• No results found

Chemical Interactions between tool and Aluminium alloys in metal cutting

N/A
N/A
Protected

Academic year: 2022

Share "Chemical Interactions between tool and Aluminium alloys in metal cutting"

Copied!
41
0
0

Loading.... (view fulltext now)

Full text

(1)

INOM

EXAMENSARBETE MATERIALDESIGN, AVANCERAD NIVÅ, 30 HP

STOCKHOLM SVERIGE 2021,

Chemical Interactions between

tool and Aluminium alloys in metal cutting

LOUISE ERKERS

KTH

SKOLAN FÖR INDUSTRIELL TEKNIK OCH MANAGEMENT

(2)

Abstract

Aluminium applications in automotive will increase by 25 % over the next decade, mainly driven by the electrification and the reduction of fuel consumption. This diploma work aims to increase the understanding of the chemical interactions between aluminium alloys and typical tool systems in metal cutting. First the temperature at the tool-chip interface was estimated by FEM simulations, using the software AdvantEdge. Thereafter the chemical interaction of the tool-chip interface was calculated at the simulated temperature using the Thermo-Calc software. The thermodynamic data and descriptions of the multicomponent systems used where found in the literature, assessed by other authors, but critically reviewed for the use in this work. The results of the FEM simulations indicate that the temperature for machining aluminium with PCD and cemented carbide is between 60-80 % of the melting temperature of Al7wt %Si alloy. The calculations of the chemical interaction in turn results in that several hard precipitates can stick to or transform on the surface of the workpiece or tool-chip interface, for example SiC, Al4C3 and evidently diamond from the tool. This work concluded that more predictive modelling is needed to refine the results and the results from this work needs to be

confirmed with experiments. The results show that the modelling can predict the reaction phases at the tool-chip interface, this can be used as input for the tool wear mechanisms during development of tooling solutions.

Key words: Aluminium, aluminium machining, PCD, tool-chip interface, Thermo-Calc, AdvantEdge, automotive

(3)

Sammanfattning

Aluminium användningen inom fordonsindustrin förväntas öka med 25 % under det närmsta

decenniet, främst på grund av elektrifiering men också för att för att minska bränsleförbrukning. Målet med detta examensarbete är till att öka förståelsen för de kemiska interaktionerna mellan

aluminiumlegeringar och typiska verktygssystem vid metallskärning, framförallt vid bearbetning av aluminiumlegeringar innehållande kisel med ett TiN-belagt PCD-verktyg och icke-belagda verktyg.

Detta gjordes genom prediktiva FEM-simuleringar av temperatur, med hjälp av mjukvaran AdvantEdge. Parallellt med detta skapades databaser för simulering av den kemiska interaktionen mellan skär och bearbetningsmaterial i programvaran Thermo-Calc. De termodynamiska data och beskrivningarna av de termodynamiska system som används var bedömda av andra författare men kritiskt granskade för användning i detta arbete. Resultaten av FEM-simuleringarna gav den beräknade temperaturen för bearbetning av aluminium med PCD ligger någonstans mellan 60-80 % av

smälttemperaturen för Al7wt % Si-legering. Beräkningarna av den kemiska interaktionen resulterar i sin tur i att flera hårda utskiljningar kan fastna på eller transformera på ytan mellan arbetsstycket och verktyget, till exempel SiC, Al4C3och diamant från verktyget. Resultaten från detta arbete visar att det går att förutsäga fasomvandlingar mellan skär och arbetsstycket, samt att detta kan användas som indata för skärförslitning under utvecklingen av verktygslösningar.

Nyckelord: Aluminium, aluminiumbearbetning, PCD, verktygs-chip-gränssnittet, Thermo-Calc, AdvantEdge, fordonsindustri

(4)

Table of content

1. AIM AND SCOPE 1

2. INTRODUCTION 2

2.1. AL-ALLOYS IN THE AUTOMOTIVE INDUSTRY 2

2.2. MACHINING PROCESS 3

3. LITERATURE STUDY 6

3.1. PART 1.ALUMINIUM MACHINING 6

3.1.1. MACHINING TOOL -PCD 6

3.1.2. TOOL-CHIP INTERFACE 7

3.1.3. SIMULATION OF FORCE AND TEMPERATURE 9

3.2. PART 2.CHEMICAL INTERACTIONS AND THERMODYNAMIC MODELLING 10

3.2.1. CALPHAD 11

4. METHOD 12

4.1. PART 1.FEM MACHINING SIMULATIONS 12

4.2.PART 2.TOOL-WORKPIECE INTERACTION 13

4.1.1. CREATING THERMODYNAMIC DATABASE 13

4.1.2. CALCULATIONS OF THE CHEMICAL INTERACTIONS 16

5. RESULTS 17

5.1. PART 1.FEM MACHINING SIMULATIONS 17

5.2. PART 2.TOOL-WORKPIECE INTERACTION 19

6. DISCUSSION 24

6.1. PART 1.FEM MACHINING SIMULATIONS TEMPERATURE 24

6.2. PART 2.TOOL-WORKPIECE CHEMICAL INTERACTION 25

6.2.1. AL-C-CO 25

(5)

6.2.2. AL-C-SI 25

6.2.3. AL-CO-SI 25

6.2.4. AL-N-TI 26

6.2.5. AL-N-SI 26

6.2.6. AL-CO-TI 26

6.2.7. GENERAL 26

6.3. ETHICAL AND ENVIRONMENTAL ASPECTS 27

7. CONCLUSION 29

7.1. FURTHER WORK 30

8. ACKNOWLEDGEMENT 31

9. REFERENCES 32

(6)

1

1. Aim and scope

The aim of this diploma work is to increase the understanding of the chemical interactions between aluminium alloys and typical tool systems in metal cutting. That includes predictive simulations of temperature at the tool-chip workpiece interface and thermodynamic modelling of the chemical interaction at the predicted temperature.

To reach the objectives finite element method (FEM) simulations of the machining process,

longitudinal turning, to find the theoretical temperature between tool and workpiece is performed. In parallel a thermodynamic database with systems of the aluminium alloys together with the tool materials is created based on literature data to be used as a tool to model the chemical interaction of the tool-chip workpiece interface at the, by FEM, predicted temperature. These results can then be used to understand the wear of the tool and the surface defects on the workpiece caused by the tool wear.

(7)

2

2. Introduction

The conventional method of evaluation of wear and tool life in machining is to 95 % empirical testing, including, sometimes, several hours of machining tests and, wear and surface measurements [1]. As an alternative route to this unsustainable method, predictive computer modelling is used, aiming to reduce the amount of material and time needed for tooling design. With well-developed software, powerful computers and mathematical models describing various physical phenomena simulations of the metal cutting can be done, reducing the cost and amount of physical experiments needed.

However, simulations rely on stored physical data and thus the simulation can only reach the quality level of the available input data it relies on. Simulations and mathematical models would not exist if it were not for the already existing experimental data, usually obtained by trial and error that has been done throughout history [2].

Polycrystalline diamond (PCD) tools are characterised by their excellent thermal conductivity, low thermal expansion and by its low friction between the tools rake face and the chip, these properties are compatible with machining aluminium. Machining experiments to optimize aluminium machining using PCD could last for months and consumes vast amounts of aluminium. In this context, predictive computer modelling is of special importance in the development phase of tooling solutions for

aluminium machining based on its capacity to modelling the process behaviour, e.g., cutting force, process temperature and reaction particles on the tool-chip interface [3].

This diploma work is focused on predicting the temperature at the tool-chip interface by using finite element method (FEM) and creating a thermodynamic database to calculate the chemical interaction between the tool and workpiece at the predicted temperature. The reason for creating databases is twofold i) commercial databases most often do not give the stability equations used for the calculations and thus it is impossible to know how the chemical interactions are modelled ii) the elements needed for tool-workpiece calculations are not available, to the knowledge of the author, in neither commercial nor public databases since some ternary systems are not assessed. To do reliable thermodynamic predictions it is vital to know the lattice stability expressions and background research, exactly which parameters are used to provide sound, science-based understanding of the results.

2.1. Al-alloys in the automotive industry

Aluminium has been used in vehicles for decades, mainly in components, and the need for light weight metals increases with the increasing demand on fuel efficiency. However, today the transition to electric and hybrid cars further drives the demand for aluminium and it is currently increasing with 3.5

% of compound annual growth rate (CAGR) and is predicted to continue to increase with 25 % over the coming decade [4]. In conventional petrol vehicles, aluminium is used in several components, e.g.

the cylinder blocks, and for structural platform parts. Aluminium has great thermal conductivity and is therefore also suitable for heat sink components [1]. With the electrification of the automotive industry comes new challenges and possibilities. With heavyweight batteries there is an even higher demand on lowering the weight of other parts of the cars [1].

(8)

3 2.2. Machining process

In a machining process, the material is removed to create a desired shape. It could be to make the piece thinner, rounder, smoother or create necessary cavity. There are three main types of machining:

turning, milling, and drilling [5]. Milling is a type of machining that uses rotary cutters to remove material off a surface, there can be several cutting inserts mounted on the rotary head, fig. 1a. The cutting head in milling usually moves perpendicular to its axis, the cutting takes place on the circumference of the cutting head. Drilling is the operation of making a circular hole, usually with a drill bit as a rotary cutting tool, fig. 1b. Opposite to milling, in drilling the tool moves along its rotation axis.

Figure 1. a) Types of milling [5], b) Types of drilling [5]

In turning, which this work is focused on, the material to become a component rotates and the cutting tool is fixed, fig. 2a. The amount of removed material thus depends on the rotation speed of the

workpiece, the cutting depth, and the forward feed, which are the main cutting parameters, fig. 2b. The tool properties and the cutting parameters are a function of the machining materials. In the turning process there is one single tool to remove material.

a. b.

(9)

4

Figure 2. a) Types of turning [5], b) Schematic turning process, f is feed, d is depth of cut, Do is the staring diameter, Dm is the diameter after cutting [6].

Beyond the cutting parameters and the tool material, the geometry is of importance for the optimal machining result. In fig. 3 the geometry of a tool is presented where the rake face is the side in contact with the chip and the flank face is faced towards the newly cut workpiece, the angle of these to affects the cutting and feed force, and the temperature in the tool-chip interface [7]. The edge radius of the cutting edge affects the parameters as the angles of the rake and flank face do. The cutting force does not exceed 2GPa when cutting aluminium and the force decreases with increasing cutting speed [8].

Figure 3. a) Cross section of the tool-chip interface b) Schematic figure of a machining tool [3].

Aluminium has for a long time been considered a material that is easy to machine due to its ductility, low hardness and low yield strength. A frequent wear type in aluminium machining is build-up edge (BUE) where the machined metal sticks to the cutting edge of the tool. Hence, the metal is now cut by the built-up edge and thus this changes the cutting conditions, a schematic figure of a BUE is shown in fig. 4. Consequently, the effective edge geometry, specifically the rake angle and the relief angle, are

Primary motion (-C’) d f

chip

Single-point tool

Continuous feed motion (-Z) Do

Vav

a. b.

a. b.

(10)

5

changed, and the edge radius as already mentioned, affects the cutting and increases the feed force, and usually increases the temperature. The BUE continuously builds up and breaks and deposits on the chip and workpiece surface, creating an undesired surface roughness that might need further

processing. When machining aluminium with high silicon content, such that is used by the automotive industry, there it not only an adhesive effect but an abrasive wear mechanism. The workpiece surface can also be affected when the tool is worn, particles or small pull-outs from the worn tool surface can stick to the newly cut surface of the workpiece and possibly chemically react with the surface and create undesired carbides or other hard particles [3].

Figure 4. Built-up edge (BUE) and how it deposits on the chip and workpiece [9].

(11)

6

3. Literature study

This report focuses on modelling the conditions at the tool-chip interface when machining aluminium alloys used in the automotive industry. This section covers the background information of aluminium machining, the preferable tools used, the tool-chip interface, the simulations of temperature and force in machining at the tool-chip interface, and simulations of thermodynamic chemical interactions.

3.1. Part 1. Aluminium machining

In this section the theory behind machining of aluminium is covered. Specifically, the properties of the polycrystalline diamond (PCD) tool used for aluminium machining, the heat generation at the tool- chip workpiece interface and how it affects the machining and surface of the workpiece. Predictive computer modelling using the finite element method (FEM) is also covered.

3.1.1. Machining tool - PCD

According to the graph presented by Klocke [3], in fig. 5, the ideal cutting tool material has high wear and heat-resistance as well as high durability and bending strength. Polycrystalline diamond (PCD) tools, that are covered in this section, are presented in the graph as DP, which has high wear and heat- resistance but low durability and bending strength. PCD tools are also characterised by their excellent thermal conductivity, low thermal expansion and by its low friction between the tools rake face and the chip. In aluminium machining, these properties permit using high cutting speeds with PCD because of the low friction, high wear resistance and its excellent thermal conductivity [3].

Figure 5. Schematic classification of cutting tool materials [3].

(12)

7

Polycrystalline diamond is a composite material, usually with about 80-90 % diamond and using cobalt as a catalyst or binder. PCD tools are produced by high pressure temperature sintering (HPHT) in the diamond stability region. A PCD tool is usually fabricated with PCD at the cutting edge and the rest of the tool is cemented carbide, as can be seen in fig. 6. Typically, PCD is brazed on the cemented carbide carrier. For cutting aluminium, PCD tools are suitable because of the low affinity of

aluminium to carbon. PCD tools are of special significance for machining aluminium alloys with high amounts of silicon. The reason to this is it that Al-Si alloy has different hardness in its structure, with hard silicon precipitates in an aluminium matrix. Because of the abrasive effect from the silicon particles, the cemented carbides wear quickly [3]. PCD tools have up to 80 times longer tool life compared to cemented carbides when machining silicon containing aluminium alloys.

In the article by Bushlya et al. [10], where they investigated the wear of PCD tools when machining metal matrix composites (MMC) of cast Al-Si alloy strengthened with SiC, they found that there was no abrasion of diamond by SiC particles but the high energy impact caused by the MMC

reinforcement lead to fracture, cracking and removal of diamond fragments. They also found diffusional wear of cobalt on PCD, where cobalt was replaced by silicon and aluminium from the MMC Al-Si alloy.

Figure 6. PCD insert with PCD in the cutting edge, the dark grey area and cemented carbide in the light grey area [11].

3.1.2. Tool-chip interface

The tool-chip interface is of interest because the heat in metal cutting is one of the limiting factors of the machining processes, and most heat is generated at the tool-chip interface. This usually negatively impacts on the tool wear as it rapidly increases and thus the dimensional accuracies deteriorates and the cost of production increases. The material that is removed from the workpiece in a machining process is called chip. All the metal removed is plastically deformed which requires large amount of energy. The chip, given in fig. 7, where the area k-l-m-n is the chip before machining, and this

becomes the deformed chip in the area p-q-r-s after it has been machined. It is in the cross-section O-D where most of the deformation happens, this is called the primary shear zone. When the chip slides over the surface of the tool, O-B, heat is generated by friction (secondary shear zone). The tool-chip interface is of importance when machining because that is where the highest temperature of the

(13)

8

process is reached. Heat is also generated at the point of rupture where the metal is physically separated from workpiece to chip, point O, however, the energy required for the rupture is

insignificant in proportion to that required to plastically deform the metal that is removed as the chip.

[8].

Figure 7. Metal cutting diagram by Trent [8].

The three heat generating mechanisms (primary O-D, secondary O-B, and separation O) all add up to the peak temperature which according to Trent et al. [8] has its peak in the zone where the tool and chip have full contact O-B [12]. The heat generated by the friction can be reduced with the help of cutting fluids, but the other two sources of heat will not be reduced. According to Klocke, over 50 % of the mechanical energy is converted to heat in the primary shear zone O-D [3].

To measure the exact temperature at the tool-chip interface is difficult because of the close contact.

Experimentally it can be measured with a thermocouple in the workpiece that measures the

temperature right before it is cut off and destroyed by the tool. The temperature can also be measured with the thermocouple on the tool close to the interface. There are also ways to calculate the

temperature theoretically using the finite element method [13] [14].

Vc

(14)

9 3.1.3. Simulation of force and temperature

As mentioned before, the conventional trial and error testing for tooling design is costly and time consuming. By first using predictive models for machining processes, before any physical testing is done, can be more cost effective. To get a useful output of a model it is important to have as accurate input data as possible. Fig. 8 covers the whole chain of machining process for modelling. Input data for modelling a machining process can be the workpiece material properties, cutting parameters, tool material and geometry. The modelling of the input data can be carried out through an analytical, numerical, empirical, or AI-based model or what is becoming more usual, a hybrid combining different models. From these models output data of fundamental variables is obtained and the conclusions drawn from the output data can be industry relevant outcomes, which could shorten lead times for new products or processes. A numerical method that is widely used to solve engineering problems is the numerical finite element method (FEM) [15].

Figure 8. A flow chart of the predictive modelling of machining [15].

FEM is used to solve mathematical models and engineering problems. FEM is a numerical method for solving partial differential equations. To solve an engineering problem the system is divided into smaller, simpler parts, these are called finite elements. Boundary values for each element is set which

(15)

10

results in a system of algebraic equations. This method then approximates the unknown function of the domain. Then the equations that model the finite elements are assembled to a larger system that models the full problem. The finite element method then uses variational methods to approximate the solution through minimizing an associated error function [16].

Previous work in this area has been done by Javidikia et al. [7] where they investigated machining of the alloy 6061-T6 with a cemented carbide tool with experiments and used the results to validate their finite element model of the same alloy and tool using the software DEFORM, the compared results was of chip thickness, cutting force and thrust force. They found that the estimated maximum temperature was at the tool-chip interface did not relate much to the edge radius of the tool, but the average tool tip temperature was dependant on the edge radius. The estimated peak temperature they found was 473 ºC with the cutting speed 950 m/min and edge radius of 0.02 mm. They also found that estimated the average temperature increased with higher cutting speeds, because of the higher

frictional and plastic works.

Lata and Hitesh [17] investigated the temperature at the tool-chip interface experimentally for different combinations of workpiece material and tool material. They found that for an aluminium alloy (IS 733 Gr 63400Aluminium) workpiece machined by a tungsten carbide, the temperature was about 380 ºC measured with the tool-workpiece thermocouple method, with the cutting speed 550 m/min and depth of cut 2.0 mm. They also found for the same aluminium alloy with a PCD tool, the temperature at the same speed and depth of cut to be around the same temperature of 380 ºC.

In an article by Santos et al. [18] they found that through machining experiments for the aluminium alloy 1350-O together with a cemented carbide with the cutting speed 600 m/min, feed rate 0.2 mm/rev and depth of cut 1 mm the temperature to be 417 ºC and for the same set-up but a depth of cut of 3 mm the temperature found was 429 ºC, both measured by the tool-workpiece thermocouple method. They concluded that the temperature increases with increased cutting speed. They also concluded that the depth of cut affects the temperature.

3.2. Part 2. Chemical interactions and thermodynamic modelling

To model the chemical interactions at the tool-chip interface at the prevalent temperature, first a thermodynamic database, with expressions of the Gibbs energy for the entire system is needed. This section covers the necessary components for thermodynamic modelling. In computational

thermodynamics (CT) the equilibrium state is the output data. This is obtained by modelling the Gibbs energy for every phase present in the system [19].

The thermodynamics models have adjustable parameters that are used to fit experimental and ab initio data. With continuous research and development, new experimental and theoretical data becomes available, this makes it possible to update the models and model parameters for more accurate

thermodynamic descriptions. This can be done through the widely used CALPHAD method described in the next section.

(16)

11 3.2.1. CALPHAD

CALPHAD is short for CALculation of PHAse Diagrams and is the method used in CT. The CALPHAD approach is based on a consistent description of the phase diagram and thermodynamic properties to predict the stable phases and their thermodynamic properties without having to perform a whole new set of experiments to collect data. In addition, it is possible to predict the metastable states by excluding any stable phase from the calculation. The philosophy of CALPHAD is to produce models that represent the thermodynamic properties of various phases to permit prediction of properties of multicomponent systems from the binary and ternary subsystems [20].

Thermodynamics describes the equilibrium state of a system, which is essential for simulations of phase transformation as all systems try to reach this state. In CT, the thermodynamic properties are described by Gibbs energy as a function of pressure, temperature, and composition. To succeed with the CALPHAD method it is crucial to have physical realistic mathematical models for the Gibbs energy for each phase. All consistent experimental and theoretical data that can be used to derive the Gibbs energy expressions are used to fit the model parameters [21].

The development of a thermodynamic description is usually referred to as an “assessment”. It describes binary, ternary or higher order systems where the authors collect and review all existing literature of the system and together with their own calculations or experiments model the Gibbs energies of the phases in the system. These assessments can be used in software to model

multicomponent systems of thermodynamics using the CALPHAD method for extrapolation [22] [23].

(17)

12

4. Method

This part covers the methods used in this project. The work is divided in two parts, the first one, part 1, covers how FEM machining simulations of the theoretical peak temperature span was produced, in the second part, part 2, the chemical interaction is theoretically predicted by thermodynamic calculations at the FEM-predicted peak temperature and temperature span (thus from part 1).

4.1. Part 1. FEM machining simulations

In this work the AdvantEdge software is used. The software has built in calculations for temperature, force, heat rate and much more. In this work the AdvantEdge property data of workpiece materials and tool materials is used. In this work, only turning was considered.

To calculate the temperature at the tool-chip interface with the software AdvantEdge the work was divided in two parts. The first part consisted of replicating the work done by Javidikia et al. [7] who used the DEFORM software. In this work the replication was carried out with the software

AdvantEdge. In their work the aluminium alloy 6061-T6 was used together with a tool of uncoated cemented carbide, the tool and material properties data used by them and also used in this work are listed in table 1. The replication was done to confirm the results and built-in equations in AdvantEdge.

In the second part the workpiece material was the same as in the replication, shown in table 1, but the tool material was changed to a PDC (medium conductivity) tool from AdvantEdge. For both models the same process parameters were used, these are given in table 2.

Table 1. Mechanical and thermal properties of workpiece and tool materials used in article by Javidikia et al. [7].

Properties AA6061-T6 Uncoated cemented carbide

Density ρ (kg/m3) 2700 11,900

Young’s modulus E (GPa) 58.5 612

Poisson’s ratio υ 0.33 0.22

Conductvity k (W/mºC) 167 86

Specific heat capacity c (J/kgºC) 896 337 Thermal expantion coefficient α (1/ºC) 23.5x10^-6 4.9x10^-6

(18)

13 Table 2. Process parameters and simulation parameters.

Process parameters Simulation paramteres

Feed 0.2 mm Avg. Length of cut ratio 10.0

Cutting speed 950.0 m/min Max- number of nodes 24000

Depth of cut 2.0 mm Max. Element size 0.1 mm

Length of cut 5.0 mm Min. Element size 0.02 mm

Initial temperature 20 ºC Fraction of radius 0.6

Friction coefficient 0.5 Fraction of feed 0.1

Mesh refine 8

Mesh coarse 1

4.2. Part 2. Tool-workpiece interaction

In this part the method to simulating the chemical interaction between tool and workpiece is described.

To simulate the chemical reactions, there is a need for reliable a thermodynamic database, thus part 4.2.1 concerns the construction of a thermodynamic databases for ternary interaction in the Al-Si-C- Co-N-Ti system. Thereafter simulations using the database in the software Thermo-Calc to calculate the reactions at the temperature span given by the FEA simulation. Thermo-Calc is a software for phase diagram, phase equilibrium and phase transformation calculations [24], in this work it is used to calculate phase equilibria of ternary systems. In this case the simulations were separated into two parts i) first ternary calculations were made with pure Al (no Si) – PCD (thus carbon and cobalt) and ii) thereafter multicomponent systems simulating aluminium-silicon alloy.

4.1.1. Creating thermodynamic database

In this section the systems considered for building a thermodynamic database in this project are presented and the reasons why a specific assessment was chosen. When creating thermodynamic databases, the thermodynamic descriptions need to be compatible with each other. Hence, occasionally it can be more suitable to use an assessment of a system because it conforms with another system also present in the same database. However, it is emphasised that throughout the work every system has been critically reviewed and selected based on the purpose of this work. The entire system is a representation of machining of aluminium alloys with high silicon content, machined by a PCD tool

(19)

14

with or without with TiN coating. The system is divided into subsystems of binaries and ternaries to examine the chemical interactions.

Al-C-Co

For the system Al-C-Co a critical assessment of the ternary system by Zheng et al. [25] was used because it is the most recent thermodynamic assessment of this system using the CALPHAD method.

The binaries included in this ternary are Al-C, Al-Co and C-Co. Al-C assessed by Gröbner et al. [26], , although there is a more recent assessment of Al-C by Deffrennes et al. [27], which showed that in the assessment by Gröbner et al. there is an underestimation of experimental heat content data for

temperatures above 1000K. In this report temperatures above 1000K are not considered and therefore the binary by Gröbner et al. is used since it is used in the ternary assessment of Al-C-Co. The system Al-C by Gröbner et al. was missing a description of diamond. Thus, the stability was added using the description from the SGTE pure elements database [28]. The binary Al-Co used was assessed by Wang et al. [29] who re-assessed the ternary system Al-Co-W because of an inverse miscibility gap at 4800K in the binary Al-Co which could intervene with obtaining a trustworthy description of the Al- Co-W system. The binary C-Co used, was assessed by Fernández Guillermet [30].

Al-C-Si

The ternary Al-C-Si system assessed by Gröbner et al. [26] was convenient and compatible with the description of Al-C by the same author and furthermore with the Al-C-Co system by Zheng et al. [25].

The binaries included in this ternary are Al-C, Al-Si and C-Si, these are all assessed by Gröbner et al.

[26].

Al-Ti-Co

The system Al-Co-Ti assessed by Dupin [31] was used as it is the newest assessment of the system.

The binaries included in this ternary are Al-Ti, Al-Co and Ti-Co. Al-Co assessed by Stein et al. [32], Ti-Al assessed by Saunders [33], and Co-Ti was assessed by Davydov et al. [34]. The binaries used in this section are used because of their fit to the ternary which they are present in, the binaries are critically reviewed. Concerning the Al-Co-Ti by Dupin there is not sufficient data to add silicon to the system. The only data that is accessible is for the binaries Al-Si assessed by Gröbner et al. [26], Ti-Si assessed by Ma et al. [35] and Co-Si assessed by Zhang et al. [36].

Al-N-Ti

The system Al-N-Ti was critically assessed by Zhang et el [37] who also assessed one of the binaries, Al-N, there is an older assessment of the ternary Al-N-Ti by Chen and Sundman [38], but according to Zhang et al. it has shown inconsistencies in the Al-Ti system. The assessment of the Ti-Al system was assessed by Witusiewicz et al. [39]. The binary Al-N used was assessed by Zhang et al. [40]. The binary Ti-N was assessed by Zeng et al. [41]. There is not sufficient data to add silicon to the system.

The only data that is accessible is for the binaries Al-Si assessed by Gröbner et al. [26], N-Si assessed by Ma et al. [42] and Ti-Si assessed by Ma et al. [35], there are no ternaries available for the system which makes it insufficient. All four elements Al-N-Ti-Si are present in the commercially available titanium database (TCTi2), but the parameters for the description is not public.

(20)

15 Al-Si-Co & C-Co-Si

The ternaries Al-Co-Si and C-Co-Si have not yet been assessed thermodynamically to the knowledge of the author. No assessments were done in this work. However, there are experimental data for Al-Si- Co by Richter et al. [43] and Raghavan [44] that could be compared with the phase diagram

extrapolated from the binaries. The binary Co-Si has been assessed more than once and the assessment that is used in commercially available databases such as TCTi2 and TCNi10 is assessed by Choi. [45], however, there is a more recent assessment by Zhang et al. [36] that pointed out that the low

temperature stable phases αCo and εCo also appear in the high temperature region in Chois model.

Therefore, the assessment by Zhang et al. is used in this work.

In this work three database files were created and were used for different ternaries with the mentioned additions called LE-Al-C-Co, LE-Al-N-Ti and LE-Al-Co-Ti, in addition silicon was also added to the Al-C-Co thus LE-Al-C-Co-Si. All the systems used in these databases are assembled in table 3.

Table 3. An overview of the systems used in this work and the authors who have assessed the systems.

System Assessed by Assessment used in this work by

Database name C-diamond STGE database [28] STGE database [28] LE-Al-C-Co-Si Al-C Gröbner et al. [26], Deffrennes

et al. [27], Ohtani et al. [46], Connétable et al. [47]

Gröbner et al. [26] LE-Al-C-Co-Si

Al-Co Wang et al. [29], Stein et al.

[32], Dupin and Ansara [48]

Wang et al. [29] LE-Al-C-Co-Si Stein et al. [32] LE-Al-Co-Ti C-Co Fernández Guillermet et al.

[30]

Fernández Guillermet et al. [30]

LE-Al-C-Co-Si Al-Si Gröbner et al. [26], Dörner et

al. [49]

Gröbner et al. [26] LE-Al-C-Co-Si C-Si Gröbner et al. [26] Gröbner et al. [26] LE-Al-C-Co-Si Co-Si Choi. [45], Zhang et al. [36],

Ishida et al. [50]

Zhang et al. [36] LE-Al-C-Co-Si Al-Ti Witusiewicz et al. [39],

Saunders [33], Okamoto [51], Kattner et al. [52]

Witusiewicz et al. [39] LE-Al-N-Ti Saunders [33] LE-Al-Co-Ti Al-N Zhang et al. [40], Lukas [53],

Hillert and Jonsson [54]

Zhang et al. [40] LE-Al-N-Ti Ti-N Zeng and Schmid-Fetzer [41],

Jonsson [55], Ohtani and Hillert [56]

Zeng and Schmid-Fetzer [41]

LE-Al-N-Ti

Co-Ti Davydov et al. [34], Cacciamani [57]

Davydov et al. [34] LE-Al-Co-Ti

N-Si Ma et al. [42] - -

Ti-Si Ma et al. [35], Seifert [58] - -

Al-C-Co Zheng et al. [25] Zheng et al. [25] LE-Al-C-Co-Si Al-C-Si Gröbner et al. [26] Gröbner et al. [26] LE-Al-C-Co-Si

(21)

16 Al-N-Ti Zhang et al. [37], Gao et al.

[59], Chen and Sundman [38]

Zhang et al. [37] LE-Al-N-Ti

Al-Co-Ti Dupin [31] Dupin [31] LE-Al-Co-Ti

Al-Co-Si No assessment found - -

C-Co-Si No assessment found - -

4.1.2. Calculations of the chemical interactions

As already mentioned, the chemical interactions were calculated at 60 % and 80 % of the melting temperature of the aluminium alloy 6061-T6 and the span was also assumed to cover aluminium alloys with high silicon content (around 7wt %) as well as for pure aluminium. The calculation of the melting temperature of pure aluminium and the Al-7wt % Si was done using the commercially available TCAL7 database in Thermo-Calc referring the Gröbner et al. assessment [26] for the binary Al-Si.

Thereafter, the chemical interactions between diamond and pure aluminium were calculated using the LE-Al-C-Co database and with Al7W %Si alloy was simulated using the LE-Al-C-Co-Si database.

(22)

17

5. Results

5.1. Part 1. FEM machining simulations

The results presented in fig. 9-11 show the temperature distribution of two simulations. In fig. 9a the results of the replication of the article by Javidikia et al. [7], and in fig. 9b the result using the same workpiece material as in Javidikia et al. [7] and using a AdvantEdge built in PCD tool material.

Figure 9. In a. replication of results from paper by Javidikia et al. [7], peak temperature ~ 480 ºC b. same workpiece material as in Javidikia et al. but PCD tool from AdvantEdge, peak temperature

~377 ºC.

The simulated peak temperature for the replication with the aluminium alloy 6061-T6 machined with an uncoated cemented carbide tool is around 480ºC, fig. 9a. The concentration of the highest

temperature is on the secondary share plane (segment O-B, fig. 7), which is also called tool-chip interface. Meaning that the simulation achieved the same results as in the article by Javidikia et al. [7].

On the other hand, when to tool was changed to PCD, fig. 9b, the maximum temperature is around 377 ºC and the maximum temperature is in the O-region, in fig. 7, and not on the tool-chip interface. In fig.

10, a magnification of the tool-side of the interface, where it is possible to see the temperature gradient of the tool. The temperature gradient in the PCD, 10b is further into the tool which indicated that the built-in model in AdvantEdge could recognise the differences in thermal conductivity of the tools.

a. b.

(23)

18

Figure. 10. Temperature gradient in tool a. Cemented carbide from Javidikia et al. [7] b. PCD from AdvantEdge. The peak temperature is 480 ºC for cemented carbide compared to 377 ºC for

PCD and the heat affected area in the PCE tool is larger.

Figure 11. Workpiece material, aluminium alloy 6061-T6 a. machined with uncoated cemented carbide tool from Javidikia et al. b. machined with PCD tool material from AdvantEdge.

In fig. 11 the workpiece and chip are presented where the temperature is more concentrated to the point between the workpiece and chip for PCD in fig. 11b. Where in fig. 11a the temperature in concentrated on the whole tool-chip area. The force is also presented in fig. 11, which is the same for both simuations. The temperature at the primary shear zone (section O-D, fig. 7) is the same for both cemented carbide and PCD.

a. b.

a. b.

Force X: 440 N Y: 220 N

(24)

19

5.2. Part 2. Tool-workpiece interaction

In fig. 12-16 the results from part 2, the chemical interaction, is presented. For every system, the chemical interaction has been calculated at two temperatures based on the melting temperature of the aluminium alloy 6061-T6, used in the first part Tmelt = 616 ºC (889.15 K), at 60 % of the melting temperature and at 80 % of the melting temperature which are close to the two peak temperatures in part 1.

Figure 12. Ternary stable phase-diagram of Al-C-Co from assessment by Zheng et al. [25] at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt) from LE-Al-C-Co.

Ternary stable phase-diagram of Al-C-Co from assessment by Zheng et al. [25] at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt) is given in fig. 12 for the stable system with graphite and then

metastable with diamond in fig. 13. The blue line in both pictures starts from the composition of the PCD a.

b.

(25)

20

tool (with the composition of 10wt % Co (mole fraction = 0.022 C) and 90wt % C (mole fraction = 0.978)) indicated by a blue dot and ends at pure aluminium. All the red lines crossed by the blue line are the potential phase transitions when machining pure aluminium with an uncoated PCD tool.

Figure 13. Ternary metastable phase-diagram of Al-C-Co from LE-Al-C-Co with assessment by Zheng et al. [25] and addition of diamond from STGE [28] at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt). The blue line in both pictures drawn from the right dot, with the composition for the tool with 10wt

% Co (mole fraction = 0.022 C) and 90wt % C (mole fraction = 0.978) to the left dot with pure aluminium.

a.

b.

(26)

21

Figure 14. Ternary phase-diagram of Al-C-Si from assessment by Gröbner et al. [26] at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt) from LE-Al-C-Co-Si database.

The ternary Al-C-Si representing machining of aluminium-silicone alloy using a diamond tool. There are differences between 60 %Tmelt and 80 %Tmelt in both fig. 14a and 14b. In 14a there is a region of

Al4SiC4+Al8SiC7+Diamond_A4 that is not present in 14b. At 80 %Tmelt, in fig. 14b instead there is a phase region that is not present at 60 %Tmelt, namely Al4SiC4+Al8SiC7+FCC_A1.

a.

b.

(27)

22

Figure 15. Ternary phase-diagram of Al-N-Ti from assessment by Chen et al. at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt) from LE-Al-N-Ti. The blue line in both pictures drawn from the right dot,

with the composition for the coating of the tool with 50at % N and 50at % Ti to the left dot with pure aluminium. The red lines crossed by the blue line are the potential phase transitions when machining pure aluminium with a PCD tool coated with TiN. In fig. 15b, the phase Ti3AlN appears, although not in

a region crossed by the blue line.

The ternary Al-Co-Ti representing the chemical interaction that could arise between Al-Co-Ti when machining of aluminium alloy with a PCD tool coated TiN, if the coating of the tool is worn so that cobalt is exposed to the machining material. There are no differences between 60 % Tmelt and 80 % Tmelt in fig.

16a and 16b.

a.

b.

(28)

23

Figure 16. Ternary phase-diagram of Al-Co-Ti from assessment by Dupin at a. 370 ºC (60 % of Tmelt) and b. 493 ºC (80 % of Tmelt) from LE-Al-Co-Ti.

a. b.

(29)

24

6. Discussion

In the following sections first the FEM simulations and then the chemical interactions will be discussed and finally a general discussion.

6.1. Part 1. FEM machining simulations – temperature

The replication of the Javidikia et al. [7] was successful, presented in fig. 9a, 10a and 11a, with a small deviation in maximum temperature 480 ºC in this work compared to 473 ºC in Javidikia et al.’s work.

The built-in equations were validated from an outside source and this confirmed that the software can accurately reproduce the results for cemented carbides. In the next step, using the same workpiece material data as in the Javidikia et al. paper but changing the tool material to PCD, taken from the AdvantEdge database, presented in fig. 9b, 10b and 11b. The results from this simulation differed significantly from the cemented carbide simulations and resulted in a lower peak temperature and the heat was distributed into a bigger part of the tool. This is reasonable in theory because of the thermal conductivity which is higher in PCD and because of the lower friction between the chip and PCD inserts compared to cemented carbides, however, how much this would lower the peak temperature is hard to say without calibration through experiments. It is important to note that the software

AdvantEdge is not calibrated for PCD but for cemented carbides, which most likely will affect the results from this simulation.

The two peak temperature results are in the range of 60-80 % of the melting temperature of aluminium alloys with high silicon content at the tool-chip interface. This has been used as input for calculating the chemical interaction temperature. The primary shear zone (O-D, in fig. 7) appears to have the same temperature when comparing uncoated cemented carbide versus PCD in fig. 11. This could be because this area is less affected by the thermal conductivity or the friction of the tool, but it is a function of the tool geometry and cutting speed (which were constant). It is an additional evidence that the FEM simulations used is coherent with the theoretical machining process/chip formation.

In the article published by Lata and Hitesh [17] they compare the temperature of machining the aluminium alloy IS 733 Gr 63400Aluminium with both tungsten carbide and PCD tools, where they find nearly the same temperature for the same machining conditions for both tools, which contradicts the theoretical results from this work. Despite the uncertainty of the calibration for the software used in this work, there is literature supporting that the lower friction of PCD tools generates less heat than carbides, hence the temperature should be lower for machining with PCD than machining with tungsten carbides.

The two peak temperatures constitute the input data in part 2, where the chemical interactions are calculated at the predicted temperatures. The lower peak temperature, 60 % Tmelt which was about 380 ºC, from the PCD simulation is seemingly lower than temperatures from experiments with PCD such as Lata and Hitesh [17] who found the experimental temperature above 400 ºC. On the other hand, is the higher peak temperature, 80 % Tmelt, about480 ºC, from the cemented carbide simulation is likely higher than any temperature obtained in PCD experiments. However, this range covers the potential temperatures in PCD machining and is therefore used to calculate the chemical interactions.

(30)

25

6.2. Part 2. Tool-workpiece chemical interaction

As already mentioned, the predictions of the tool-chip interaction will only be as accurate as the data on the ingoing systems are that the calculations are based on. Thus, first in this section a discussion of the pros and cons of the selected systems to construct the thermodynamic databases given. There are as mentioned, many combinations of ternaries for the system Al-Si-C-Co-N-Ti to consider and there are also lacking data for some ternaries.

6.2.1. Al-C-Co

The LE-Al-C-Co system, in fig. 12 in the results section, the stable phase diagram with graphite by Zheng et al. [25] and fig. 13, the metastable phase diagram with diamond (added in this work) [28].

The metastable phase diagram is relevant because the PCD tool consists of diamond, but it is unknown whether the diamond transforms to graphite when it is worn at the tool-chip interface. The stable and metastable phase diagrams are identical from 0 % C to about 43 % C and differs where C content exceeds 43 % where in the stable phase diagram graphite is present and in the metastable phase diagram diamond is present at the current temperature.

The blue line drawn in the phase diagrams crosses the phases that could potentially appear depending on the composition. Given the phases, depending on how the tool is worn there will be various compositions on the surface of the workpiece and chip. Pull-outs of the tool edge could stick to the workpiece surface at the elevated temperature caused by machining, then the composition in that point is equal to the right dot of the blue line in fig. 12 and 13. This would mean a composition of graphite, kappa and FCC for the stable phase diagram and diamond, kappa and FCC for the meta stable phase diagram. The kappa-carbides is an ordered cubic phase, with perovskite E21-type structure and FCC is the metallic face centred cubic cobalt matrix.

6.2.2. Al-C-Si

The data from the assessment of Al-C-Si by Gröbner et al. [26] was added to the same custom

database as Al-C-Co. The same binary for Al-C by Gröbner et al. [26]was used in both systems. In 14a there is a region of Al4SiC4+Al8SiC7+Diamond_A4 that is not present in 14b. At 80 %Tmelt, in fig.

14b there is a phase region that is not present at 60 %Tmelt, it is Al4SiC4+Al8SiC7+FCC_A1. There is an uncertainty in the exact temperature at the tool-chip interface and therefore difficult to say without experiments which of the phases in 14a or 14b that are most accurate.

6.2.3. Al-Co-Si

The ternary system Al-Co-Si is not assessed to the knowledge of the author. The binaries for this ternary exist, Al-Si, Co-Si and Al-Co. It is only Co-Si that is not present in the database Al-Co-C-Si created here. Although the binaries are most likely not enough to extrapolate an adequate ternary phase diagram of the system Al-Co-Si since ternary phase will be missing.

(31)

26 6.2.4. Al-N-TI

The assessment by Chen et al. [38] was used to calculate the Al-N-Ti system given in fig. 15. This diagram can be used to represent machining of pure aluminium with a TiN-coated tool. The blue line in the phase diagrams crosses all phases that could appear thermodynamically when the coating is worn of the tool and potentially chemically reacts with the aluminium. However, no regard has been taken to reaction with the underlying carbide. TiN from the tool can stick to the new surface of the component being machined and transform to AlN and Al2Ti or Al3Ti. These hard particles could affect further processing or machining.

6.2.5. Al-N-Si

No assessment of the Al-N-Si system is to be found to the knowledge of the author, therefore creating a user database including Al-N-Ti-Si was not possible. Although all four elements are present in the Ti-database TCTi2 available from Thermo-Calc Software AB. This can be used, although with some caution since the calculations behind can be tweaked to fit better with Ti in high amounts, which the database is constructed for. But at least it gives an indication of the chemical interactions.

6.2.6. Al-Co-Ti

The Al-Co-Ti assessment [31], in fig. 16, is complex with many phase transformations depending on composition. This system describes the interaction of aluminium machined by a coated PCD tool, where the coating is worn so that parts of the PCD is exposed. The Co could the react with the aluminium creating harder particles on the surface of the workpiece.

6.2.7. General

If these systems were combined into Al-Si-C-Co-N-Ti in one database, a thorough review of

dissimilarities and resemblances of the binaries that are described in different assessments are needed to make it coherent. As well as the description and naming of the phases that are present in more than one system, which applies to many phases for example there is the phase AL5CO2 in the LE-Al-C- Co-Si database is named AL5CO2_D811 in the LE-Al-Co-Ti even though it is the same phase.

However, the Gibbs energy expressions do also differ as they are based on different assessments and thus also the energy expressions have to be compared and evaluated.

Comparing the ternary phase diagrams of the Al-C-Co system both the metastable with diamond and the stable with graphite in fig. 12 and 13 there is no new phases appearing in the given interval.

However, there are several phases that can form at the tool-chip interface, such as kappa or Al4C3 for example. Another aspect that can contribute to the degradation is if graphite nucleates that will enhance the graphitisation of the diamond. Then the chemical interactions move from the metastable to the stable phase diagram probably resulting in an increased degradation of the tool. On the other hand, at the Al-rich side, aluminium carbides might enhance the abrasive wear of the Co-

binder/catalyst. Another scenario is reactions between the cobalt binder and the aluminium resulting in, most often, brittle intermetallic phases but this needs to be further investigated. Clearly the phases

(32)

27

shown in fig. 12-16 are equilibrium diagrams nevertheless they will give guidance to the driving forces prevailing locally for chemical interactions.

The calculated temperatures are high relative to the melting point thus atomic mobility temperatures;

possibly therefore it is more likely that there is a phase transformation at 80 % of the melting temperature rather than with 60 % of melting temperature. But this needs to be verified

experimentally. Referring to the work of Bushlya [10] they observed Co-silicide formation at the tool- chip interface when machining Al-MMC with PCD, hence proving chemical reactions at the tool-chip interface. It is an interesting question whether it will be the Co-binder in the diamond tool that will react with the aluminium, fig. 12 and 13 or in case of an Al-Si alloy if carbon will primarily react with the Si Si+C=> SiC as indicated in fig. 14 in the Al-Si-C system. Thus, forming abrasive carbides.

The potential phases that could arise from BUE and tool wear that sticks to the workpiece when machining aluminium alloys with PCD are usually harmful for further processing, but this needs to be investigated. The hard carbides that could occur on the workpiece surface will wear any cutting tool if the workpiece is further machined. The same would happen even if there is no phase transformation and the pull-outs of diamond material from the PCD tool would slide over the tool surface and wear it.

The plausible scenarios in this theoretical work needs to be verified with experiments in order to understand and verify the possible predictions. Without experiments it is difficult to know if or how much of the diamond tool that is worn and if it transforms to graphite that might react with the machining material. It is not likely that the diamond phase would transform to graphite in reaction with the workpiece, since the highest pressure in machining is 2 GPa [8] which is not in the diamond stability region. If the tool is worn in a way that creates small pull-outs, these will still be in the diamond phase in contact with the workpiece material. If the diamond is transformed into graphite it might react with the chip and create hard carbides like Al4C3, but the abrasiveness versus diamond is low.

6.3. Ethical and environmental aspects

One of the reasons for the increased use of aluminium in the automotive industry is to reduce the weight of vehicles and in turn reduce the fuel consumption. The electrification of the car industry is also a way to reduce fossil fuel consumption and as for the conventional vehicles, reducing weight is crucial to increase efficiency. Both conventional and electric automotive uses light weight metals to reduce weight, and aluminium is widely used as a lightweight metal because of its strength to weight ratio and price.

Aluminium is recyclable but with the increased demand of aluminium, recycling existing aluminium is not enough and therefore forces the extraction of bauxite and producing pure aluminium from Bauxite is highly energy consuming, using the Bayer process. A life cycle analysis is needed to determine if the fuel reduction by lighter vehicles compensates for the use of new ore-based aluminium. In the life cycle analysis, it is important to also investigate where the aluminium is produced and from what energy sources. Because of the high energy consumption, one also needs to consider if the energy is disposed in a way so that other industries or even the local civil populations get insufficient energy supplies because of the aluminium production.

(33)

28

As mentioned earlier, the usage of predictive modelling is a way of reducing lead times and costs for processes or products, especially for machining aluminium with PCD, since the testing would take months and require vast amounts of aluminium before the tool is worn. Hence, the methodology and the database created in this work will contribute to the use of both recycled and virgin aluminium.

(34)

29

7. Conclusion

• The FEA simulated peak temperature in the machining process is in the range of 60-80 % of the melting temperature of the alloy 6061-T6 for PCD and cemented carbide tools,

respectively.

• The simulation with PCD showed the lower temperature, close 60 %, which is coherent with literature comparing PCD with cemented carbide, it is recommended that a calibration should be done increase the precision of the simulation of PCD in the FEM software.

• The heat affected zone is larger for the PCD tool than for the cemented carbide tool.

• The calculations of the chemical interaction in turn results in that several hard precipitates can stick to or transform on the surface of the workpiece or tool.

o At the Al – PCD interface, carbides that can form are: Al4C3 & Kappa, and intermetallic phases: Al3Co, Al5Co2 & Al9Co2

o At the Al+Si – PCD interface, carbides that can form are: Al4C3 & SiC, and ternary phases: Al4SiC4 and Al8SiC7

o And evidently diamond from the tool.

• Even though there are no or few differences in phases present at 60 or 80 % of the melting temperature, it is more likely that a phase transformation occurs in the tool-chip interface with the higher temperatures because there is more atomic mobility in the material.

• It is unknown whether the PCD diamond particles or tool itself transform to graphite during use. This in turn gives, according to the present work, different chemical degradation paths since the chemical reactions are different.

• The systems Al-Co-Si and C-Co-Si are not yet thermodynamically assessed to the knowledge of the author, therefore should the results from the thermodynamic database LE-Al-C-Co-Si calculated with silicon be treated with some caution, since the database is incomplete.

• Creating a thermodynamic database for the whole system Al-Si-C-Co-Ti-N requires a

thorough review of different assessments of the same binaries used in different ternaries and a review of the naming and description of all phases.

(35)

30

7.1. Further work

There are many possibilities for further work related to this project. The software used is not calibrated for PCD and therefore the result of the temperature is not entirely reliable. FEM simulations using a custom PCD tool with public properties data could be done or using another software that is calibrated for PCD.

The thermodynamic databases in this project should be refined and silicon should be added to the databases if that data is assessed in the future. Also reviewing and adding the ternaries together to multicomponent database covering the whole system of Al-Si-C-Co-Ti-N. And later, experiments are necessary to fully understand the wear of the tools and what surface defects that wear causes on the workpiece.

(36)

31

8. Acknowledgement

I wish to thank Denis Boing at Sandvik Coromant for mentoring and immense patience with teaching me all about the machining process in such a pedagogical way. I would like to thank Susanne Norgren at Sandvik Coromant for believing in me and accepting me to this thesis. I wish to thank Susanne for motivating me and suppling me with knowledge and information. My two supervisors at Sandvik Coromant, Susanne and Denis, have not only supplied me with their individual help, but as a team they have been incredibly coherent which I am greatly thankful for. I would also like to thank Fredrik Haglöf at Sandvik Coromant for his support and knowledge, helping me to solve any database and Thermo-Calc related problems. Finally, I wish to thank Prof. Malin Selleby at KTH - MSE for her supervision and mentoring, providing me with infinite knowledge in an understandable way.

(37)

32

9. References

[1] DuckerFrontier, “2020 North Americ Light vehical aluminum component and outlook,” 2020.

[2] J. Banks and C. M. Sokolowski, Principles of Modeling and Simulation: A Multidisciplinary Approach, vol. 31, ohn Wiley & Sons, Hoboken, 2009.

[3] F. Klocke, Manufacturing Process 1, Springer Heidelberg Dordrecht London New York, 2011.

[4] IEA Inernational energy agency, “Technology Roadmap - Electric and plug-in hybrid electric vehicles,” IEA.

[5] Sandvik Coromant, “Metal cutting knowledge,” [Online]. Available:

https://www.sandvik.coromant.com/en-gb/knowledge/pages/default.aspx. [Accessed 1 February 2021].

[6] A. Y.-C. Nee, Handbook of Manufacturing Engineering and Technology, Springer-Verlag London, 2015.

[7] M. Javidikia, M. Sadeghifar, V. Songmene and M. Jahazi, “On the impacts of tool geometry and cutting conditions in straight,” The International Journal of Advanced Manufacturing

Technology, 15 January 2020.

[8] E. M. Trent and P. K. Wright, Metal Cutting Fourth Edition, Butterworth–Heinemann, 2000.

[9] Z. Wang, V. Kovvuri, A. Araujo, M. Bacci, W. N. P. Hung and S. T. S. Bukkapatnam, “Built- up-edge effects on surface deterioration in micromilling processes,” Journal of Manufacturing Processes, vol. 24, no. 2, pp. 321-327, 2016.

[10] V. Bushlya, F. Lenrick, O. Gutnichenko, I. Petrusha and Osipov, “Performance and wear mechanisms of novel superhard diamond and boron nitride based tools in machining Al-SiCp metal matrix composite,” Wear, Vols. 376-377, no. A, pp. 152-164, 2017.

[11] Sandvik Coromant, “Turning inserts grades non ferrous material,” [Online]. Available:

https://www.sandvik.coromant.com/sv-se/products/turning-inserts-grades-non-ferrous- materials/pages/default.aspx. [Accessed 12 January 2021].

[12] E. M. Trent, “Metal cutting and the tribology of seizure III temperatures in metal cutting,” Wear, pp. 64-81, 8 June 1988.

[13] E. M. Trent, “METAL CUTTING AND THE TRIBOLOGY OF SEIZURE: I Seizure in metal cutting,” Wear, pp. 29-45, 8 June 1988.

[14] E. M. Trent, “METAL CUTTING AND THE TRIBOLOGY OF SEIZURE: II MOVEMENT OF WORK MATERIAL OVER THE TOOL IN METAL,” Wear, pp. 47-64, 8 June 1988.

(38)

33

[15] P. J. Arrazola, T. Özel, D. Umbrello, M. Davies and I. S. Jawahir, “Recent advances in modelling of metal machining processes,” CIRP Annals - Manufacturing Technology, vol. 62, 2013.

[16] D. L. Logan, A first course in the finite element method, Cengage Learning, 2011.

[17] S. Lata and R. R. Hitesh, “Investigation of Chip-Tool Interface Temperature: Effect of

Machining Parameters and Tool Material on Ferrous and Non-Ferrous Metal 1,” Materials today Proceedings, vol. 5, no. 2, pp. 4250-4257, 2018.

[18] M. C. Santos, A. R. Machadom and M. A. S. Barrozo, “Temperature in Machining of Aluminum Alloys,” in Temperature Sensing, Intechopen, 2018.

[19] P. Perrot, A-Z of Thermodynamics, Oxford University Press, 1998.

[20] Calphad, “calphad.org,” [Online]. Available: https://calphad.org/. [Accessed 10 12 2020].

[21] M. Hillert and M. Selleby, “Computerized Thermodynamics for Materials Scientists and Engineers,” Stockholm.

[22] “thermocalc.com,” [Online]. Available: https://thermocalc.com/about-us/methodology/the- calphad-methodology/. [Accessed 10 12 2020].

[23] H. Lukas, G. S. Fries and B. Sundman, Computational Thermodynamics The Calphad Method, Cambridge University Press, 2007.

[24] Thermo-Calc, “thermocalc.com,” [Online]. Available: https://thermocalc.com/products/thermo- calc/. [Accessed 11 12 2020].

[25] W. Zheng, S. He, J. Wang and H. Mao, “Thermodynamic Evaluation of the Co-Al-C System by Coupling Ab Initio Calculations and CALPHAD Approach,” no. 39, 2018.

[26] J. Gröbner, H. L. Lukas and F. Aldinger, “Thermodynamic Calculation of the Ternary System Al-Si-C,” vol. 20, no. 2, 1996.

[27] G. Deffrennes, B. Gardiola, M. Allam, D. Chaussende, A. Pisch, J. Andrieux, R. Schmid-Fetzer and O. Dezellus, “Critical assessment and thermodynamic modeling of the Al–C system,”

Caphad, vol. 66, 2019.

[28] A. T. Dinsdale, “SGTE data for pure elements,” Calphad, vol. 15, no. 4, pp. 317-425, 1991.

[29] P. Wang, W. Xiong, U. R. Kattner, C. E. Campbell, E. A. Lass, O. Y. Kontsevoi and G. B.

Olson, “Thermodynamic re-assessment of the Al-Co-W system,” Calphad, vol. 59, pp. 112-130, 2017.

[30] A. Fernández Guillermet, “Thermodynamic Properties of the Fe-Co-C System,” Zeitschrift für Metallkunde, vol. 79, no. 5, pp. 88-95, 1998.

[31] N. Dupin, Private communication, 2020.

(39)

34

[32] F. Stein, C. He and N. Dupin, “Melting behaviour and homogeneity range of B2 CoAl and updated thermodynamic description of the Al–Co system,” Intermetallics, vol. 39, pp. 58-68, 2013.

[33] N. Saunders, “System Al-Ti,” in COST 507 Thermochemical database for light metal alloys Volume 2, European Communities, 1992, pp. 89-94.

[34] A. V. Davydov, U. R. Kattner, D. Josell, R. M. Waterstrat, W. J. Boettinger, J. E. Blendell and A. J. Shapiro, “Determination of the CoTi congruent melting point and thermodynamic reassessment of the Co-Ti system,” Metallurgical and Materials Transactions A, vol. 32, pp.

2175-2186, 2001.

[35] X. Ma, C. Li and W. Zhang, “The Thermodynamic Assessment of the Ti—Si—N System and the Interfacial Reaction Analysis,” Journal of Alloys and Compounds, vol. 394, no. 1-2, pp. 138- 147, 2005.

[36] L. Zhang, Y. Du, H. Xu and Z. Pan, “Experimental investigation and thermodynamic description of the Co–Si system,” Calphad, vol. 30, no. 4, pp. 470-481, 2006.

[37] Y. Zhang, P. Franke and H. J. Seifert, “CALPHAD modeling of metastable phases and ternary compounds in Ti-Al-N system,” Calphad, pp. 142-153, December 2017.

[38] Q. Chen and B. Sundman, “Thermodynamic assessment of the Ti-Al-N system,” Journal of Phase Equlibria, vol. 19, 1998.

[39] V. T. Witusiewicz, A. Bondar, U. Hecht, J. Zollinger, L. V. Artyukh and T. Y. Velikanova, “The Al–B–Nb–Ti system: V. Thermodynamic description of the ternary system Al–B–Ti,” Journal of Alloys and Compounds, pp. 86-104, April 2009.

[40] Y. Zhang, P. Franke, D. Li and H. J. Seifert, “Lattice stability of metastable AlN and wurtzite-to- rock-salt structural transformation by CALPHAD modeling,” Materials Chemisty and Physics, pp. 233-240, 1 December 2016.

[41] K. Zeng and R. Schmid-Fetzer, “Critical assessment and thermodynamic modeling of the Ti-N system,” Materials Research and Advanced Techniques, vol. 87, no. 7, pp. 540-554, 1996.

[42] X. Ma, C. Li, F. Wang and W. Zhang, “Thermodynamic assessment of the Si–N system,”

Calphad, vol. 27, no. 4, pp. 383-388, 2003.

[43] K. W. Richter and D. Tordesillas Gutiérrez, “Phase equilibria in the system Al–Co–Si,”

Intermetallics, vol. 13, no. 8, pp. 848-856, 2005.

[44] V. Raghavan, “Al-Co-Si (Aluminum-Cobalt-Silicon),” Journal of Phase Equilibria and Diffusion volume , vol. 29, pp. 57-59, 2008.

(40)

35

[45] S.-D. Choi, “Thermodynamic analysis of the Co-Si system,” Calphad, vol. 16, no. 2, pp. 151- 159, 1992.

[46] H. Ohtani, M. Yamano and M. Hasebe, “Thermodynamic Analysis of the Fe-Al-C Ternary System by Incorporating ab initio Energetic Calculations into the CALPHAD Approach,” ISIJ International, vol. 44, no. 10, pp. 1738-1747, 2004.

[47] D. Connétable, J. Lacaze, P. Maugis and B. Sundman, “A Calphad assessment of Al-C-Fe system with the carbide modelled as an ordered form of the fcc phase,” Calphad, vol. 32, no. 2, pp. 361-370, 2008.

[48] N. Dupin and I. Ansara, “Thermodynamic assessment of the system Al-Co,” Rev Metall-Cahiers D Info Technol, vol. 95, pp. 1121-1129, 1998.

[49] P. Dörner, E.-T. Henig, H. Krieg, H. L. Lukas and G. Patzow, “Optimization and calculation of the binary system Al-Si,” Calphad, vol. 4, no. 4, pp. 241-254, 1980.

[50] K. Ishida, T. Nishizawa and S. M. E, “The Co-Si (Cobalt-Silicon) system,” Journal of Phase Equilibria, vol. 12, no. 5, pp. 578-586, 1991.

[51] H. Okamoto, “Al-Ti (aluminium-titanium),” Journal of phase equlibria, vol. 14, p. 121, 1993.

[52] U. R. Kattner and J.-C. C. Y. A. Lin, “Thermodynamic Assessment and Calculation of the Ti-Al System,” Metallugical Transactions A, vol. 23, pp. 2081-2090, 1992.

[53] H. L. Lukas, I. Ansara, A. T. Dinsdale and M. H. Rand, “System Al-N,” in Thermochemical database for light metal alloys, Volume 2, Luxembourg, European Commission, 1998, p. 65.

[54] M. HIllert and S. Jonsson, “An assessment of the Al-Fe-N,” Metakk. Trans A, vol. 23A, p. 3141, 1992.

[55] S. Jonsson, “Assessment of the Ti-N system,” Zeitschrift für Metallkunde, vol. 87, p. 691, 1996.

[56] H. Ohtani and M. HIllert, “A thermodynamic assessment of the Ti-N system,” Calphad, vol. 14, no. 3, pp. 289-306, 1990.

[57] G. Cacciamani, R. Ferro, I. Ansara and N. Dupin, “Thermodynamic modelling of the Co–Ti system,” Intermetallics, vol. 8, no. 3, pp. 213-222, 2000.

[58] H. Seifert, “System Si-Ti,” in COST 507 Thermochemical database for light metal alloys Volume 2, Luxembourg, Eurpoean Commisson, 1998, pp. 266-269.

[59] J. Gao, C. Li, N. Wang and Z. Du, “Thermodynamic analysis of the Ti-Al-N system,” Journal of University of Science and Technology Beijing, Mineral, Metallurgy, Material, vol. 15, no. 4, pp.

420-424, 2008.

(41)

TRITA ITM-EX 2021:72

www.kth.se

References

Related documents

46 Konkreta exempel skulle kunna vara främjandeinsatser för affärsänglar/affärsängelnätverk, skapa arenor där aktörer från utbuds- och efterfrågesidan kan mötas eller

The increasing availability of data and attention to services has increased the understanding of the contribution of services to innovation and productivity in

Närmare 90 procent av de statliga medlen (intäkter och utgifter) för näringslivets klimatomställning går till generella styrmedel, det vill säga styrmedel som påverkar

I dag uppgår denna del av befolkningen till knappt 4 200 personer och år 2030 beräknas det finnas drygt 4 800 personer i Gällivare kommun som är 65 år eller äldre i

If distant shadows are evaluated by integrating the light attenuation along cast rays, from each voxel to the light source, then a large number of sample points are needed. In order

relations generally are of higher quality for groups of congeneric molecules, for example, nitrogen heteroaromatics, [2] it has also been shown that more general relationships exist;

This study aimed to validate the use of a new digital interaction version of a common memory test, the Rey Auditory Verbal Learning Test (RAVLT), compared with norm from

In chemical engineering it is often used as a first estimation of the kinetics of a given reaction and in the pulp and paper industry it has been used in studies of for example kraft