University of South Australia School of Advanced Manufacturing and Mechanical Engineering

DEFECT DETECTION IN HIGH PRESSURE DIE CASTING PRODUCT USING IMAGE PROCESSING TECHNOLOGY

by Wahab, Md Said in Student 10: 100008279

Under the supervision of Dr. Lingxue Kong

A thesis submitted in partial fulfilment of the requirement for the degree of Master of Engineering (Advanced Manufacturing Technology)

July, 2003

Disclaimer

I declare the following to he my

(l\\

n \\'UrK. unless nthlT\\'isc rl'i"c-rl'Jlcl·'1.

l,:ni\'ersity's policy on plagi;lrism.

;1'·

,klt:}-. ;

h-_ il,,-

Abstract

In this thesis an Automated Visual Inspection (AVI) system for ckl'eel dL'lc'L'li"i1 ('( 111~h

Pressure Die Casting (HPDC) product has been dcveloped, The sy,lelll i, l'i1;lblin~

1(\

detect external structural defects of the product particularly for crack and Iwk. This lhesi, also describes detailed procedures based on image processing enhancement/preprocessing,

segmentation,

codingifeature

technnl,,~}

extractinn

IIhich include and

I III age

analysis/classification/interpretation, The commercial Matlab image processing kls been used to implement these procedures, An intelligent approach based on

mllrpllOlog~

and

fuzzy logic is proposed to detect such structural defects on the surface.

The proposed method has been implemented and tested on a number of 111'1)(' I'rllducl. The results suggest that the method provide an accurate identification tn lhe dekelS and be extended for further application, The results also suggest that the system 1I('rk lIell enough to help die casting manufacturer to improve the value or die Glsling I,rllduc\ lil;11 they can obtain from automated inspection system,

A few suggestions have been discussed for future research to address the limiwlillns of this research.

II

Acknowledgments

I wish to express m)' sincerest thanks to my advisor. Dr. Lingxue f--:.!lng rllr hi, gllld;lllc'C', paticncc, and support throughout this thcsis, especially through the tillle, \'. hc'lI thL' writing bccamc di fficul t.

Special thanks to my wife Anizam Mohamed Yusofand my childrcn !l1ms) id;lh & 1);lni;tI for their love and encouragement. I dedicate this thcsis to my parcl1b It)r their ullll;lgging faith in mc, \Vithout thcir dcvotion, support. patiencc, and encourageillent thr()lIglh llit thc' pcriod of study. this work would not have bcen possible,

III

Contents 1. Introduction .............................................................. . 1.1. The background of the problem .............................................. . 1.2. Problem addressed to die casting .................................... . 1.3. Suggested solution .................................................... ..

.'"

IA. Aim/objective of the project.....................................................

4

1.5. Scope and limitation.. .........................................................

4

1.6. Thesis roadmap.................................................................... 5

2. Litcl'atLlI'c Rcvic'v ........................................................ 6 2.1. Die Casting Process.............................................................

6

2.1.1. Hot chamber machine..........................................

7

2.1.2. Cool chamber machine..........................................

g

2.2. The advantages of die casting process...................................

10

2.3. Application of die casting............................................ ............. 12 2A. Die casting defect.....................................................

13

2.5. Current inspection quality control process............................

17

2.6. Overview of Human Vision andlvlachine Vision .............................. 18 2.6.1. Human Vision................................................

. ......... 18

2.6.2. Machine Vision................................................

19

2.7. An artificial intelligence in vision system for decision making ........... ..

,),,1

2.8. Summary ............................................................. .

24

3. Imagc Acquisition and Storagc ........................................ 25 3.1. Introduction ............... .

IV

6. Conclusions and "ccoI11mcndations .................................. M, 6.1. Introduction ..................................................... . 6.2. Achievement. ................................................... . 6.3. Recommendation for future work ............................................... 7()

Appcndixcs Appendix A

Typical defects in die casting ............................................ 75

Appendix B

Comparison of human vision. machine vision and compliter \·isioll.:;n

Appendix C

Matlab Code....................................................

\,1

..

~~

Figures and Tables Figures

Figure 2.1

Hot chamber machine .... '" .................................................... 7

Figure 2.2

Cool chamber Machine ......................................................... 9

Figure 2.3

Commercial High Pressure Die Casting machine ........................... 9

Figure 2.4

Process economic for die casting .............................................. 10

Figure 2.5

Die casting end user 2003 ........................ '" ....................... " .. 12

Figure 2.6

Sample of crack defect. ......................................................... 15

Figure 2.7

Sample of hole I ................................................................. 16

Figure 2.8

Sample of hole 2 ................................................................. 16

Figure 2.9

Industrial manufacturing cell with vision system .......................... 20

Figure 3.1

Vision inspection system procedure .......................................... 26

Figure 3.2

Front lighting system ............................................................. 28

Figure 3.3

Backlighting system .................. " .............. ' ......................... 28

Figure 3.4

Structured lighting system ...................................................... 29

Figure 3.5

(a) Image matric presented in pixel (b) image appear in the screen and (c) graycale value corespond to the intensity of the image ................ 31

Figure 4.1

(a) Original image. and (b) image after wiener filter. ...................... 36

Figure 4.2

A histogram computed from image after wiener process .................. 39

Figure 4.3

Threshold is set at> 60 ......................................................... 40

Figure 4.4

Threshold is set at> 109 ....................................... .

Figure 4.5

A negative image after morphological operation ........ .

Figure 4.6

Background of the image...... ... ... ......... ............................

48

Figure 4.7

Image after background subtraction process..............................

48

VII

... .41 . ... 47

Figure 4.8

Final image after segmentation operation ................................... -1')

Figure 4.9

Pseudocolor index image for matrix A ......................................

~

Figure 4.10

Pseudocolor index image for three objects ..................................

~2

Figure 4.11

Ellipse orientation............................................................. (,()

Figure 5.1

A final image after decision making procedure ............................. ()-1

Figure 5.2

Result from analysis . image 1.................................................. hh

Figure 5.3

Result from analysis image 2 .................................................. 67

~

I

~

Tables

Table 4.1

Sub morphological operation .................................................. -13

Table 3. J

Type of extracted feature from the imagc .................................... 5""

Table 3.3

Extracted feature for object 1 ................................................. 56

Table 3.4

Extracted feature for objcct 2 ................................................. 5()

Table 3.5

Extracted feature for object 3 ................................................. 57

...

\,111

Nomenclature

Symhol

Explaination

1->

The conjugate of matrix A

A'

The transpose of matrix A

A " B

The matrix procluet of A ancl B

/

','

'is not an clement of E

'is an clement of

c

'is a subset of

La

. IX

I fl tro .'" uction This chapter will begin by giving a rather detailed explanation of the

b~ll'kgr
LOW 'PRESSURE

10

"""-----,\

\SAND 1

_J

"'ell '", LOW

W

0::

PRESSLJRf

D·'

_ _ _.....1-::-_

1

10

.......I.I ..,--____ L. _____- '

I 3

2

10 10 10' NUMBER OF COMPONENTS

10

5

10°

Figure 2.4 Process economic/or die casting [5}

Figure 2.4 shows a chart for economic selection of several casting process based on the cost per component against the number of components. It has sho\\J1 that the HPDC process is the most economic process for high volume production (10 3 and more).

10

Chapter 2. Literature Review

followed by low pressure, permanent mold and finally sand casting. Other advantages are

\51:

High-speed production Die casting provides complex shapes within closer tolerances than many other mass production processes. Little or no machining is required and thousands of identical castings can be produced before additional tooling is required.

Dimensional accuracy and stability Die casting produces parts. while maintaining close tolerances.

Sirengih and weighl Die cast parts are stronger than plastic injection moldings of the same dimensions. Thin wall casting is stronger than those possible with other casting methods. In addition. because die casting is not consisted of separate parts welded or fastened together. the strength is that of the alloy rather than the joining material.

Mulliplefinishing techniques Die cast parts can be produced with small textured surfaces, and they are easily plated or finished with a minimum surface preparation.

5iimplitied assemhly Die castings provide integral fastening elements, as bosses and studs. Hoks can he cored and made to tap drill sizes, or external treads can be cast.

11

Chapter 2. Litera/ure Review

2.3

Application of Die Casting

Each of the metal alloys available for die casting offers particular advantages to the components. Therefore, the selection of materials for components varies depending on their function and properties of the components. Figure 2.5 shows the application of several alloys in different industry.

Die Casting End Users 2003

~::r I

t..ler~~:, ) . - . :

~.. r ..=;·~·;;: .-.('!-~;!. L~--

.

=-;·.\'... ·r

. :.1.

..

-J

~ Il;:n::·-.~"I:~ ~.~.iiiiiiiiiiiii__

_l1·~···~.;\/·~J-.U·-::·. ~·~.·I~ .. ~ ~ ···..:u!...":, It;:· ....~ .fl'· . :""el~.

1

I!L I!!

1,--·----'---·--'---

:

\/.:"'.: 'S",'" :,,":.111';"'.';. --:':'

r ..,·.j.·,, .::'::" . "."

.

i. IlJ AI :;003

III h

;,0003 :2 '-'1g :1003

,"U--------------------

.... .::::h;:":r:.~ ;

":':J11I~·.

U:_'I~

....... ~::.

I:..·tl·.~

.._ _ __

1· ...-II:......t·.:::·'I'·::~···'I·~ ~

Aulvl"u~ .'':- At:~h lC:It.t4

..

I!

.j

Aut'':f.'':':''f'C''~: II\'!', AU"); .....): ,'C :';IP'~"Ii-t

-

u

,.:.n :. . 'j.:-.

:'>;:'.'-«"it-aln ~:nl'r

Ms .i";' . . . . . . .'. ., . . . . . . . _. . (;

~

_~

-F . . . . "

i

-:;:=1

T

!,

:

______ ... :

••

~_

,

..... ' ....

~ n::;:~::.: •••

*

i

***141

.. .gas'

0'-..:

Percentages

Figure 2.5 Die casting end user 2003 [5J

Some of the applications for these alloys are [5]:

Aluminum Die casting Aluminum is lightweight and while processed high dimensional stability of complex shapes and thin walls can be achieved. Aluminum has good corrosion resistance and

12

Chapter 1. Literatllre Revie'W

mechanical properties, high thermal and electrical conductivity, as well as strength at high temperature. Examples are the transmission housing in motor industries and the base for a Teleprinter.

Zinc Die casting

Zinc is the easiest alloy to cast, it offers high ductability, high impact strength and

IS

easier platable. Zinc is economical for small parts and has a low melting temperature which prolongs the die life; Examples are the toys, small size components and the components that have a good surface finish such as chrome-plated.

Brass Die casting

This alloy possesses high hardness, high corrosion resistance and the highest mechanical properties of alloy cast. It offers excellent wear resistance and dimensional stability. with strength approaching that of a steel part; Example is the plumber hardware such as water taps and shower mixer.

Magnesium Die casting

The easiest alloy to machine, magnesium has an excellent strength-to-weight ratio and is the lightest alloy commonly die cast; Example is a mobile phone case.

2.4

Die Casting Defects

Die casting defects can be classified into external and internal defects. External Jdects are normally related to surface defects which can be visually identified \\hereas the internal defects are those defects that built internally and can be seen \\hen certain operation has been carried out such as X-Ray, Ultra sound or any destructive process. 13

Chapter 2. Literature RevieJ.1"

Both internal and external defects also have many different categories depending on geographical areas and when or how the defect has been accrued.

Some ofthe typical defects are as follows [10];

Surface defect:

Those defects are commonly called cold flow

Laminations:

Defect from layers of metal forming during the process

Gas Porosity:

Internal porosity from trapped gas of various kinds.

Blisters:

Surface manifestation of trap gas

Flow porosity:

Surface or internal porosity from poor pressure condition

Shrink porosity:

Porosity from the volume change as the metal changes state.

Sinks:

Surface depression from shrinkage porosity problems

Leakers:

Porous sections of the casting

Cracks:

Visible and not-so-very visible cracks in the casting

Inclusions:

Foreign material that may cause machining problems.

Holes:

Similar to crack, but it is a hole.

Sample images of the above defects are listed in Appendix A.

2.4.1

Defectfor detection

As mentioned in chapter one, only two types of defect will be analyzed in this project. that is, cracks and holes. Cracks along with sinks and leakers are among the defects that have shrinkage problems as their common underlying cause. However. while shrinkage is the root problem for a great many cracks, there are a number of other factors that can also cause a crack.

14