design, 31, pp.303-315, 1999.

[Ye 2001] Ye, X. G., Fuh, J. Y. H., Lee, K. S., A hybrid method for recognition of undercut features from moulded parts, Computer-Aided Design, 33, pp.102334, 2001.

162

References
[You 1995] You, C. F. and Chu, C. H., An automatic path generation method of NC rough cut machining from solid models, Computers in Industry, 26(2), pp.161-173, 1995.

[Yu 1996] Yu, D. Y., Duan, Z. C., Zhang, W. and Liu, J. N., Flat-end cutter path generation for a sculptured surface for 3-axes NC machining, Journal of Materials Processing Technology, 57, pp.65-69, 1996.

[Yuan 1993] Yuan, Z. S., Li, D. Q., Chen, X. and Ye, X. G., Integrated CAD/CAE/CAM system for injection moulding, Computing & Control Engineering Journal, 4 (6), pp.277-279, 1993.

[Zhang 1997] Zhang, Y. F., Lee, K. S., Wang, Y., Fuh, J. Y. H. and Nee, A. Y. C., Automatic slider core creation for designing slider/lifter of injection moulds, CIRP International Conference and Exhibition on Design and Production of Dies and Moulds, pp.3338, Turkey, 1921 June 1997.

[Zhang 2002] Zhang, Y. F., Liu, H. H. and Lee, K. S., Automated Generation of Lifters for Injection Moulds, International Journal of Advanced Manufacturing Technology, 19, pp.537543, 2002.

163

You might also like