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
AK A KN N OW O WL L E E DG D GM ME EN N TS T S
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.
AB A BS S TR T RA AC C T T
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.
SU S U MM M MA AR RY Y
A
AKKNNOOWWLLEEDDGGMMEENNTTSS ... 1
A ABBSSTTRRAACCTT..........................................................................................................................................................................................................................................................................................................................22 SUSUMMMMAARRYY ... 3
ININTTRROODDUUCCTTIIOONN... 4
P PAARRTT II:: PPRROOJJEECCTT SSCCEENNAARRIIOO ... 5
1.PRESENTATION OF COMPANIES... 5
1 1..11)) HHALALMMSSTTAADD UUNNIIVVEERRSSIITTYY... 5
1.1.22)) UUDDDDEEHHOOLLMM TTOOOOLLIINNGG AABB... 5
1.1.33)) SSTRTREECCOONN AA//SS... 6
2.THE PROJECT... 7
2 2..11)) TTHHEE PPRROOJJEECCTT EENNVVIIRROONNMMEENNTT... 7
2.2.22)) DDESESCCRRIIPPTTIIOONN OOFF WWOORRKK... 7
P PAARRTT IIII:: LLIITTEERRAATTUURREE SSTTUUDDYY... 8
1.A FEW POLISHING METHODS... 8
2.MECHANICAL POLISHING... 8
2.2.11)) PPOOLLIISSHHIINNGG EEQQUUIIPPMMEENNTT ... 8
2.2.22)) PPOOLLIISSHHIINNGG TTEECCHHNNIIQQUUEESS,,““RRUULLEESS”” ... 9
3.3.POLISHING MECHANISMS... 11
3 3..11)) HHISISTTOORRYY... 11
3.3.22)) AABBRRAASSIIOONN AANNDD PPOOLLIISSHHIINNGG... 11
3.3.33)) CCHAHARRAACCTTEERRIIZZAATTIIOONN OOFF AA PPOOLLIISSHHEEDD SSUURRFFAACCEE... 14
PAPARRTT IIIIII:: PPOOLLIISSHHIINNGG SSTTAANNDDAARRDDSS ... 16
1.NECESSITY OF STANDARDS... 16
2.MEASUREMENTS... 16
2.2.11)) MMEEAASSUURREEMMEENNTTSS DDEEVVIICCEESS... 16
2 2..22)) MMEEAASSUURREEMMEENNTTSS AANNDD RREESSUULLTTSS... 17
PAPARRTT IIVV:: DDEESSIIGGNN OOFF EEXXPPEERRIIMMEENNTTSS ((DDOOEE))... 18
1.PRESTUDY... 18
1.1.11)) TTHHEE MMAACCHHIINNEE... 18
1.1.22)) TTHHEE PPAARRAAMMEETTEERRSS... 19
1 1..33)) CCHOHOIICCEE OOFF TTHHEE MMAATTRRIIXX... 20
1.1.44)) AASSSSIIGGNNMMEENNTT OOFF TTHHEE FFAACCTTOORRSS TTOO TTHHEE CCOOLLUUMMNNSS... 21
2.EXPERIMENTS AT STRECON... 23
2.2.11)) PPAARRTT 11::PPRREEPPAARRAATTIIOONN OOFF TTHHEE EEXXPPEERRIIMMEENNTTSS... 23
2 2..22)) PPAARRTT 22 ::FFLOLOWW OOFF WWOORRKK... 24
3.ANALYSIS OF THE RESULTS... 25
4.RESULTS AND DISCUSSION... 25
C COONNCCLLUUSSIIOONN OOFF TTHHEE PPRROOJJEECCTT ... 31
FUFUTTUURREE ... 31
PEPERRSSOONNAALL CCOONNCCLLUUSSIIOONN ... 32
R REEFFEERREENNCCEESS... 33 A
APPPPEENNDDIIXX........................................................................................................................................................................................................................................................................................................................3355
I I NT N TR RO OD DU U C C TI T IO ON N
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.
PA P AR RT T I I : : PR P R OJ O JE EC CT T S SC C EN E NA AR R IO I O 1. Presentation of companies
1.1.11)) HaHallmmssttaadd UUnniivveerrssiittyy
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.
1.1.22)) UdUdddeehhoollmm TToooolliinngg 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
1.1.33)) StStrreeccoonn 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.
2. The project
2
2..11)) ThThee prproojjeecctt eennvviirroonnmmeenntt
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.
2.2.22)) DeDessccrriippttiioonn ofof wwoorrkk
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
PA P AR RT T I I I: I : LI L IT TE ER RA AT TU UR R E E S S TU T UD D Y Y
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]
2
2..11)) PoPolliisshhiinngg eqequuiippmmeenntt [[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.
Al2O3: 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.
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).
2
2..22)) PoPolliisshhiinngg tetecchhnniiqquueess,, ““rruulleess””
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
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
3. 3 . Polishing mechanisms
3
3..11)) HiHissttoorryy
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 (<1900): Polishing is a cutting process.
At the beginning of the 20th century (1921): Beilby proposed the flow hypothesis which is that asperities might reach melting temperature during contact rubbing; high spots of the surface are molten and flow into the roughness valley. An amorphous smeared layer, the so called “Beilby layer”, would cover the surface.
Beilby also meant that material removal rate correlate with melting point and not with hardness.
1971: Samuels declined the flow hypothesis. According to him; polishing is a cutting process close to abrasion.
In fact, experiments were made and material removal could be observed by measuring weight loss during polishing. So, even if there is a melting process during polishing, there is a cutting process or abrasion because of this material removed.
3.3.22)) AbAbrraassiioonn anandd ppoolliisshhiinngg
Step by step polishing
A step by step polishing shows clearly a rapid decrease of the roughness values already in the first seconds of the polishing process.
After the elimination of the deepest furrows of the previous grinding process, a significant decrease cannot be observed anymore.
Polishing scratches are visible on the figure 1 below, indicating the abrasive nature of the polishing process.
Figure 1: Development of the surface topography by step-by-step polishing [9]
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]
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]
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
3.3.33)) ChChaarraacctteerriizzaattiioonn ooff aa popolliisshheedd ssuurrffaaccee
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
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.
PA P AR RT T I I II I I: : PO P OL LI I SH S HI IN NG G S ST TA AN ND D AR A RD D S S
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.
2.2.1)1) MMeeaassuurreemmeennttss dedevviicceess
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
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.
2
2..2)2) MMeeaassuurreemmeennttss anandd RReessuullttss
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
PA P AR RT T I I V: V : D DE ES SI I GN G N O OF F E EX X PE P ER RI I ME M EN N TS T S ( (D D oE o E) )
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 1.1.11)) ThThee 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
1.1.22)) ThThee ppaarraammeetteerrss
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.
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
1.1.33)) ChChooiiccee ooff tthhee mmaattrriixx
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
1
3,4 2
6,7 5
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
1
1..44)) AsAsssiiggnnmmeenntt ooff tthhee ffaaccttoorrss ttoo tthhee cocolluummnnss
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
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
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.
2.2.11)) PaParrtt 11:: PrPreeppaarraattiioonn ooff tthhee eexxppeerriimmeennttss
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)
