-
OPTIMIZATION OF THE POLISHING
PROCEDURE USING A ROBOT ASSISTED POLISHING EQUIPMENT
av
Marielle Gagnolet •••• 2008 01 25
Handledare: Sabina Rebeggiani Examinator: Bengt-Göran Rosén
Ett examensarbete utfört enligt kraven för Högskolan i Halmstad
för en Magisterexamen i Teknisk Produkt- och
Produktionsförbättring
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 1/43
AAKK NNOOWWLL EEDDGGMM EENNTTSS
I would like to thank all the people who helped me during these
5 months, to adapt myself in an other country and also to carry out
my project; especially:
� Mr. Bengt-Göran Rosén, Professor and director of the lab, for
his warm welcome to Halmstad University, and for sharing his
knowledge to help me.
� Ms. Sabina Rebeggiani, PhD student, for her patience, for
always being available when I needed, and for helping me in this
project.
� Mr. Zahouani, Professor at the ENISE for giving me the
opportunity of carrying out my internship in Halmstad
University.
� Mrs. Levy, English teacher at the ENISE, for being always
available in case of problems.
� Mr. Frédéric Cabanettes, who helped me with all the procedure
to come here and for being available along these 5 months.
� Mr. Stefan Rosén and Ms. Karin Westerberg from Toponova AB for
their instructions and guidance in performing optimal measurements
on the stylus and the interferometer.
� Mr. Kim Lorenzen, Mr. Jens Grønbæk, Mr. Lars Sørensen (from
Strecon A/S) and all the employees of Strecon A/S for their welcome
to Strecon and their collaboration in this project.
� Mr. Alf Sandberg from Uddeholm Tooling AB, for his trust
leaving us leading this project.
� Mr. Zlate Dimkovski, Mrs. Bibbi Johansson, Mrs. Monica
Lindström and all the people of the university for their help along
this semester.
� All the students I met during my stay in Sweden and who made
this experience become unforgettable and humanly enriching for
me.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 2/43
AABBSSTTRRAACCTT
Today, manual polishing is the most common method to improve the
surface finish of mould and dies for e.g. plastic injection
moulding, although it is a cumbersome and time-consuming process.
Therefore, automated robots are being developed in order to speed
up and secure the final result of this important final process.
The purpose of this thesis is to find out some clues about the
influence of different parameters for the polishing of a steel
grade called Mirrax ESR (Uddeholm Tooling AB) using a Design of
Experiment. The report starts with a brief description of
mechanical polishing (the techniques and polishing mechanisms) and
ends up with the optimization of the polishing procedure with a
polishing machine, the Strecon RAP-200 made by Strecon A/S.
Even if all the runs of the Design of Experiment couldn’t be
carried out, the surfaces studied revealed some information about
the importance of the previous process (turning marks not removed)
and about the link between the aspect of the surfaces and the
roughness parameters.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 3/43
SSUUMM MM AARRYY
AAKK NNOOWWLL EEDDGGMM EENNTTSS
.......................................................................................................................................
1
AABBSSTTRRAACCTT..........................................................................................................................................................................................................................................................................................................................22
SSUUMM MM AARRYY
...........................................................................................................................................................
3
II NNTTRROODDUUCCTTII
OONN.................................................................................................................................................
4
PPAARRTT II :: PPRROOJJEECCTT SSCCEENNAARRII OO
........................................................................................................................
5
1. PRESENTATION OF
COMPANIES........................................................................................................................
5 11..11)) HHAALLMMSSTTAADD UUNNIIVVEERRSSIITTYY
.......................................................................................................................
5 11..22)) UUDDDDEEHHOOLLMM TTOOOOLLIINNGG AABB
....................................................................................................................
5 11..33)) SSTTRREECCOONN
AA//SS........................................................................................................................................
6 2. THE
PROJECT...................................................................................................................................................
7 22..11)) TTHHEE PPRROOJJEECCTT EENNVVIIRROONNMMEENNTT
................................................................................................................
7 22..22)) DDEESSCCRRIIPPTTIIOONN OOFF WWOORRKK
.........................................................................................................................
7
PPAARRTT II II :: LL II TTEERRAATTUURREE
SSTTUUDDYY.......................................................................................................................
8
1. A FEW POLISHING
METHODS............................................................................................................................
8 2. MECHANICAL POLISHING
................................................................................................................................
8 22..11)) PPOOLLIISSHHIINNGG EEQQUUIIPPMMEENNTT
........................................................................................................................
8 22..22)) PPOOLLIISSHHIINNGG TTEECCHHNNIIQQUUEESS,, ““
RRUULLEESS””
........................................................................................................
9 33.. POLISHING
MECHANISMS...............................................................................................................................
11 33..11)) HHIISSTTOORRYY
.............................................................................................................................................
11 33..22)) AABBRRAASSIIOONN AANNDD PPOOLLIISSHHIINNGG
.................................................................................................................
11 33..33)) CCHHAARRAACCTTEERRIIZZAATTIIOONN OOFF AA
PPOOLLIISSHHEEDD
SSUURRFFAACCEE....................................................................................
14
PPAARRTT II II II :: PPOOLL II SSHHII NNGG
SSTTAANNDDAARRDDSS............................................................................................................
16
1. NECESSITY OF
STANDARDS............................................................................................................................
16 2.
MEASUREMENTS...........................................................................................................................................
16 22..11)) MMEEAASSUURREEMMEENNTTSS
DDEEVVIICCEESS...................................................................................................................
16 22..22)) MMEEAASSUURREEMMEENNTTSS AANNDD
RREESSUULLTTSS..........................................................................................................
17
PPAARRTT II VV:: DDEESSII GGNN OOFF EEXXPPEERRII MM EENNTTSS
((DDOOEE))............................................................................................
18
1.
PRESTUDY.....................................................................................................................................................
18 11..11)) TTHHEE MMAACCHHIINNEE
.....................................................................................................................................
18 11..22)) TTHHEE
PPAARRAAMMEETTEERRSS...............................................................................................................................
19 11..33)) CCHHOOIICCEE OOFF TTHHEE MMAATTRRIIXX
......................................................................................................................
20 11..44)) AASSSSIIGGNNMMEENNTT OOFF TTHHEE FFAACCTTOORRSS TTOO
TTHHEE
CCOOLLUUMMNNSS...............................................................................
21 2. EXPERIMENTS AT
STRECON...........................................................................................................................
23 22..11)) PPAARRTT 11:: PPRREEPPAARRAATTIIOONN OOFF TTHHEE
EEXXPPEERRIIMMEENNTTSS
.....................................................................................
23 22..22)) PPAARRTT 22 :: FFLLOOWW OOFF WWOORRKK
....................................................................................................................
24 3. ANALYSIS OF THE
RESULTS...........................................................................................................................
25 4. RESULTS AND
DISCUSSION............................................................................................................................
25
CCOONNCCLL UUSSII OONN OOFF TTHHEE PPRROOJJEECCTT
................................................................................................................
31
FFUUTTUURREE
..............................................................................................................................................................
31
PPEERRSSOONNAALL CCOONNCCLL UUSSII
OONN.............................................................................................................................
32
RREEFFEERREENNCCEESS....................................................................................................................................................
33
AAPPPPEENNDDII
XX........................................................................................................................................................................................................................................................................................................................3355
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 4/43
II NNTTRROODDUUCCTTII OONN Currently in my 4th year at the
ENISE, I made this year my second industrial placement. After a
first experience in Andrezieux-Bouthéon (France) in the car
industry working in the production, I decided this year to have an
experience in a research lab and also to go abroad. It was for me
the opportunity to discover a working environment very different
from what I already knew, to deal with another language and so
improve my English. It was also for me a challenge to go there
alone, meet other people from different countries, and of course to
discover another country and its culture, both in every day life
and at work. Attracted by Scandinavian countries and given the
partnership between the ENISE and the Halmstad University in
Sweden, I made this internship in its mechanical lab, managed by
Professor Bengt-Göran Rosén. This lab deals with a few PhD
students, working for different companies like Volvo, Uddeholm
Tooling AB… I worked with one of them, Sabina Rebeggiani, on a
project about the optimization of a polishing machine.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 5/43
PPAARRTT II :: PPRROOJJEECCTT SSCCEENNAARRII OO
1. Presentation of companies
11..11)) HHaallmmssttaadd UUnniivveerr ssii ttyy
Halmstad is located on the Western coast of Sweden and counts
86.000 permanent inhabitants. Tourists, businessmen, members of the
armed forces and the students ensure the prosperity of the
entertainments and the culture throughout the year. The university
offers to Halmstad new prospects interesting for the future, where
knowledge and competences play an increasingly important part. The
university offers at the same time single and famous programs and
lectures. Nearly 7000 students study Medical and Social Sciences,
Economy, Behavioral Science, Health as well as Engineering. The
majority of the programs are directed towards the innovation and
the entrepreneurship. Many projects and theses emerge, with terms,
from a productive cooperation between the students and the
industry.
11..22)) UUddddeehhoollmm TTooooll iinngg AABB ((UUTTAABB)) The
first firm of UTAB for steel production was found in 1668 in
Stjärnfors in Värmland (Sweden). From small mills and workshop
hammers in the late 17th century, UTAB grew into a steel, power and
forest products group.
Nowadays, UTAB is the world’s leading supplier of tooling
material and related services; it is a multinational company which
acts locally in more than 100 countries. UTAB is working for tool
manufacturers, tool users and their suppliers about steel in
general, that is to say from the production of different kind of
steel, to their treatment and their fields of application.
Halmstad
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 6/43
11..33)) SSttrr eeccoonn AA//SS
Strecon A/S is a Danish company located in Sønderborg (Denmark).
It is specialized in development, engineering, manufacture, sale
and service of tooling solutions based on the strip-winding
technology.
The principle of this technology is to wind a thin strip of
high-strength steel (0.1 mm thickness) around a core of hardened
tool steel or tungsten carbide while maintaining a controlled
winding tension. An optimal stress distribution is obtained by
varying the winding tension from strip layer to strip layer, and
the state of stress in the wound coil section of the product is
equal to a conventional pre-stressing system with "several hundred"
stress rings. This strip-winding principle gives the product a
loadability that in many cases is 50% higher than conventional
compression ring solutions.
If the strip-winding technology is the main product of Strecon,
the company also has a strong competence in:
� FEM modeling of precision forging and diamond synthesis tools
and processes.
� In-house developed models and material tests to optimize the
FEM simulations.
� In-house manufacture of key processes, e.g. strip-winding,
turning, grinding, and finishing.
With this project of a polishing machine, Strecon is trying to
acquire knowledge in high surface finishing, the polishing of tool
steel.
They built a first prototype this year and are working with this
in order to build another machine in 2008.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 7/43
2. The project
22..11)) TThhee pprr oojj eecctt eennvvii rr oonnmmeenntt
UTAB produces among other things a steel grade called Mirrax ESR
(old name “Stavax Supreme”) which is commonly used for plastic mold
steel. Mirrax ESR is an Electro Slag Remelted (ESR) material
hardened and tempered to a hardness of 50HRC. Given the use of this
steel, the polishing operation is essential. Currently, manual
polishing is the most common method to improve the surface finish
of mould steel, although automated robot processes are being
developed to speed up and secure the final result of the polishing
process. A most desirable outcome from a polishing process is to
achieve a consistent surface finish from tool to tool in order to
have a reliable production process.
Moreover, Strecon is building a polishing robot: the Strecon
RAP-200 (Robot Assisted Polishing) which is still under development
(Figure 1). The purpose of UTAB in this project is to find an
automated polishing process for their steel, able to give
mirror-like surfaces. With reference to Strecon, this project
should increase their knowledge of this process (influence of the
different parameters) and allow them to optimize their polishing
machine.
22..22)) DDeessccrr iipptt iioonn ooff wwoorr kk
Title of the project: Optimizing the polishing procedure for
plastic mould steel by using a robot assisted polishing
equipment
The objective of this project is to more closely evaluate a
robot assisted polishing equipment by examining the surface quality
during various polishing procedures for the plastic mould steel
Mirrax ESR. The final objective is to give guidelines on optimum
polishing procedures for this steel grade, which is of a very high
quality for moulds but difficult to polish.
Different steps should schedule this project:
� Literature study on polishing techniques and polishing
mechanisms
� Characterization of polishing standards
� Study of Design of Experiments in order to prepare experiments
at Strecon
� Experimental polishing work performed at Strecon A/S in
Sønderborg, Denmark
� Surface characterization after various preparation steps via
grinding and polishing at Halmstad University
� Analysis of the results of the Design of Experiments with the
Software Modde 8 [Umetrics Inc, www.umetrics.com]
Figure 1: Strecon RAP-200 Picture from Strecon
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 8/43
PPAARRTT II II :: LL II TTEERRAATTUURREE SSTTUUDDYY I started
this project by reading many scientific articles, in order to
acquire knowledge about the polishing process since it was a
process almost unknown for me. First, I studied the different
techniques used within industry to then focus more on the
mechanical polishing which is directly linked to the project.
Finally, I also tried to find information on polishing mechanisms
from a metallographic point of view.
1. A few polishing methods There are many different polishing
techniques used in the industry, such as:
� Laser polishing [1]
� Chemical mechanical polishing [2], [3]
� Magnetic polishing [4]
� Electropolishing [5], [6]
� Electron Beam Irradiation [7]
� Electrical Discharge machining [8]
� Mechanical polishing...
These techniques are presented in Appendix I, apart from
mechanical polishing which is the technique used in this project.
Therefore it will be more explained in the next chapter.
2. Mechanical polishing Mechanical polishing is the removal of
material to produce a scratch-free, specular surface using fine
abrasive particles. [12]
22..11)) PPooll iisshhiinngg eeqquuiippmmeenntt [[1122]]
Polishing is typically done using polishing cloths, specially
designed lapping plates, or stones. Polishing with a cloth or
lapping plate requires the use of free abrasive, and is a very low
damage process when performed properly. There are 4 basic types of
abrasives:
� SiC: Rough lapping, rarely used for applications that require
smooth surface finishes. � Al 2O3: Sharp, angular structure. Used
when fine surface finishes are required. � B4C: Harder than the
other abrasives (except from diamond); blocky crystal
structure.
Excellent removal rates; used for fast removal with moderate
surface quality needed. � Diamond: Hardest material, sharp, angular
structure. High removal rate and high
surface finishing quality.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 9/43
Figure 1: Polishing stones [source: www.joke.de]
Figure 2: Examples of polishing carriers and free abrasive
suspensions [source: www.joke.de]
For rough polishing, the abrasives are in the form of stones:
abrasives are hold by a soft or hard bond (for example resin bond).
These stones (Figure 1) are directly in contact with the surface to
polish (with or without lubrication depending on the stone).
For finer polishing, free abrasive suspensions are used with
polishing cloths, polishing plates, or other carriers. A soft or
hard material is used as a carrier such as cast iron, copper,
composites, wood etc. A paste mixed with abrasives is applied on
the material and the movement of the carrier implies removal of
material (Figure 2).
22..22)) PPooll iisshhiinngg tteecchhnniiqquueess,, ““ rr
uulleess”” Manual mechanical polishing is still the best way to get
mirror-like polished surfaces with complicated geometry since the
operators can adjust the process according to the surface look.
That is why each polisher, thanks to his experience, has his own
technique of polishing. Nevertheless, hereunder are given some
operative suggestions for correct polishing [source: Lucchini
Sidermeccanica & Zanola, www.lucidaturastampi.it] which can
sometimes be hardly reproducible with a machine. Of course, the
preliminary process before polishing (grinding, turning...) is as
important as polishing process itself and in both cases, basic
rules are:
� Use cleaned polishing tools: the cleanness at every phase of
the polishing processing is of great importance. While passing to a
finer-grained abrasive medium, a good cleaning of the treated piece
and of the operators hands is necessary, in order to avoid
undesired abrasive particles or dust in the following steps.
Joke WS Diamond paste
polishing plate
wood polishing cloth
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 10/43
� When using a finer grain size, it is better to polish in a
shifted direction of 45° in comparison with the previous one, until
the surface presents only the defects due to the current polishing
direction (see figure 3).
Right at the moment when all the imperfections of the previous
processing are gone, it is a good practice to keep on polishing for
further 10% of the already spent time before passing to a
finer-grained abrasive medium. That is to remove the surface layer
warped by the mechanical stress due to previous mechanical
preparation steps. Change of polishing direction is important even
to prevent the forming of depressions and unevenness.
� Pressure and heat should be kept as low as possible, because
it could badly influence
the structure and the hardness of the material. When possible,
the use of conspicuous quantity of cooling liquid is suggested.
� When polishing big and flat mould surfaces, avoid the manual
use of disk, to reduce
the risk of getting an extended irregularity of shape. Polishing
costs and wear and tear on tools can be reduced just following
these specific rules:
� Polishing movement should start from corners, edges, i.e. from
less reachable areas. � Polishing pressure should be fitted
according to the tool hardness and to the grain size
of the paste. � Polishing should be executed in room without
draughts and dusts. Hard dust particles
can easily contaminate and spoil an almost finished surface. �
Each tool should be used for just one type of paste and be kept in
air tight containers. � The paste should be laid on the tool in
case of manual polishing, on the piece in case
of machine polishing. � It is necessary to be careful and to
possibly protect edges and sharpened corners in
order not to round them. � Hard removing of material needs hard
polishing tools and coarse-grained paste.
Problems in mechanical polishing: It is very difficult to know
when the polishing operation is finished, in fact, it is important
not to polish too much in order to avoid overpolishing (Such
problems are discussed in Appendix II).
45º
Figure 3
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 11/43
33.. Polishing mechanisms
33..11)) HHiissttoorr yy Polishing is an old manufacturing
process but the mechanisms behind it have been and are still
subjects of enquiry for a considerable period of time.
� According to Hooke, Newton, Herschel (
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 12/43
But also, on figure 2, interrupted scratches can be observed.
There are two possible reasons why such a fact occurs:
- The cutting depth provided by the grain can vary while it is
pushed against the material.
- The scratches can gradually be covered by a flow of material
of the surface top layer.
� Abrasion
If is difficult to know if there is a melting of material during
polishing, but it is sure that abrasion occurs. So, based on the
assumption that the material removal is predominantly of abrasive
nature, the chip formation can be divided into the following three
steps:
1. The polishing grain comes in touch with the material surface,
which will be elastically deformed. Due to the relative motion of
the two friction partners, on the one hand, shear stresses arise on
the workpiece surface and on the other, compressive stresses are
generated due to pressure applied by the polishing grain over the
surface.
2. As soon as the yield strength of the
material is exceeded, the bulk material deforms plastically.
This leads also to material accumulations beside the scratching
trace.
3. Then, the tensile strength is locally
exceeded and the result is a chip formation.
Figure 2: Detail of interrupted scratches [9]
Graphic representation of the chip formation, where Vc is the
cutting speed [9]
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 13/43
� Grinding
We can try to understand the polishing process thanks to
grinding process. We will then try to find the link between
grinding and polishing. In a grinding process, we can see that
abrasion occurs but not all the time. In fact, there are three
modes of interactions between the grit paper and the workpiece:
1. Lapping: The grain rolls between the specimen and the
preparation disk which implies the creation of small cavities with
strong deformation. (Figure 3)
2. Grinding: The grain is fixed and acts as a machine
tool, i.e. the rake angle is correct for cutting a chip, so it
must be positive or 0. (Figure 4)
3. Plowing: If the rake angle is negative, there is no chip
formation anymore and only a
groove is made in the specimen surface: material is displaced
into a ridge, on each side of the groove. (Figure 4)
� Link between polishing and grinding First, the equipment for
polishing is different: instead of grit paper, we can have for
example polishing cloths, but we still have a grain in contact with
the workpiece and we still produce scratches, and that is why we
can link these two processes. The choice of polishing cloth also
depends on which step of polishing it is:
0 - +
Positive
angle Negative angle
CUTTING PLOWING
Figure 3: Traces of Lapping [9]
Figure 4: Grinding (cutting process) and plowing [10]
[10]
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 14/43
For rough polishing (Ra>>1µm); hard cloths are used since
it implies a higher load and larger chips, whereas for final
polishing (Ra < 1µm), softer and more resilient cloths will make
smaller scratches and less deformation on the surface. We can bring
out four main differences between polishing and grinding:
- A lower contact pressure is applied during polishing - The use
of different type and size of abrasives and their holding methods -
Type and diamond mesh size of diamond disc - When free abrasives
are used, the grains are not fixed
33..33)) CChhaarr aacctteerr iizzaatt iioonn ooff aa ppooll
iisshheedd ssuurr ffaaccee The best way to understand polishing
mechanisms is probably to analyze polished surfaces.
Figure 1: Schematic sketch of the dislocation substructures
observed beneath a surface
polished on 6µm diamond abrasive [11]
A 3D-sketch representing the microstructure of a polished
material is shown on figure 1. Three different structures can be
distinguished:
- The sub grain structure: The grains are small and tend to be
equiaxed in shape. - Recrystallized grain: Grains are larger than
those in the sub grain structure. It is better
to describe this structure as partly recrystallized because sub
boundaries and dislocations can be observed in the electron
microscope, and that parts of the
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 15/43
boundaries of recrystallized grains were not often sharp. These
grains are mostly identifiable by the presence of annealing
twins.
- The slab cell structure: Cells free from dislocations and
separated by diffuse
boundaries composed of arrays of dislocations and dislocation
tangles. The cells usually have a length of several micrometers and
a thickness of approximately 0,1-0,2µm.
Of course the type of structure varies with the type of
finishing process and sometimes from place to place across a
particular polished surface. As it is shown in figure 2, the
structure of the outermost layers changes with increasing fineness
of polish, the number of sub grains and recrystallized grains
decrease, and the structure consisted almost entirely of slab-sharp
cells.
Figure 2: Summary of the structures observed in surfaces of
Copper abraded and polished by
various means [11]
According to Samuels, these sub grains bring into disrepute the
hypothesis of a Beilby layer since the grains are too large to be a
part of a melting layer.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 16/43
PPAARRTT II II II :: PPOOLL II SSHHII NNGG
SSTTAANNDDAARRDDSS
1. Necessity of standards After a first meeting in Denmark at
the end of September, it was decided to measure some polishing
standards from two different companies: Zanola and Bales (Figure
1). These standards are a way to show the polishing skills of a
company and not only with roughness values but also showing the
aspect of the surfaces.
Thanks to these standards we could understand what level of
polishing we should try to reach with this new machine.
2. Measurements To characterize these standards, I first had to
measure them, in order to compare them to the coming machine
polished surfaces. First, the instruments available to measure
surface roughness are discussed and then the one I used; secondly,
the results of these measurements.
22..11)) MM eeaassuurr eemmeennttss ddeevviicceess Halmstad
University uses the equipment of a company located near the
university, Toponova AB. Among other things, they have a mechanical
stylus and an interferometer. These two devices allow to measure
roughness by different ways (optical and contact). The university
has a SEM (Scanning Electron Microscope) which allows taking
pictures of a sample but does not give numerical information about
the roughness.
� Stylus : (Figure 2) It can perform 2D (and 3D measurements by
putting profiles together). It consists of a small tip moving
across the surface. The vertical translation of the stylus, while
it is going over the surface, is converted into a signal. This
signal is then exported into a processor to be converted into a
number and a visual profile.
Figure 1: Standards from Bales on the left and Zanola on the
right
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 17/43
� Interferometer : (Figure 3) It can perform 3D measurements. It
is an optical method using the interference of light: this
instrument is measuring the wavelength of light and distances. Two
beams are reflected in different ways on two surfaces not parallel
to each other. Some beams cancel and others augment each other
giving rise to a pattern of alternate dark and light fringes. Their
spacing and shape depend on reflector and on the regularity of the
surface.
� Comparison : Unlike the stylus, the interferometer is a
non-contact technology which means that we cannot damage the
surface. Moreover, a lot of noise is disturbing the results when
polished surfaces are measured with the stylus (the resolution is
too low). That is the reason why I used the interferometer to
measure my surfaces.
22..22)) MM eeaassuurr eemmeennttss aanndd RReessuull ttss I
measured the three best surfaces of Zanola and Bales (20
measurements of each surface). The results were analyzed with the
software MountainsMap [MountainsMap premium 4.1,
www.digitalsurf.fr]. Appendix III shows the results of the Zanola
samples (a robust Gaussian filter of 250µm was used) for three
different steps of polishing: SZ13, SZ15 and the best polished
surface SZ17 (SZ17 is machine polished). The Bales sample cannot be
used since the results showed surfaces which don’t seem to be
mechanically polished. In fact it looks like a melt surface for two
of the three different steps of polishing (Figure 5), so we cannot
compare these samples to polished surfaces:
Figure 3:Interferometer Figure 2: Diagram of a stylus
SPI A1 3µm SPI A1 6µm SPI A1 15µm
Figure 5: Problems with Bales standards
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 18/43
PPAARRTT II VV:: DDEESSII GGNN OOFF EEXXPPEERRII MM EENNTTSS
((DDooEE))
1. Prestudy Design of Experiment (DoE) is a structured,
organized method that can be used to determine the relationship
between the different factors (Xs) affecting a process and the
output of that process (Y) [Source: www.camo.com]. In this DoE, the
response will be the roughness of the surface (Ra value or other
surface parameters). There are different steps to perform a DoE
:
� Brainstorming to decide the parameters and their level �
Choice of an appropriate matrix, assign factors and interactions to
columns � Running experiments � Analysing results � Running
confirmation experiments
11..11)) TThhee mmaacchhiinnee When we had the meeting at
Strecon at the end of September, Kim Lorenzen and Jens Grønbæk
presented their equipment (Figure 1) to us and the results of their
first trials with the machine, with different stones, different
time of polishing, etc. This polishing machine allows working with
flat cylindrical workpieces.
Rotation of the disk
Figure 1 : The polishing machine at Strecon
Vibration direction
Moving direction of the stone on the surface
Bar
Stone/ Carrier
Robot
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 19/43
11..22)) TThhee ppaarr aammeetteerr ss Then the parameters which
influence the polishing process were listed (see the table
below):
Process variable Machine Holder Tooling Material Other Initial
roughness x Force x Vibration pulstime x Vibration length x
Stiffness of ‘bar’ x-dir x Stiffness of ‘bar’ z-dir x Rotation
speed x Feed x number of operations Lubrication x Grinding stone
(grit size, density, binding (hardness), grit material)
x
Diamond brittle, monocrystalline, grain size
x
Carrier material (wood, felt etc)
x
Material residual stresses, hardness, direction/position, bar
dimension
x Mirrax ESR
There are many parameters and of course it is not possible to
keep all of these to make a great DoE. That is why we had to make
some choices; which parameters we considered as the most important.
After some discussions and thanks to some articles about polishing
machines [13], [14], [15], we agreed that we should have at the
most 5 parameters for this DoE; if we took more, we risked to have
wrong or incomplete results (especially concerning interactions
between different parameters). Furthermore, some parameters are
fixed or difficult to change; for example the material is Mirrax
ESR and it is quiet difficult to change the bar. So finally, after
studying, comparing and thinking about these parameters, it
transpired that some parameters had to appear in the DoE: the
force, the grain size, the carrier, the rotation speed, the number
of polishing operations and the vibration pulse time of the bar.
Moreover, concerning the initial roughness, we are not sure that a
smoother surface would give better results with a given experiment
than a rougher one, so it is important to check it.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 20/43
Finally we had 7 parameters instead of 5; but, the rotation
speed and the number of polishing operations both represent the
contact time between a point on the disk and the polishing tool so
we chose one of them. Then, diamond grain size and carrier are
usually linked, we decided to set of « couples » of a grain size
and a carrier (soft, hard...) The final parameters were :
� Initial roughness � Force � Vibration pulse time � Diamond
grain size and carrier � Number of operations OR Rotation speed
11..33)) CChhooiiccee ooff tthhee mmaattrr iixx The experiments
deal with 5 parameters, and we don’t know if each parameter has a
linear evolution with the response, which is the roughness of the
surface (in a first time Ra). It means that for each variable,
there will be not 2 different values (a high and a low value) but
three because of this possible non linearity. This is a very
important step in the choice of the matrix: only a 3-level matrix
is possible. Moreover, it would be a waste of time and money to
achieve a full design of experiments with so many parameters: 5
3-level parameters mean 3^5=243 trials so a fractional factorial
design is compulsory for example Taguchi designs could be suitable
[16]. Taguchi Designs are often used for DoE in the industry; it is
the results of many years of study, and all the software of DoE
Analysis deal with this kind of matrix. That is why I chose to use
it. A tool useful for the choice of the right Taguchi matrix is the
degree of freedom: This number depends on the parameters, their
level and on the interactions studied. The interactions studied
were between:
� Force/ Diamond grain size & carrier � Initial roughness/
Diamond grain size & carrier � Rotation speed/ Diamond grain
size & carrier
Calculation of degree of freedom, n: For each 3- level
parameters: n=2 For each interaction: n=4
↪↪↪↪n total = 5x2 + 4x3 = 22 In the Taguchi designs, different
orthogonal arrays are proposed, depending on the number of
parameters, the level of parameters (ex: L9= four 3-level
parameters and 9 trials; L27= thirteen 3-level parameters and 27
trials). To choose a Lk Orthogonal Array, it is compulsory to
respect the following rule:
degree of freedom (n) < k
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 21/43
1
2 3,4 5 6,7
9,10 8
11 12,13
Dots: parameters Lines: interactions Numbers: Nb of column
Figure 2 : Linear graph of L27 Orthogonal Array
This condition implies the L27 Orthogonal Array (Figure 1) is
compulsory.
Parameter Nb runs 1 2 3 4 5 6 7 8 9 10 11 12 13
1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 2 2 2 2 2 2 2 2 2 3 1 1 1
1 3 3 3 3 3 3 3 3 3 4 1 2 2 2 1 1 1 2 2 2 3 3 3 5 1 2 2 2 2 2 2 3 3
3 1 1 1 6 1 2 2 2 3 3 3 1 1 1 2 2 2 7 1 3 3 3 1 1 1 3 3 3 2 2 2 8 1
3 3 3 2 2 2 1 1 1 3 3 3 9 1 3 3 3 3 3 3 2 2 2 1 1 1
10 2 1 2 3 1 2 3 1 2 3 1 2 3 11 2 1 2 3 2 3 1 2 3 1 2 3 1 12 2 1
2 3 3 1 2 3 1 2 3 1 2 13 2 2 3 1 1 2 3 2 3 1 3 1 2 14 2 2 3 1 2 3 1
3 1 2 1 2 3 15 2 2 3 1 3 1 2 1 2 3 2 3 1 16 2 3 1 2 1 2 3 3 1 2 2 3
1 17 2 3 1 2 2 3 1 1 2 3 3 1 2 18 2 3 1 2 3 1 2 2 3 1 1 2 3 19 3 1
3 2 1 3 2 1 3 2 1 3 2 20 3 1 3 2 2 1 3 2 1 3 2 1 3 21 3 1 3 2 3 2 1
3 2 1 3 2 1 22 3 2 1 3 1 3 2 2 1 3 3 2 1 23 3 2 1 3 2 1 3 3 2 1 1 3
2 24 3 2 1 3 3 2 1 1 3 2 2 1 3 25 3 3 2 1 1 3 2 3 2 1 2 1 3 26 3 3
2 1 2 1 3 1 3 2 3 2 1 27 3 3 2 1 3 2 1 2 1 3 1 3 2
11..44)) AAssssiiggnnmmeenntt ooff tthhee ffaaccttoorr ss ttoo
tthhee ccoolluummnnss In our DoE, we need five columns on the
thirteen available. Nevertheless we must not assign a factor to a
column by random, there are some rules. In fact some columns of the
matrix can be a parameter but can also allow the study of an
interaction so when possible, it is important not to assign a
factor to a column of an important interaction. The linear graph of
the matrix (Figure 2) gives a way to fill it:
Figure 1 : Taguchi Design ; L27 Orthgonal Array
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 22/43
Finally, Figure 3 shows the matrix which will be used for the
experiments: Parameter
Nb runs 1 D
2 B
3 DxB
5 E
6 DxE
8 A
9 DxA
11 C
1 1 1 1 1 1 1 1 1 2 1 1 1 2 2 2 2 2 3 1 1 1 3 3 3 3 3 4 1 2 2 1
1 2 2 3 5 1 2 2 2 2 3 3 1 6 1 2 2 3 3 1 1 2 7 1 3 3 1 1 3 3 2 8 1 3
3 2 2 1 1 3 9 1 3 3 3 3 2 2 1
10 2 1 2 1 2 1 2 1 11 2 1 2 2 3 2 3 2 12 2 1 2 3 1 3 1 3 13 2 2
3 1 2 2 3 3 14 2 2 3 2 3 3 1 1 15 2 2 3 3 1 1 2 2 16 2 3 1 1 2 3 1
2 17 2 3 1 2 3 1 2 3 18 2 3 1 3 1 2 3 1 19 3 1 3 1 3 1 3 1 20 3 1 3
2 1 2 1 2 21 3 1 3 3 2 3 2 3 22 3 2 1 1 3 2 1 3 23 3 2 1 2 1 3 2 1
24 3 2 1 3 2 1 3 2 25 3 3 2 1 3 3 2 2 26 3 3 2 2 1 1 3 3 27 3 3 2 3
2 2 1 1
A Force B Initial roughness C Vibration pulse time D Diamond
size & carrier E Nb of operations/ Rotation speed
Figure 3 : Matrix of the DoE
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 23/43
2. Experiments at Strecon I was at Strecon from the 4th to the
6th of December in order to carry out the Design of Experiment. For
this purpose, the work there had to be divided in two distinct
parts:
� The preparation of the DoE, that means to decide of a first
step of polishing in order to get 3 different initial roughnesses
for the DoE.
� The performing of the DoE.
22..11)) PPaarr tt 11:: PPrr eeppaarr aatt iioonn ooff tthhee
eexxppeerr iimmeennttss We started with making trials to determine
a first step of polishing. This step is compulsory to get three
different roughnesses with some turned and ground disks (9 runs
with turned disks with Ra1, 9 ground disks with Ra2 and 9 ground
disks with Ra3; figure 1); we measured them, made some trials with
two different diamond stones, different number of operations…
Figure 1 : Turned disk (Ra 1) It was also a good introduction
before the DoE, to observe how they were working with this machine
and how they were measuring the surfaces. At the end of the first
day, the first step of polishing was decided (the stone used,
figure 2, the number of operations and all the other
parameters)
Figure 2: JOKE MF 600 (stone used)
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 24/43
22..22)) PPaarr tt 22 :: FFllooww ooff wwoorr kk The second
phase was the Design of Experiments. We hadn’t time to prepare all
the disks; nevertheless, we decided to perform the Design of
Experiments with the turned disks. The 3 levels for each parameter
were decided: VALUE: 1 2 3 A Force (kg) 2 5 10 B Initial roughness
(Ra; µm) 0.1 0.24 0.3 C Vibration (pulse/min) 0 1000 2000 D Diamond
size & carrier (Figure 3) 15µm & hard wood 15µm & soft
wood 9µm & soft wood E Rotation Speed (Rpm) 180 400 700
Figure 3: Diamond paste 15µm (left) and soft carrier (right)
All the disks dealt with 6 bands (three on each side), and were
marked in order not to be mixed up with what we did on every band.
While I was there, we had time to make the 9 runs on the turned
disks. Appendix IV shows the matrix of the DoE with all the values
of the parameters. Figure 4 and 5 are pictures taken during the DoE
on the turned disk: on figure 4 diamond paste is applied on the
disk and on figure 5, polishing is performed.
Figure 4: Application of the diamond paste Figure 5: Design of
Experiment run 5
carrier
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 25/43
3. Analysis of the results After the experiments at Strecon, I
could start measuring the first 9 areas polished. I measured it
with the interferometer with 15 measurements at magnification 10x
and could then analyse it with help of MountainsMap. I didn’t use
any filter for my surfaces because after comparing the results with
and without filters, the difference was not significant so I
decided to use a form removal process with polynomial of order 2.
For each surface I made a sheet showing the surface and the
roughness values, like the one shown on figure 1 in order to make
the analysis to be easier (all the results are presented in
Appendix V)
Figure 1: Extract of the results obtained with MountainsMap
4. Results and discussion The results were not those expected at
the beginning of this project because of technical problems: at the
moment I was writing this report, we had some bad news from
Strecon. They broke their measurement device so they couldn’t
finish preparing the other disks and so, we couldn’t run the rest
of the tests before the end of my project. Now, we just have four
parameters because the nine polished surfaces done had the same
initial roughness. So the problem is that we have 9 points of 27
(See figure 1). With the points we have, it seems that only the
influence of the diamond size and carrier could be studied in more
detail, since the remaining parameter (force, vibration pulse time
and rotation speed) have fixed values in relation to each
other.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 26/43
So we don’t have a standard Design of Experiments which means
that the data cannot be studied with commonly used software.
Instead, we tried to use Simca and Excel to get some information.
Unfortunately we didn’t get evidences, even about the influence of
the diamond grain size and carrier, but we can still say that the
Ra value cannot really describe a level of polishing. Checking the
surface aspect gave more information; in fact it is easy to
distinguish different kind of surfaces. For example, some of them
are scratched in many directions as shown in figure 2 (especially
the three surfaces with a rotation speed of 180 Rpm, a force of 5
kg and 2000 vibrations per minute). Whereas the others reveal
distinct grooves as we can see in figure 3.
180 400 700 Rotation speed (Rpm)
Force (kg)
10
5
2
Vibration pulse time (Vib/min)
0
1000
2000
Figure 1:Representation of the parameters studied in the Design
of experiments performed
15µm & hard wood 15µm & soft wood 9µm & soft
wood
Figure 2: Trial 13 (180 Rpm, 5 kg, 2000 vib/min, 1 5µm &
soft wood)
Figure 3: Trial 5 (400 Rpm, 10 kg, 0 vib/min, 15 µm & hard
wood)
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 27/43
Figure 4: Disk 4 after DoE
From the previous pictures we could think that the surfaces like
figure 2 are the best. When compared to the visual aspect, this was
not the case; the shiniest one (Area 2 on figure 4) was actually a
surface like the one shown in figure 3.
Figure 3 reveals a problem; are the marks on the surface from
the turning process? The initial surface (after the first step of
polishing) doesn’t show clearly these turning marks (figure 5) but
when we compare the profiles from trial 5 and 13 with the profile
from the turned surface (before the first step of polishing), it
seems that it is in fact turning marks.
µ m
-2.5
-2
-1.5
-1
-0.5
0
0 .5
1
0 50 100 1 50 200 2 50 300 350 40 0 450 50 0 550 6 00 650 70 0
µm
Len gth = 7 17 µm Pt = 1 .14 µ m Sca le = 3 .5 µm
Trial 13 (180 Rpm, 5kg, 2000vib/min, 15µm&soft wood), disc 4
area 1 µm
-2.5
-2
-1 .5
-1
-0 .5
0
0.5
1
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 µm
Length = 706 µm P t = 1 .3 µm Sca le = 3 .5 µm
µm
-2.5
-2
-1.5
-1
-0.5
0
0 .5
1
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 µm
Length = 715 µm Pt = 1 .46 µm Scale = 3.5 µm
Trial 5 (400 Rpm, 10kg, 0 vib/min, 15µm&hard wood), like
disc 4 area 2
Initial surface µm
-1 .5
-1
-0 .5
0
0 .5
1
1 .5
2
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800
µm
Length = 848 µm P t = 3.1 µm Scale = 3 .5 µm
Turned surface
Figure 5: Comparison between the surfaces, turning marks
100 µm
100 µm
100 µm
100 µm
Area 1
Area 2
Area 3
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 28/43
Depending on the surfaces, the turning marks are more or less
detectable, and the distance between the valleys confirms the
hypothesis that the marks from the turning process were not
removed. Probably, the first grinding step was not long enough.
Moreover, the graphs on Excel or Simca couldn’t show evidences;
but we could still find some parameters like the mean density of
furrows which gave some results:
The two shiniest surfaces are those with the lowest mean density
of furrows (green squares in figure 6) and also the lowest number
of islands (see results in Appendix V).
500
550
600
650
700
750
800
850
900
0 2 4 6 8 10 12
Mea
n D
ensi
ty o
f F
urr
ow
s [c
m/c
m2]
D4A6 D4A5 D4A4D4A3 D4A2 D4A1
D3A6 D3A5 D3A4TURN (1 surface) Initial (1 surface)
Figure 6 : The mean density of furrows of the different
areas
In fact, if we make a classification of the surfaces depending
on their gloss (visual comparison of the surfaces), we can see more
clearly the link between these parameters (see figure 7):
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 29/43
Figure 7: Comparison between Number of Islands and Mean Density
of Furrows with visual classification of the Gloss
On this graphs we can see the link between the gloss of the
surface and the number of islands, where the classification are
very close. It is less clear with the mean density of furrows but
still we can see that three “groups” of surfaces are always
present. These groups can also be seen on a plot representing the
ratio Svk/Sk as we can see in figure 8:
D4A5
D3A5
D4A1
D3A6
D3A4
D4A2
D4A3
D4A6
D4A4
720
740
760
780
800
820
840
0,5 1,5
gloss
D3A5
D4A1
D3A6
D4A6
D4A4
D3A4
D4A2
D4A3
D4A5
0
1
2
3
4
5
6
7
8
9
10
0,5 1,5
number of islands
D4A5
D3A5
D4A1
D3A6
D3A4
D4A2
D4A3
D4A6D4A4
200
300
400
500
600
700
800
900
1000
0,5 1,5
D4A3D4A5 D3A6
D4A1
D4A2
D4A6D3A5
D3A4
D4A4
0
0,2
0,4
0,6
0,8
1
1,2
0 1 2 3 4 5 6 7 8 9 10
Svk
/Skl
Figure 8: Ratio Svk/Sk
Gloss Number of Islands Mean Density of
Furrows
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 30/43
Figure 10 shows an Excel analysis of these data, which confirms
with numerical values these correlations between the discussed
parameters and the gloss of the surface.
Mean density of furrows NoI Svk Spk Sk Svk/Sk Spk/Sk Spk/Svk
Mean density of furrows 1
Number of islands 0,64062 1
Mean volume of islands -0,81550 -0,775
Mean height of islands 0,55847 0,723
Mean surface of islands -0,71044 -0,931
Mean height / surface ratio 0,55191 0,9426
Svk -0,09277 0,2419 1 Spk 0,04367 0,6467 0,292 1
Sk 0,01810 0,6738 0,468 0,895 1 Svk/Sk 0,11534 -0,587 -0,245
-0,902 -0,95 1 Spk/Sk 0,17900 -0,116 -0,478 0,02 -0,42 0,3204 1
Spk/Svk 0,05339 0,5651 -0,034 0,944 0,777 -0,863 0,1737 1 Smmr
0,12398 0,4118 0,824 0,394 0,352 -0,147 0,0079 0,11244
Vvv (10 80 2) -0,01376 0,4694 0,923 0,554 0,749 -0,559 -0,549
0,26655 Sr2 0,14259 0,7168 0,221 0,965 0,935 -0,936 -0,112 0,9268
Std -0,26444 0,3841 0,246 0,793 0,663 -0,76 0,0559 0,75102
Isotropy 0,31193 0,2006 0,422 -0,014 -0,18 0,3382 0,416 -0,177
Diamond stone &
carrier -0,30520 -0,292 0,456 0,263 0,134 -0,073 0,1453 0,14562
speed -0,21392 0,1194 -0,477 0,442 0,341 -0,544 0,0695 0,62079
pressure -0,52813 -0,529 0,467 -0,329 -0,07 0,1566 -0,545
-0,4732 pulse 0,74723 0,3561 -0,047 -0,111 -0,32 0,4209 0,5543
-0,1235 Gloss 0,68367 0,9185 0,275 0,601 0,696 -0,587 -0,241
0,52376
Figure 10: Extract of the correlation table of the data
Finally, concerning the experimental part of the DoE, we can say
that a high rotation speed has a bad influence because of the use
of free abrasives. In fact, we noticed during the experiments that
the diamond paste is thrown away while the disk is rotating
(centrifugal force), for 700Rpm but even for 400Rpm.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 31/43
CCOONNCCLL UUSSII OONN OOFF TTHHEE PPRROOJJEECCTT According to
the previous results, and at this stage of the analysis, we can say
that:
� It is not possible to describe a polished surface only with a
Ra value. � We need to perform all the experiments to make a DoE
Analysis and to evaluate the
influence of each parameter. � The turning marks reveal that the
first step of polishing is not good enough: We
thought we had removed the turning marks when measuring the
initial surface with a stylus but the results showed us the
contrary.
� A high rotation speed is a problem because of the moving of
the diamond paste. � The shiniest surfaces were those with the
turning marks which mean that on these
surfaces, we probably removed more material since we reached the
marks from the turning process.
� The surface qualities we got during the tests with the machine
seem, at the moment, to be very far from the surface qualities of
the polishing standards.
FFUUTTUURREE My project couldn’t be finished on time so we can
discuss how it should be organized in a future work on this project
and what to focus on:
� As it was said in the conclusion, the most important thing is
to achieve a complete Design of Experiment to finally have all the
results.
� About the results we got, we couldn’t really see the turning
marks on the initial
surfaces before the DoE, even if they were measured.
Nevertheless it is important to know when one step of polishing is
over, i.e. when the marks from the previous process are
removed.
� This project highlighted a point which could be a problem for
the machine: the
repeatability. In fact, after measuring the surfaces after the
first polishing step, we could see significant differences between
the Ra-values of some disks, so it must be interesting to know if
it is a problem of the machine or of the accuracy of the
measurement devices or both.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 32/43
PPEERRSSOONNAALL CCOONNCCLL UUSSII OONN During these five months
of work in Halmstad, I learned a lot on multi-aspect: the work, the
language, the Swedish culture and the culture of many European
countries.
� First, I could improve considerably my knowledge in a subject
I didn’t know very well; the study of surfaces, how to measure a
surface, I discovered other measurement devices, how to treat the
measurements. In this purpose, I had the opportunity to use
softwares like MountainsMap to analyse the measurements and also
MATLAB, and Simca for treatment of data. Then, my subject was about
polishing of steel. The literature study allowed me to prepare and
to know the subject before starting working on this project and to
acquire knowledge about what kind of polishing processes exist and
more exactly what mechanical polishing is, since I didn’t really
know in details what this technique was. I also read a lot about
the Design of Experiments, which is very useful in the industry and
so I could learn how to carry out this kind of experiments and how
important it is to prepare carefully the DoE, to be there for the
experiments… Moreover, I could see that working for companies which
are not located at the same place is really difficult, since it
means to communicate by e-mails or by phone and it makes the
project go slowly. I also discovered the work in a research lab
which is very different from the work in a company, and it is for
me the best way to know where I want to be in my future job. I
liked working in the research lab since you are not really in a
company but you still have a contact with industrial world.
� For this first placement abroad, I had to work always in
English, so it was a great experience since it is a constant work
to listen and understand people and also to share my ideas in
another language even if sometimes it can be difficult. And
everyday I was speaking English: at work, in the street, and at
home since I was the only French living with 13 other people from
Netherlands, Germany, Poland, Austria, Lithuania, Thailand.
� Finally, this stay abroad was a fantastic human experience: I
met people from many different countries, co-workers and students.
The student house where I lived allowed me to meet a lot of people
and to share our culture every day. We discovered the Swedish
culture together and also Scandinavian countries doing trips in the
cities around, in Sweden, Norway, and Denmark. I think this
experience made me be more open minded on the world and people who
don’t have the same way of life and to see that there are a lot of
differences even between European countries, thing that I wasn’t
really aware of before. To conclude, it was a special experience,
unforgettable, and very enriching on many points.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 33/43
RREEFFEERREENNCCEESS
� Different techniques of polishing: [1] Laser polishing of
parts built up by selective laser sintering (2007) International
journal of machine tools & Manufacturer 47 A. Lamikiz, J.A.
Sanchéz, L.N. López de Lacalle, J.L. Arana [2] Mechanics,
Mechanism, and modeling of the chemical mechanical process Jiun-Yu
Lai (Massachussets institute of technology) [3] Material removal
mechanism in Lapping and Polishing C.J Evans, E. Paul, D,Dornfekd,
D.A. Lucca, G. Byrne, M.tricard, F.Klocke, O.Dambon, and B.A.
Mullany [4] Magnetic polishing of three dimensional die and mold
surfaces (2007) International Journal of Advanced Manufacturing
technology Jeong-du Kim & Ill-hwan Noh [5] Basic studies for
the electro polishing facility at Desy CARE Conf 05-029-SRF N.
Steinhau-Kühl; R. Bandelmann, K. Escherich, D. Keese, M. Leenen, L.
Lilje, A. Matheisen, H. Morales, P. Schmüser, E. Schulz, M. Seidel,
J. Tiessen [6] VESTA Sterile Technology GEA process equipment
division, Tuchenhagen [7] A new polishing method of metal mold with
large-area electron beam irradiation (2007) Journal of Materials
processing technology Y. Unoa, A. Okadaa, K. Uemurab, P. Raharjo ,
S. Sano , Z. Yuc, S. Mishima [8] A study on the mirror surface
machining by using a micro-energy EDM and the electrophoretic
deposition polishing (2007) International Journal of Advanced
Manufacturing technology Biing Hwa Yan, Kun Ling Wu, Fuang Yuan
Huang, Chun Chieh Hsu
� Polished surfaces: [9] Influence of the polishing process on
the near-surface zone of hardened and unhardened steel (2005)
Sciencedirect, volume 258 issues 11-12 F. Klocke, O. Dambon, G.G.
Capudi Filho [10] The True Microstructure of Materials
Materialographic Preparation from Sorby to the Present Kay Geels,
Struers A/S, Copenhagen, Denmark [11] The Nature of Mechanically
Polished Surfaces of Copper (1998) Elsevier science Inc. Materials
characterization D. M. Turley and L. E. Samuels [12] Lapping and
Polishing Basics Applications laboratory report 54, South Bay
Technology
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 34/43
� Automated polishing [13] Automated polishing of Die steel
surfaces (2002) International Journal of Advanced Manufacturing
technology vol 19, no. 4 J.P. Huissoon, F.Ismail, A.Jafari [14]
Short papers: development of an automatic mold polishing system
(2005) IEEE Transactions on automation science & engineering
Vol 2 Ming J.Tsai, Jou-Lung Chang, and Jian-Feng Haung [15]
Intelligently automated polishing for high quality surface
formation of sculptured die (2002) Journal of Materials processing
technology J.H. Ahn, M.C.Lee, H.D.Jeong, S.D.Kim, K.K.Cho
� Design of experiments: [16] Taguchi Techniques for quality
Engineering_ Loss function, Orthogonal Experiments, Parameter and
Tolerance Design Phillip J.Ross, McGraw-Hill Book Company; 279
pages
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 35/43
AAPPPPEENNDDII XX
APPENDIX I: A few polishing
methods..............................................................................
36 APPENDIX II: Problems in mechanical
polishing.............................................................
39 APPENDIX III Results of standards measurements
(Zanola)...........................................40 APPENDIX IV:
Matrix of the
DoE.......................................................................................41
APPENDIX V: Results of the nine trials of the
DoE...........................................................42
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 36/43
APPENDIX I: A few polishing methods
1. Laser polishing [1] Laser polishing is usually employed for
simple geometries. Nevertheless, complex geometry can also be
polished, but with a five-axis machine and a correction of the
focal position instantaneously. Laser polishing is based on the
melting of a microscopic layer and a fast re-solidifying of the
melted material. The main parameters which influence the final
result are: the laser beam feed rate, laser power and the laser
beam diameter → There is also Laser surface texturing which is one
of the most recent processes using laser technology. It is a metal
additive process. With this process, surfaces tend to have a bad
quality.
2. Chemical Mechanical Polishing [2],[3]
Chemical Mechanical Polishing, or CMP, is an old wet slurry
technology. During the CMP process, a wafer surface is polished for
planarization using a slurry and a polishing pad. The abrasive
particles in the slurry grind against the sample surface, loosening
material. The chemicals in the slurry then etch and dissolve the
material. This process is very common in the semi conductor
industry, for the polishing of small components like integrated
circuits.
3. Magnetic polishing [4] During magnetic polishing, the
polishing pressure in the tool-workpiece interface is produced by
the permanent magnet included inside the polishing tool. Two stages
are necessary to get a good quality: - A first-stage polishing of
initial rough surfaces which uses the abrasive wheel for material
removal function. - A second-stage polishing, i.e., fine polishing,
after the first rough polishing, which uses the magnetic brush for
material removal function instead of abrasive wheel.
Picture: www.agc.co.jp
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 37/43
4. Electropolishing [5],[6]
In basic terms, the object to be electropolished is immersed in
an electrolyte (typically phosphoric and sulphuric acid) and
subjected to a direct electrical current. The object is maintained
anodic, with the cathodic connection being made to a nearby metal
conductor. In electropolishing, the metal is removed ion by ion
from the surface of the metal object being polished.
During electropolishing, the polarized surface film is subjected
to the combined effects of gassing (oxygen), which occurs with
electrochemical metal removal, saturation of the surface with
dissolved metal and the agitation and temperature of the
electrolyte. Electropolishing selectively removes microscopic high
points or "peaks" faster than the rate of attack on the
corresponding micro-depressions or "valleys."
5. Electron beam irradiation [7]
Electron beam irradiation is a technology using the phenomenon
of explosive electron emission. At first, a magnetic field is
generated by the solenoid coil mounted on the outer side of the
chamber. At the moment when the magnetic field takes a maximum
intensity, pulse voltage is loaded to a ring shape anode. In the
chamber, the electrons start to move towards the anode.
Simultaneously, the electrons move spirally because of the applied
Lorentz forces. Next, argon atoms are ionized by the repetitious
collision with electrons, which generates plasma near the anode.
When the plasma intensity takes a maximum, pulse voltage is applied
to the cathode. The electrons are accelerated by high electric
field due to electric double layer formed near the cathode, and the
explosive electron emission occurs. Then, EB with high energy
density is irradiated to the workpiece surface. This process is a
fast method, and also allows having a better corrosion resistance
after irradiation.
Picture: www.delstar.com
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 38/43
6. Electrical discharge machining [8]
Electrical discharge machining (EDM) is a thermal process that
involves melting and vaporisation of the workpiece electrode. It is
widely used in the aerospace, mould making and die casting
industries for manufacturing plastic moulds, forging dies and die
casting dies made from hardened tool steels. The EDM process uses
electrical discharges to remove material from the workpiece. Each
spark produces a temperature of between 10,000-20,000°C.
Consequently, the workpiece is subjected to a heat affected zone
(HAZ), the top layer, which consists of recast material. The
thickness, composition and condition of this layer depend on the
discharge energy and the set-up of the workpiece, tool electrode
and dielectric fluid.
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 39/43
APPENDIX II: Problems in mechanical polishing (Lucchini
Sidermeccanica & Zanola) Major problems during the polishing
may be found when an excessive polishing, called “Overpolishing” is
carried out. Actually it may happen that the longer a surface is
polished, the worst its conditions become. The “Overpolishing” is
linked to two distinct phenomena that are called “orange peel” and
“pitting”.
Figure 1: Roughness trend in comparison with polishing duration
time
1) Orange peel This phenomenon results from the overcoming of
the yield point in micro areas of an inhomogeneous surface.
Depressions develop on the piece surface at level of softer
micro-areas; this is commonly called “Orange peel”. The causes of
this phenomenon can be:
- Excessive pressure performed during the polishing processing.
- Irregular areas of residual austenite more sensitive to
permanent deformations than other tempered structure. -
Chemical-structural inhomogeneity of the initial material -
2) Pitting The very small cavities that can be observed in a
flat surface during the polishing phase commonly derive from
inclusions that are removed from the surface during the polishing
process. Generally the removed particles are sulphides or oxides,
which, not only differ in hardness and stiffness from the metallic
matrix surrounding them, but also are characterized by a middle/low
adhesion to the metallic material. The main factors responsible of
the forming of pitting are:
- polishing duration and pressure - steel pureness (especially
for what concerns hard non metallic
inclusions) - the type of used polishing tool - the abrasive
medium
The differences in hardness between matrix and non metallic
inclusions are the main cause of pitting. During the polishing
process, the matrix will be more rapidly removed than the hard non
metallic particles (Figure 3). Gradually the polishing hits the
hard particles until these ones come off the material which causes
cavities.
OVERPOLISH
OVERPOLISHING
Figure 3: Inclusions’ behaviour during
polishing
Figure 2: Surface with Orange peel aspect
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 40/43
APPENDIX III: Results of standards measurements (Zanola)
SZ13 polishing 3µm
SZ17 polishing 0,25µm
SZ15 polishing 1µm
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 41/43
APPENDIX IV: Matrix of the DoE
1 2 3 5 6 8 9 11
Nb trials Disc band
D_Diamond paste
B_initial Ra DxB E_Rpm DxE
A_ force DxA
C_vibration pulse
1 5 4 15 hard 1 1 180 1 2 1 0 2 5 5 15 hard 1 1 400 2 5 2 1000 3
5 6 15 hard 1 1 700 3 10 3 2000 4 3 4 15 hard 2 2 180 1 5 2 2000 5
3 5 15 hard 2 2 400 2 10 3 0 6 3 6 15 hard 2 2 700 3 2 1 1000 7 7 4
15 hard 3 3 180 1 10 3 1000 8 7 5 15 hard 3 3 400 2 2 1 2000 9 7 6
15 hard 3 3 700 3 5 2 0 10 6 1 15 soft 1 2 180 2 2 2 0 11 6 2 15
soft 1 2 400 3 5 3 1000 12 6 3 15 soft 1 2 700 1 10 1 2000 13 4 1
15 soft 2 3 180 2 5 3 2000 14 4 2 15 soft 2 3 400 3 10 1 0 15 4 3
15 soft 2 3 700 1 2 2 1000 16 8 4 15 soft 3 1 180 2 10 1 1000 17 8
5 15 soft 3 1 400 3 2 2 2000 18 8 6 15 soft 3 1 700 1 5 3 0 19 6 4
9 soft 1 3 180 3 2 3 0 20 6 5 9 soft 1 3 400 1 5 1 1000 21 6 6 9
soft 1 3 700 2 10 2 2000 22 4 4 9 soft 2 1 180 3 5 1 2000 23 4 5 9
soft 2 1 400 1 10 2 0 24 4 6 9 soft 2 1 700 2 2 3 1000 25 8 4 9
soft 3 2 180 3 10 2 1000 26 8 5 9 soft 3 2 400 1 2 3 2000 27 8 6 9
soft 3 2 700 2 5 1 0
initial roughness: 1--> ground disks with 2 polishing steps
2--> turned disks 3--> ground disks with 1 polishing step
9 trials performed
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 42/43
APPENDIX V: Results of the nine experiments
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 43/43
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 44/43
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 45/43
-
GAGNOLET Marielle 4GM
REPORT OF INDUSTRIAL PLACEMENT
From 09/10/2007 till 01/25/2008
Page 46/43