Optical Appearance of Aluminium, Martin Aggerbeck PhD Th. (Tech Univ. Denmark, 2010)

8/10/2019 Optical Appearance of Aluminium, Martin Aggerbeck PhD Th. (Tech Univ. Denmark, 2010)

http://slidepdf.com/reader/full/optical-appearance-of-aluminium-martin-aggerbeck-phd-th-tech-univ-denmark 1/119

Technical University of Denmark

Department of Management EngineeringDepartment of Mechanical Engineering

Optical Appearance of AluminiumAluminiums optiske fremtoning

Handed in: 2010-09-10

By:

Martin Aggerbeck, s042362

Supervisors:Torben Lenau, DTU-MAN

Rajan Ambat, DTU-MEK

In collaboration with:

Danish Technological InstituteBang & Olufsen

DTU Fotonik









Martin Aggerbeck DTU MAN & MEK 2010-09-10Optical Appearance of Aluminium

1

Acknowledgements

This is the thesis of the Design and Innovation master program at the Technical University ofDenmark. The Design and Innovation program is run by DTU-MAN and this project hasbeen produced in close collaboration with DTU-MEK, where most of the work has beencarried out due to the nature of the project. This thesis counts for 30 ECTS points for the period from 1st of March to 10th of September 2010.

First of all I would like to thank my supervisor at DTU-MEK, Rajan Ambat for believing inme and getting me into this project and for his help, support and encouragement throughoutthe project. And a big thank to my supervisor at DTU-MAN, Torben Lenau for his hugesupport throughout the project even though I left another project with Torben Lenau tomake this project.

Thanks to the Danish Agency for Science, Technology and Innovation for their economicalsupport of the project.

A big thank to my industrial contacts Kristian Rechendorff and Klaus Pagh Almtoft from theTribology Centre at Danish Technological Institute and Ib Kongstad and Flemming Jensenfrom Bang & Olufsen, they have been very kind and keen to participate and skilled whenanswering technical questions and joining discussions.

I have been several times at DTU Risø at the Department of Photonics Engineering where Ihave been met with great hospitality and a lot of help from Jørgen Schou and Stela Canulescu,thanks to them both.

A special thank to Ph.D.-student Juliano Soyama at DTU-MEK, with whom I have had a weekly meeting for discussing each other’s project.

A big thanks to all the people, that have helped me with equipment, discussions, setting upexperiments and so much more: Manthana Jariyaboon, Helle B. Brandt, Steffen Munch, LailaLeth, Flemming B. Grumsen from DTU-MEK and Christian Ravn from IPU. Also LarsPedersen and John Troelsen at the workshop at DTU who are always helpful and extremelyefficient.

Thanks to Søren Bredmose Simonsen, Ph.d. student at Haldor Topsøe and again KristianRechendorff from Danish Technological Institute, who has proofread this thesis.

Finally thanks to Per Møller who is always ready for a discussion and throwing twenty ideas atthe table at the same time.

Technical University of Denmark, 2010-09-10.Supervisors:Torben Lenau, DTU-MAN, Section for Innovation and SustainabilityRajan Ambat, DTU-MEK, Section for Materials and Surface Engineering

__________________Martin Aggerbeck




2

2 Resumé Danish)

Tre aluminium handelslegeringer (1050, 5754 og 6082), med forskelligeoverfladebehandlinger (kombinationer af ætsning og polering) er blevet undersøgt medLOM, GDOES og fotospektrometri, for at belyse, hvordan og hvorforlegeringssammensætning og overfladebehandling påvirker aluminiumemners optiskefremtoning. Der er fundet frem til at refleksionsfarven er legeringsafhængig og atrefleksionsintensiteten både er afhængig af overfladebehandling og legeringens renhed.Undersøgelser har vist at et emne med høj overfladeruhed efter væskepolering bliver sortunder anodiseringsprocessen. Indledende forsøg med PVD film af Al og Ti på støbt Al erudført for at undersøge sammenhængen mellem Ti-koncentrationen og emnernes optiskefremtoning. Opståede problemer under anodiseringen startede undersøgelser af filmen medLOM, SEM og XRD. Få fotospektrometrimålinger er udført på de u-anodiserede PVD f ilm.

Igangværende forsøg af et vandmærke under det anodiserede lag og et mokumé gane-produkter blevet opstartet.

3 Abstract

Three commercial aluminium alloys (1050, 5754 and 6082), with different surfacetreatments (combinations of etching and polishing) has been investigated by LOM, GDOESand photospectrometry, to study how and why alloy composition and surface treatmentaffects the optical appearance of an Al specimen. It is found that reflection colour is alloydependent and reflection intensity depends on both surface treatment and alloy purity.

Studies have shown that a specimen with high surface roughness caused by water polishingturns black during anodising. Introductory experiments have been made with PVD coatingsof Al and Ti on cast Al to investigate, the connection between Ti-concentration and opticalappearance. Problems occurring when anodising the coatings have initiated examinations ofthe coatings by LOM, SEM and XRD. A few photospectrometry measurements have beendone on the non-anodised PVD coatings. Ongoing experiments of a watermark under theanodised layer and a mokumé gane product have been initiated.




3

Abbreviations

The following abbreviations is used throughout this thesis:AFM Atomic Force Microscopy

BEI Backscatter Electron Imaging

CRT Cathode Ray Tube

DC-diode Direct Current two-electrode

EDS Energy Dispersive X-ray Spectroscopy

LOM Light Optical Microscope

PVD Physical Vapour DepositionRFGDOES

Sometimes just GDOES

Radio Frequency Glow Discharge

Optical Emission SpectroscopySDOM Standard Deviation Of Mean

SEI Secondary Electron Imaging

SEM Scanning Electron Microscope

TEM Transmission Electron Microscope

XRD X-Ray DiffractionTable 4.1: Overview of abbreviations.




4

Table of Contents

1 Acknowledgements ......................................................... 1

2 Resumé Danish)............................................................. 2

3 Abstract ........................................................................ 2

4 Abbreviations ................................................................ 3

5

Introduction ................................................................... 6

6 Theory .......................................................................... 8

6.1 Optics ........................................................................................................ 8

6.2 Materials ................................................................................................... 9

6.2.1

Aluminium ........................................................................................... 9

6.2.2

Recycled cast aluminium..................................................................... 13

6.3 Surface modification processes for aluminium ............................................. 14

6.3.1

Polishing by buffing and water polishing ............................................. 14

6.3.2 The anodising process ........................................................................ 15

6.3.3 Physical Vapour Deposition (PVD)....................................................... 19

6.4 Characterisation techniques....................................................................... 25

6.4.1 Scanning Electron Microscope (SEM) .................................................. 25

6.4.2 (Radio Frequency) Glow Discharge Optical Emission Spectroscopy((RF)GDOES)................................................................................................. 27 6.4.3 XRD................................................................................................... 27 6.4.4

Spectrophotometer measurements....................................................... 28

7

Materials and preparation techniques .............................. 29

7.1 Materials used.......................................................................................... 29

7.1.1 Used Al alloys ................................................................................... 29

7.1.2 Titanium grade 1 ............................................................................... 30

7.2 Surface finishing treatments on commercial alloys ....................................... 30

7.2.1 Polishing using a buffing wheel ........................................................... 31

7.2.2 Anodising treatment........................................................................... 32

7.2.3 Naming of commercial alloy specimens ............................................... 33

7.3

Synthesis of aluminium based plasma coatings............................................ 34

8

Characterization techniques ........................................... 36

8.1.1 Optical microscopy and SEM.............................................................. 36

8.1.2 GDOES ............................................................................................. 37 8.1.3 XRD................................................................................................... 38

8.1.4 Photospectrometry ............................................................................. 39

8.1.5

Measurement errors on photospectrometry ......................................... 42

9 Results and discussion ................................................... 43

9.1 The effect of alloy and treatment on optical properties ................................ 43

9.1.1

Anisotropy in optical measurements on a surface ................................. 43




5

9.1.2

Presentation of the anodised specimens...............................................44

9.1.3

LOM examination of the anodised specimens ......................................45

9.1.4 Element enrichment in the surface........................................................46

9.1.5 Alloy / treatment for anodised specimens ............................................48

9.1.6

Full, diffuse and specular reflection .....................................................51

9.1.7 Visual effect of anodising....................................................................56

9.1.8 Summary on alloy and treatment affect on optical properties................60

9.2 Water polished specimens .........................................................................62

9.2.1 Black anodising..................................................................................63

9.2.2

The etching experiment on water polished specimen ............................65

9.2.3

Summary on water polishing part........................................................70

9.3 The effect of Ti-concentration on optical properties in aluminium coating.......70

9.3.1

Examination of the PVD coating ..........................................................71

9.3.2

Visual impact of titanium concentration on full reflection.......................74

9.3.3

Full vs. diffuse reflection......................................................................76

9.3.4

Change of substrate material ..............................................................78

9.3.5 Anodising the PVD coatings ................................................................78

9.3.6 Reason for bad anodising...................................................................80

9.3.7 Summary on PVD part ........................................................................83

9.4 Watermark under the anodised layer .........................................................84

9.5 Mokumé gane ..........................................................................................85

10 Conclusion 87

11

Future Work 88

12

References 89




6

Introduction

Aluminium is a key part of the green technology of the future. The possibility ofmanufacturing low weight products and thereby increasing the energy efficiency, e.g. in thetransport sector is very important. Moreover aluminium is very easy and cheap to recycle,however, the recycled (secondary) aluminium contains a high amount of impurities, making itless ef ficient to use than the newly extracted (primary) aluminium. Inter-metallic particles ofimpurity elements are formed during casting and thermo-mechanical treatment in thesecondary aluminium. This is the source of problems and secondary aluminium is thereforemore obvious for casting than as wrought aluminium, since cast aluminium is not as vulnerable to impurities.

For many products, when one uses aluminium for designing (e.g. Bang & Olufsen), the use of

the casting process is important because complicated shapes can be easily fabricated.However, the cast components have a more inhomogeneous microstructure compared to thecomponents fabricated by a machining process from thermo-mechanically treated aluminium(sheets, bars, etc.). Additionally, the cast components have a number of defects especially the porosity is important. Due to these facts and the chemical composition of the cast aluminium, problems occur with surface modification especially anodising for decorative application.Unfortunately, cast aluminium turns black and contagious when anodised caused by thesilicon added for easing the casting process.

The problems can hopefully be solved by coating the cast components by an aluminiumcoating, either commercially pure or added alloying elements. The physical vapour deposition (PVD) process can make the coating, which then can be anodised, so that the adverse effect ofthe cast components can be overcome. The coating can be used for acquiring both gooddecorative optical and / or corrosional properties. However, in order to achieve this, themicrostructure and composition of the coating should be controlled; PVD parameters shouldbe adjusted properly, and the coating microstructure needs to be investigated in detail.Furthermore, the used plasma technique also allows the introduction of additional alloyingelements, which are not possible to alloy with aluminium by a melting process. Thereby theoptical and corrosional properties of the coating can be varied. The present work focuses onthe optical properties of aluminium specimens as a function of chemical composition andsurface treatment.

This project is a part of the work carried out under the innovation consortium project, IdeAlsurfaces, started in March 2010. The consortium consists of six industrial partners and fiveknowledge institutes in Denmark, Belgium, Germany and Italy and is partly founded by the Danish Agency for Science Technology and Innovation.

The overall objective with IdeAl is to develop a generic technology platform based onaluminium based coatings deposited on a variety of materials by PVD or laser cladding incombination with the anodising process. The aim is to produce coatings with good decorative,optical properties, and with satisfactory corrosion wear resistance.




7

The IdeAl project has two focuses:

1. Environmentally friendly corrosion protection of aluminium parts, among otherthings by avoiding the use of chromium and increasing the possibility of usingrecycled aluminium.

2.

Design of durable optical effects using aluminium based plasma coatings on a wide variety of substrate materials, e.g. recycled aluminium, steel and other metals, polymers, composites and ceramics.

The goal for this report is to investigate to which extent alloy composition and surfacetreatments affect the optical appearance of aluminium, and gaining experience with analysesmethods of the structure of a PVD coating and its optical properties.

The findings in this project will be further exploited and strengthened during the Ph.D. program stating in October, hopefully developing new ways to decorate aluminium surfaces.

This thesis is divided into two major parts:

1. Alloy composition and surface finishing’s effect on optical properties of anodisedspecimens (sections 9.1 + 9.2)

This part should give fundamental contribution for a future standard method of analysingoptical properties of aluminium specimens.

2. The effect of titanium concentration in aluminium coating on optical properties(section 9.3)

This part should give experience with the use of varying compositions with the PVD process

and how the change of concentration of only one element affects the optical properties of thespecimen.

Two minor parts in the ends of the report presents ongoing experiments making a watermarkbeneath the anodised layer and a remote control decorated by the mokumé gane technique.

The work on this project has been conducted in collaboration with Danish TechnologicalInstitute (hereafter TI), DTU Department of Photonics Engineering and Bang & Olufsen(hereafter B&O).

This project is partly based on experience from a master thesis from 2008, Anodisering af aluminium med optimerede egenskaber (Anodising of aluminium with optimized properties) byH. Tobias Holt and Rasmus Hauberg Møller [1] and a Ph.D. thesis from 2009 Advanced Anodizing Technology by Naja Tabrizian-Ghalehno [2].




8

6

Theory

6 1

Optics

Optics is a huge scientific field. Here, only the relevant topics will be briefly described.

Light can be viewed as a beam, as an electromagnetic wave or as particles. Light interacts withsurfaces and generates various visual effects, which are determined by the interaction of thelight waves with the material. This causes effects such as scattering, reflection, refraction,interference, absorption and diffraction depending on the nature of the surface and thematerial. In this project the focus is on the interaction of the visible spectrum of light with thesurface causing various visual effects. The visible spectrum of light is not distinct but variesfrom person to person and therefore from source to source [3] [4]. The visual spectrum isapproximately from 400 nm (violet) to 750 nm (red).

When a beam of light is hitting a surface the reflection can be described as a function of [5][6]:

• Specular reflection: The direct reflection, where the incident angle equals thereflected angle, e.g. a mirror.

• Diffuse reflection: The light scattered in the surface topography, e.g. a piece of paper.• Ambient reflection: The light scattered from the surroundings without the current

light source taken into account.

Diffuse reflection comes as a result of the surface roughness reflecting the light in alldirections in a specular way, but in such small areas, that it is not perceived as specularreflections but as a diffusing surface. Diffusivity therefore increases with surface roughness with magnitude comparable to the wavelength of light. In Figure 6.1 is illustrated specular anddiffuse reflection, where the blue arrows are the incoming rays of light, and the red arrows arethe reflected rays of light.

Figure 6.1: Illustration of specular and diffuse reflection (source: http://www.physicsclassroom.com).

From Figure 6.1 it is seen that the diffusivity of a material is increased by the surfaceroughness.




9

6 2 Materials

The IdeAl consortium, and therefore this project, focuses on lightweight application ofaluminium and aluminium based coatings. Both wrought and cast aluminium is also used asunderlying substrate for the aluminium coatings. Also other metals, ceramics, polymers andcomposites are to be used as substrate in the long term.

6 2 1

Aluminium

Aluminium is extracted from bauxite1 and it requires several steps and a big amount of energyin formation of alumina (Al2O3) (extracted aluminium is called primary aluminium). On theother hand it only takes about 5 % of the energy for extracting new aluminium to recycle usedaluminium (called secondary aluminium) [7].

Aluminium has become remarkably popular and it is the most used non-ferrous metal with an

annual consumption of 25 million tons [7]. The success is primarily due to the big diversity ofsemi products and functionality in castings and extrusions [8] and due to the relationshipbetween density and Youngs modulus, which is as high as the one for stainless steel, while atthe same time aluminium structures turns out to be up to 50 % lighter than equivalentstructures made of stainless steel [7]. Aluminium is the lightest of all common metals (2700kg/m3) making it excellent for e.g. the transport industry improving the fuel consumptionsignificantly.

When pure aluminium is exposed to atmospheric air a thin passivating oxide layer of alumina(Al2O3) is formed on the surface. After creating the layer all over the specimen the process willstop. Alumina is very hard and highly resistant to corrosion in atmospheric and wetenvironments and therefore the alumina is effectively protecting the underlying aluminium.The natural oxide film is 1-3 nm [8] and can therefore easily wear off, but a thicker layer ofalumina can be created in a controlled way by the anodising process which is described indetails on page 15.

1 Bauxite is a form of rock and consists of several minerals including Al(OH3), -AlO(OH),

AlO(OH) and iron and titanium oxides [6]




10

Aluminium has a high reflectivity, which does not vary significantly with wavelength in the visible spectrum of light. Therefore aluminium is often used in mirrors, reflector screen,optical equipments and the like [4]. The optical spectrum of the reflected light fromaluminium is presented in Figure 6.2 along with silver, platinum, gold and iridium.

Figure 6.2: Representation of the reflectivity as function of wavelength(source: http://www.furuyametals.co.jp).

As seen in and Figure 6.2 aluminium has a relatively uniform reflection of the visible lightaround 91-92 %. This is quite unique together with silver, which is the reason that thesematerials are the ones used for making mirrors. In general metals have an area of reflectionaround 40-60 %2, as seen for iridium and gold in the shorter wavelengths.

Pure aluminium is very soft and mechanically weak and therefore not useful for manyapplications. This is why aluminium alloys are added various amounts of additional elementsas shown in Table 6.1. Aluminium alloys also invariably contains impurities such as Fe, Si andMn to a certain extent. Making extra pure aluminium is very costly, therefore some level ofimpurities is accepted in all aluminium alloys. Problem can occur during surface modificationor corrosion can appear due to these impurities and their electrochemical behaviour.

2 From a discussion with Jørgen Schou, Senior Researcher at DTU Fotonik




11

Various aluminium alloys have been divided into groups according to their primary alloyingelements. In the EN 573 and EN 1780-standards the casting alloys have a group numberfollowed by four others while wrought alloys have the group number followed by threenumbers. A quick overview of the main groups are seen in Table 6.1:

Primary alloyingelement(s)

Cast alloys Wrought alloys Typical applications

None 1xxxx 1xxx Foil, sheets, decorative

Copper (Cu) 2xxxx 2xxx Aircraft industry

Manganese (Mn) 3xxx Cans, building radiatorsSilicon (Si) 4xxxx 4xxx Heat exchangers

Magnesium (Mg) 5xxxx 5xxx Cans, transportation, building

Magnesium + silicon 6xxx Transportation, building

Zinc (Zn) (+ copper) 7xxxx 7xxx Aircraft industry, radiators

Other elements 8xxx Foil and aircraft industryTable 6.1: The main group of the aluminium alloys.

The 8xxx series contains often Fe+Si and Li.

There are different reasons for the use of different alloying elements. In the following theeffect of some of the elements is described:

Iron is mainly an unwanted but common impurity. Normally it gives grey or black stripes ifthe Fe-Si ratio is greater that 7:1. The stripes are a result of inter-metallic compoundsbetween iron and silicon.

Silicon is both seen as impurity and alloying element. It is normally used in cast alloys to easethe casting process. Even small amounts of silicon can give shadow formation in the surfaceafter anodising.

Titanium is a grain refiner. It can cause grey or black stripes as with iron.

Magnesium and manganese enhances the mechanical properties of the alloy. Magnesium doesnot have a large impact on the decorative properties up to 7 % where manganese content upto 1% will vary the surface from clear to brown to clearly spotted. 0,3-0,5% of manganese willcause a brown colouring due to formation of MnO2. Manganese also enhance the weldabilityas silicon.

Chromium content of 0,3% causes a yellowish surface.




12

From a decorative point of view the following limits are advised:

Element Advised Concentration limit

Iron (Fe) 0,5%

Silicon (Si) 2-3%

Magnesium (Mg) 7%

Manganese (Mn) 0,5-0,8%

Zinc (Zn) 6-8%

Chromium (Cr) 0,3%

Titanium (Ti) 0,3%

Copper (Cu) 1-2%

Table 6.2: Advised limits from a decorative point of view [9].

There are many applicable alloy groups for this project but here only the most relevant wrought alloys will be described, 1000- , 5000- and 6000- alloys that is.

The 1000-series

The 1000-series consist of alloys with a purity of at least 99 % with Si and Fe as the major

impurities. These alloys are highly corrosion resistant due to the low amount of alloyingelements [7], have high heat and electrical conductivity and are very soft and thereby good workability.

The 1000-series is mainly used for electrical and chemical fields plus decorative applications.

The 1050-alloy is generally pure with various extents and very much used for decorative andother applications where high levels of mechanical properties are not needed. The 1050 alloygives a good compromise between mechanical property, ductility, and decorative appearance[7]. Because of its purity of 99,5% it is very soft compared to the other used alloys.

The 5000-series

The main alloying element in the 5000-alloys is Mg (normally up to 5 %) added in order toenhance the mechanical properties. Apart from Mg the 5000-series is often alloyed with Mn(additional mechanical strength), Cr (corrosion protection) and Ti (grain refiner).

Heating can cause precipitation of Al3Mg 2 intermetallic compounds, commonly called the - phase. Precipitations happen in the grain boundaries and are anodic with respect to the grainsand therefore there is a risk of intercrystalline corrosion especially when the Mg-level is high,under significant strain hardening, and at high service temperatures [7]. For the same reasoncaution is recommended when using more than 3 % Mg if the alloy is used above 66 °Cbecause of the risk of stress corrosion [10].

The 5754-alloy is mainly alloyed with Mg, and with the combination of Mn or Cr the alloy issuited in the building sector, transport and mechanical industries. [7]




13

The 6000-series

The main alloying elements in the 6000-alloys are Mg and Si. They form magnesium silicide(Mg 2Si) making very fine precipitates that increases the strength of the alloys. The 6000-series have many very good properties e.g. mechanical, corrosional and visual. Therefore the6000-series is widely used for extrusions around 6 million tons is used per year [7].

Due to its favourable properties, the 6000-series is used in a large variety of products in thetransport industry and other mechanically requiring applications.

The 6000-seies have a slight tendency to intercrystalline corrosion but normally not to a veryhigh extent. These alloys do not experience stress corrosion.

The alloy 6082 is widely used in commercial vehicles, railcars, shipbuilding, in the mechanicsindustry, and as forging stock [7].

6 2 2 Recycled cast aluminium

In 2008 according to [11] the amount of recycled aluminium covers 20-30 % of the totalspend aluminium in Europe. The recycled aluminium comes mainly from the buildingindustry, the transport sector, mechanical and electrical engineering and householdappliances. In 2004 the amount of collected scrap was 70 % of the available resources [7].Because of the vast amount of impurities the recycled aluminium is more or less only used forcasting [11].

About 60 % of all cast aluminium today is used in the automotive industry as enginecomponents [7]. Many of these alloys contain a lot of silicon and also at least 3 % copper. Thereasons for the use of silicon are to enhance the ability of filling the form and reduce thesuction of aluminium when it hardens after the casting process. As seen on Figure 6.3 the Al-

Si phase diagram has a eutectic point around 12,6 wt% of Si. Most cast alloys are around thisamount of Si creating a eutectic microstructure.

Figure 6.3: Al-Si phase diagram [1].

Copper is also added to ease the casting process.




14

The 40000-series

The 40000-series is the most commonly used of the cast aluminium alloys. The series has agood castability due to its low eutectic point temperature (577 ºC).

The 46500-alloy contains about 2-4 % copper and is especially suitable for pressure castingand is much used for complicated geometries. When heated, particles of CuAl 2 will precipitate along the grain boundaries, which will enhance the mechanical properties.Unfortunately the CuAl2 particles will act cathodic to the aluminium. Therefore the materialaround the precipitates will slowly corrode. Consequently, the adhesion between the grains will slip and the material will fail [1].

6 3

Surface modification processes for aluminium

A number of surface treatment processes can be used for aluminium and its alloys; however,only the important ones used in this project will be described in detail.

6 3 1

Polishing by buffing and water polishing

Prior to the anodising process some of the specimens may have scratches, pits and otherinhomogeneous appearances that can be removed by polishing. There are differed betweenmechanical and chemical polishing, which are used for different roughness and finishing.Polishing can e.g. give a smooth mirror like surface for shiny products or subsequent coating process, or a uniformly rough surface making it suitable for e.g. painting.

In the present project, two types of mechanical polishing have been used, which will bedescribed in detail:

6 3 1 1

Wheel polishing

The wheel-polishing machine is using a cloth wheel for polishing. A wheel of round cloth- pieces is made and sewed together. The wheel is fixed on a rotating machine so that the cloth wheel is vertical and the machine spins usually more than 1200 rounds per minute [8]. A polishing wax is added and the machine can now be used for polishing.

Difference in the alloy can give different results when buffing due to their alloying elementsand hardness from e.g. heat treatments. In general, harder specimens are tougher to polish while the process heat softer specimens more. The temperature easily reaches 200 °C and maylocally rise up to 500-1000 °C [8].

The number of threads in the mop may vary, and a higher number of treads gives a harder buff.On the wheel are cloth disc separators adjusting the buffing wheel to become a bit softer.

The process is carried out by first polishing perpendicular to the largest scratches on thespecimen. When these are removed, the specimen is turned so that the next round of polishingis done perpendicular to the scratches just made when removing the big scratches.

Due to the heat from the polishing process Beilby [12] has suggested that the material fromthe peaks of the aluminium is displaced into the valleys making an amorphous and / or veryfine crystalline layer, called the Beilby layer.






16

In Table 6.3 the suggested thickness of the anodising layer can be seen according to the use ofthe final product:

Thickness Application areas

25 µm Surfaces exposed to an extensive form of corrosion or wear. Especially outdoorapplications in highly corrosive environments.

20 µm

Surfaces exposed to heavy or normal outdoor applications, e.g. for buildingmaterials, vehicles, ships, etc.Indoor surfaces exposed to chemicals or moist-containing atmosphere inconnection with production plants, process lines and equipment used in e.g. thefood industry.

15 µmSurfaces exposed to medium indoor wear, e.g. door handles or decorativeoutdoor areas.

10 µmSurfaces used indoors and outdoors in dry clean atmospheres e.g. reflectors,fittings, decorative plates for vehicles, and sports equipment.

5 µm Only suitable for surfaces used indoors.Table 6.3: Overview of suitable application areas as a function of the thickness of anodised aluminium

layer [9].

The anodised alumina layer is hard, scratch and corrosion resistant and can be used fordecorative applications uncoloured and can be coloured in a wide variety of colours.

The anodising process is normally divided into the several subprocesses, as seen in Figure 6.5, which will be gone through in the following:

Figure 6.5: Schematic anodising process.

6 3 2 1 Degreasing

During the degreasing phase oil, lubricants and other greasing agents are removed from thesurface. This can be done e.g. by immersing the specimen into a soap solution or simply usingethanol.

6 3 2 2

Etching

The etching process is often done in sodium hydroxide at 60 °C. This removes the natural passivating oxide and a thin layer of material from the surface as the microstructure of thislayer is different from the bulk. The etching process levels out larger surface imperfectionsbut it also creates a mat pattern to the surface, and is therefore not used for shiny specimens.During this process some precipitated products of alloying elements could be formed on thesurface, e.g. for copper containing alloys, copper enrichment is possible at the surface.

As seen in Figure 6.5 the etching process is not always done but is e.g. skipped when makingshiny surfaces.




17

6.3.2.3

De smutting

The purpose of the de-smutting phase is to remove smut (precipitates) from the etching process sticking to the surface of the alloys. It is normally done in a solution of concentrated

nitric acid. Pure aluminium gives less smut than highly alloyed aluminium. Nitric acid alsoremoves any enriched copper on the surface in the case of copper containing alloys.

6.3.2.4 Anodising

The anodising creates a film of aluminium oxide (alumina) on the surface by anelectrochemical reaction. The properties of the anodised layer depend on the substratecomposition, bath chemistry, temperature etc., and varying these parameters will change the properties of the anodised layer. Anodising is a commonly used process, normally performedon aluminium, titanium and magnesium.

Anodising is an electrolytic process where the specimen to be anodised is coupled as anode in

a chemical bath, and potential is applied so that the metal will dissolve from the surface, which subsequently is converted into oxide on the surface. The cathode in anodising is oftenaluminium or titanium, the latter being more expensive but also much more resistant to theelectrolyte and therefore hold much longer. The electrolyte in the electrochemical cell can besulphuric acid, chromium acid, boric acid and oxalic acid and others, or even a combination oftwo or more electrolytes [13]. Sulphuric acid is most common since it is effective and cheap.

The anodising converts the surface of the aluminium into alumina (Al2O3), which is a hard,brittle, very corrosion resistant transparent ceramic. During this conversion the material willexpand and one third of the total anodised layer will then be above the original surface of thespecimen [14]. The converted layer consists of pores in hexagonally shaped cells with a central pore perpendicular to the surface as illustrated in Figure 6.6.

Figure 6.6: Idealised anodising structure (source: electrochem.cwru.edu).




18

The pore diameter varies but it is typically about 15 nm and 1000 times thinner than thethickness of the layer [7].

The process starts at the surface and runs perpendicular into the specimen. Consequently, thesurface of the anodised layer will be exposed to the anodising agent for the longest time. Thiscan create a powdering effect where the pores do not have vertical walls, but they are incliningaway from the centre of the pore.

At the bottom of the pores is a thin dense barrier layer , which is only 0,1-0,2% of the total filmthickness. The barrier layer is non-porous and conducts the current from the electrolyte tothe specimen bulk building further the oxygen layer [8].

The anodising process is controlled by the voltage, current density, temperature and time. Theimpact of these will briefly be described [2]:

Voltage: When increasing the voltage the pore diameter is increased and thereby the numberof pores per cm2 decreases.

Current density : Low current density produces soft, porous and thin films.

Temperature: Low temperature gives less pores per cm2 but bigger pores and thicker cell walls. The growth speed increases with the temperature.

Time: The thickness of the coating is mainly controlled by the time. This is the case until amaximum thickness where the powdering effect will stop the layer from getting thicker. Theinclining walls of adjacent pores will simply meet and follow the pore growth downwards intothe specimen.

6 3 2 5 Colouring

The colouring process is normally done in one of the following three ways [9]:• Absorption colouring is the cheapest, simplest and most commonly used method. In this

process the specimen is immersed into a liquid solution containing the dye that will gointo the pores of alumina. This process can be used for a wide variety of colours.

• In electrolytic colouring a metal or metallic salt is precipitated in the bottom of the pores. The variety of colours are not as great as with dyes, but these colours are moreresistant to uv light and the colour therefore holds for a much longer time.

• Integral colouring colours the alumina instead of in the pores between the alumina. Inthis process the electrolyte of the anodising process is containing one or more organicacids that will colour the oxide layer. A big advantage is the anodising and colouring is

done at the same time in the same bath.




19

Figure 6.7 illustrates absorption colouring and integral colouring.

Figure 6.7: Colouring methods: Left: Absorption colouring (the Al2O3 is transparent).Right: Integral colouring.

6 3 2 6 Sealing

After anodising and the possible colouring process the surface has a high capillary effect as aresult of the porous surface making it extremely susceptible to fingerprints, oil, dirt etc.Additionally if dyes have been added into the pores, the dyes will tend to leave the pores ifthese are not closed. To avoid this, the specimen is sealed normally by dipping the specimeninto hot demineralised water around 100 °C. Sealing can also be done e.g. with steam (from water), chromate and dichromate.

During the sealing process the upper most ceramic alumina is converted together with the water into the softer mineral called böhmite:

Al2

O3

+ H2

O

2AlO(OH)2

During this conversion the surface will expand and close the top of the pores in the surface.

The sealing also stop particles from getting into the pores acting as corrosive particles.

During the sealing process the optical appearance can be impaired due to the closing of thecapillary pores. This is called sealing bloom. This can influence the interaction of light on thesurface. Also change in chemistry during the sealing process can affect the optical properties.

Temperature, pH, water characteristics and amount of silica in water are essential factors [2].

6.3.3

Physical Vapour Deposition PVD)

The PVD process is used for depositing a thin film of commercially pure aluminium oraluminium mixed with a controlled amount of other elements.

Since PVD is a plasma process the concept of plasma is described followed by an outline ofthe PVD process.




20

6 3 3 1

Plasma

By raising the temperature of a gas, the electrons will be removed from their atom ormolecule. This state of matter is called an ionised gas or plasma

Plasma is consisting of three types of particles:

• Neutral particles – atoms and molecules• Ions• Electrons

Because of the charged particles the plasma as a whole will be electrically conductingalthough the amount of ions often will constitute less than 0,1% of all the particles.

When heating creates the plasma it is called thermal ionization. In this case the plasma is produced by electrical discharges in a vacuum chamber, this is sometimes called electricalionisation.

The plasma is produced in an electric field between a positive anode and a negative cathode. Whenever an electron is removed from a neutrally charged atom/molecule, a negativelycharged electron and one positively charged ion is formed. The electron will be acceleratedtowards the anode and on its way it is likely to hit one or more electrons making more charged particles – again hitting other electrons – thereby maintaining the plasma.

The initial electron will hit electrons that will hit a whole subsequence of electrons. For thisto happen the initial electron most have a lot of energy. Apart from hitting electrons freeingfrom their atoms, energy is also absorbed by excitation of electrons that during decaying emitsa photon. This energy is wasted according to the maintenance of the plasma.

The initial electron is generated when positively charged ions bump into the target material ofthe cathode.

In Figure 6.8 the voltage drop between the anode and cathode can be seen.

Figure 6.8: Voltage drop over anode/cathode region [1]. Left: Before plasma is ignited. Right: Plasma isswitched on. Note that the gradient is mainly close to the electrodes.

Normally Vp is 10-30 V and V0 is 300-500 V.

Notice that the actual potential of most of the plasma (V p) is higher than the potential of theanode due to the fact that the electrons have a high velocity compared to the other particles.The electrons will therefore more often leave the plasma than the ions.




21

6 3 3 2

The PVD process

In this project the focus has been primarily on the non-reactive PVD processes. The non-reactive PVD-process is subdivided into thermal evaporation processes where the target is

heated to evaporation, and the sputtering process, which is the focus of the following part.The sputtering process is using an ionized gas, usually Ar, hammering into the target, which isthen atomized, i.e. sublimated. Because of the low pressure inside the chamber the atomizedtarget material cover everything inside the chamber including the chamber walls, the holders,and the substrates as seen in Figure 6.9. The PVD process is a line-of-sight process meaningthat only surfaces in straight line-of- sight from the target will be coated.

Figure 6.9: Everything in the chamber is coated during the process.Notice the Ar inlet in the top of the chamber.

A BIAS voltage can be coupled to the specimen holder making the specimens slightlynegatively charged e.g. 40 or 80 V. The negative charge attracts the Ar-ions, which therebyslightly sputter the surface of the specimens.




22

The amount of target atoms that is atomized is depending on the sputtering yield (also calledsputtering coefficient) of the target material (Figure 6.10). The sputtering yield is def ined as“the number of atoms that is removed from the target per incoming ion”.

Figure 6.10: Sputter yield / coefficient as a function of atomic number [9].

In Figure 6.10 it is seen that it is easy to sputter e.g. gold and silver but hard to sputter siliconand titanium.

6 3 3 3 Magnetron sputtering

Magnetron sputtering is basically the simple sputtering configuration with additionalmagnetic fields. The extra magnetic field is close to the target (cathode) and is perpendicular

to the electric field imposed by the electrodes as illustrated in Figure 6.11.

Figure 6.11: The principle of magnetron sputtering [15].

This forces the particles to circulate through the magnetic field instead of following themagnetic field in a straight line to the substrate (anode). Because of elongated path thenumber of collisions is raised and the plasma can be maintained at a lower pressure. Aconventional DC-diode sputter operates at 0,5 - 10 Pa where as a magnetron sputtering process can operate at 10-3 – 1 Pa. The reduced pressure leads to less contamination of the process e.g. from the Ar gas. The lower pressure is also raising the mean free path of the target particles, which is the average path the particles can move before they collide with a particlefrom the gas. The longer mean free path, the higher the energy of the target particle when itreaches the substrate and the better quality is the f inal coating at the substrate.




23

When comparing magnetron sputtering, to conventional sputtering the former process raisesthe deposition rate and the ionisation efficiencies in the plasma and lowers the substrateheating effects [16].

The non-reactive sputtering process can be combined with a reactive gas (typically oxygen ornitrogen) in the chamber, turning it into a reactive sputtering process. This is e.g. used to makecoatings of: TiN, TiO2, TiAlN and CrN.

One of the problems with magnetron sputtering is the low usage of the target, which for atstandard cathode is generally less than 30 % [15]. Inserting an extra magnet in the centre ofthe cathode and unbalancing the magnets to one another have improved the utilizationsignificantly. In Figure 6.12 the type 1 has a stronger centre magnet while type 2 has a strongerring of magnets in the perimeter of the cathode.

Figure 6.12: Schematic representation of the plasma confinement observed in conventional andunbalanced magnetrons [16]. Type 1: Stronger centre magnet.

Type 2: Stronger ring of magnets in the perimeter.

The type 2 unbalanced magnetron sputtering is the most used and gives plasma that is notstrongly confined to the target region but is also flowing towards the substrate.

6 3 3 4

Coating structure model

There are important considerations to be done according to the microstructure of thedeposited thin film when coating with PVD. It is important to know the crystal structure and whether the coating is tightly closed or if it contains small pores. For making an overview ofthe structure of thin PVD coatings Thornton [17] made a structure model (Figure 6.13).




24

This model has been developed from experiments with OFHC (Oxygen Free High thermalConductive) copper [17]:

Figure 6.13: The Thornton structure model [17].

The Thornton structure model is divided into three main zones and a transition zone:

Name StructureTemperature range,suggested in [17]

Zone 1 Porous structure T/TM

Zone TDensely packedfibrous grains

0,3

Zone 2Tight columnarstructure

0,5

Zone 3 Equiaxial structure

Table 6.4: Structure zones in the Thornton model.TM is the melting temperature of the coating material.

In several cases the Thornton model has not been completely satisfying thus others have madestructure models e.g. Messier and Kelly & Arnell [16]. There have been discussions on whether the y-axis should describe the Ar pressure or if other values (if necessary combined) would be more appropriate and if this actually should be four dimensional or even more.

In relation to the structure model, the most important to this project, is being aware of thedifferent structure zones.




25

6.4 Characterisation techniques

The characterisation have primarily been carried out by four characterisation techniques:scanning electron microscope (SEM), radio frequency glow discharge optical emissionspectroscopy (RFGDOES), normally called G-DOES, X-Ray Diffraction (XRD) and photospectrometry. In this section these four methods will briefly by described.

6.4.1 Scanning Electron Microscope SEM)

For many tasks in material technology a light optical microscope (LOM) can be used, butsometimes it is necessary with a higher magnification. In this case a scanning electronmicroscope (SEM) is used. If even higher magnification is needed a transmission electronmicroscope (TEM) should be used. Here follows a description of the SEM with the mostrelevant features and detectors.

In Figure 6.14 is a schematic overview of a SEM.

Figure 6.14: Schematic overview of the component in a SEM [source: Wikipedia.org].

The source to the electrons is the electron gun, which is basically a heated filament of tungstenor LaB6, which emits electrons (a field emission electron gun can also be used). The electronsare attracted by an anode ring and are hereby accelerated towards the specimen. The lensesfocus the beam via electromagnets. The electron beam goes through the deflection coils that

control the direction of the beam and can thereby make the electron beam sweep over thespecimen surface like the electron beam of an old CRT (cathode ray tube) television screen.

A vacuum pump is connected to the microscope chamber to avoid the electrons of the beamfrom collisions with the particles in the air. The magnitude of the vacuum is among otherfactors dependant on the filament type and the sample type, but is in general the interval of10-4-10 -2 Pa for W and LaB6 filaments.




26

When the beam hits the sample two different kinds of electrons will be scattered from thesurface of the sample with energies as seen in Figure 6.15.

Figure 6.15: Energy distribution of scattered electrons.

Area A: Secondary electrons. Area B: Back scattered electrons [18].

The electrons in area A with low energy (< 50 eV) are generally the secondary electrons (thebeam is the primary electrons) and the electrons in area B with high energy (as high as theincident beam) are called the backscattered electrons.

The secondary electrons are electrons being removed from its atom by a primary electron. Anelectrical field of +400 V attracts the electrons to the secondary electron detector (Figure 6.14).The brightness of the image from this detector depends on the number of secondary electronsreaching the detector. The signal from the secondary electrons origins from only a fewnanometres into the sample surface. Because of the narrow beam it will have a large depth offield compared to e.g. a LOM. Due to these abilities the secondary electron images (SEI) isexcellent for images of topology in a specimen surface.

The backscattered electrons are the primary electrons scattered back after a collision with anatom in the sample. A donut-shaped backscatter electron detector is situated over the sample

around the incoming beam, since this is where the backscattered electron signal has thehighest intensity. Heavy elements in the sample will scatter electrons back more strongly thanlight elements, and therefore heavier elements will appear brighter on the backscatter image.The picture, are a result of the mean value of the elements in one spot, meaning that e.g. Al will appear brighter (heavier) than Al2O3 since the mean weight of pure Al is higher than themean weight of Al and O (with and without the ratios considered). The signal of thebackscattered electrons can origin from a depth of up to 100 nm into the sample, and therebymuch deeper than the secondary electrons [18]. Due to the mentioned abilities thebackscatter electron images (BEI) are normally used for studying the surface for differencesin elements, e.g. whether a particle in the surface is a different material than the substrate.

Apart from the secondary and the backscatter electron detectors an X-ray detector is oftenused. After emitting a secondary electron, the atom is left in an exitet state with a hole in oneof its shells. An electron from an outer shell fills the hole, and this can generate a characteristic X-ray with an energy corresponding to the atomic energy shell. The energy / wavelength ofthe emitted electromagnetic wave is registered by the X-ray detector. All elements have itsown characteristic X-rays and by e.g. energy dispersive X-ray spectroscopy (EDS or EDX) it is possible to quantify the amount of the elements in a given area of the sample. By thistechnique the EDS can be used to examine the composition of an area of a specimen.




27

6.4.2 Radio Frequency) Glow Discharge Optical Emission

Spectroscopy RF)GDOES)

GDOES is used to make a depth profiling of elements in a material by sputtering into the

surface while detecting optical discharge emissions.An argon gas is ionized and accelerated into a vacuum chamber and hammer into thespecimen, by which the atoms in the surface are removed from the surface into the chamber(Figure 6.16). Additionally the atoms are exited and therefore emitting characteristic wavelengths like in an EDS-analysis. The GDOES is slowly sputtering from the surface intothe specimen and thereby the GDOES can be used both for substrate analysis, to get a pictureof the element distribution from the surface and into the specimen.

Figure 6.16: Schematic overview of the GDOES chamber (Source: The graphical abstract of [19]).

After an experiment it is possible to quantify the result, so that it is possible to calculate the weight percentage of each element.

By initial experiments with a given material it is possible to know the depth of the crater as afunction of the time.

6.4.3

XRD

X-ray dif fraction (XRD) uses X-ray waves, which are sent towards a material. It is mainly theelectrons of the material that reflect the X-ray beam, and the beam is thereby diffracted [20].

The diffraction pattern is specified by the wavelength of the beam,, the spacing between thelattice planes, d, and the incident angle, . The interference from the diffraction is completelyconstructive when Bragg’s law is fulfilled [3]:




28

2d sin = n where is the incident angle to the surface, n is an integer and is the wavelength of the incoming beam. A schematic illustration of Bragg’s law is presented inFigure 6.17

Figure 6.17: Schematic illustration of Bragg's law (source: commons.wikimedia.org).

XRD can be used for several purposes e.g. identifying crystallographic phases and lattice typesand parameters.

6 4 4 Spectrophotometer measurements

A spectrophotometer is used for measuring the intensity of light in the visible and near visiblearea as a function of wavelength. The results from a spectrophotometer is giving a relative value of the intensity over the visible spectrum, thus it is possible to derive the generalintensity and the colour of the light from the graph. Spectrophotometers can be used for alltypes of sources, e.g. measuring directly into a light source, transmission through e.g. a liquid ora reflection on a material surface, which will be additionally elaborated here.

When measuring the reflection of a material, it is often preferred to use an integrating sphere.An integrating sphere is a hollow sphere covered on the inside with a white layer, with a highlydiffusing and powerful reflection. With this equipment it is possible to get a completemeasurement of both the specular and the diffuse reflection of the current surface. Since anintegrating sphere diffuses completely the incident light the result is the same no matter where a beam of light of a given intensity and colour enters the integrating sphere. [21]




29

7 Materials and preparation techniques

In the following chapter the used materials and the preparation techniques employed in the present project is described.

7 1

Materials used

The primary materials for the project have been commercial aluminium alloys and aluminiumbased plasma coatings.

7 1 1 Used Al alloys

During the project, four wrought alloys were used and one cast alloy. The details of thesematerials are given below:

7 1 1 1

Alloy 1050

The 1050 alloy has the following composition:

Alloy Mg Si Fe Cu Mn Zn Ti V

1050 0,05 0,25 0,4 0,05 0,05 0,05 0,03 0,05Table 7.1: The composition of the 1050-alloy [22].

Specimens of 50*50*1 mm has been cut out of a sheet of 1050 alloy supplied by LMG(Lemvigh-Müller) in Denmark.

7 1 1 2

Alloy 5754

The 5754-alloy has the following composition:

Alloy Mg Si Fe Cu Mn Zn Ti Cr

5754 2,6-3,6 0,4 0,4 0,1 0,5 0,2 0,15 0,3Table 7.2: The composition of the 5754-alloy [22].

Specimens of 50*50*1 mm has been cut out of a 5754 alloy sheet by an unknown supplier.

7 1 1 3 Alloy 6082


Alloy Mg Si Fe Cu Mn Zn Ti Cr

6082 0,6-1,2 0,7-1,3 0,5 0,1 0,4-1,0 0,2 0,1 0,25Table 7.3: The composition of the 6082-alloy [22].

Specimens of 50*50*1 mm has been cut out of a 6082 alloy sheet supplied by LMG.




30

7 1 1 4 Alloy 6401


Alloy Mg Si Fe Cu Mn Zn Ti6401 0,35-0,7 0,35-0,7 0,04 0,05-0,2 0,03 0,04 0,01

Table 7.4: The composition of the 6401-alloy [23].

Specimens of 75*40*5,5 mm in alloy 6401 ALMINOX alloy were delivered from B&Osupplied by WkW-Erbslöh Automotive, Velbert, Germany. ALMINOX is a special flawlessalloy that is only produced by Erbslöh [24].

The specimens were polished by buffing on one side before sent to TI.

7 1 1 5 Alloy 46500

The cast 46500 alloy has the following composition:

Alloy Mg Si Fe Cu Mn Zn Ti Pb Sn Ni

46500 0,15-0,55 8,0-11,0 0,6-1,2 2,0-4,0 0,55 3,0 0,2 0,35 0,25 0,55Table 7.5: The composition of the 46500-alloy [25].

The main alloying elements in the 46500-alloy are Si and Cu.

Specimens of 75x25x4 mm in alloy 46500-alloy was delivered from B&O supplied from PDCTeknik, Denmark. The specimens were polished on one side before sent to TI.

7 1 2

Titanium grade 1

Titanium grade 1 has the following composition:

Alloy C Fe H N O Ti

Ti Grade 1 0,1 0,2 0,015 0,03 0,18 99,5Table 7.6: The composition of the titanium grade 1 [26].

The grade 1 titanium is a commercially pure alloy.

The titanium is used as cathodes in the PVD process. With varying power on the twocathodes it is possible to vary the amount of Ti in the PVD coating. It this way a series of

specimens with different amounts of Ti has been produced to investigate how this changesthe optical properties of the coating.

7 2

Surface finishing treatments on commercial alloys

Here is the summary of various surface treatments employed throughout the project forinvestigating the effect of surface finishing on optical properties before and after anodisingon the commercial 1050, 5754 and 6082 alloys.




31

7 2 1 Polishing using a buffing wheel

The buffing is done on a Polette 6NE from REF-Motor A/S at DTU (Figure 7.1). Themachine runs 3000 rounds per minute and uses a standard 150 mm cloth disc.

Figure 7.1: The used buffing machine.

As polishing agent is used a wax block, which is designated as coarse grained . The actual polishing agent is unknown.

The polishing was examined and accepted based on visual inspection.

To transport heat away from the specimen, a small aluminium block was cut with a recess in

order to hold the specimens while polishing. By using the block the chance of generating the previously described Beilby layer was reduced.

7 2 1 1 The water polishing machine

Water polishing was done on a Vapormate 3 delivered by ABS-Hesø I/S at DTU (Figure 7.2).

Figure 7.2: The used water polishing machined.




32

The top of the machine was opened and the specimen was placed on a cloth on the floor ofthe chamber. The jet of water and particles was forcefully applied perpendicular to thespecimen surface and generally done in easy motions while ensuring that the whole specimenis being polished more or less equally.

7 2 2 Anodising treatment

The anodising of the required aluminium specimens is carried out using a sulphuric acid bath.The specimen is mounted on a titanium holder and a steel cylinder is used as the otherelectrode. The experiment has been done in standard beakers with the following details:

• Degreasing with ethanol• 4 minutes etching in a 60 g/L Alficlean solution at 60-62 ºC. Alficlean is a NaOH-

based cleaning and mild etching agent used at B&O• 2 minutes rinsing in a beaker with demineralised water• 2 minutes in a 55 ml/L HNO3 (from 65 % HNO3)•

2 minutes rinsing in a beaker with demineralised water• 30 minutes anodising in 100 ml/L H2SO4, at 21 ºC with 14 V with agitation• 2 minutes rinsing in a beaker with demineralised water.• 30 seconds rinsing with demineralised tap water• 1 minutes blow drying at low level• 1 minutes blow drying at high level• Keep in dessicator

The anodising was done in a double-sided glass connected to a silicon oil cooler to maintainthe temperature. The temperature was continuously monitored to maintain the temperatureat a stable level. In general the cooling pump was set at 18,5 ºC as soon as the anodising process was started to maintain the sulphuric acid at 21ºC. See Figure 7.3 for a graphicaloverview of the anodising setup.




33

Figure 7.3: Graphical overview of the anodising setup.

7 2 3

Naming of commercial alloy specimens

Three different alloys has been used with two different polishing methods (or without polishing), been etched (or not) and been anodised (or not). The naming of the specimens isfollowing this template:

Indexnumber

Surfacetreatment

Alloy Etching?Notanodised?

xx n / B / W yyyy E %

Table 7.7: Template for naming the specimens.

Here xx is the sequential number given to the specimen. The n/B/W means that the specimenhas received “no polishing”, Buffing or W ater polishing. The alloy is written in the place of yyyy. If the specimen has been etched the letter E is written after the alloy. If the specimen hasnot been anodised, there is a % at the end. Not etched and not anodised specimens do nothave the given mark.

Example:

27B1050E is specimen number 27 that has been polished with the buffing machine, it is the1050 alloy and it has been etched and anodised.

The specimens are presented continuously in the results and discussion section due to thehigh amount of specimens.




34

7 3

Synthesis of aluminium based plasma coatings

7 3 1 1

The PVD machine

The PVD process was done at the Tribology Centre at TI in Århus on a CemeCon 800/8, which is shown in Figure 7.4.

Figure 7.4: The CemeCon 800/8 PVD machine at TI in Århus.

The specimens are cleaned before coating. This is done by 30 seconds in an ultra sonic bath at40 ºC with acetone, followed by cleaning in ethanol, and blow-drying with clean air.

The specimens are mounted in the chamber along with two different sized pins, a small sheet

of glass and silicium. These are used for following tests of the coating and stored for laterreferences.

For the Ti-experiments two targets are used, where the left is an aluminium 1050 target. Theright target is either a pure titanium target or a Ti target with Al1050-pins drilled into it tolower the surface area of Ti. The latter is used for getting an even lower amount of Ti in thecoating.




35

The standard process has been a coating with pure 1050-alloy. Meaning that two 1050-targets were used both with 1000 W.

Left target power Right target power Heat Starting pressure Argon flow Substratebias power Cathode voltage

1000 W 1000 W 0 W 6 mPa 200 ml/min - 40 V ≈ - 500 V

Table 7.8: Standard setup for the PVD-process.

When a Ti-containing target replaces the right target, it is set at a cathode power between100-300 W and then the left target power is raised to 2 kW to maintain the expected coatingthickness.

6401 targets have also been used with 1000 W as cathode power.




36

8 Characterization techniques

8 1 1

Optical microscopy and SEM

For the study of microstructure there were used a LOM Olympus GX41 (with Altra 20 SoftImagining System for connecting to the computer) and the JSM-5900 SEM from JEOL(Figure 8.1) was used both at DTU.

Figure 8.1: The JSM-5900 SEM at DTU.

SEM pictures are taken with an acceleration voltage of 10 kV and a spot size of 35-40. EDSanalysis are done at 15 kV and calibrated with a nickel specimen.




37

8 1 2

GDOES

For an elemental analysis across the thickness of the anodised layer (depth profiling), GlowDischarge Optical Emission Spectroscopy (GDOES) was used. The equipment used was theGD Profiler 2 from Horiba Jobin Yvon as seen in Figure 8.2.

Figure 8.2: The GD Profiler 2 at DTU.

The examinations were done on a hard coatings 0 -program, which is specified in Table 8.1.

Program name Pressure Power Pulse? Module Phase

Hard coatings 0 650 Pa 35 W No 7,6 V 3,8 VTable 8.1: Setup details on the GDOES.

The examinations were done throughout the thickness of the anodised layer but in this reportmainly the first 200 nm will be in focus.

The measurements have been quantified according to weight percentage in the Quantum XP2 program from HORIBA Jobin-Yvon. However the machine was not calibrated for Mgand Zn, and therefore the results does not include these elements.

The thickness of the GDOES measurements is based on a value calculated in [2], wherecomparisons with SEM pictures and other measurements concluded that the GDOES

sputtering removes approximately 37 nm/s when sputtering alumina.




38

8 1 3 XRD

For XRD measurements the D8Discover from Bruker AXS has been used (Figure 8.3).

Figure 8.3: The XRD setup. On the right is the X-ray source, in the middle is the specimens sticked tothe holder and to the left is the detector.

A continuous measurement has been done moving 0.03 degrees every nine seconds. Theturning of the specimen and the detector has been done with a two-to-one rate, meaning thatthe detector turns twice as fast as the specimens. This is called a locked couple scan. Thegenerator has been set to 40 kV and 40 mA.

The measurements have been analysed in the EVA-software from Bruker.




39

8 1 4 Photospectrometry

The optical measurements on the specimens have been measured at DTU Fotonik at Risøusing the equipment seen in Figure 8.4.

Figure 8.4: The photospectrometry setup. On the picture is shown the lamp (blue box), the integratingsphere (grey box) and the photospectrometer (black box).

A Mikropack HL-2000 white halogen light source (Mikropack GmbH, Germany) isconnected via an optical fibre to the top of a small aluminium cup. The cup is coated inside with a white hemisphere called an integrating sphere. The optical fibre lights from the top ofthe integrating sphere and down through a hole in the bottom of the cup (10 mm indiameter). In one side of the cup, there is an optical fibre connecting to a photospectrometer(Avantes, AvaSpec-2048), which is connected to a PC via USB. The PC has the AvaSoft 7.4software installed.

Prior to measurements the program is calibrated, to make a black reference the lamp shutteris closed and the cup is placed on a black piece of paper. The shutter is opened, and with aspecial white calibration specimen the white reference is recorded in the program as the 100

% reflection spectrum. By this calibration the following measurement is done with respect tothe calibration spectrum, thus the spectrum of the lamp is taken into account in the spectrumof the reflection curves.

As previously described, the reflection of a material can be described as a result of specular,diffuse and ambient reflections. Since the aluminium cup is shutting of the surroundings, thereflection is here only from diffuse and specular reflections in the material.

Before each measurement the specimen is visually examined to make sure that the area ofmeasurement is not contaminated.

Each specimen is measured two times, preferably on both sides unless the one side is not ofsatisfying surface quality. The two measurements are plotted in the same figure and the




40

difference is estimated to validate the measurement error. An example of a validation plot canbe seen on Figure 8.5.

Figure 8.5: Example for a validation graph.

Figure 8.5 shows two measurements for the specimen 27B1050E (alloy 1050 specimen, whichhas been polished by buffing, etched and anodised), which have been measured on each side of

the specimen. By examination of the graphs, the measurements are determined to vary half a percentage point. These measurements are therefore accepted as representative for thespecimen and the measurement error is estimated to be ± 0,5 %-point. The mean graph is also plotted on the figure and this is the graph used for representing the measurements onspecimen 27B1050E on the following made figures.

Notice in Figure 8.5 that there is a bump around 640 nm. This is caused by the equipment andnot by the specimen. This has been tested on several other specimens than the ones for this project.

In general for all the specimens, the characteristic 3 (colour) of the two graphs from themeasurements are the same, but with varying intensity. The difference in intensity is in general

under two percent, and the ones with a higher variation are supplemented with additionalmeasurements. An average of the measurements is calculated and used for plotting.

By replacing a small part of the integrating sphere with a black part, the specular reflectioncan be removed from the reflection graphs measuring only the diffuse reflection. Bysubtracting the diffuse reflection from the full reflection (diffuse and specular) thecontribution from the specular reflection can be calculated.

3 By characteristic is meant the course or shape of the graphs

400 500 600 700

65

66

67

68

69

70

71

Measurement on the other

side of the specimen R e f l e c t i o n [ % ]

Wavelength [nm]

Validation graph example

27B1050E

Used mean graph

Measurement on oneside of the specimen






42

Figure 8.7: Full and diffuse reflection for a mirror and the water blasted specimen 50W1050%.

Figure 8.7 shows that the setup is working as expected.

Adjacent averaging in the Avasoft program has been used for smoothing the graphs.

8 1 5 Measurement errors on photospectrometry

From the validation plots a measurement error has been specified so that the differencebetween the validation graphs is the measurement error in the photospectrometrymeasurements. The measurement errors are presented in Table 8.2:

Specimens Measurement error

Anodised specimens ± 0,5 % points

Non-anodised specimens ± 2,0 % pointsPVD coated specimens ± 0,5 % points

Table 8.2: Measurement errors for photospectrometry experiments.

Notice that the measurement errors stated in Table 8.2 is only for the intensity, not for thecharacteristic (colour), which has an insignificant measurement error.

400 500 600 700

0

10

20

30

40

50

Specular reflecting specimen (mirror)

R e f l e c t i o n [ % ]

Wavelenght [nm]

Diffuse specimen (50W1050%)

Full (thick) vs. diffuse (thin) reflection

Specular and diffuse reflecting specimen




43

9 Results and discussion

This section is divided into three major parts about the effect of alloy composition andtreatment on optical properties (9.1), water polished specimens (9.2) and 9.3 concerning theeffect of titanium concentration in a PVD coating (9.3). In the end, two minor sections dicussthe possibility of making a watermark underneath the anodised layer (9.4) and the use ofmokumé gane (9.5), which both are in progress.

9 1

The effect of alloy and treatment on optical properties

This section is divided into several subsections. The first subsections only focus on anodisedspecimens being polished by buffing and non-polished specimens.

9 1 1 Anisotropy in optical measurements on a surface

On a rolled sheet, the grains of the specimens will not be equiaxial but visibly elongated alongthe direction of the rolling. To test whether the grain orientation affects the opticalappearance, introductory measurements were made in the same spot with varying angleaccording to the rolling direction. In Figure 9.1 the measurements of angle dependency onspecimen 22n5754E (alloy 5754, not polished, etched and anodised) is presented. On the x-axis is the wavelength from 400-750 nm. On the y-axis is the reflection intensity, which is a percentage of the reflection from the calibration specimen.

Figure 9.1: Angle dependency for the reflection on specimen 22n5754E.

400 500 600 700

55

60

65

180 degrees

90 degrees

45 degrees


Wavelength[nm]

Angle dependency on sample 22n5754E

0 and 180 degrees follow the rolling direction

0 degrees




44

In Figure 9.1 it is observed that there might be only a slight angle dependency on the specimen with the used equipment. The reflection is highest for 0 and 180 degrees (following the grainselongated axis) and lowest for 90 degrees (perpendicular to the elongated axis). Thedifference in the measurements is up to two percent for the non-polished specimens.

When polishing, the elongated grains are not visible anymore and the angle dependency isremoved.

9 1 2 Presentation of the anodised specimens

The anodised specimens are presented in Table 9.1.

Anodised specimens

Alloy 1050 Alloy 5754 Alloy 6082

No polishing, with etching 36n1050E 22n5754E 23n6082E

Buffing, no etching 24B1050 25B5754 26B6082

Buffing, with etching 27B1050E 28B5754E 29B6082E

Table 9.1: Overview of the used anodised specimens.

The relationship between the surface treatment of different alloys and the optical effects isinvestigated. Therefore as seen in Table 9.1, the specimens number 36+22-29 are variationsof the alloys 1050, 5754 and 6082 plus no polishing and buffing with and without etching.




45

9 1 3 LOM examination of the anodised specimens

The surface conditions of the anodised specimens have been examined with a LOM (Figure9.2).

Figure 9.2: Surface conditions of the nine anodised specimens –the scale bar in the picture is 100 m.

Figure 9.2 shows clearly that the non-polished specimens have significant impressions fromthe rolling process, which are removed during the polishing process. The structure of the1050-alloy is smoother than the 5754- and 6082-alloys that contain holes and pores in thesurface both with and without both buffing and etching. The surface of 5754 and 6082 alloysseems to be similar.




46

9 1 4

Element enrichment in the surface

Tabrizian-Ghalehno [2] suggest that the different colours of different aluminium alloyscould be due to enriched concentration of alloying elements (nanometer scale) in the surfaceof the specimens. This enrichment was detected to peak around 30 nm from the surface. With the performed optical measurements, it is now possible to correlate these with theelement concentration by using GDOES. An example of the GDOES measurements is seen inFigure 9.3.

Figure 9.3: GDOES measurement for an alloy 6082 specimen, which has been polished and not etchedbefore anodising.

Figure 9.3 presents the amount of elements in wt% as a function of the depth under the

specimen surface. This figure is representative for the GDOES measurements for the 5754and the 6082 alloys (see appendix).

As presented in Figure 9.3 has the same enrichment as proved in [2], although it does not peakexactly at 30 nm.

As presented in Figure 9.4 the 1050 alloy has the same tendency of element concentration inthe surface as the other alloys. However, the non-etched specimen has as high concentrationas the other alloys, whereas this is much lower for the etched specimens.

0 50 100 150 200

0

5

10

15

Ti+Cu

HSi

Mn

N W t %

26B6082

Depth (nm)




47

Figure 9.4: GDOES measurements for the 1050 alloy specimens.

0 50 100 150 200

0

5

10

15

0 50 100 150 200

0

5

10

15

0 50 100 150 200

0

5

10

15

Ti Si

AlO

Mn

H

N

S

W t %

36n1050E

Fe+Cr

W t %

Ti+Cu Si

H

Mn

O

N

24B1050

CuTiMn H

S

N

Si

W t %

Depth (nm)

27B1050E

Alloy 1050 for 0-200 nm




48

9 1 5

Alloy / treatment for anodised specimens

In Figure 9.5 the photospectrometry measurements are shown in three plots4 where the top plot is only with alloy 1050, the middle is with alloy 5754 and bottom is with alloy 6082. In

each of the plots the black graph is the specimen that has not been polishing and etched. Thered graph is for buff ing and no etching and the blue is for buff ing followed by etching.

In Figure 9.5 the middle and the bottom plots are quite similar, with the red and blue graph with the same intensity, and the black graph about five percent points below the others. Inthe top plot, the red graph is the one with the lowest reflectivity and still with blue a littleabove the black graph. From this finding it can be deduced that alloys 5754 and 6082 visuallyreact in the same way after the performed treatments, while 1050 does not. This is discussedin the end of this section. Additionally it is found that the etching does not affect thereflection intensity of 5754 and 6082 alloys, since the red and blue graphs has the sameintensity in the middle and lower plot, however, the etching increase the reflection intensityof the 1050-alloy.

If looking at the characteristic (intensity vs. wave length reflected) of the graphs, the colour ofthe surface deduced from the graph is the same for the same alloy with no dependency to thetreatment. Meaning that, the colour of the specimens is alloy-dependent, while the intensityof the reflection is treatment-dependent.

In Figure 9.6 the same nine graphs have been plotted, but this time sorted so that graphs in thesame plot have received the same treatment. Thus, here the variation in a plot is the alloy.

In these plots it is observed that the top plot (no polishing, with etching) and the lowest plot(buffing, with etching) are similar. The reflection of 1050 has higher intensity than the othertwo, which has the same intensity. The middle graph is standing out since 1050 has the sameintensity as the other alloys in the interval around 600-650 nm.

When looking at the characteristic (colour) of the graphs 1050 is having relatively higherintensity in the shorter wavelengths and lower intensity in the longer wavelengths. Thisapplies for all three plots. Visually this means that 1050 will appear more bluish (or perceivedas a cleaner white/grey) than the other alloys. This is also the case when looking at thespecimens with the eye.

Additionally, 5754 has higher intensity in both short and long wavelengths compared to 6082.This could in principle give 5754 and 6082 slightly different tints, although this is not seen

with the eye.

As seen in Figure 6.2 (page 10) aluminium has a high reflectivity compared to other metals.From this it can be expected that the intensity of the reflection in aluminium alloys will bereduced with the amount of alloying elements. Therefore 1050 would be expected to have thehighest reflectivity of the three alloys, which is seen for the etched specimens. However, thiscorrelation is not present for the not-etched specimens, where they have the same reflectivity.

4 Notice that the word plot is consistently used for a drawn co-ordinate system with graphs.

Meaning that the word graph is used for a single graph on a plot.




49

When looking at the GDOES measurements for alloy 1050 (Figure 9.4), the concentration ofthe alloying elements is much higher for the non-etched specimen, and therefore this can bethe reason for the lower reflection.

Figure 9.5: Photo spectrometric plots for alloy 1050, 5754 and 6082 with three different pre treatmentsbefore anodising.

400 500 600 700

65

70

400 500 600 700

55

60

65

400 500 600 700

55

60

65

Buffing, no etching

No polishing, etching


Alloy 1050Buffing, etching


Alloy 5754 Buffing, no etching


Buffing, etching


Wavelength [nm]

Alloy 6082

Sorted by alloy

Buffing, no etching


Buffing, etching




50

Figure 9.6: Photo spectrometric plots for three different treatments before anodising.

400 500 600 700

55

60

65

70

400 500 600 700

60

65

400 500 600 700

60

65

70

6082

5754


No polishing, with etching

Sorted by treatment

1050


Buffing, no etching

6082

5754

1050


Wavelength [nm]

Buffing, with etching

6082

5754

1050




51

9.1.6 Full diffuse and specular reflection

As described, measurements can be made, with only the diffuse reflection. This has been doneto get an overview of how big a contribution is made from the diffuse and the specularreflection, and how this varies with alloy and pre treatment followed by anodising.

The measurements of the full (or total) and only the diffuse reflections are presented inFigure 9.7 and Figure 9.8. In the same plot, two graphs of the same colour represent the samealloy and pre-treatment. The thick graphs are the f ull reflection (diffuse + specular), the thingraphs are for the diffuse reflection. The distance between these two graphs is the specularreflection.

In Figure 9.7 it is observed that the black graphs on all the plots has the smallest distance toeach other, especially for 5754 and 6082 alloys. This means that the non-polished specimens

are relatively more diffuse reflecting than the polished specimens. Since the buffing is done toobtain a more shiny surface, this is as expected.

The graph for buffing, no etching (red) has the greatest distance for alloy 1050, meaning thatthis specimen has the highest specular reflection. However, the etching of the polishedspecimens does not have an influence on the reflection type of the 5754 and 6082 alloys. Thismight be because the etching agent in these experiments (Alficlean) is mild, compared to thecaustic etching where large amounts of material are removed from the surface.

In the middle and the lower plot the graphs are looking more or less the same, thus again 5754and 6082 act very much the same to the treatments. This corresponds with the LOMinvestigations presented in Figure 9.2.

When looking at Figure 9.8 the etched specimens (top and bottom plots) are comparable, whereas the non-etched 1050- alloy specimen (middle plot) is standing out once more withthe lowest diffuse ref lection.

When looking at each single pair of graphs (same colour, thick and thin in the same plot), it isclear that the characteristic of the reflection is changed, meaning that the colour of thereflection is different for diffuse and specular reflection. The distance between the graphs isincreasing with the wavelength, meaning that the characteristic (colour) of the graphs changes when comparing the full and the diffuse reflection.

To make this fact clear, Figure 9.9 shows the specular reflection of the specimens where thediffuse graph has been subtracted from the full reflection graph. Here the specular reflectionis more reddish compared to the diffuse reflections in Figure 9.7. Additionally the graphsshow that the characteristics of the graphs are the same for the polished specimens, butcompared to the non-polished specimens they are different. This means that the colour of thespecular reflection of these specimens vary with the surface treatment, and not with the alloyas for the full reflection.




52

The difference in colour between the diffuse and the specular reflection could be due to thesurface structure of the specimens. As earlier mentioned the buffing gives a roughness around1 µm, which is close in magnitude to the light, thus this can affect the light reflection. It couldalso be due to scattering of the light caused by the alumina. Scattering will affect the blue light

the most leaving more red lights in the specular reflection. Whether this is a general property for materials or it is a special property for these specimens(or for anodised aluminium) has not been investigated further.






54

Figure 9.8: Measurements of full and only diffuse reflection, sorted by treatment.

400 500 600 700

50

55

60

65

70

400 500 600 700

50

55

60

65

70

400 500 600 700

50

55

60

65

70

6082

5754


1050


Full (thick) vs. diffuse (thin) reflection

Sorted by treatment

5754

60821050


Buffing, no etching

5754

6082

1050


Wavelenght [nm]

Buffing, etching




55

Figure 9.9: Only the specular reflection, sorted by treatment

(calculated as full minus diffuse reflection).

400 500 600 700

0

5

400 500 600 7000

5

10

15

400 500 600 700

0

5

10

6082

5754



1050

Specular reflection

Sorted by treatment


6082

57541050

Buffing, no etching


Wavelenght [nm]

6082

5754

1050

Buffing, etching




56

9 1 7 Visual effect of anodising

Another set of specimens was prepared with the same treatment but without anodising(Table 9.2):

Not anodised specimens


No polishing, with etching 41n1050E% 42n5754E% 43n6082E%

Buffing, no etching 44B1050% 45B5754% 46B6082%

Buffing, with etching 47B1050E% 48B5754E% 49B6082E%

Table 9.2: Overview of the used not anodised specimens.

This was to investigate how the anodising is affecting the optical properties of aluminium ingeneral. And hopefully thereby to explore which part of the reflection is f rom the alumina and which is from the substrate.

The results of the anodised and non-anodised specimens are plotted together (Figure 9.11and Figure 9.12). Graphs with the same colour (on the same plot) are the same alloy and same pre-treatment. The thick graphs are for the anodised specimens and the thin graphs are forthe non-anodised specimens.

In figure Figure 9.11 the results are presented in the same way as Figure 9.5, sorted by alloy.

In the top plot (1050-alloy) the no polishing, etching -specimen (black) falls in intensity afteranodising. The polishing, etching -specimen (blue) rise in intensity after anodising. And the polishing, no etching -specimen (red) is both falling and rising varying with wavelength.

In the middle plot (5754-alloy) the two polished specimens rise in intensity after anodising.The non-polished both rising and falling varying with the wavelength.

In the bottom plot (6082-alloy) the polished, no etching (red) graph has the same intensity, theblue is rising and the black falls significantly.

The results do not seem to have any clear patterns as an effect of the treatment.

When looking at each single pair of graphs (same colour, thick and thin in the same plot), it is

clear that the characteristic (colour) of the reflection is changed, when anodising thespecimens. The thin graphs (except the blue in the bottom plot) for the specimens beforeanodising is compared to their thick counterpart, and they all have relatively less intensity inthe shorter wavelengths and relatively higher intensity in the longer wavelengths. This meansthat the specimens will look more red and less blue before the anodising compared to after.Since the only difference from the previous measurements is the added layer of alumina, thislayer is expected to absorb more red than blue light. However, the explanation for this is notclear, which needs more detailed experiments. Notice that in the previous example the overallintensity of the reflection has not been taken into account, it is only relative to themeasurement for the same specimen with and without anodising.




57

The findings described above is clearly seen with the eye as observed in Figure 9.10 where two6082-alloy specimens have been compared. It is seen that the unanodised specimen to the lef thas a much more warm and reddish tint, than the anodised specimen to the right.

Figure 9.10: Alloy 6082 with etching, before anodising (left) and after anodising (right).

As mentioned the specimen of alloy 6082 with polishing and etching (blue in bottom graph)does not follow this pattern, but the other way around. Additionally, the specimen of alloy6082 with polishing and no etching (red) does not have the same significant change as theother graphs. Consequently, this can be alloy dependent, and if that is the case, it is probablydue to higher amount of silicon or lower amount of magnesium since this is where the 6082-alloy is clearly different from the 5754-alloy.

From Figure 9.12 an extra point must be added. The alloy 1050 has the highest reflection forall the specimens before the anodising, no matter the pre treatment. In both the top and themiddle plot the 6082 alloy specimens has the highest reflection compared to 5754-alloy, which is not the case in the bottom one. Additionally the 6082-alloy in the bottom graph hasa much different characteristic compared to the other alloys with the same treatment.

The alloy 6082 specimen, that has been polished and etched seems to be abnormal and

another specimen should be prepared to validate these measurements. This is supported bythe fact than on no other plot is observed an effect of the etching on alloy 6082.




58

Figure 9.11: Reflection curves for anodised and not anodised specimens, sorted by alloy.

400 500 600 700

60

65

70

75

400 500 600 700

50

55

60

65

70

400 500 600 700

50

55

60

65

70

Polishing, no etching

Polishing, etchingno polishing, etching


Alloy 1050


Alloy 5754

no polishing, etching

Polishing, etching


Anodised (thick) vs. not anodised (thin)

Sorted by alloy


Wavelength[nm]

Alloy 6082

no polishing, etching

Polishing, etching





59

Figure 9.12: Reflection curves for anodised and not anodised specimens sorted by treatment.

400 500 600 700

50

55

60

65

70

75

400 500 600 700

50

55

60

65

70

400 500 600 700

50

55

60

65

70

5754

6082

R e f l e c t i o n [

% ] No polishing, etching

1050

R e f l e c t i o n [ % ] Polishing, no etching 57546082

1050

57546082


Wavelength[nm]

Polishing, etching1050

Anodised (thick) vs. not anodised (thin)

Sorted by treatment




60

9 1 8 Summary on alloy and treatment affect on optical properties

LOM investigation of the anodised specimens showed that non-polished specimens havesignificant impressions from the rolling process, and that the structure of the 1050-alloy isfiner than the 5754- and 6082-alloys that contains holes and pores in the surface.

GDOES analysis showed an enrichment of alloying elements within the first 50 nm of thesurface of the anodised specimens.

The 1050-alloy has the highest reflectivity compared to 5754 and 6082 due to its higher purity. Colour is alloy dependent, while intensity is treatment dependent. This might be dueto the fact that the surface treatment determines the roughness, therefore diffused reflectionand in turn the intensity. Etching with Alficlean increases reflectivity of 1050 compared tothe other alloys, probably due to the fact that the low levels of intermetallics are removed, andthereby giving a higher reflectivity. However, this is not easy for other alloys due to the higher

density of distributed intermetallic particles. More detailed work on correlation withmicrostructure and roughness before and after etching is needed to understand the exactreason. It is interesting to see that in all cases, the more complex alloys 5754 and 6082 actsimilar to the treatments, including that they are not affected by etching with Alficlean.

Buffing increases the specular reflection for all the alloys.

The specular reflection is reddish for all the specimens the colour varies with the surfacetreatment, and not with the alloy as for the full reflection. It is suggested that alumina scattersthe blue light causing the specular reflection to be reddish. It has not been checked whether areddish specular reflection is a common property for specular reflecting materials.

Specimens will look more reddish and less bluish before anodising compared to after. Since

the alumina layer is the only difference, it must be the source of the change, e.g. because thealumina absorbs more red than blue light.




61

In Table 9.3 is presented an overview:

Alloy 1050 5754 6082

Standard reflection(anodised), intensity

Highest Same as 6082 Same as 5754

Standard reflection(anodised), colour

Bluish white Reddish white Reddish white

Etching effect Higher intensity None None

Buffing effect, totalreflection

Higher intensity when etched

Higher intensity Higher intensity

Buffing effect, specularreflection Higher intensity Higher intensity Higher intensity

Specular colour vs. diffusecolour

More reddish5 More reddish5 More reddish5

Anodising effect, intensity Unknown Unknown Unknown

Anodising effect, colour More blue More blue Slightly more blue

Intensity before anodising Highest Medium Lowest

Intensity change whenelement enrichment inspecimen surface

LoweredEnrichments

present for allmeasurements

Enrichments present for allmeasurements

Colour effect of elementenrichment in specimensurface

NoneEnrichments

present for allmeasurements

Enrichments present for allmeasurements

Table 9.3: Summary of findings on optical variations.

It was planned to compare the results with SEM pictures, and specimens were put into the

SEM, unfortunately, these specimens charged too much for taking pictures. The specimensmust be carbon coating and taken into the SEM again. Unfortunately, the charging wasdiscovered too late and therefore this has not been done.

5 The effect is amplified when buffing




62

9 2

Water polished specimens

To investigate the effect when making an extra rough surface, water polishing was used. InTable 9.4 the anodised specimens are presented and in Table 9.5 the not-anodised specimensare presented.

Anodised specimens


Water polishing, no etching 30W1050 (31W5754) (32W6082)

Water polishing, with etching 33W1050E 34W5754E 35W6082E

Table 9.4: Water polished specimens that have been anodised.

The parenthesis around specimen 31+32 mean, that these are not made.

Not anodised specimens


Water polishing, no etching 50W1050% 51W5754% 52W6082%

Water polishing, with etching (53W1050E%) (54W5754E%) (55W6082E%)

Table 9.5: Water polished specimens that have not been anodised.The parenthesis around specimens 53-55 means that these are not made.






64

Figure 9.14: SEM picture of water polished, etched but not anodised specimen.

Figure 9.15: SEM pictures of water polished, etched and anodised specimen.

The pictures show a very chaotic structure with small structures of material everywhere.There do not seem to be a big difference in the microstructure of the specimens, although it isimportant to remember that the anodised specimen is having the previously described porousanodised structure. As opposed to the non-polished and buffed specimens where the poresare unidirectional perpendicular to the surface, here the pores will possibly go in all directions,going into the surface, out from the surface and parallel to the surface. This might trap thelight inside the specimen surface structure instead of being reflected back. This phenomenonis also known from zinc powder, which appears black even though it is actually light grey. Onthe other hand, the pores are about 15 nm wide and they do therefore not affect the light.






66

As expected based on experience from B&O the specimen went from black to light grey afteranodising, which is seen in Figure 9.16.

Figure 9.16: Etching experiment. The left side has been carbon coated for the SEM.

As observed in Figure 9.16 only five seconds of etching have a big impact on the appearanceafter anodising. Hereafter the specimen is slowly turning lighter. This is also supported by photospectrometry measurements, which is presented in Figure 9.17.

Figure 9.17: The etching experiment.

Even though it is hard to see in the picture (Figure 9.16), it is shown in Figure 9.17 that theintensity increases with the amount of etching.

400 500 600 700

10

20

30

40

50


Wavelength[nm]

The etching experiment

0 sec

5 sec

10 sec

20 sec

30 sec




67

SEM pictures were taken of the specimen, where Figure 9.18 is of etching0, Figure 9.19 is ofetching5 and Figure 9.20 is of etching30 (etchng10 and etching20 can be seen in appendix). Itis seen for etching0, etching5 and etching30 that the structure starts chaotic but is etchedmore and more homogenous by the NaOH giving round craters in the surface as seen in

etching30.

Figure 9.18: SEM of etching0.




68

Figure 9.19: SEM of Etching5.

Figure 9.20: SEM of etching30.




69

In the visual appearance in Figure 9.16 the biggest change is from etching0 to etching5, and when examining the SEM pictures of these, it is discovered that the biggest folds andcorrugations of the structure is etched away in the first f ive seconds.

GDOES were used to investigate whether an extremely high enrichment of alloying elementsis present in the surface after water polishing (Figure 9.21). The etching0 is seen in the top leftcorner, followed by 5 and 10 seconds of etching going right. The etching 20 and etching 30 isin the lower row.

Figure 9.21: GDOES measurements on etching experiment.

From Figure 9.21 it is found that the enrichment in the surfaces is similar for all the

specimens. However, when correlating the measurements, it is experienced that the amountof enrichment of elements peak somewhere between etching5 and etching20. Notice that Cuis higher at etching5 than etching10, meaning that Cu peaks before the other elements. Thismight be due to the fact that Cu has a quite high sputter coefficient as seen in Figure 6.10(page 22).

As already mentioned, the rolling process enriches the alloying elements in the surface. Whenetching, the aluminium is etched more than the other elements, and therefore a ditch willappear around these non-aluminium-particles, which consequently will be dug free over time. When sputtering it will be “easier” to sputter particles if they have already been partly dugfree. This progress occurs until a point, where the particles are etched completely free and the particles leaves the surface during the etching process prior to the sputtering. This is the

0 50 100 150 200

0

5

0 50 100 150 200

0

5

0 50 100 150 200

0

5

0 50 100 150 200

0

5

0 50 100 150 200

0

5

S

W t %

Depth (nm)

Ti

Cu Mn

Si

H

N

Etching0

W t %

Depth (nm)

Ti

Cu

Mn

Si

H

N

Etching5

Fe+Cr

Ti

W t %

Depth (nm)

Cu

Mn

Si

H

N

Etching10

Ti

Cu

W t %

Depth (nm)

Mn

Si

H

N

Etching20

Mn

Si

H

N

W t %

Depth (nm)

S

Etching30

Etching experiment -zoom




70

evolution that can be followed on the GDOES graphs. This finding is probably mostinteresting from an academic point of view.

9.2.3 Summary on water polishing part

When water polished specimens are anodised (no matter of alloy and etching in Alficlean),they turn black. This effect would be nice to control as long as UV-light and wear does notaffect the colour in an unfortunate way.

The black anodising can be turned grey by just five seconds of etching from a rough etchingagent as 6 % NaOH, and the specimen is turning brighter with the time of etching, at least inthe first 30 seconds.

When etching, the structure goes from extremely chaotic and rough to more homogeneousand slowly the roughness is also evened out.

9.3

The eff ect of Ti concentration on optical properties in

aluminium coating

Notice that all measurements in the following sections have been performed on non-anodisedspecimens.

In Table 9.7 an overview of the PVD specimens is presented:

PVD specimens

Specimenname

TI batchname

CompositionLayerthickness

Right targetAlso steelsubstrate?

Pure1050 b2165 Pure 1050 22 – 28 µm 1000 W (as left)

Pure6401 b2164 Pure 6401 22 – 28 µm 1000 W (as left)

AlTi3 b2194 3,0 ± 0,1 wt% Ti 24 µm Mixed, 100 W Yes

AlTi4 b2178 4,3 ± 0,1 wt% Ti 7 µm Pure Ti, 100 W

AlTi5 b2188 5,4 ± 0,1 wt% Ti 16 µm Pure Ti, 100 W Yes



AlTi9 b2179 9,3 ± 0,1 wt% Ti 6 µm Pure Ti, 200 W

Table 9.7: Overview of PVD coated specimens. Notice that AlTi8 is rounded down in the specimenname even though it mathematically should be rounded up. SDOM (Standard Deviation Of Mean) is

given as error on compositions based on six measurements, three in two different sites.

All thickness measurements have been done at TI by cross section pictures on a SEM or by acoating thickness test. The composition is measured with EDS in the SEM at DTU. Notice

that the rest of the composition is Al1050, not chemically pure Al.




71

Some coatings are only done on cast aluminium and some are both done on cast aluminium-and steel-substrates as marked in Table 9.7.

9 3 1 Examination of the PVD coating

The PVD coatings have been examined with a LOM, with a representative example in Figure9.22.

Figure 9.22: Representative LOM cross section picture of the AlTi5 specimen.

In Figure 9.22 the top black part is the resin and the mixed lower part is the cast aluminium, with the coating as the light clean part between these. As mentioned this picture isrepresentative for the specimens, however, some specimens turned out to have quite large particles underneath the coating and some are even cracking the coating as seen in Figure 9.23.

Figure 9.23: Example of particles underneath and penetrating the coating. Specimen: AlTi7.




72

In general the coatings covers the substrate sufficiently, but some places as seen in Figure 9.23the coatings does not have as good adhesion to the substrate as e.g. in Figure 9.22. Somerepresentative LOM pictures can be seen in the appendix.

A SEM picture of the AlTi4 is presented in Figure 9.24.

Figure 9.24: SEM picture of cross-section of AlTi4. Taken at Århus University.

In Figure 9.24 the columnar structure described in the Thornton model (page 23) is clearlyseen. From the image the column size is measured to be approximately 0.2 to 1.0 µm wide. It isnot possible to see whether it is a zone 1 or a zone 2 structure. There is no visible spacesbetween the columns, but it is not certain whether this is due to a zone 2 structure, or whether

it is because the spaces are not discovered e.g. because they are located elsewhere or they aresmaller than what is detectable in the picture.




73

The specimens have been analysed by XRD. A representative example (AlTi3) can be seen inFigure 9.25. The other analyzed specimens gave similar results.

Figure 9.25: XRD measurements for AlTi3. Green lines are for Al and blue lines are for Si.

From Figure 9.25 it is seen that most of the peaks are from Al and Si. The Al can both be from

the coating and from the substrate of cast aluminium. On contrary the Si must be signals fromthe substrate.

Around 2 = 35º there is a peak calculated to be a beta peak of the strong peak at around 2 =38º. There are two unidentified peaks in this graph, which tend to be present in the othergraphs as well. There is an extra unidentified peak for AlTi4, and only for this. All XRDmeasurements can be seen in appendix including two plots for steel substrates. The steelsubstrates do not seem to affect the tendencies of the measurements.

Both unidentified peaks are present in the 1050-alloy measurements, meaning that thetitanium does not cause these. The peak at 2 = 37º is present for the steel substrate as well, which support the fact that that this is originating from the coating.

The XRD measurements show that α-Al and Si phase is found, and one or more unidentified phases. From these measurements no definitive conclusions can be made.




74

9 3 2 Visual impact of titanium concentration on full reflection

To find out how the Ti-concentration is affecting the optical properties of the coating, thetitanium specimens were measured with the photospectrometer at Risø. The results can beseen in Figure 9.26 with the clean 1050-alloy as reference.

Figure 9.26: Photospectrometry of AlTi coatings if different concentrations.

In Figure 9.26 the clean 1050 have the highest intensity followed by AlTi5, AlTi4 andAlTi8+AlTi9 together. Although the AlTi5 is higher than the AlTi4, it is a trend that theintensity of the reflection falls with the amount of Ti in the coating.

400 500 600 70075

80

85

AlTi8

AlTi9

AlTi4

AlTi5

R e f l e c t i o n

[ % ]

Wavelength[nm]

pure 1050

AlTi coatings on casted Al






76

9 3 3

Full vs diffuse reflection

The full reflection (thick graphs) and the diffuse reflection (thin graphs) have been comparedfor samples AlTi5 and AlTi8 in Figure 9.28.

Figure 9.28: Photospectrometry for full (thick) and diffuse (thin) reflection on the AlTi5 and AlTi8specimen.

As previously seen AlTi5’s full reflection is of higher intensity than the one for AlTi8. Thesame turns out to be the case for the diffuse reflection. Additionally, the diffuse reflection is“on level” at the shorter wavelengths but has got lower intensity in the longer wavelengths,meaning the specular reflection is reddish compared to the diffuse reflection in Figure 9.28.This is the same tendency as the previously comparisons of the full and diffuse reflections in

section 9.1.6 for the other specimens. When looking at the specular reflection for the same two specimens in Figure 9.29 it isobserved that the specular reflection is bigger for the AlTi8 than the AlTi5.

400 500 600 70070

75

80

85

90

AlTi8

AlTi5


Wavelenght [nm]

Full (thick) vs diffuse (thin) reflection




77

Figure 9.29: Specular reflection for AlTi5 and AlTi8.

In Figure 9.29 it is also found that AlTi8 has the most reddish specular reflection.

This is verified when looking with the eye (Figure 9.30):

Figure 9.30: Picture of AlTi5 (left) and AlTi8 (right).

400 500 600 700

0

5

AlTi8


Wavelenght [nm]

AlTi5

Specular reflection for AlTi5 and AlTi8




78

At Figure 9.30 the lamp reflection is both sharper and slightly more reddish on the AlTi8 tothe right. Although this is much more clear with the eye, than on this picture.

These are the only AlTi-specimens measured for only the diffuse reflection. Figure 9.29indicates that the specular reflection is increasing with the amount of titanium butmeasurements of more specimens needs to be made.

Additionally this shows that while the specular reflection is increased the diffuse reflection isdecreased with the concentration of Ti. This might be due to the effect on the surfaceroughness of the coating, which should be investigated further.

9 3 4 Change of substrate material

To check whether the substrate material has an effect on the reflection curve for the coating,a few measurements were made with coatings on cast aluminium and steel (Figure 9.31).

Figure 9.31: Photospectrometry on steel and cast Al specimens with the same AlTi5-coating.

Figure 9.31 shows two measurements on a cast aluminium substrate and two measurementson a steel substrate. As expected, Figure 9.31 indicates that there is no difference in thereflections.

Notice that the coating of AlTi5 is 16 µm, so it should be expected that only the mechanical properties of the substrate and not the optical properties should have an effect, therefore theresults of Figure 9.31 is expected.

9 3 5

Anodising the PVD coatings

B&O has anodised a layer of about 7µm on the PVD coatings. This has been done on severalspecimens both with and without titanium as alloying element. Only the clean 1050-alloy has

400 500 600 700

80

85Steel

Steel

Casted Al


Wavelenght [nm]

Casted Al

Test on different substrates

AlTi5






80

The same is observed on the AlTi5 specimen in Figure 9.33. A closer examination with the eyeshows small pores in the coatings of all the three specimens.

It is not expected that the anodising should be better than achieved on the AlTi5 specimen,but the anodising of the 6401 coating is not expected and implies that the coating is notsatisfactory prior to the anodising.

Due to this problem, no more AlTi specimens have been anodised, instead time has beenspend on understanding the reason for the bad anodising of the 6401 coating.

9 3 6 Reason for bad anodising

The bad anodising can be the result of two reasons:

• A porous coating is created during the PVD, making it possible for the anodisingelectrolyte to reach the substrate.

• Elements in the coating are located or making inter metallic particles in an

unfavourable way.These two possibilities will be discussed in the following:

9 3 6 1 Electrolyte reaches the substrate

If there is a passage through the coating, the sulphuric acid will get in contact with thesubstrate material. When the substrate material is cast aluminium, the silicon might darkenthe surface as described for cast aluminium specimens. If the substrate is e.g. steel, thesulphuric acid will corrode the steel instead and probably not leave a good result.

The open structure of the coating can either be due to larger pores, cracks or particles as seenin LOM photo in Figure 9.23 (page 71) or because of minor pores between the grains in thesurface structure. B&O examined cross sections with LOM of the coatings after the anodisingand found no large pores or cracks. A representative picture can be seen in Figure 9.35.

Figure 9.35: LOM of specimen AlTi5. This picture is taken after the anodising.The anodising layer is almost impossible to see, but is marked with the upper red line.




81

Figure 9.35 presents the anodised layer (very dark, 3 µm thick) and the rest of the coating, which is measured to be 10 µm thick.

Since no big flaws in the surface was found, a Thornton zone 1 structure (page 23) should beexpected, if contact with the substrate is causing the problem. On the other hand, if it then will be possible to obtain a zone 2-structure all over the coating, the coating will be closed andthe substrate material will not be in contact with the sulphuric acid.

9 3 6 2 Particles in the coating

If the structure is closed then the problem is in the coating. This will be inter-metallic particles gathered in the columnar structure, mainly in the grain boundaries. This willinterfere with the anodising process and might make the anodising bad. Particles in thecoating should be present in the XRD measurements, which have neither been validated nordiscarded.

The possibility of a surface structure that is very chaotic giving a black anodising as seen withthe water polished experiments was investigated. SEM investigations of alloy 1050 and alloy6401 and some of the AlTi specimens (top view) have been made, and they look identical.Figure 9.36 presents the structure of the surface of a 6401-alloy specimen before anodising.The same structure is seen for all the specimens (including 1050) making it less plausible thatthe surface structure is causing a black anodising, since the 1050-alloy can be anodised.

Figure 9.36: SEM picture of 6401-alloy specimen before anodising.

The problem with the bad anodising has not yet been solved, although new tests have beeninitiated.




82

9 3 6 3

Tests on PVD coatings

To find out what is actually causing the problem, a new specimen is being made with 6401-alloy for both the substrate and the coating. In this way, if the anodising turns out bad thenthe problem is caused by particles or bad structures in the coating. If the anodising turns good,then the problem is due to contact with the substrate and thereby a non-closed coating.

After this result an experimental series will be started testing by varying heat power, specimenbias voltage and target power. Until now 6401 substrate has been coated with 6401-alloy ascoating with the following PVD settings:

Specimen name Heat power Specimen bias voltage Target power

B2187(standard)

0 W - 40 V 1000 W

B2203 500 W - 40 V 1000 W

B2204 1500 W - 40 V 1000 W

B2205 0 W - 80 V 1000 W

B2206 0 W - 40 V 2000 W

Table 9.8: Overview of experimental series for avoiding bad anodising.

As seen in Table 9.8 the heat power, bias voltage and target power has been changed from the“standard” program (b2187). Since this series is made on 6401 as substrate, this can mainly beused if the testing on the standard program turns bad, meaning that the problem is caused by particles in the coating.




83

9 3 7 Summary on PVD part

The total reflection intensity depends negatively on the Ti concentration. The colour does

only depend slightly on the Ti concentration, but the dependency is not obvious.

Measurements show that a higher Ti-concentration increase the specular reflection, meaningthat the roughness is less with higher Ti-concentration.

Substrate does not affect the optical properties due to the thickness of the coating.

It would have been more interesting to make measurements after anodising, but due to thebad anodising of the 6401-coating, the main priority was to get better in controlling the PVDcoating. The bad anodising can be due to:

• Electrolyte reaches the substrate• Particles in the coating

LOM pictures show in general good adhesion but somewhere it is wrecked by particles placedunder the coating.

SEM pictures reveal the expected columnar structure, although it is not possible to register pores for the electrolyte to reach the substrate.

XRD measurements prove that an aluminium phase and silicon phase are present, includingone or more unidentified phases.

It has not been concluded, which of these reasons is causing the bad anodising, but further

tests has been initiated.




84

9 4

Watermark under the anodised layer

Since the PVD coating growth morphology is affected mainly by the surface morphology ofthe substrate, a memory effect is observed where the tiny details of the surface morphology ofthe substrate will be transferred to the coating. This means that even a tiny scratch on thesurface will be preserved through the coating and therefore the scratch can be observed on thecoating surface unlike in the case of electroplating where the scratch will be covered withmetal deposition. This was also reported in [1] where a specimen of cast aluminium was polished and coated with 30 µm of 1050 aluminium. Now the polishing marks could not beseen but after the anodising process the polishing grooves were coming clear and visible. Asillustrated in Figure 9.37 the PVD layer follows the topography of the underlying specimen,this is called the memory effect.

Figure 9.37: The memory effect. PVD layers on top of a product surface [1].

Due to the memory effect, if the specimen has a scratch, the PVD coating will follow thisscratch. As indicated in Figure 9.37 the memory effect wears off when the coating getsthicker.

This effect can e.g. be used for writing the product company name into a product, but it canalso be used as theft protection. Today computers, phones and other equipment at bigcompanies e.g. DTU has been branded the company name by melting the plastic of the product as seen in Figure 9.38.

Figure 9.38: Telephone at DTU with branded name as theft protection.




85

Instead of a theft protection in a way that actually harms the aesthetics of the product thistype of theft protection has the potential of meeting the aesthetics of the rest of the product.At the same time the watermark is placed underneath the anodised layer and thereby it is not possible to remove the watermark without wrecking the product.

An initial experiment has been started, where a sheet has been laser engraved with varyinglaser settings. The deepest of the laser engraving has been measured to be about 22 µm. Thespecimen is going to be coated with 30 µm of 1050-alloy and anodised with about 7 µmafterwards.

9 5

Mokumé gane

An ancient forging method has recently been reinvented. The process is called mokumé ganeand was used for swords, jewellery and ornaments in the ancient Japan. The technique relieson experience, but Ian Ferguson has written a partly theoretically founded book on thesubject [27].

The technique is using two different alloys alternating in thin layers on top of each other. Thealloys are selected in a way so that they react differently to a selected pigment. It is importantthat the adhesion of the layers does not imply too much diffusion between the two alloys. Thematerial is rolled out to a plate or a sheet, which is made uneven, whereafter the sheet is polished into an even surface again (Figure 9.39).

Figure 9.39: Mokumé gane: Alternating layers of two different alloys [27].The polishing line is seen as a thin black line cutting through the bumps.

The required amount of polishing will vary in different parts of the specimen. Thereby it will vary which alloy is at the surface in different areas of the specimen. After anodising, colouringand sealing, the specimen will vary in appearance in different areas as seen in Figure 9.40.




86

Figure 9.40: Aluminium bowl produced with the mokumé gane technique [27].

As seen in Figure 9.40 it is possible to use the mokumé gane technique on aluminium. Themokumé gane gives the product a more alive and sort of pottery look.

An experiment has been started, where the mokumé gane technique is going to be used. Aremote control from B&O has been drilled for recesses with about 200 µm. In these recessesTI has started to coat different layers of aluminium alloys for about 20µm at the time.Afterwards B&O will anodise and colour it, hopefully, making it into the first mokumé ganeremote control. Until now there has not been done any considerations on which alloys to use,but it has to be alloys that react differently to dyes.

Even though two remote controls have received the same treatment, this technique is sosensitive, that the products appearance will vary as known from pottery. Now the customercan choose which remote control he likes the most.




87

1 Conclusion

1.

The experimental technique for determining the optical properties of the surfaces hasbeen tested on various surfaces and standardized for the investigation. The techniqueseems to be very much useful for establishing the optical properties of the surfaces withand without anodising.

2. The optical properties of the surface depend on the surface finishing and the alloycomposition. The intensity of reflection is most dependent on the surface finishing due tothe big change in diffused reflection as a function of roughness (at least f or rolled sheets), while the colour of the reflection depends most on alloy composition.

3. Anodising of the sample also shows an effect on the optical properties turning the

specular reflection reddish and the diffuse reflection bluish, which might be due toscattering of the light by the aluminium oxide layer. It has not been possible to concludehow the total reflection intensity is affected by anodising.

4. GDOES depth profiling of the anodised layer show enrichment of alloying elements inthe surface, which might be affecting the intensity of the ref lection of pure alloys as 1050.

5. For the PVD coated specimens, increased quantity of Ti in the coating decreases theoverall reflection intensity, while the specular reflection is increased. Additionally it isshown that increased aluminium purity increases the reflection intensity.

6.

X-ray diffraction of the PVD coated samples showed that the Ti in the coating present aselemental Ti without any second phase formation, since unidentified phases are alsofound for the clean 1050 coating. There might still be phases, which are too small to bedetected by the XRD.

7. Water-polished specimens have a very rough surface structure making it totally matt.Surprisingly, when anodising a water-polished specimen turns black, which can be avoidedby etching the specimen before anodising. An explanation of the phenomenon has notbeen within the scope of this project.




88

Future Work

1. The described bad anodising of the PVD coatings is an essential problem for the furthermovement of the IdeAl project. As described a series of specimens have been made to findthe reason for the bad anodising and to investigate how to avoid it.

2. More detailed investigation on the relation between optical appearance, surface finishing,chemical composition, and microstructure of the alloy. This will be carried out by preparing binary and ternary alloy systems with different alloying elements in PVDcoatings.

3. Effect of heat treatment and phase formation in the coating on change in optical

properties of the anodised layer in presence of various alloying elements in the coating.

4. Variation of refractive index of the anodised layer for the similar alloys with and withoutsurface treatment to understand the role of undissolved particles in the anodised layer onoptical appearance.

5. Standardised experiments using coloured anodised layers containing particles withknown size, to understand the scattering effect as a f unction of particle size.

6. More detailed electron microscopic investigation of the microstructure of the alloys /coatings and surface roughness by AFM to be correlated with optical appearance.

7. Investigations of laser cladded coatings and their influence on optical properties in detail.




89

2 References

[1] H. T. Holt and R. H. Møller, Anodisering af aluminium med optimeredeegenskaber . Master thesis, DTU MEK: , 2008.

[2] N. Trabrizian-Ghalehno, Advanced Anodizing Technology. Ph.D., DTU-MEK: ,

2009.

[3] E. Hecht, Optics, 4th ed., Addison Wesley, 2001.

[4] R. J. D. Tilley, Colour and the Optical Properties of Materials: An Exploration of

the Relationship Between Light, the Optical Properties of Materials and Colour .

Wiley-Blackwell, 1999.

[5] Course: 02569, Light and Materials, “ N02-b-Phong-2008.”

[6] “Wikipedia.” [Online]. Available: http://www.wikipedia.org/. [Accessed: 10-Aug-2010].

[7] C. Vargel, Corrosion of Aluminium, 1st ed. Elsevier Science, 2004.

[8] S. Wernick, R. Pinner, and P. G. Sheasby, The Surface Treatment and Finishing of

Aluminium and Its Alloy, Volume 1, 6th ed. ASM International, 2001.

[9] P. Møller and L. P. Nielsen, Advanced Surface Technology, Part 1, 1st ed. Møller

& Nielsen, 2010.

[10] “Aluminum Metallurgy.” [Online]. Available:

http://www.secowarwick.com/pressrel/articles/aluminummetallurgy.htm.

[Accessed: 08-Aug-2010].

[11] H. Dam, L. Gerward, O. Leistiko, T. Lindemark, A. Nielsen, and O. T. Sørensen,

Materialebogen, 1st ed. Nyt Teknisk Forlag, 2008.

[12] J. Stokes, Theory and Application of the High Velocity Oxy-Fuel (HVOF) Thermal

Spray Process. Dublin City University, 2008.

[13] SkanAluminium, Aluminium Overfladebehandling. Oslo: Skanaluminium, 1972.

[14] U. J. Andersen, Anodisering af aluminium. Danish Technological Institute,

Kemiteknik, 1980.

[15] G. Br äuer, B. Szyszka, M. Vergöhl, and R. Bandorf, “Magnetron sputtering -Milestones of 30 years,” Vacuum, vol. 84, no. 12, pp. 1354-1359, Jun. 2010.

[16] P. J. Kelly and R. D. Arnell, “Magnetron sputtering: a review of recent

developments and applications,” Vacuum, vol. 56, no. 3, pp. 159-172, Mar. 2000.

[17] J. A. Thornton, “Influence of substrate temperature and deposition rate on structure

of thick sputtered Cu coatings,” Journal of Vacuum Science and Technology, vol.

12, no. 4, pp. 830-835, Jul. 1975.

[18] J. , “Invitation to the SEM World,” JEOL, 2006.




90

[19] T. Nelis, M. Aeberhard, M. Hohl, L. Rohr, and J. Michler, “Characterisation of a

pulsed rf-glow discharge in view of its use in OES,” Journal of Analytical Atomic

Spectrometry, vol. 21, no. 2, p. 112, 2006.

[20] P. Møller, Overfladeteknologi, 1st ed. Kbh.: Nyt Teknisk Forlag, 1998.

[21] “Encyclopedia of Laser Physics and Technology - integrating spheres, Ulbricht

sphere.” [Online]. Available: http://www.rp-

photonics.com/integrating_spheres.html. [Accessed: 10-Aug-2010].

[22] J. R. Davis, Aluminum and Aluminum Alloys (Asm Specialty Handbook),

Illustrated edition. ASM International, 1993.

[23] “6401.” [Online]. Available: http://www.euralliage.com/6401.htm. [Accessed: 31-

Aug-2010].

[24] “ALMINOX - WKW.automotive - wkw.de.” [Online]. Available:

http://www.wkw.de/en/competence/material/alminox.html. [Accessed: 01-Sep-2010].

[25] “Indian & U K Standard ( Chemical Composition of Aluminium Alloys.” [Online].

Available: http://www.docstoc.com/docs/39173549/Indian-and-U-K--Standard-(-

Chemical-Composition-of-Aluminium-Alloys. [Accessed: 20-Aug-2010].

[26] “ASM Material Data Sheet.” [Online]. Available:

http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MTU010.

[Accessed: 03-Sep-2010].

[27] I. Ferguson, Mokume Gane. Krause Publications, 2004.





Appendix contents

1.

GDOES measurement for non-polished and buffed specimens Pages: B-D

2. GDOES measurements for water polished specimens Pages: E-H

3. SEM pictures of etching10 and etching20 Pages: I-J

4. LOM pictures of the AlTi specimens Pages: K-Q

5. XRD measurements of PVD coatings Pages: R - W



1.

GDOES measurements for non-polished and buffed specimens



0 50 100 150 2000

5

10

15

0 50 100 150 200

0

5

10

15

0 50 100 150 200

0

5

10

15

S Al

O

Ti Si

Mn

H

N W t %

22n5754E

S

Fe+Cr

W t %

Cu+TiSi

Mn

H

N

25B5754

Ti+CuSi

H

Mn

N W t %

Depth (nm)

28B5754E




0 50 100 150 200

0

5

10

15

0 50 100 150 2000

5

10

15

0 50 100 150 200

0

5

10

15

Ti+Cu+Cr

HSi

Mn

N23n6082E

Ti+Cu

HSi

Mn

N W t %

26B6082


CuTi

Mn

Si

H

N W t %

Depth (nm)

29B6082E



2.

GDOES measurements for waterpolished specimens





0 5000 10000 15000 20000

0

10

20

30

0 50 100 150 200

0

5

10

15

20

MnCN

S

W t %

34W5754E -full analysis

O

Cu+Cr Ti Si

N

Mn

S

w t %

Depth [nm]

34W5754E -zoom

C



0 5000 10000 15000 20000

0

10

20

30

0 50 100 150 200

0

5

10

15

MnSiN

S

W t %

35W6082E -full analysis

O

Cr TiSi

N

Mn

Al

w t %

Depth [nm]

C35W6082E -zoom



3.

SEM pictures of etching10 andetching20





4.

LOM pictures of the AlTi-specimens















5.

XRD measurements of PVD-coatings













Optical Appearance of Aluminium, Martin Aggerbeck PhD Th. (Tech Univ. Denmark, 2010)

Documents