Development of MMC process for high performance aluminium components

(1)

Report No. 2018-008

Development of MMC process for high performance

aluminium components

(2)

(3)

RISE SWECAST Status

Öppen

Projekt nr Projekt namn

1220 Development of MMC process for high performance aluminium components

Författare

Marie Fredriksson

Sammanfattning

Metalmatriskompositer (MMC) är material som kombinerar en metallmatris med keramiska förstärkningar i form av fibrer, partiklar eller flingor. Det finns en leverantör av gjuten MMC i Sverige som förser bilindustri med bromsskivor av aluminium. Materialet har potential att ersätta andra tyngre metaller och därmed bidra till att minska utsläpp från fordon.

Projektet MACS har fokuserat på att förstå materialet och dess svagheter, så väl som att hitta andra lämpliga gjutmetoder och undersöka bearbetningsparametrar då partiklarna försvårar bearbetningsprocesser.

Abstract

Metal matrix composites (MMC) are materials which combine a metallic matrix with ceramic reinforcements in the form of fibers, particulates or flakes. There is one supplier of MMC castings in Sweden providing brake discs in aluminium for automotive industry. The material has the potential of replacing other heavier metals and thereby contributing to lower emissions in vehicles.

The MACS project has aimed towards understanding the material and its imperfections, as well as finding other viable casting methods and examine the machining parameters since the particles aggravates cutting operations.

(4)

(5)

Index

1 INTRODUCTION ... 1

2 MMC – METAL MATRIX COMPOSITE ... 1

3 RESEARCH WORK WITHIN THE PROJECT ... 3

3.1 APPENDIX 1:LITERATURE REVIEW:RECYCLING AND RECLAMATION OF SIC PARTICLES IN ALUMINIUM MMC ... 3

3.2 APPENDIX 2:PERMANENT MOULD CASTING OF MMC ... 3

3.3 APPENDIX 3: SUMMARY OF EXPERIMENTAL PROCEDURES PERFORMED AT AUTOMOTIVE COMPONENTS FLOBY ... 3

3.4 APPENDIX 4:VISCOSIMETRY, THERMO-PHYSICAL PROPERTIES AND RHEOCASTING ... 4

3.5 APPENDIX 5: METALLMATRISKOMPOSITER I MASSTILLVERKADE KOMPONENTER OCH PRODUKTER ... 4

3.6 CASTING SIMULATION OF METAL MATRIX COMPOSITES ... 4

4 IMPLEMENTATION ... 4

5 REFERENCES ... 5

(6)

(7)

1

1 Introduction

This report summarizes the work within the project MACS, Development of MMC process for high performance aluminium components. The project has been divided into workpackages addressing different issues and the results will be attached as appendices.

The project has been funded by the strategic innovation program Lättvikt, a collaboration between VINNOVA, Swedish Energy Agency and Formas.

Participating parties: Automotive Components Floby, Comptech, Fundo Components, Husqvarna, Markaryds Metallarmatur, SecoTools, Lund University, Jönköping University and RISE SWECAST (former Swerea SWECAST).

2 MMC – Metal matrix composite

Metal matrix composites (MMC) are materials which combine a metallic matrix with ceramic reinforcements in the form of fibers, particulates or flakes. MMC began to develop in the 1960´s. Since then the material has gained popularity in several areas due to its excellent properties when it comes to mechanical strength, wear and thermal conductivity for instance. MMC with light metal alloys also provides a high strength to weight ratio.

In castings, alloys of aluminium and magnesium are often used as matrix materials and the reinforcement may consist of for example silicon carbide (SiC), aluminium oxide (Al2O3) or titanium diborides (TiB2). In this project the metallic

matrix consists of the aluminium alloy AlSi9Mg0,6 and the ceramic reinforcements are 20 vol% SiC.

An important factor regarding MMC is wetting. Wettability is defined as the ability of the melt to spread on a solid surface, i.e. how extensive the contact between the melt and particle is, Figure 1 [1]. If the contact angle Θ is equal to zero perfect wetting is achieved and if the contact angle is equal to 180° no wetting exists.

Figure 1 Schematic picture of a drop on a substrate for measuring contact angle [1]

In order to calculate the contact angle the specific energies γ are used according to the Young-Dupree equation [2]:

𝛾𝑠𝑣 = 𝛾𝑠𝑙+ 𝛾𝑙𝑣 cos 𝛩

where γsv is the surface energy between solid and vapor, γsl the surface energy

(8)

2

spreading of this drop on the solid will only occur if the free energy of the system is decreased.

Unfortunately, aluminium always has an oxide layer on the surface due to the affinity to oxygen. This oxide layer is disadvantageous when it comes to wetting behavior. In addition, the particle itself may also have an oxide layer, for instance silicon oxides on SiC particles and also a gas layer covering the particle [1]. Very few ceramics are naturally wet by aluminium, hence requiring assistant in order to increase wetting.

In order to promote wetting there are a few methods that several authors describe; the addition of alloying elements, coating of the particles and heat treatment of the ceramic particles [1,3]. Alloying elements can affect wetting either by promoting chemical reaction between melt and particle or by modification of the oxide covering the particle [3]. Alloying elements can affect the wetting angle positively, for example the use of Lithium for wetting of SiC by aluminium or magnesium [4]. Decreasing the wetting angle will promote reactivity between metal and substrate. Magnesium is a known element that promotes wetting by reducing the interfacial energy of the melt. It reacts with the oxides, SiO2 in the

case of SiC particles, on the particle surface because the high affinity to oxygen thus promotes wetting. In the alloy used in this project magnesium is added at a level of 0,6 wt% in order to promote good wetting.

Al2O3 particles which already are considered to be ideal particles in aluminium

alloys due to good interfacial compatibility however, may react with magnesium forming MgO or spinels (MgAl2O4) degrading the particles [55]. Magnesium is

one of the most important alloying elements for Al-Si casting alloys since it is a strengthening constituent that enables heat treatment of the matrix material, precipitating Mg2Si.

Coating of the particles is another viable method. The purpose is to cover the particles with a coating that will react with the matrix and create good wetting. The coating may also remove the gas layer covering the particle [6]. It is also common to heat treat the particles in order to remove impurities and alter the oxide layer on the surface of the reinforcements [1]. The oxide layer on SiC particles also prevent the particle to react with the aluminium forming aluminium carbides, Al4C3. The Al-SiC system is thermodynamically unstable resulting in

[77]:

4 Al+ 3 SiC → Al4C3 + 3 Si

SiC is thermomechanically unstable in aluminium in temperatures above 1000 K (ca 726 °C) [55]. Casting however, is often performed at 740 °C and even a bit higher. The reaction above may be depleted if the silicon level in the aluminium alloy is increased.

Wetting is influenced by temperature as shown in xx where the contact angle has been measured after 15 min in different temperatures for the Al-SiC system [88].

(9)

3

Figure 2 Measured contact angle θ after 15 min at different temperatures for Al-SiC [8]

3 Research work within the project

The results from this research project are attached to this overview. This is a short summary of its content.

3.1 Appendix 1: Literature review: Recycling and reclamation of

SiC particles in aluminium MMC

This review gives a summary of how MMC materials can be recycled and how the particles and the metal may be reclaimed. Recycling of MMC is not really a problem. Melt cleanliness is however of great importance since regular cleansing fluxes may not be used since this would be detrimental to the wetting of the particles.

There are several ways of separating the particles from the melt, either mechanically or chemically making it possible to reuse the constituents for other purposes.

3.2 Appendix 2: Permanent mould casting of MMC

Pressurized casting processes should give better properties than gravity casting due to faster solidification and the possibility to compress porosity. This report summarizes trials on tilt casting of MMC with post-processing as HIP and heat treatment. HIP process of permanent mould castings has the possibility of exceeding the properties for squeeze cast MMC. It is however an expensive process. The combination of heat treatment and HIP did not show the potential and further work should be carried out in order to optimize the processes.

3.3 Appendix

3:

Summary

of

experimental

procedures

performed at Automotive Components Floby

At Automotive Components Floby there is a process for producing MMC from AlSi9Mg0,6 and SiC particles, which has been used in this project. The facility is a prototype foundry where brake discs have been developed and produced.

(10)

4

In the MACS project several demonstrators have been produced in order to try the technique and the material in other applications than the brake disc; a connecting rod for a chain saw and a carrier disc for a cutting machine.

3.4 Appendix 4: Viscosimetry, thermo-physical properties and

rheocasting

Material characterization of MMC has been done within the project. Thermo-physical properties are presented as well as trials with the rotational bob method for measuring viscosity. Rheocasting, which is a semisolid casting method, has been performed and compared to liquid high pressure die casting.

3.5 Appendix 5: Metallmatriskompositer i masstillverkade

komponenter och produkter

This report presents work conducted in forging on cast raw material as well as machining parameters for MMC. There has also been a study on the economical aspects of producing a brake disc in MMC by Monte Carlo simulations. A master thesis has been presented during the project – Masstillverkning av komponenter i MMC baserat på gjutning och varmsmide by Kristoffer Lundkvist and Daniel Jarlmo-Måård, Lund University. A scientific paper on machining has been published [9].

3.6 Casting simulation of metal matrix composites

The part on simulation is the subject of submission of scientific publication and will therefore not be attached to this report. This is a short summary of the work: In this project the possibilities of simulating gravity casting of MMC in two different commercialized softwares for foundry industry; Flow3D CAST and ProCast, have been investigated. The experimental work was based on high speed filming of the casting process in a specially designed casting mould with a plate of quartz glass mounted. A common aluminium casting alloy was used as a reference to the particle reinforced alloy with 20 vol% SiC.

Different viscosity models of how the particles affect melt behavior were analyzed. Flow 3D CAST was used to validate the model and ProCast to simulate the model based on the parameters calibrated in Flow3D CAST. The results showed that some of the parameters describing surface conditions and the contact between melt and mould, as contact angle and surface roughness, influence the results to a great extent. These parameters were calibrated in order to get consistency between experiment and simulation both for the alloy with and the alloy without particles. Acceptable correlation between experiment and simulation was achieved after calibration.

A Master thesis has been conducted in this part: Modelling and Simulation of Mold filling in gravity casting of Aluminium and MMC alloys by Akhil Manne and Pramod S Hiregoudra, Jönköping University.

4 Implementation

The results from this project have shown that it is possible to produce high performance aluminium components which could replace heavier metals like for

(11)

5

example steel or iron. This will be reported in publications as well as in industrial education in order to disseminate these results.

It is however an expensive process today which needs to be optimized. Automotive Components Floby has shown that they can produce brake discs of sufficient quality in a prototype machine. In order to improve production economy however, it is important to adjust this process for serial production. Calculations according to Monte Carlo method performed by Lund University have shown that it is possible to decrease production cost of brake disc substantially.

5 References

1. Hashim J. et al, The wettability of SiC particles by molten aluminium alloy, Journal of Materials Processing Technology 119 (2001) 324-328 2. Makkonen L., Young´s equation revisited, J. Phys. Condens. Matter 28

(2016)

3. Mortensen A et al, Solidification Processing of Metal-Matrix Composites, Journal of Metals, February, 1988

4. Mortensen A., Interfacial phenomena in the solidification processing of metal matrix composites, Materials Science and Engineering, A135 (1991) 1-11

5. Pai B.C. et al, Role of magnesium in cast aluminium alloy matrix composites, Journal of materials science 30 (1995) 1903-1911

6. Zhang M.X., Crystallographic study of grain refinement in aluminum alloys using the edge-to-edge matching model, Acta Materialia 53 (2005) 7. Sritharan T. et al, A feature of the reaction between Al and SiC particles in

an MMC, Materials characterization 47 (2001) 75-77

8. Laurent V. et al, Wettability of SiC by aluminium and Al-Si alloys, Journal of material science 22 (1987) 244-250

9. Bushlya V. et al, Performance and wear mechanisms of novel superhard diamond and boron nitride based tools in machining Al-SiCp metal matrix composite, Wear 376-377 (2017) 152-164

6 Appendix

Appendix 1: Literature review: Recycling and reclamation of SiC particles in aluminium MMC

Appendix 2: Permanent mould casting of MMC

Appendix 3: Summary of experimental procedures performed at Automotive Components Floby

Appendix 4: Viscosimetry, thermo-physical properties and rheocasting Appendix 5: Metallmatriskompositer i masstillverkade komponenter och produkter

(12)

Swerea SWECAST AB Status

Öppen

Project No. Project name

1220 Development of MMC for high performance alumiminium components

Author

Utgåva

Date

Marie Fredriksson 2017-02-06

Address Telephone Fax E-mail

Swerea SWECAST AB Nat 036-30 12 00 Nat 036-16 68 66 swecast@swerea.se P O Box 2033 Int +46 36-30 12 00 Int +46 36-16 68 66 www.swereaswecast.se

SE-550 02 Jönköping, Sweden

Literature review: Recycling and reclamation of SiC

particles in aluminium MMC

Summary

This review summarizes the different techniques and processes available for recycling and reclamation of metal matrix composites. Recycling of MMC:s means processing the material and reuse it as a composite. Once the composite is too contaminated in order to be reused, it can be reclaimed and thereby still have some value. Reclamation is the process of separating the reinforcement from the matrix and possibly use the constituents for other purposes.

Sammanfattning

Denna litteratursökning sammanfattar olika tekniker och processer för att återanvända och återvinna metallmatriskomposit. Återanvändning i det här fallet är synonymt med att man processar materialet och återanvänder det som komposit. När kompositen är för förorenad för att återanvändas kan den återvinnas och därmed inneha ett fortsatt värde. Vid återvinningen separeras förstärkningsmaterialet från matrisen och möjligen används beståndsdelarna till andra syften.

(13)

Swerea SWECAST AB

2

1 Recycling of aluminium

Aluminium is the most common metal on the face of the earth. There is no lack of aluminium, but to produce aluminium from bauxite by electrolysis is highly energy-intense and therefor harmful to the environment. To produce aluminium from scrap metal, that is recycled materials, only takes 5 % of the energy required to produce aluminium from raw material. It is therefore of great importance to always consider recycling when working with components of aluminium.

According to EU:s End of life vehicles directive 95 % of a vehicle shall be recyclable from 2015. This is an increase of 10 % since the directives from 2007 [1]. Considering the limited resources on earth it is most likely that more products will get similar directives.

The scrap may come from EOL (End of life) products, but also from foundry scrap within the foundry.

1.1 Recycling and reclamation

Recycling of MMC:s means processing the material and reuse it as a composite. Once the composite is too contaminated in order to be reused, it can be reclaimed and thereby still have some value. Reclamation is the process of separating the reinforcement from the matrix and possibly use the constituents for other purposes.

2 Reclamation

2.1 Separation of particles from matrix metals

Reinforcements are stable within composites. Separation is only possible if there is a change in energy at the interface between reinforcement and matrix metal. When it comes to MMC the term wetting is readily recurrent. Wetting is the process of when a fluid is covering another fluid or a solid substrate, in this case a solid substrate.

Figure Schematic picture of wetting (Wikimedia)

The schematic picture shows a droplet on a solid substrate. If wetting is good, the contact angle Θc is small and vice versa. The arrows show the surface energies

between the different materials; LG is liquid/gas, SG is solid/gas and SL is solid/liquid. When it comes to MMC and separation of particles, the energies can be changed either mechanically or chemically.

(14)

Swerea SWECAST AB

3

Matrix metal and reinforcements have their own microstructures; hence they are not mixtures as an alloy or chemical compound. Therefore the separation of the particles from matrix metal only requires separation of the interfaces [2].

Reinforcements in air are more stable than reinforcements in metals, hence there is a driving force to separate the particles if some external effort is added. If separation occurs the separation will most likely start at an exposed particle at the surface of the composite.

2.1.1 Mechanical separation

Mechanical separation is done in the melted phase, but is not as common as chemical separation. There are a couple of processes that are possible. Squeezing the particles out from the composite will leave some remaining matrix metal in the compressed particles. Filtering the particles from the melt is another viable method. Some processes rely on gravity for separation of the constituents.

2.1.2 Chemical separation

Chemical separation is similar to fluxing processes used to remove slag and oxides from conventional aluminium melt. A material which can create an interface of lower energy to the particle than the matrix is required. These materials will at mixing infiltrate at the interfaces and replace the matrix metal at the interface [2]. The material used needs a low solubility in the metal so that the separation may begin. These materials are molten salts and alkali metals, like for example sodium Na, potassium K, sodium chloride NaCl and potassium chloride KCl. These salts can effectively remove ceramic particles by wetting them, but the salt itself is not wetted by the liquid aluminium [3].

D.M. Schuster et al have investigated two types of processes, injection with gas via impeller and rotary salt furnace [4]. In the first process inert argon gas was injected into the melt by a graphite shaft with a graphite impeller. Salt was added to the melt and at the same time argon was flushed through the melt, bringing all particles to the surface. Complete dewetting was achieved after 10-20 min in a 300 kg furnace [4]. The other process investigated is a technique applicable for large volume recovery. The investigation was made in a 15 tonnes rotary salt furnace. Overall recovery of aluminium was 80 % [4].

3 Recycling of MMC

Depending on where the scrap origins from, the recycling process may vary. Considering direct cast scrap and foundry scrap up to a certain degree of dirty material, the metal may be remelted using different techniques [7]. Recycling of MMC for direct reuse as its original form is the main cost driver and should be considered as primary recycling. If this is not possible, the reclamation process is to be considered [7].

3.1 Aluminium carbides

Metal matrix composites with particles can be recycled as a product, for example as conventional returns into the melt. Recycling means, in opposite to reclamation, that the material is reused as composite. In order to get an economical process most foundries need to remelt the scrap in-house and that is why there has to be a process for it. Care must always be taken during melting

(15)

Swerea SWECAST AB

4

considering the melt temperature since the formation of aluminium carbides by the reaction:

4Al + 3SiC ↔ Al4C3 + 3Si [4]

The formation of aluminium carbides may be controlled by controlling time, temperature and silicon content. The problem will enhance with lower silicon levels. Above 9 wt% silicon the formation can be avoided at regular temperatures for foundries [4].

3.2 Recycling process from Duralcan

At melting, conventional fluxes cannot be used to clean the melt since it would separate the particles from the matrix materials similar to what happens at chemical separation at reclamation. Duralcan, one of the companies producing MMC:s, has developed a practice for fluxing and degassing of MMC:s without dewetting the particles [4]. According to this process virgin material is added to the furnace and kept at 700 °C and argon gas is used at a flow of at least 0.12 l/min. If there is enough room scrap may be added as well at this time. When the melt starts to sink continue to add scrap metal. The melt is supposed to rest for 40-60 min in order to let dross rise to the surface. Skim the surface. Now it is time to agitate the melt in order to distribute particles that have set in the furnace with a dry paddle. Start mechanical stirring at 300-450 rpm. Do not create a vortex, i.e. a swirl. If there are particle remaining at the bottom, use the paddle to agitate them into the impeller. Set the argon gas at 2 l/min and insert the wand in the melt. Increase to 4 l/min and flux for 20 min. These parameters can vary according to the authors. Remove wand and stop the agitator. Let the melt rest for 40-60 min, skim the melt and start the agitator for 10 min. After this procedure the melt is ready to use [4]. This procedure is stated for a melt of 135 kg.

3.3 Influence of recycling on the material

Understanding what influences the physical and mechanical properties of MMC:s is a challenge, since they are sensitive to several factors – type of reinforcement, manufacture, recycling history, re-melt parameters like temperature, holding time and number of times of recycling and so on [6].

Ref. [5] has conducted investigations on the recycling process on selected properties as conductivity, density, hardness, tensile strength and microstructure. The material was recycled up to 9 times and every time tests were made. There was no evident change in microstructure, but significant change in properties. Elongation was reduced by 25 %, electric conductivity by 30 % and hardness by 10 % for example. According to the article no melt treatment was conducted on the material. Test samples and ingots for further re-melting were cast. The authors open up for improved recycling process by proper selection of liquid metal treatment [5].

Fan et al presented a study of the chemical reaction between silicon carbide particles and the matrix during multiple re-melting using DSC (Differential Scanning Calorimetry) [6]. A comparison between pure Al, 5wt%Si and Al-8wt%Si was conducted when it comes to re-melting at different temperatures. The material was produced using infiltration of preforms. Some conclusions are that the interfacial reactions increase with temperature, but when a certain degree of interfacial reaction has occurred it eventually cease. It was also discovered that when 8 % Si was added, the chemical reaction between Al and SiC was completely suppressed even at 1050 °C. The authors states that by controlling the

(16)

Swerea SWECAST AB

5

temperature and if possible using an alloy with a silicon level above 8 %, no reaction will occur [6].

4 Referenser

[1] ELV directives, European Union

[2] Y.Nishida, Introduction to Metal Matrix Composites

[3] V. Kamavaram et al, Recycling of aluminum metal matrix composite using ionic liquids: Effect of process variables on current efficiency and deposit characteristics, Electrochimica Acta 50 (2005) 3286-3295

[4] D.M. Schuster et al, The recycling and reclamation of metal-matrix composites, JOM, may 2003 (Duralcan)

[5] A. Klasik et al, Effect of multiple remelting on selected properties of dispersed reinforced aluminum matrix composite, Journal of KONES Powertrain and Transport Vo. 15, No 3, 2008

[6] T. Fan et al, Chemical reaction of SiCp/Al composites during multiple remelting, Composites: Part A 34 (2003) 291-299

[7] Y. Tang et al, Recycling of composite materials, Chemical Engineering and Processing 51 (2012) 53-68

(17)

Permanent mould casting of MMC

Marie Fredriksson, RISE SWECAST and Bo Mattsson, Fundo Components

(18)

(19)

RISE SWECAST Status

Open

Project No. Title

1220 Minskad materialanvändning vid tillverkning av gjutna komponenter

Sammanfattning

Gjutmetoder för metallmatriskompositer (MMC) idag är framför allt trycksatta metoder för att minimera porositet och säkerställa god distribution av partiklar som fås pga. fin mikrostruktur vid snabb stelning.

Gravitationsgjutning är dock en enkel och effektiv process för att producera gjutgods och därför har begränsade försök utförts med vaggjutningsprocessen för att verifiera genomförbarheten.

Då porositet kan vara ett problem i gravitationsgjutningsprocesser för MMC har även HIP, hot isostatic pressing, utförts för att undersöka potentialen för materialet.

Prover från HIP uppvisar mycket bättre brottförlängning an gravitationsgjutna prover i gjuttillstånd så väl som prover från ”squeeze casting” (tryckgjutning). Att använda värmebehandling är en bra metod för att öka brottgräns och sträckgräns, men brottförlängningen är fortsatt låg. En kombination av HIP och T6 undersöktes, men då värmebehandlingsprocessen inte var optimerad i detta projektet kunde inte den fulla potentialen bekräftas.

Abstract

Casting methods used for metal matrix composites (MMC) today are foremost pressurized methods in order to minimize porosity defects and ensure good distribution of particles due to fine microstructure at fast solidification.

Gravity casting is however simple and effective processes to produce castings and therefore limited trials have been conducted in tilt casting process to verify the feasibility.

Since porosity may be a problem in gravity casting processes for MMC also HIP, hot isostatic pressing, has been conducted to investigate the potential of the material.

HIP samples show much better elongation to fracture than as cast gravity cast samples as well as squeeze cast samples. Using heat treatment is a good method to increase UTS and YS, but elongation is still low. A combination of HIP and T6 was investigated, but since the heat treatment process was not optimized in this project the full potential was not confirmed.

(20)

(21)

Index

1 INTRODUCTION ... 1 2 EXPERIMENTAL PROCEDURE ... 1 3 RESULTS... 3 3.1 MECHANICAL PROPERTIES ... 3 3.2 METALLOGRAPHIC EXAMINATION ... 4 3.2.1 Microstructure ... 4 3.2.2 Fracture surfaces ... 5

3.3 FLUIDITY USING LOOP ... 11

4 CONCLUSIONS... 12

5 REFERENCES ... 12

Appendix

(22)

(23)

1

1 Introduction

Casting methods used for metal matrix composites (MMC) today are foremost pressurized methods like squeeze casting. This facilitates the possibility of producing casting with low porosity content and thereby higher quality. This project has investigated the possibility of using non pressurized casting methods like gravity casting in permanent moulds and the effect of post processing in order to increase mechanical properties.

2 Experimental procedure

A mobile tilt casting machine was used to produce samples from AlSi9Mg0,6 reinforced by 20 vol% silicon carbide particles (SiC). The material was produced by melting AlSi9Mg0,6 ingots in an electrical heated furnace. Melting temperature was approximately 730 °C. SiC particles were heated and mixed with nitrogen during 1 hour. 20 vol% of SiC were then stirred into the melt with a device forcing the particles into the melt. After incorporating the particles another stirring device was used to keep the particles from settling due to higher density of the particles compared to the melt aluminium. The melt was then poured from the melting furnace to the holding furnace. Casting temperature is normally 745-750 °C.

During the tests there were also measurements on fluidity using the patented device “LOOP” from Bryne AB, Sweden, Figure 1.

Figure 1 LOOP device from Bryne AB, Sweden [1].

The LOOP itself is made of an alkaline earth silicate wool and will therefore hold temperature for some time. It is equiped with a plug which will be removed when the melt in the cup has reached correct temperature. The melt will flow through the channel and reach a number of steps before it solidifies, Figure 2.

(24)

2

Figure 2 Schematic picture of LOOP where the melt flows in a spiral reaching a number of steps[1].

The amount of steps reached depends on the cleanliness of the melt.

The geometry used for casting was a staircase model providing different solidification rates and thereby different microstructures. The test samples were cut out according to the red arrows in Figure 3.

Total length of the geometry was 400 mm and the different steps were 50, 120, 90 and 60 mm in length and 55, 35, 15 and 5 mm in height respectively.

The geometry of the test bars cut from the geometry is seen in Figure 4.

Figure 4 Test bar geometry

1.1 _1.2

1.3

1.4

(25)

3

Some of the samples were subjected to HIP (hot isostatic pressing) in order to decrease the amount of porosity. The HIP process uses elevated temperature and isostatic pressure from an inert gas, often Argon, in order to dense metal material by pressing porosity. The treatment was conducted by Bodycote in Germany according to the Densal process. Parameters are proprietary to Bodycote and have not been revealed. However, normal parameters when HIPing aluminium are 2-6 hrs at 510-521°C and an applied pressure of 103 MPa [2].

Heat treatment according to T6 was performed in a furnace from Nabertherm (Nabertherm Air Circulation FurnaceN120/65HA). Times and temperatures used were:

• Solution heat treatment: 520 °C during 6 hrs • Ageing: 180 °C during 2,5 hrs

DSC measurements were performed in order to find correct solution heat treatment temperature. It was then decided to use 520 °C to ensure that incipient melting would not occur. Heat treatment was not in any way optimized in this project, but was conducted to show the potential of the material.

Some of the HIP samples were also heat treated in order to investigate the combined effect of these post processes. In order to get full potential of the combination of HIP and heat treatment, time and temperature of T6 should be tested and optimized. Brunner et al [3] investigated the combination of HIP and heat treatment on aluminium alloy AlSi7Mg0,3. It was found that the best combinations of good elongation to fracture and tensile strength was achieved using HIP+T6, and the second best was HIP+T4. The parameters used for heat treatment was 540 °C in 2 hrs when solution annealing and 165 °C in 2,5 hours when ageing. This was not MMC though.

The tensile testing was performed according to SS-EN ISO 6892-1:2016 A222 on a Zwick Z050TH.

3 Results

3.1 Mechanical properties

The following graph shows ultimate tensile strength and elongation to fracture for two samples from each process, Figure 5. The samples tested are from thicknesses 35 and 5 mm respectively from the cast geometry. Since the samples are few, the certainty of the values cannot be confirmed in this project but should be further analyzed in another project where sample geometry as well as heat treatment data is optimized.

As cast and heat treated samples show low elongation to fracture and often break before the yield strength is reached. HIP samples reach yield strength and the elongation is somewhat better probably due to decrease in porosity.

Rio Tinto Alcan, the producer of the MMC Duralcan, presents ultimate tensile strength after T6 heat treatment for the alloy F3S.20S of 317 MPa and elongation of 0,4% [4]. The Duralcan samples are cast-to-size test bars, solutionized at 538 °C in 8 hrs and aged at 154 °C in 5 hrs. The alloy contains 8,5-9,5% Si, 0,4-0,6% Mg and 20% SiC, and are thereby quite similar to the alloy examined in this

(26)

4

project. Samples subjected to HIP and heat treatment reach similar values as Duralcan material, which shows the potential of heat treated gravity cast components with less porosity.

Figure 5 UTS versus elongation

The markings for squeeze casting respond to as cast samples cut from a squeeze cast component of approximately 20 mm thickness in the same material and are only for comparison. They seem to correspond to as cast 5 mm permanent mould samples.

3.2 Metallographic examination

3.2.1 Microstructure

Figure 6 shows the comparison between sample from thin and thick section respectively. The SiC particles are being pushed in front of the dendrites during solidification which gives a slightly more inhomogenous distribution of the particles in thick sections that solidifies more slowly.

Figure 6 Samples from thin (5 mm) and thick (35 mm) section respectively (55X) 0 50 100 150 200 250 300 350 0 1 2 3 4 Rm (M Pa) A (%) As cast 35 mm As cast 5 mm HIP 35 mm HIP 5 mm T6 35 mm T6 5 mm HIP+T6 35 mm HIP+T6 5 mm Squeeze cast

(27)

5

It can be seen that the dendrites (light areas in structure) are larger in the thicker sections since they grow during solidification, Figure 7.

Figure 7 Samples from thin (5 mm) and thick (35 mm) section respectively (220X)

Figure 8 shows how the particles are distributed between the dendrites.

Figure 8 Micrograph showing particles between dendrites (220X)

3.2.2 Fracture surfaces

The fracture surfaces have been investigated in stereomicroscope and scanning electron microscope (SEM) in order to identify the cause of fracture since the results differed between the two samples.

As cast 5 mm

Figure 9 shows a comparison of as cast samples where the left one performed better than the right one which did not even reach yield strength before fracture. A large area of the sample has a smooth, shiny surface which might be caused by oxides.

(28)

6

Figure 9 As cast sample where the right sample fractured before reaching yield strength.

Using SEM it can be seen that the left sample above contains some shrinkage porosity, Figure 10, while the right sample contains large deformed surface with oxide film, Figure 11.

Figure 11 Right sample above which did not reach yield strength showing large surface defects of cracks and oxides (SEM).

As cast 35 mm

None of the samples from as cast 35 mm reached yield strength. One of the samples shows a large cracked area in the fracture surface, covered with oxide, Figure 12.

Figure 10 Left sample above which reached yield strength showing shrinkage porosity (red markings) (SEM).

(29)

7

Figure 12 Cracked surface causing fracture (SEM).

The second sample shows similar defect, cracks caused by oxide films.

Figure 13 Cracks in fracture surface, covered by oxide films (SEM).

HIP 5 mm

These samples performed best of all samples concerning elongation to fracture. Figure 14 shows one sample with a small defect (183 µm) which probably is porosity not closing totally during HIP process. Both samples showed similar defects.

Figure 14 Small defefct in HIP sample showing good elongation to fracture (SEM)

(30)

8

HIP 35 mm

One of the samples has somewhat lower elongation to fracture than the other HIP samples. A larger flat surface in one of the corners caused the fracture, Figure 15. This area was probably created by oxide film.

Figure 15 Flat surface causing fracture in the sample (SEM)

The other sample shows an oxide defect as well. It is smaller which results in a better elongation.

(31)

9

T6 5 mm

Heat treated samples show high ultimate tensile strength but poor elongation. One each of the samples from thick and thin area respectively did not reach yield strength.

Figure 17 shows a crack which might have caused the failure in one of the samples. This sample did reach yield strength

Figure 17 Large crack in fracture surface (SEM)

The sample not reaching yield strength contains shrinkage porosity as well as cracks in the fracture surface, Figure 18.

Figure 18 Cracks and shrinkage porosity in fracture surface (SEM)

T6 35 mm

The sample not reaching yield strength contains smooth surfaces with oxides as well as cracks in the fracture surface, Figure 19.

(32)

10

Figure 19 Left sample showing crack in fracture surface. Right sample showing flat surface from oxides (SEM)

The second sample shows some smaller shrinkage defects as well as oxide defects, Figure 20.

Figure 20 Left sample showing shrinkage porosity and right sample showing surface with oxides (SEM)

HIP and T6 5 mm

The thicker samples performed somewhat better than the thinner samples when it comes to heat treated HIP samples. Figure 21 shows the sample not reaching yield strength. A larger oxide defects has formed close to the surface of the sample.

(33)

11

HIP and T6 35 mm

The other thin sample shows porosity and clustering particles.

Figure 22 Porosity in sample fracture surface (SEM)

HIP and T6 35 mm

The sample showing best elongation contains very little defects, but show clustering of particles, Figure 23 (left).

The other sample shows oxide defects as well as clustering, Figure 23 (right).

Figure 23 Left sample showing clustering of SiC particle and right sample showing oxide defect (SEM)

3.3 Fluidity using LOOP

The fluidity test in LOOP is summarized in Table 1.

Fluidity tests should not be used as a quantitative measure but it is obvious that the fluidity decreases with time during these trials, i.e. fewer steps are being filled in the LOOP. More testing should be conducted in order to see if it is increasing amount of oxides or if the SiC particles somehow change the viscosity of the melt.

(34)

12

Table 1 Results from LOOP fluidity test

Sample No. Time No. of steps in LOOP 1 10:00 16,8 2 14 3 13,2 4 11:20 12 5 10,5 6 11,8 7 13:45 8,8

4 Conclusions

Casting MMC in non-pressurized processes shows to be a feasible method. It is however important, as always in castings, to optimize the geometry and use a clean melt of good quality. Samples with the lowest elongation to fracture show extensive oxide defects. It is not possible to clean MMC melts the conventional way with fluxing since this would be detrimental on the wetting effect between melt and particle. Good routines for melt cleanliness and avoidance of contact with air is therefore of great importance.

Thinner castings with faster cooling are preferable in order to ensure that particle distribution is as good as possible.

HIP is a good, but of course quite expensive, way of reducing porosity to get better mechanical properties in the casting. Tiryakioglu and Campbell [2] stated that bifilms (folded oxide films) cannot be closed to a satisfactory extent and therefore UTS and YS may increase, but not the elongation. In our investigations elongation was actually readily improved. The cause of the fractures was indeed foremost oxide films. Combining HIP with heat treatment is possible but needs to be optimized concerning times and temperatures.

5 References

1. Bryne AB, Sweden (www.bryne.se)

2. Tiryakioglu M., Campbell J., The effects of HIP on bifilms in aluminium castings, World Foundry Congress 2006

3. Brummer M. et al, Heat treatment of aluminum castings combined with hot isostatic pressing, Proceedings of the 12th international conference on aluminium alloys, 2010, pp. 1095-1100

4. Alloying aluminium with high technology, Duralcan brochure, Rio Tinto Alcan

(35)

Dokumentnamn

MACS , slutrapport AC Floby

Utfärdare (avdelning, namn, user-id) Datum

2018-09-11

Sida

5100 , Stefan Kristiansson , skristi1 1 (8)

Sökväg

Ref: Dokument-ID: 0625 Version: 2

SUMMARY OF EXPERIMENTAL PROCEDURES PERFORMED AT AUTOMOTIVE COMPONENTS FLOBY

Index

1 INTRODUCTION ... 1 2 EXPERIMENT ... 1 3 RESULTS AND DISCUSSION ... 4 4 CONCLUSIONS ... 8

1 Introduction

Swedish industry faces a major challenge in meeting future and multifaceted competition. Requirements such as finding materials that result in weight saving, reduced emissions, resource use etc. become clear.

In this project we have focused on using MMC, i.e. a particle reinforced (SiCp) aluminium alloy.

Samples of this material have been made out of brake discs for passenger cars with good results. Strength, weight reduction (up to 50%), fuel saving / reduced CO2 emissions and

decreased particulate emissions, have been some of the focusing terms. With this as a basis, there would be conditions for finding more applications.

Historically, the production of brake discs has been done with Squeeze casting. We want to try and evaluate alternative methods such as for example forging.

2

Experiment

Carrier disc, connecting rod

Together with the project participants (Husqvarna) we have identified 2 demonstrators, Figure 1:

- Carrier disc for cutting machine - Connecting rod for chain saw

(36)

Dokumentnamn

2018-09-11

Sida

Both products are linked to handheld products where weight and performance are significant properties.

The first raw parts have been obtained by gravity casting in simple sand moulds.

The carrier disc has been cast with a certain machining allowance after which it has been processed immediately.

The connecting rods were cast with a machining allowance adapted for subsequent forging operation and machining. The purpose of the forge is to minimize the effect of any pores or inclusions in the material as well as giving the product its necessary strength properties.

Multi tool MMC brake disc

We have continued to develop our existing concept i.e. Squeeze cast MMC disc. We have developed a multitool where we have the ability to cast raw parts up to about diameter 410 mm, Figure 2.

Figure 2 Multi tool developed within the project

(37)

Dokumentnamn

2018-09-11

Sida

Ventilated brake discs

Another step in this development is to try out ventilated discs.

Samples have been produced in 3D printed sand molds, partly with 3D printed sand cores to be tested in our existing casting machine, Figure 4 and Figure 5.

Figure 4 3D printed sand moulds for ventilated brake discs

Figure 5 3D printed cores for ventilated brake discs

MMC, thermal and mechanical properties

In order to get a deeper knowledge of our material, we have, together with Swerea Swecast, identified and put together a sample matrix where mechanical and thermal strength tests have been conducted, Table 1.

(38)

Dokumentnamn

2018-09-11

Sida

Table 1 Test matrix for material properties of MMC

We have also identified a number of additives to try in the future with the purpose of increasing the material's MOT (Maximum Operating Temperature). These tests will be performed after the project has ended.

In order to quickly verify the material samples a tool in which we can squeeze the raw parts of test rods has been developed, Figure 6.

Figure 6 Model for test rods

3 Results and discussion

Carrier disc

Raw parts for carrier discs have been produced.

Shape and visible outcome of carrier disc look good. Weight reduction of 35-40% should be possible.

Material analysis show that a small sedimentation of the SiC particles has been observed, but the distribution of particles is consistent and good.

(39)

Dokumentnamn

2018-09-11

Sida

However, estimates show that the price for an MMC carrier disc compared to the current carrier disc from steel is too high.

Connecting rod

Raw parts for forging crankshaft have been produced for verifying mechanical properties. Some variants have been produced; sand cast raw parts as well as water cut materials from squeeze cast brake discs in order to ensure optimum material precision, Figure 7.

Figure 7 Raw parts for connecting rods

Some forging tests have also been conducted, Figure 8.

Figure 8 Forged connecting rods

Husqvarna has done a number of tests on mechanical properties, i.e. fatigue tests, which show that the MMC samples currently do not meet the required strength requirements.

Ventilated brake discs

This was the first test to produce ventilated discs. The test has been done in a 3D printed sand mould, Figure 9.

(40)

Dokumentnamn

2018-09-11

Sida

Figure 9 Ventilated brake discs

The next step is to test with 3D printed sand cores in our casting machine in the squeeze casting process.

(41)

Dokumentnamn

2018-09-11

Sida

Thermal and mechanical properties – results

(42)

Dokumentnamn

2018-09-11

Sida

Mould/tool for raw parts test rods

First casting and material tests have been conducted. Further casting and optimizations are necessary to ensure that we can produce test bars with the correct quality.

Figure 10 First trials with test rod tool

4 Conclusions

Interest of saving weight, finding durable materials and reducing particulate emissions is an upward trend. There would therefore be a good future for increased demand for products in MMC materials.

It is necessary to try to develop the thermal and mechanical properties of the MMC material to fit into a wider spectrum.

Furthermore, it is important to continue the development of alternative methods of

production, including the production of materials (casting methods, forging etc.) as well as processing (optimization of cutting tools, cooling methods, etc).

(43)

Viscosimetry, thermo-physical properties and rheo

casting of MMC

Index

1 INLEDNING/INTRODUCTION ... 1 2 EXPERIMENT WORK AND RESULTS ... 1 2.1 VISCOSIMETRY ... 1 2.2 THERMO-PHYSICAL PROPERTIES ... 2 2.3 RHEOCASTING AND HPDC ... 3 2.4 MECHANICAL TESTING ... 7

2.4.1 Tensile testing ... 7 2.4.2 Compression workability testing ... 8

2.5 ADDITIONAL WORK ... 12

(44)

1

1 Introduction

In the current section the work performed at Jönköping University for the Vinnova funded project MACS is summarised.

The experimental method and results are described in a sequential order starting with material charachterisation, followed by casting experiments and a study of the mechanical charactersisation. Finally the additional work made for the project is presented.

2 Experiment work and results

2.1 Viscosimetry

Efforts were made to measure the viscosity of the Al-SiC slurry. The method use was the rotating bob method. The conventional bobbin is a cylinder but for a slurry this will measure wall slip instead of the internal friction that viscosity represents. To measure this a bobbin with a symmetric cross as cross-section was developed and used with the intent of creating a bobbin made of the slurry, figure 1. (It should ne noted that efforts were made using both the conventional bobbin and this cross shaped bobbin.) The measurements gave values that would represent viscosity but with a very large uncertainty 100-300%, as well as a drift with time for higher rates of rotation, figure 2.

The reason for this large variation was not uniquely determined but it is likely due to slurry inhomogeneity that are difficult to overcome due to the time-scale of the measurement (slow heating and long measurement times). This was also suggested by the weak drift towards higher viscosity at longer measurement durations for the higher rate of rotation that may indicate improved mixing. Based on this the standard estimated is equally good to use for the slurry and the use of the measured data does not add to quality of simulations. This effort for additional tests was thus abandoned.

(45)

2

Figure 2. The torque signa from the rotating bob measurements showing variations in the range 100 to 300% from the mean as well as drift at higher rotational rates with time.

2.2 Thermo-physical properties

The thermo-physical properties for the brake disc material were determined for the Al-20%SiCp material provided by AC Floby.

Thermal expansion was roughly linear and constant with temperature and of the magnitude 22x10-6 [m/(m K)], figure 3.

Specific heat followed the conventional appearance, figure 4, for an aluminium alloy. Being a cast material there is always some equilibrations from the as cast stat between 200°C to 400°C due to segregation from solidification.

Thermal diffusivity drops with increasing temperature, figure 4.

Thermal conductivity, that is a calculated entity based on density, thermal expansion, specific heat and thermal diffusivity, shows variations with temperature, but these are influenced by the homogenisation processes and only a weak trend towards increased conductivity with temperature can be noted, figure 5.

Figure 3. Thermal expansion as function of temperature. Equation in the figure is the equation for the fitted linear trendline giving the thermal expansion coefficient for the Al-20%SiCp material

0 5 10 15 20 0 100 200 300 To rq u e (% ) Time (sec)

100 RPM 800C

0 5 10 15 20 0 50 100 150 200 To rq u e (% ) Time (sec)

10 RPM 800C

y = 2,218E-05x - 8,565E-04 0,00E+00 2,00E-03 4,00E-03 6,00E-03 8,00E-03 1,00E-02 1,20E-02 0 100 200 300 400 500 600 d L/ L0 Temperatur, °C dL/L0

(46)

3

Figure 4. Specific heat for the Al-20%SiCp material showing some non-equilibrium activity in the range from 200C to 400C.

Figure 5. Thermal diffusivity of the Al-2+%SiCp material.

Figure 6. The calculated entity thermal conductivity for the the Al-20%SiCp material

2.3 Rheocasting and HPDC

Effort to rheocast the Al-SiCp materials were made using a vertical high pressure die casting set-up, figure 7.

0,800 0,900 1,000 1,100 1,200 1,300 1,400 0 100 200 300 400 500 600 C p , J /( g* K) Temperatur, °C Specifik värmekapacitet 50 55 60 65 70 0 200 400 600 mm ^2/ s Temperatur, °C Diffusivitet 155 160 165 170 0 100 200 300 400 500 600 W/( mK) Temperatur, °C

Konduktivitet

(47)

4

Figure 7. The rheocasting set up used for the experiments. Left is the machine and right is the upper and lower die halves illustrating the die cavity used.

The process of casting started, after roughly 15 mins of stirring, with fully liquid casting of typically four shots to get high pressure die cast samples of the Al-20%SiCp material. Following this rheocasting of Al-Al-20%SiCp material was conducted. Once roughly half of the melt was consumed the weight of the cast material was assessed and the melt was diluted to Al-15%SiCp and stirred for 15 minutes. The process was then repeated with typically four shots of fully liquid material and followed by rheocasting.

In figure 8 the RheoMetal process used is illustrated. First the so called EEMs are cast. In our case we use a manual process and store the EEMs in an oven at 120°C to replicate the conditions in a factory. The EEMs are then stirred into the ladle and the slurry is generated. The EEM amount typically is 7% by weight but for the current study 6% by weight was used arbitrarily to compensate for the fact the SiCp particles were present. Under normal conditions the temperature after slurry fabrication would be 612°C±1°C but it was less stable for the composite slurry and varying from 622°C to 630°C. For each shot a unique slurry was made as the RheoMetal process is a “slurry on demand” process.

Once the slurry was made the process for the fully liquid casting and the rheocasting process was the same. The melt or slurry was poured into the shot sleeve. The sleeve then moves and docks with the tool and the piston injects the melt or slurry into the die cavity, figure 8.

(48)

5

Figure 8. The rheocasting process illustrated.

It should here be noted that the process study went as far as being able to produce reasonable samples, but internal soundness can be significantly improved optimising the slurry and the casting parameters further.

In figure 9 the casting directly out of the die with the biscuit is showed, as well as the castings produced after the biscuit was removed. The samples were directly used for tensile testing. For compression testing smaller cylinders were EDMed out of some of the tensile bars for workability testing. The EDM processing revealed that there were severe porosity issues, figure 10.

(49)

6

In figure 9. The cast components are shown. Left is the cast component (from a test with 7% by weight EEM) used in the current study for the process initial set-up. To the right the cast components are shown with the good liquid cast component at the top, above the ruler, and the rheocast component at the bottom. The component to the left components in the right-side picture are the Al-15%SiCp and those to the right Al-20%SiCp.

Figure 10. EDMed cylinders cut in half illustrating the porosity problem

To ascertain that testing was to be made on good samples, Archimedes principle was used to screen the samples made by the EDM process. This revealed that the good sampled had the following densities, Table 1.

(50)

7

Table 1. Densities of the EDMed cylinders where S indicates semisolid and L indicates fully liquid casting. 20 is 20%SiCp and 15 is 15%SiCp

Cast sample ID EDM ID 1 EDM ID 2

L20-3 2.7105 kg/m³

S20-3 2.6715 kg/m 2.7034 kg/m³

L15-2 2.7222 kg/m³.

S15-3 2.7523 kg/m³ 2.6812 kg/m³

2.4 Mechanical testing

Two types of mechanical testing were made, tensile testing of the as-cast tensile bars and compresssion testing of the EDMed cylinders.

2.4.1 Tensile testing

Tensile testing was made at room temperature in the as cast state, figure 11. Common for RheoMetal processed material is that it is slightly softer and more ductile compared to fully liquid cast material, which is true for 20%SiCp.

Figure 11. Tensile test curves at room temperature, Solid lines are rheocast materials and dashed lines are fully liquid casting. Green illustrated an approximate Youngs Modulus

0 50 100 150 200 250 0 0,2 0,4 0,6 0,8 1

Str

ess

MP

a

Strain %

L20:2 L20:1 S20:1 S20:2 S20:3 S20:4 S20:5 S20:6 L15:1 L15:2 S15:1 S15:2 S15:3 S15:4 S15:5 S15:6

(51)

8 2.4.2 Compression workability testing

The EDMed samples were compressed using purpose build tool for the current project, figure 12. This tool was assembled in the Universal testing machine used for the tensile testing. It uses a 2-pillar stand to guarantee a parallel contact between the upper and lower tools. Each tool was individually heated and controlled by a thermocouple. The sample was heated in between the two tools and as the set temperature was achieved for the sample testing began. The samples at different state of compression are shown in figure 13.

Figure 12. The purpose built-too used for the current project

Figure 13. The appearance of the cylinders after different temperatures and degrees of upsetting

For the compression testing friction is very important to assess the flow stress. At elevated temperature achieving and maintaining both the sample temperature and zero friction was very difficult and therefore sticking condition was decided to be the best choice. The in-homogeneous work can be relatively well estimated but there will still be a minor error as the simplistic approach used includes friction but will not properly represent the barrelling taking place.

The resulting from stress curves for the tested conditions are shown in figure 14a-f. In the testing the compression was aborted at 30% and 60 % height reduction. For each temperature two different strain rates (cross-head speeds were used)

(52)

9

Figure 14a. Flow stress at 400oC low strain rate (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

Figure 14b. Flow stress at 400oC (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

(53)

10

Figure 14c. Flow stress at 450oC slow strain rate (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

Figure 14d. Flow stress at 450oC high strain rate (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

(54)

11

Figure 14e. Flow stress at 500oC low strain rate (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

Figure 14f. Flow stress at 500oC high strain rate (L: Liquid sample, S Semisolid sample, number 15 and 20 indicate %SiCp)

From the figures the materials behave as expected in most cases as a nearly ideal plastic material at high temperature and increasing the strain rate (cross-head speed) increases the flow stress. There is the odd sample deviating, see for instance sample S20-4 in figure 14c which shows a different behaviour and is likely due to a stronger shape deviation making the assumptions for the evaluation of the flow stress fail to some degree and therefore should be disregarded from. There are different degrees of this deviation but there are curves in figure 14a-c that can be seen.

The material can be regarded as a visco-plastic material, equation (1) and for these it is possible to define the Zener–Hollomon parameter, Z, equation (2), also known as the temperature compensated strain rate as a measure of the deformation process to understand the nature of the deformation process.

(55)

12

Equation (1)

Equation (2)

Taking the stress strain curves and evaluating the Zener-Hollomon parameter allows the evaluation of the activation energy, Table 2. In table 2 a clear trend from unreinforced to the 15%SiCp and 20%SiCp is seen as an increase of the activation energy with the fraction of reinforcement. There are also differences between the Liquid and the Semisolid materials but these are not so easily interpreted and requires further investigation.

Table 2, Comparison of the activation energies for the studied material and that of an unreinforced wrought aluminium

Material Activation energy

L15 287,75 kJ/mol

S15 364,29 kJ/mol

L20 450,28 kJ/mol

S20 419,47 kJ/mol

Wrought Al-Mg-Si-Cu 178-237kJ/mol

2.5 Additional work

2.5.1 DSC for heat treatment settings

As part of the study a discussion on the maximum solution temperature was started and a special DSC run was made to assess the events occurring during heating of the Al-20%SiCp material, figure 15.

In Figure 15, a small initial melting can be seen at 539oC followed by the major melting event at 559oC. Solution heat treatment should thus be kept well below 539o_{C. This was also checked to asses the difference between the Liquid cast}

material and the Semisolid cast material, figure 16, for which the traces were identical.

𝜀̇ = (𝑠𝑖𝑛ℎ(𝛼𝜎))𝑛𝜀̇exp (− 𝑄 𝑅𝑇)

𝑍 = 𝜀̇ exp (− 𝑄 𝑅𝑇)

(56)

13

Figure 15. DSC trace with temperature indications for the Al-20%SiCp material to identify melting for the assessment of the maximum solution treatment temperature

Figure 16. DSC trace with temperature indications for the Al-20%SiCp material in the Liquid cast state and the Semisolid cast state suggesting that there is no difference between the processing routs in terms of the melting.

2.5.2 Tool insert in additive manufacturing

In the project an insert for die casting tool was designed and manufactured through additive manufacturing. The main difference to allow for a squeeze casting like process, figure 17a. The printed lower tool half is shown in figure

249 °C _{339 °C}

426 °C _{559 °C}

(57)

14

17b. It should be noted that this tool was printed as a solid block which is unsuitable for additive manufacturing. The machine time was 27 hours.

Figure 17a The model for the casting of the demonstrator

(58)

1

MMC 2016–2018

─

Metallmatriskompositer i masstillverkade komponenter och produkter Jan-Eric Ståhl och Volodymyr Bushlya, Industriell produktion, LTH, Lunds universitet

I samarbete med Lanny Kirkhorn, Fredrik Schultheiss och Jakob Johansson

Sammanfattning

Föreliggande rapport beskriver forskningsarbetet utfört av Industriell Produktion, LTH, Lunds univer-sitet inom ramarna för projektet: ”Utveckling av MMC för högpresterande aluminiumkomponenter” Vinnova 2015–05072 som leds av Marie Fredriksson vid Swerea SWECAST i Jönköping. Redovisade resultat bygger också på tidigare kunskaper och tidigare projekt som forskargruppen vid Lunds uni-versitet genomfört. Resultaten från detta arbete visar att det kan vara fullt möjligt att masstillverka komponenter i MMC under rätt förutsättningar. Det finns goda möjligheter för att lyckas realisera MMC i bruksprodukter även om ytterligare forskning och utveckling krävs. En sekventiell tillverkning bestående av gjutning av enkelt ämne, smidning till nära slutlig form och en viss omfattning av skä-rande bearbetning kan ge en konkurrensmässig tillverkning i förhållande till etablerade koncept. In-dustriellt kommer det att krävas investeringar i ny utrustning samtidigt som nya produkter måste konstrueras och anpassas till MMC och dess tillhörande tillverkningsteknologier.

Innehållsförteckning

1. Koncept, ansatser och bakgrund ... 2

2. Ämnesgjutning av MMC ... 3

2.1 Ämnesframtagning inför smidning ... 3

2.2 Resultat och genomfört arbete kring ämnesgjutning ... 4

3. Varmsmide av MMC ... 4

3.1 Optimal smidestemperatur ... 4

3.2 Deformationsgrad och smidestemperatur ... 5

3.3 Resultat och genomfört arbete avseende varmsmide ... 6

4. Skärande bearbetning av MMC ... 6

4.1 Verktyg och verktygsförslitning vid svarvning ... 7

4.2 Ytjämnhet, ytintegritet vid svarvning ... 9

4.3 Verktygsförslitning vid fräsning ... 11

4.4 Ytjämnhet vid fräsning ... 11

4.5 Resultat och genomfört arbete avseende skärande bearbetning ... 12

5. Framtagning av demonstratorer ... 12

5.1 Framtagning av fräskropp i MMC ... 13

5.2 Framtagning av vevstake ... 14

5.3 Framtagning av en fullskalig bromsskiva för personbil ... 15

6. Ekonomiska överväganden, slutsatser och fortsatt arbete ... 16

7. Erkännande ... 16