on the mechanisms of volume change related casting defects

On the Mechanisms of Volume Change Related Casting Defects Formation in Lamellar Graphite Iron Péter Svidró

DOCTORAL THESIS IN ENGINEERING MATERIALS SCIENCE STOCKHOLM, SWEDEN 2017

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

On the Mechanisms of Volume Change Related Casting Defects Formation in Lamellar Graphite Iron Péter Svidró

Doctoral Thesis, 2018 KTH Royal Institute of Technology School of Industrial Engineering and Management Department of Material Science and Engineering SE-100 44 Stockholm, Sweden

ISBN: 978-91-7729-590-7

Akademisk avhandling som med tillstånd av KTH i Stockholm framlägges till offentlig granskning för avläggande av Teknologie doktorsexamen

torsdagen, den 26 April 2018, kl. 10:00 i sal B3, Brinellvägen 23, KTH, Stockholm, Sverige

This thesis has been conducted in cooperation with

Materials & Manufacturing – Foundry Technology School of Engineering, Jönköping University

SE-55111 Jönköping, Sweden

Abstract

As part of the development of more demanding casting components in lamellar graphite iron, the present thesis is devoted to discuss the phenomena related to the volume change related casting defect formation. Special efforts were dedicated to interpret the solidification of the primary phase in hypoeutectic lamellar cast iron. This is an area that has been neglected in the research literature of cast iron, but that is relevant for the volume change related to the casting defect formation. A series of both known and novel experimental methods were included as well as known and novel interpretation methods. Thermal analysis using two thermocouples displaced in the same thermal filed in combination with multiaxial linear displacement measurement were used to study the columnar to equiaxed transition during solidification. The applied interpretation methods indicate the transition as an interval. This transition is influenced by the variation of process parameters such as the pouring temperature and inoculation. Other significant observations were obtained when the multiaxial volume change measurement on a cylindric sample demonstrated an anisotropic character of the casting skin movement during solidification. In complex shaped cast geometries, a shrinkage porosity or a metal expansion penetration were provoked. These defects were found in connection with the casting surface. Specifically, the casting skin was either missing or formed with a wide extradendritic space which was caused by the dendrite coarsening. In the studied complex shaped casting samples, the dendrite coarseness was well related to the local solidification time. An extensive literature survey gave a clear indication on the problems related to the volume change measurements during solidification. Thus, this led to the development of a novel measuring instrument and a new interpretation of the volume change. The new measuring device separates the solidification from the influence of granular molding materials by introducing a sample holder with properties close to the solidifying cast iron. A thermal analysis based on temperature differences, and volume change calculations based on dilatation measurements was used on a thermally stable spherical sample. The released latent heat in relation to the volume change is interpreted as a function of the local stress state in the extradendritic liquid. The temperature of the coherent columnar zone is always below of the temperature of the bulk in the equiaxed zone. In the beginning of the solidification and during the later stage of the solidification interval, the columnar zone is contracting faster relative to the equiaxed bulk. This causes a compressive stress in the extradendritic liquid. At a stage, when the temperature and composition of the columnar zone reaches the eutectic state, the columnar zone start to expand.

This causes an overall volume increase, which is interpreted as the reason for a tensile tension development in the extradendritic liquid. The alternation between the compressive and tensile stresses during the solidification within the same solidifying zone is considered as the possible driving force for the shrinkage respective metal expansion formations. These observations are considered to be valid even for the case of complex shaped casting solidifications. Thus, the expansion of the columnar zone causing a tensile stress in the liquid will drive the gaseous atmosphere in between the extradendritic area. Furthermore, the contraction of the columnar zone causes compressive stress and squeeze of the extradendritic liquid to the casting-mold surface. Keywords Lamellar graphite iron, hypoeutectic, shrinkage porosity, metal expansion penetration, solidification, dendrite coherency, volume change, displacement.

Sammanfattning

Denna avhandling behandlar fenomen som inträffar i samband med defektbildning orsakad av volymförändringar vid stelning av gjutjärn med lameller grafit, vilket är en viktig kunskap i strävan efter att producera gjutjärn med hög kvalitet. Stelningsstudier av den primära fasen i undereutektisk gjutjärn är ej rapporterade i litteraturen. I denna avhandling introduceras både kända och nyutvecklade observationsmetoder som möjliggör undersökningar av den primära austenit fasen. Termisk analys via registrering av två svalningskurvor i den undersökta termiska domänen i kombination med deformationsmätningar av provkroppens yta med hjälp av rörelsegivare har använts för att studera omslag mellan pelar- och likaxlig kristalltillväxt av den primära austeniten. Observationerna indikerar att omslaget beror på variationer i processparametrar som gjuttemperatur samt mängd och tillsatsmetod av ympmedel. En annan viktig observation från studierna av rörelsemönstret i provkroppens yta under stelning var att pelarkristallzonen deformeras olika i olika riktningar: Detta demonstrerar deformationsförloppets anisotropi. Gjutna komponenter med en komplex geometri var benägna att bilda defekter främst i anslutning till formytan, där den lokala stelningstiden var långsammast i den undersökta domänen. Pelarkristallzonen i långsamt stelnade områden utvecklas till en grov morfologi, där grovleken är proportionell mot den lokala stelningstiden. En omfattande litteraturstudie bidrog till förståelsen om vilka problem som uppstår i samband med mätningar av volymförändringar vid stelning. Baserat på litteraturstudien utvecklades en ny mätmetod som bidrar till en reviderad tolkning av gråjärnets stelningsförlopp. Den nya mätmetoden möjliggör iakttagelsen av volymförändringar utan inverkan från den granulerade formmaterialet. En sfärisk provkropp med tunt skal av stålplåt används för undersökningen. Provkroppens yttemperatur samt den geometriska centrumpunktens temperatur mäts. Samtidigt registreras formväggrörelserna som ger möjlighet att beräkna volymförändringen av den stelnande provkroppen. Det frigjorda stelningsvärmet beräknas från temperaturskillnaderna som råder i provkroppen och i kombination med den beräknade volymförändringen beräknas det rådande trycket i restsmältaområdet mellan austenitdendriterna. Pelarkristallzonen krymper fortare både i början och i slutet av stelningsförloppet och bidrar till att tryckspänningar råder i de sist stelnande områdena. Pelarkristallzonen svalnar till en eutektisk temperatur tidigare än metalliska massan i den likaxliga zonen. Den efterföljande grafitutskiljningen i pelarkristallzonen bidrar till en volymökning av provkroppen som anses orsaka dragspänningar i restsmältaområdet mellan austenitdendriterna.

Förekomster av drag- respektive tryckspänningar under samma stelningsförlopp anses vara drivkraften för uppkomsten av en krympporositet respektive en metallexpansionspenetration i gjutgods med både enkla och komplexa geometrier. Nyckelord Gjutjärn med lameller grafit, undereutektisk, krympporositet, metallexpansionspenetration, stelning, dendritkoherens, volymförändring,

Acknowledgement

I express my sincere gratitude to:

First, Professor Attila Diószegi for giving me the opportunity to work with him and being a mentor during my studies. His guidance, patience and cooperative attitude was exemplary.

Professor Pär Jönsson for his support and encouragement.

Lennart Elmquist my assistant supervisor for the valuable discussions.

Esbjörn Ollas, Jacob Steggo, Jörgen Bloom, Lars Johansson, Peter Gunnarson and Tony Bogdanoff for their technical support.

Richárd Szepesi for the help in controller programming.

All my colleagues at the department of Materials and Manufacturing, JU.

To my parents, Adrien and József and my brother József Tamás for their helpfulness.

Eszter, my beautiful wife for all her support.

All participants of the Järnkoll 2C and VastIron projects, and the involved partners such as Volvo Group AB, Scania CV AB, Jönköping University for their precious contribution to the work, and KK Stiftelsen for the financial support.

Honorable mention [in alphabetical order]: Á.C.S. (ÁJ), attu (HG), doni (TD), gé (SzG), gyahmati (GyT), imi (nyunyókah) (KI), karesz (BK), krisz (GKL), lászló (KL), májkí (TM), pók (KP), saba (MCs), szaki (SzG), szaszák norbi, tibi (CsT) & the rest

Jönköping, 2018

The Doctoral Thesis is based on the appended papers

Supplement A

Determination of the columnar to equiaxed transition in hypoeutectic lamellar cast iron Péter Svidró, Attila Diószegi ISIJ International Vol. 54 (2014) No. 2 pp 460-465 Supplement B

Investigation of dendrite coarsening in complex shaped lamellar graphite iron castings Péter Svidró, Attila Diószegi, Mohsen Saffari Pour, Pär G. Jönsson Metals 2017, 7, 244. Supplement C

On problems of volume change measurements in lamellar cast iron Péter Svidró, Attila Diószegi International Journal of Cast Metals Research Vol. 27, 2014 - Issue 1 pp 26-37 Supplement D

Extended method of volume change measurements during solidification of lamellar graphite iron Péter Svidró, Attila Diószegi, Pär G. Jönsson Presented at the International Symposium on the Science and Processing of Cast Iron (SPCI-XI). 2017. Jönköping, Sweden (Accepted for publication in the reviewed conference proceedings) Supplement E

Determination of pressure in the extradendritic liquid area during solidification Péter Svidró, Attila Diószegi, Pär G. Jönsson and Doru M. Stefanescu Journal of Thermal Analysis and Calorimetry, DOI 10.1007/s10973-018-7088-z, 2018 Other relevant publications not included in the thesis:

Defect formation mechanisms in lamellar graphite iron related to the casting geometry Attila Diószegi, Péter Svidró, Lennart Elmquist. Izudin Dugic

International Journal of Cast Metals Research Vol. 29, 2016 - Issue 5 pp 279-285 Volumetric Changes During the Solidification of Cast Iron Attila Diószegi, Péter Svidró ASM Handbook, Volume 1A, Cast Iron Science and Technology. 2017. pp 880-930 Influence of Primary Austenite on the Nucleation of Eutectic Cells Lennart Elmquist, Attila Diószegi, Péter Svidró Key Engineering Materials Vol. 457. pp 61-66. Trans Tech Publications, Switzerland, 2011. Patent applications

1. Sampling device and method for sampling a liquid or viscous material (Tekniska Högskolan i Jönköping Aktiebolag) Péter Svidró, Attila Diószegi International publication number: WO 2017/054846 A1 International application number: PCT/EP 2015/072424

2. Method of and device for analysing a phase transformation of a material (Tekniska Högskolan i Jönköping Aktiebolag) Attila Diószegi, Péter Svidró International publication number: WO 2017/055046 A1 International application number: PCT/EP2016/071146

The author’s contribution to the different papers in this thesis Supplement A-C: literature survey, evaluation of data, calculations, microstructural analysis using optical microscopy, microstructure analysis preparation using touch screen and image editing software, performing image analysis, performing simulation with casting software, writing the major part of the papers.

Supplement D-E: literature survey, evaluation of data, calculations, development of the measuring method, building the measurement equipment, application and calibration of the sensors (LVDT, IR, TC), development and programming of the data logger, development and production of the sampler, preparation and execution of the experiments, writing the major part of the paper.

Dedicated to my sons, Ákos & Bence.

Contents

Chapter 1. Introduction .......................................................... 1

1.1. Retrospective ................................................................................... 2

1.2. Lamellar graphite iron ..................................................................... 3

1.3. Solidification .................................................................................... 5

1.3.1. Primary solidification .................................................................. 6

1.3.2. Columnar to equiaxed transition ................................................... 8

1.3.3. Eutectic solidification ................................................................ 10

1.3.4. Density change ......................................................................... 10

1.3.5. Volume change ......................................................................... 11

1.3.6. Volume change related casting defects ........................................ 12

Chapter 2. Research Approach ............................................. 15

2.1. Purpose and aim ............................................................................ 15

2.2. Research design ............................................................................. 15

2.2.1. Research strategy ...................................................................... 15

2.2.2. Research questions .................................................................... 16

2.2.3. Overview of performed research ................................................ 16

2.3. Materials and experimental layouts ............................................... 17

2.3.1. Determination of CET ............................................................... 17

2.3.2. Investigation of dendrite coarsening ........................................... 19

2.3.3. On the problems of volume change measurements ....................... 20

2.3.4. Pressure state in the extradendritic area ....................................... 21

2.4. Experimental methods ................................................................... 22

2.4.1. Sample preparation ................................................................... 22

2.4.2. Color etching ............................................................................ 24

2.4.3. Evaluation of the microstructure ................................................ 24

2.4.4. Quantitative characterization of the microstructure ...................... 25

2.4.5. Calculation of fraction solid with thermal analysis ....................... 25

2.4.6. Calculation of pressure .............................................................. 25

Chapter 3. Results and discussion ......................................... 27

3.1. Columnar to equiaxed transition in LGI ......................................... 27

3.2. Investigation of the dendrite coarsening ......................................... 30

3.2.1. Shrinkage porosity .................................................................... 30

3.2.2. Metal expansion penetration ....................................................... 33

3.3. Volume change measurements in LGI ............................................ 36

3.4. Pressure state in the extradendritic liquid area ............................... 43

3.5. Concluding discussion .................................................................... 45

Chapter 4. Conclusions ......................................................... 49

Chapter 5. Future work ........................................................ 52

References ............................................................................ 53

Chapter 1. Introduction

The hypoeutectic lamellar graphite iron (LGI) is a base alloy for the modern casting production due to its excellent material and manufacturing features. Despite the advantages, the complex shaped LGI castings are exposed to casting defect formations such as shrinkage porosity (SP) and metal expansion penetration (MEP). As it was described in the literature, the formation of these types of casting defects is related to the liquid transport within the dendrite network. The aim of this thesis is to investigate the important conditions that influence the liquid transport during solidification leading to SP or MEP formations, as is summarized in Figure 1.

Figure 1. Schematic concept of the thesis with the five supplements.

liquid pressure

macrostructure

Supple

ment B

Supplement A

Supplement C

Supplemen

t E

Supplement D

“on the problems of volum

e

mic

rost

ructure

“inv

estig

atio

n of

dendrite

coarsening

in c

ompl

ex sh

aped lamellar

grap

hite iro

n” “determination of

the columnar to equiaxedtransition in hypoeutectic

lamellar cast iron”

change measurements in

lamellar cast iron”

liquid area durin

g sol

idifi

catio

n“

pressure in th

e ext

rand

endr

itic

“determin

atio

n of

“extended method of volume

change measurements during

solidification of lam

ellar graphite iron“

liquid transportwithin the dendritic

network

2 | INTRODUCTION

Chapter presentation The aim of this chapter is to introduce the reader to the terminology and the concepts used throughout the thesis related to the casting procedure and production of lamellar graphite iron. The chapter begins with a brief retrospective on the procedure and the alloy, and continues with a short description of the theoretical background of the solidification process.

1.1. Retrospective

Casting is a shaping process. It is a prehistoric manufacturing procedure, where the earliest cast object known today is from 3500 BC and they are made of copper [1, 2]. The principle of the method is to pour molten metal into a mold cavity and to let it cool down to room temperature. The mold is a solid material with a melting point higher than the liquid metal, which is made of a cavity of the desired shape. After the cooling, the casting is removed from the mold. The main advantage of the casting process is the almost unlimited design freedom. Therefore, casting is suitable to produce complex parts. The drawback is the tendency to form process attributed defects during the solidification. Iron also has a long history. A Mesopotamian smelted iron artifact (distinguished from meteoric iron by its lack of nickel) dates from 5000 BC [3]. The first big step in the development of iron metallurgy was probably the accidental discovery of the carburization of iron in 1200 BC, which is also the commonly accepted date for the start of Iron Age. Ferrous artifacts became more common after this date [4, 5]. The earliest cast iron object finding is from 600 BC [6]. The casting procedure and the iron alloy have both come a long way. Nowadays, casting is the most preferred method to produce complex parts from metals, as it is cost-effective in high volume production. Approximately, in 90% of the manufactured goods that exist are cast parts [7]. A significant amount of research and development work lead the application of iron from jewelry to heavy-duty combustion engine parts. In 2015, the worldwide casting production reached 104,1 million metric tons. Whereof 84 million tons are ferrous (steel and iron). The production of 46,7 million tons of lamellar graphite iron (LGI) clearly shows its importance within the technical alloys [8].

INTRODUCTION | 3

1.2. Lamellar graphite iron

Today, lamellar graphite iron alloy is produced in the largest volume among the technical alloys [8]. This is not only due to the relatively low production cost of the casting procedure. The LGI also has good mechanical and thermal properties as well as good machinability and vibration damping properties. These features make the LGI the most favorable alloy to be used in the automotive and marine industry and for the production of complex shaped engine parts such as engine blocks and cylinder heads.

Technical alloys – or sometimes referred to as engineering alloys – are metallic alloys which are classified based on their main chemical constituents like Fe, Al, Cu, Zn, Mg. The technical alloys are divided into two main groups: ferrous and non-ferrous alloys. The main component in ferrous alloys is Fe and the class can be subdivided into two groups: steel and iron. The division of ferrous alloys is based on the content of the carbon in the alloy. Consequently, the Fe-C binary equilibrium phase diagram (Figure 2) is used to indicate fundamental points and events related to the material (chemical composition, temperatures) and the solidification properties (phases, phase transformations) in the alloy.

Figure 2. The stable Fe-C binary phase diagram, calculated by Thermo-Calc

2016b with the TCFE7 database [9].

hypereutecticcomposition

hypoeutecticcomposition

C [wt%]

4.30 72

1539

2500

1147

727

0

Tem

per

ature

[°C

]

g-Fe

liquid

g-Fe+liquid graphite + liquid

Fe + graphite

g-Fe + graphite

4 | INTRODUCTION

Ferrous alloys containing less than 2 wt% of carbon are named steels. Ferrous alloys with more than 2 wt% of carbon are called cast irons [10]. In the cast iron alloy, the majority of the carbon content is graphitic carbon [11]. As carbon is the most governing element of cast iron, these alloys are generally classified by the morphology of the graphite phase. The morphology of the graphite phase in cast iron is a microstructural attribution what determines the physical and the mechanical properties of the alloy. According to the morphology of the graphite, the following three main cast iron classes can be defined [7, 12, 13]:

- Lamellar graphite iron (LGI; usually referred to as gray cast iron, GCI; GJL), where the graphite phase has a shape of lamellas or flakes.

- Spheroidal graphite iron (SGI; usually referred to as ductile

iron; GJS), where the graphite phase has a spheroidal shape.

- Compacted graphite iron (CGI; usually referred to as vermicular iron; GJV), where the graphite phase has a worm-like shape.

Cast iron is a multielement Fe-C based alloy that contains a minimum of 2 wt% of carbon [14]. It’s important to emphasize the word multielement as commercial alloys are rarely binary alloys. Instead, the commercial cast iron alloys are compounds with the addition of several chemical elements like Si, P, Mn, Ni, Sn, Cu, Cr and Mo. These elements are added to the alloy to obtain designated mechanical properties of the final cast component with their impact on the formation of the iron and the graphite phases during the solidification [15–17]. An important classification of the cast iron alloys is related to their chemical compositions. The eutectic composition of a cast iron alloy is assumed at 4.3 wt% of carbon, based on the Fe-C equilibrium phase diagram (Figure 2). Cast iron alloys with more than 4.3 wt% of carbon are called hypereutectic cast irons. Furthermore, cast irons with less than 4.3 wt% of carbon are considered hypoeutectic cast irons. The important difference between the hypereutectic and hypoeutectic compositions is the solidification sequence. In hypereutectic alloys, the phase that precipitates first is the graphite phase. In hypoeutectic alloy, the phase that precipitates first is the austenite phase. The alloys that are investigated in this thesis are hypoeutectic lamellar graphite irons, as most of the cast iron production uses this type of alloy today.

INTRODUCTION | 5

1.3. Solidification

Solidification is the most important part of the casting procedure. This is the transformation of the liquid phase to a solid phase as the randomly situated atoms of the melt are arranged into a regular lattice order. The rearrangement of the atoms requires the release of energy in the form of heat. The heat from the melt is released into the mold and the surrounding atmosphere [7, 18, 19]. The temperature difference between the parts defines a thermal gradient that points from a higher temperature area to a lower temperature area. The thermal gradient is used to show the paths and directions of the heat transport during solidification in Figure 3.

Figure 3. Schematic of the thermal gradient of the solidification.

The release of heat during the solidification reflects many details of the process. Consequently, the cooling curve such as the one presented in Figure 4 is used to observe and interpret the solidification sequence [7, 20].

Figure 4. Cooling curve of a LGI alloy, parted into four sections.

melt

mold

atmosphere

900

1000

1100

1200

1300

1400

Tem

per

ature

[°C

]

t [s]

0 1000100 200 300 400 500 600 900700 800

2.

3.

1.

4.

tsol

6 | INTRODUCTION

The cooling curve of a hypoeutectic lamellar graphite iron alloy can be partitioned into four main sections. In the first section (1.), the temperature of the alloy is above the liquidus temperature. In the fourth section (4.) the temperature of the alloy is under the solidus temperature. The most important ones are section two (2.) – called the primary solidification and section three (3.) – called the eutectic solidification. The solidification starts from the boundary of sections one and two, where the phase distribution of the material is a 100% liquid fraction and a 0% solid fraction. The solidification ends at the boundary of sections three and four, where the phase distribution of the material is a 0% liquid fraction and a 100% solid fraction. The time between these boundaries is called the solidification time (tsol).

1.3.1. Primary solidification

The solidification sequence of hypoeutectic LGI starts with the crystallization of austenite. During this event, two types of primary crystals are distinguished. They are classified by their position and growth direction and they are called columnar crystals and equaxied crystals. The development stages of a crystal in chronological order are the nucleation, growth, and coarsening [21–23]. The solidification of the crystals is accompanied by a release of the latent heat of solidification. According to the general nucleation theories, the surface of the mold and impurities or substrates are the preferential nucleation sites [7, 24]. The primary austenite has a face centered cubic (FCC) crystal lattice. Therefore, the growth directions are orthogonal [25] (Figure 5).

Figure 5. Dendrite schematic showing the primary stem and two sets of

orthogonal side branches [26].

The growth of the primary austenite is a dendritic form [22, 27, 28]. Behind the tip of the primary stem, lateral arms branches are formed as secondary arms, on that tertiary arms and may branches further on [29] (Figure 6).

INTRODUCTION | 7

Figure 6. Classic concept of a dendrite [30].

The columnar crystals form first. As soon as the temperature of the melt falls below the liquidus temperature, the primary solid phase - austenite (-Fe) - starts to nucleate on the mold wall (Figure 7a). The primary stem of the dendrite grows in the opposite direction of the thermal gradient, in a columnar manner. The columnar crystals grow in a spatial competition with each other and perpendicularly to the surface of the mold (Figure 7b). The network of the columnar crystals forms the columnar zone. This zone is comparable with a container and it is usually referred to as the casting skin [27, 30, 31]. The next step in the solidification sequence is the nucleation of the equiaxed crystals. This event occurs in the overcooled liquid front ahead of the columnar crystals, probably on floating impurities or substrates [32, 33] (Figure 7a). The growth conditions of the equiaxed crystals are different from the columnar ones. The equiaxed crystals in this stage of the solidification are floating in the melt. Their position in the liquid is controlled by the gravity and the thermal convection. They are surrounded by the liquid and the heat flow of their solidification is equal in every growth direction. Consequently, the dendrite arms develop equally in every direction [30, 31] (Figure 7b). The agglomerate of the equiaxed crystals are called the equiaxed zone. As the solidification advances, the columnar and the equiaxed crystals are filling up the space and the crystals collide with each other [34]. This point is the dendrite coherency, which is also identified as the columnar to equiaxed transition (CET) [35–40] (Figure 7c).

growth direction of the primary stem

8 | INTRODUCTION

Figure 7a. Nucleation of crystals.

Figure 7b. Growth of dendrites.

Figure 7c. The dendrite coherency.

Figure 7d. The final microstructure.

Figure 7. Schematic of the macrostructure development. The solid phase is indicated with white, the liquid phase is indicated with gray color. Figure 7a shows the nucleation of the primary crystals. Figure 7b shows the growth of the primary crystals. When the crystal density becomes more compact, the theoretical boundary of the austenite crystals is indicated with dashed line in Figure 7c. Within a crystal, the austenite is still in mixture with a considerable fraction of liquid. The light gray colored regions around the primary dendrites are the intradendritic liquid, the dark grey colored region between the crystals is the extradendritic liquid. In Figure 7d, the thick line represents the theoretical boundary between the columnar and the equiaxed zones within the final macrostructure.

1.3.2. Columnar to equiaxed transition

At the point of the dendrite coherency, several properties of the system change decisively. From the crystal growth point of view, the crystals collide with each other. Therefore, the longitudinal growth mechanism is blocked. According to previous studies [41, 42], a further development of the dendrites is only possible by coarsening. This is also known as the Ostwald ripening. This is a surface energy driven morphological change controlled by diffusion.

INTRODUCTION | 9

There is a large interfacial area between the initial dendrites and the melt, that creates a thermodynamic imbalance between the solid and the liquid phase. The intention of the system is to reduce the interfacial free energy between the phases. The reduction is achieved by a diffusion process where the solid phase has dissolved and reprecipitated. The deposition process results in the reduction of the total interfacial area [41–43]. The dendrite coarsening in compact shaped LGI castings are published in several articles [44, 45]. The initial dendrite skeleton, which is formed at the dendrite coherency, is important from a thermal conductivity point-of-view. Before the dendrite coherency, the equiaxed and columnar crystals are not interconnected. Therefore, the path for the heat that released by the equiaxed crystals is interrupted by the liquid towards to the exterior. After the coherency, the interconnection of the dendrites connect the transport paths of the heat in the solid phase. This represents an improvement of the thermal conductivity of the body [46, 47]. An important aspect is the change of the feeding properties. The structure of the dendrite network is an important factor of the liquid flow, as well as the feeding. According to the feeding theory [48], the liquid feeding and the mass feeding are the mechanisms that supply the solidifying phase with liquid before the dendrite coherency. The flow of the liquid is easy between the randomly distributed dendrite crystals (Figure 7a and Figure 7b). After the dendrite coherency, the driving feeding mechanisms change to an extradendritic- or a burst feeding. The coherent dendrite network obstructs the flow of the liquid in the remaining space between the dendrite arms and the crystals. This space is the inverse of the dendrite network, namely the interspace. The parts of the interspace must be distinguished. The liquid between the dendrite arms that surrounds the crystals is situated in the intradendritic space (indicated with a light gray color in Figure 7c). The liquid between the crystals, flows in the extradendritic space (indicated with a dark grey color in Figure 7c). They are diverse in temperature, in fluidity properties, and in chemical composition – due to the segregation phenomena [15, 49]. With the advancement of the solidification, the interspace is continuously reduced that makes the flow of the liquid even more hindered. In other words, the permeability of the dendrite network decreases significantly after the dendrite coherency. The zones in the casting that solidifies last are called the last to solidified (LTS) or the last to freeze (LTF) zones.

10 | INTRODUCTION

1.3.3. Eutectic solidification

According to the general theory of the solidification [31], the composition of the residual liquid changes due to the solute redistribution between the phases during the dendritic growth of the primary austenite [50]. The excess of carbon in the solid phase is rejected into the intradendritic liquid. Once the liquid region rich in carbon reaches the required undercooling, graphite precipitation starts by nucleation on pre-existing inclusions in the liquid. For lamellar graphite iron these inclusions are reported as complex sulfides (MnS, XS), which nucleated on complex oxides of Al, Mg, Si, Ti and Zr [51-54]. The eutectic phase grows as graphite and austenite, with the solidification front in contact with the melt. The graphite protrudes into the liquid ahead of the austenite, so the graphite phase is the leading phase in the eutectic growth (Figure 8) [55-56].

Figure 8. Sketch of the growth mechanism in flake graphite. In the center, there

is a primarily precipitated graphite. It thickens by the diffusion of C atoms through the austenite layer [57].

1.3.4. Density change

Density (ρ) is defined as mass (m) per unit volume (V) and it is expressed as follows:

V

m (1)

with the unit of kgm . The density of the LGI in the solid phase is higher compared to the liquid phase. During the cooling, the density of the alloy increases due to the liquid contraction. The density reduction indicated with 1 in Figure 9 is caused by a graphite precipitation. The density of the graphite at 2.3-2.7 10 kgm is almost one third of the iron’s 7.9 10 kgm . The density reduction, indicated with 2 in Figure 9, is caused

by a solid-state transformation. The density variation plays an important role in the modelling and simulation of casting processes. Therefore, large amounts of research work have been conducted to find the accurate way of measurement of this material property [58].

C

C

C

γ

γ

INTRODUCTION | 11

Figure 9. Density of a LGI alloy (‘K’) as a function of temperature [59].

The chemical composition of the alloy ‘K’ shown in Table 1.

Table 1. Chemical composition of alloy “K” measured by optical emission spectrometer (OES) on coin samples. [59].

Element C Si Mn P wt (%) 3.80 1.74 0.53 0.03

1.3.5. Volume change

The density variation of the alloy results in the change of the specific volume. The specific volume (v) is defined as the reciprocal of the density and expressed as follows:

1

m

Vv (2)

with the unit of m kg . The change of the specific volume of the alloy ‘K’ is shown in Figure 10. The reduction in the volume - contraction is explained by the increment of the density of the metallic phase. The increment of the volume - expansion is explained by the graphite precipitation at the time of the eutectic solidification (indicated with 1 in Figure 10) and the solid-state transformation (indicated with 2 in Figure 10) [52].

6.85

6.89

6.93

6.67

7.01

7.05

Den

sity

[10

3 k

gm

-3]

Temperature [°C]

0 1250500 1000750250

1

2

12 | INTRODUCTION

Figure 10. Specific volume of a LGI alloy (‘K’) as a function of temperature [59].

The volume change that accompanies the solidification of hypoeutectic lamellar graphite iron is even more complicated in sand molds. The total volume change is a summary of the shrinkage of the alloy and the movement of the mold wall [60, 61]. The overall result of the volume change during the cooling is a shrinkage. This is a well-known phenomenon and it is managed by using an oversized design of the mold. In foundry practice, this is the ‘shrinkage allowance’ or also referred as the ‘patternmaker’s allowance’ [24]. During cooling, the structural transformations are represented on a macroscale as the displacement of the surface of the casting. Therefore, the volume change represents the character of the solidification.

1.3.6. Volume change related casting defects

The randomly recurring casting defects attributed to the volume change are the shrinkage porosity (SP) and the metal expansion penetration (MEP). The common feature of these defects is that they form during the solidification, more specifically after the CET. They are situated in the last to freeze zones, close to, or on the metal-mold interface [62, 63]. The shrinkage porosity is a material deficit, and can be identified as an internal defect (Figure 11).

1.422

1.429

1.436

1.444

1.451

1.458S

pec

ific

volu

me

[10

-4 m

3kg

-1]

Temperature [°C]

0 1250500 1000750250

1

2

INTRODUCTION | 13

Figure 11. Scanning Electron Microscope picture of a shrinkage porosity. The

developed dendrite arms are covered with oxide layer [64].

According to the literature, these types of casting defects are situated on the border of the dendrite crystals. The surface of the dendrites is usually covered with an oxide layer, which means that the dendrite was in direct contact with the atmosphere outside the casting cavity. The development stage of the dendrites inside the pore is comparable to the surrounding dendrites, it follows that the pore was formed after the CET. It occurs probably due to insufficient feeding [62, 64, 65]. The metal expansion penetration is a material surplus, it can be identified as an external defect (Figure 12).

Figure 12. Color-etched surface of a casting with metal expansion penetration defect. The red line represents the original metal-mold interface. The area above the red line shows the bulk microstructure that contain both primary and eutectic phases, as is expected from the hypoeutectic composition. The area below the red line shows sand particles from the mold which are burned into the castings surface and the MEP defect (on the left). The penetrated metal seems to have an eutectic structure, without primary precipitated austenite.

200 μm

14 | INTRODUCTION

Some authors found, that a MEP defect is formed as a liquid is squeezed out from the casting and the defect tends to occur at the end of the solidification [66, 67]. Observations from other authors indicate the MEP defect formation at the beginning of the eutectic solidification [68]. An important study that considers the effect of the dendrite network structure on the MEP defect formation mechanism reports, that the MEP defect solidified without primary precipitated austenite [63, 69]. However, the alloy was a hypoeutectic LGI. It was concluded, that the penetration of the liquid between the sand grains of the mold occurred after the CET.

Chapter 2. Research Approach

Chapter presentation This chapter describes the materials, the research methods and the research activities those have been applied in the thesis. The chapter begins with the introduction of the aim of the work and continues with the specification of the experimental procedures.

2.1. Purpose and aim

The performance of a cast component is specified by the microstructural features such as the output of the solidification. The process involves several phenomena like phase transformations, density changes, mass transport, fluid flow, coarsening, contraction and expansion. The majority of these events are present in the cooling curve and appears also as a macroscale displacement. Therefore, the thermal analysis and the volume change measurements are the most important tools in the prediction of the microstructure, in the investigation of the solidification, and the defect formation process. The aim of the present work is to obtain novel knowledge about the mechanisms and substantial conditions of casting defect formation during the solidification of hypoeutectic lamellar graphite iron. Further, to enhance the accuracy of the volume change measurements and the interpretation of the results.

2.2. Research design

2.2.1. Research strategy

The research begun with the collection of relevant works in the field of casting defect formations and volume change of cast irons to become familiar with the research area. Knowledge gaps were identified, and topics of interest were defined. Research questions were formulated. Experiments were set up, data were analyzed than conclusion were made. The workflow applied in the thesis is summarized in Figure 13 [70].

Figure 13. Schematic of the research approach [70].

literaturereview on

the topic ofinterest

researchquestions

design andexecution of experiments

dataevaluation conclusion

16 | RESEARCH APPROACH

The collection of the information is an iterative cycle of monitoring and evaluation of publicly available works that are relevant in the research field [71]. The information resources for journal and conference articles, books, e-books, manuscripts were online databases like the Materials Science & Engineering Database (including METADEX) and the library of the Jönköping University.

2.2.2. Research questions

The general theories of the solidification and the formation of volume change related casting defects were studied. The knowledge of the microstructure development of LGI in compact shaped castings can be extended to complex shaped castings. Also, there is lack of information on how the development of the micro- and macrostructure and the CET interacts in the volume change during solidification. A critical literature review on the known volume change measurement techniques helped to classify the possible experimental variables, to understand limitations and to identify the state of the art experimental methods. This thesis aims to answer the following research questions:

RQ 1 How does the CET influences the development of the dendrite network?

RQ 2 How does the development of the dendrites influences the

morphology of the dendritic interspace?

RQ 3 Does the development of the macrostructure has any influence on the volume change during the solidification?

RQ 4 Which condition controls the direction of the residual liquid

transport within the dendrite network?

2.2.3. Overview of performed research

The aim of Supplement A was to find answer to RQ 1. The progress of CET in LGI were investigated with three different type of determination techniques. The study involved thermal analysis, and the evaluation of expansion force values together with the released latent heat of the solidification. The purpose of Supplement B was to find answer to RQ2 by the investigation of the morphological diversity of the primary dendrites in complex shaped castings.

RESEARCH APPROACH | 17

The morphological characterization techniques that have been developed in previous works were applied in the study in LGI samples with SP and MEP defects. The purpose of Supplement C was to find an answer to RQ3 by the study of the development of the macrostructure and its relation to the volume change. A comparison of the different volume change measurement techniques was presented. Furthermore, a known experimental layout was used by combining multiaxial volume change measurements with expansion force measurements. The aim of Supplement D and Supplement E was to find answer to RQ4 by the introduction of a new measurement method. The volume change was calculated from linear displacement values. Thereafter, the cooling curve analysis was conducted by the Fourier thermal analysis method. Finally, volume change values and the thermal analysis results were combined and introduced into the Clausius-Clayperon equation algorithm to calculate the pressure state of the extradendritic liquid.

2.3. Materials and experimental layouts

In order to determine the CET (Supplement A) and to study the problems of the volume change measurements (Supplement C), experiments were performed in a laboratory environment. The base alloy was prepared in a 60kg medium frequency induction furnace and held at 1460 ±5 °C temperature for 20 minutes. Thereafter, the melt was poured into a preheated, 40 kg capacity hand ladle while Sr based inoculation was added into the stream. Experiments were also performed to investigate the dendrite coarsening (Supplement B) and to study the pressure state of the extradendritic liquid (Supplement D and Supplement E). These were performed in an industrial environment. The alloys were prepared in a 4 ton induction furnace and treated according to the production metallurgy.

2.3.1. Determination of CET

The study in Supplement A involves one base alloy (Table 2). Table 2. Chemical composition measured by OES on coin samples.

Element C Si Mn P S Cr Mo min wt (%) 3.30 1.65 0.54 0.05 0.09 0.15 0.21 max wt (%) 3.79 1.82 0.64 0.05 0.11 0.17 0.25


The amount of inoculant, the casting temperature and the time interval between the inoculation and the pouring were adjusted in individual charges. Overall, 11 charges were prepared. However, due to measurement failures three charges were excluded from the investigations. The metallurgical parameters for the individual charges are shown in Table 3.

Table 3. Metallurgical parameters of the individual charges.

Charge id 14 15 16 19 20 21 23 34 Tcasting (°C) 1440 1380 1440 1440 1440 1320 1320 1380

tinoculation (min) 0 5 10 0 10 10 10 5

wtinoculant (%)

0.05 0.10 0.05 0.15 0.15 0.05 0.15 0.10

The pouring temperature of the samples was 1440 ± 5 °C. The experiment was designed to perform a Fourier thermal analysis based on two temperature points, coupled with volume change and expansion force measurements on horizontally allocated bars (350 x Ø 50 mm). The samples for volume changes and expansion force measurements were cast simultaneously. The experimental layout is illustrated in Figure 14.

Figure 14. Schematic of the experimental assembly.

The material of the mold was croning sand, the dimensions of the drag and the cope were 45 x 100 x 407 mm. Two ‘S’ type (PtRh10) thermocouples were placed along the axis of the cylinder, perpendicular to the parting line of the mold.

1. inlet2. frame

3. quartz rod4. shell mold

5. thermocouples6. force measuring unit

7. displacement transductor

1

2 3

3

3

4

56/7

7

50 mm


The measurement point of the first thermocouple was located in the center (Tc) of the cylindric section, which is equivalent to the cylinder’s axis (inner temperature point). The measurement point of the second thermocouple was displaced by 10 mm from the axis towards to the surface (Ts) (outer temperature point). The thermocouples were protected by a quartz tube (Ø 3 / Ø 5 mm), closed in one end. Quartz push rods (Ø 10 mm) were inserted on the parting line to transmit the displacement and the expansion force towards the sensors. The mold for the volume change measurement was equipped with four push rods. The two push rods were placed as prolongation of the axis of the cylinder, with the scope to measure the longitudinal displacement. The other two push rods were inserted perpendicularly to the axis of the cylinder, to measure the radial displacement on the test bar. The second mold for the expansion force measurement was equipped with two push rods placed in similar positions as the push rods aimed for the longitudinal displacement measurements. The sensors and the push rods were assembled on a steel frame. The signals from thermocouples, displacement and force measuring sensors were recorded by a HBM-Beam U10 software system.

2.3.2. Investigation of dendrite coarsening

The study in Supplement B involves two alloys, which are defined in Table 4.

Table 4. Chemical composition of alloys, measured by OES on coin samples.

Element C Si Mn P S Cr Mo Content in SP sample wt (%)

3.28 1.96 0.64 0.03 0.06 0.26 0.05

Content in MEP sample wt (%)

3.18 1.77 0.56 0.05 0.09 0.15 0.24

The samples were cast from a 50kg preheated hand ladle, where the casting temperature was 1448 ± 5°C. In both samples, the mold material was Epoxy-SO2 hardened quartz sand. The castings were designed to provoke SP and MEP defects by creating thermal conditions similar to those found in a complex shaped cylinder head casting. The SP sample (Figure 15) combines simple geometrical domains like cylinders and plates to initiate SP. The bottom plate (240 x 120 x 20 mm) and the top plate (240 x 120 x 10 mm) is connected with three cylinders (100 x Ø 40 mm) and a vertical plate (220 x 15 x 100 mm) parallel to the cylinders. The distance between the cylinders and the vertical plate is 7 mm.


Figure 15. Isometric view of the SP sample.

The MEP sample (Figure 16) combines cylindrical geometry domains with concave surfaces to initiate a MEP defect. The cylindrical casting (80 x Ø 80 mm) includes an internal channel (45 x Ø 30 mm) with a rounded end (r 15 mm).

Figure 16. Isometric view of the MEP sample.

2.3.3. On the problems of volume change measurements

The study in Supplement C involves the two alloys, described in Table 5.

Table 5. Chemical composition of alloys, measured by OES on coin samples.

Element C Si Mn P S Cr Mo Content in Heat

I wt (%) 3.52 1.81 0.59 0.05 0.11 0.15 0.25

Content in Heat II wt (%)

3.51 1.85 0.54 0.05 0.09 0.15 0.24

The amount of the additional inoculant was 0.05 wt% for Heat I and 0.15 wt% for Heat II. Also, the experiment was performed in the same way and with the same equipment as it is described in chapter 2.3.1.

10 mm

10 mm


2.3.4. Pressure state in the extradendritic area

The study in Supplement D and Supplement E involves one alloy, which chemical composition is given in Table 6.

Table 6. Chemical composition measured by OES on coin samples.

Element C Si Mn P S Cr Mo wt (%) 3.20 1.88 0.61 0.03 0.09 0.15 0.07

The melt was poured into a 2kg capacity pre-heated hand cup. The temperature of the melt was 1420 ± 5°C. A spherical sampler was submerged into a melt in the hand-held cup for 2.5 seconds. Thereafter, the measurement was started as the push rods touched the surface of the sampler. The signals from the thermocouple, the IRs and the LVDTs were collected by a unique controller system with a ten records per second sampling rate. The schematic of the experimental layout is shown in Figure 17.

Figure 17. Schematic of the spherical sampler and arrangement of the sensors. 1, spherical sampler. 2, lid. 3, steel tube. 4, thermocouple. 5, push rod. 6, schematic representation of the field of view of an infrared (IR) thermometer.

1

2

3

4

5

5

56

6

6

10 mm


The experiment was designed to perform a Fourier thermal analysis using two temperature points simultaneously and to make volume change measurements on a spherical sample (Ø 33 mm). The experimental assembly consists of the following parts. A steel sphere made of two hemispheres with a wall thickness of 0.5 mm. The diameter of the top open is 13mm. Also, a lid made of the same material and the same wall thickness as the sphere is used. The diameter of the lid is 18mm and it is welded to a steel tube (inner Ø 2 / outer Ø 4 mm) made of the same material as the sphere and which is closed at one end. A mineral insulated N-type (Nicrosil–Nisil) thermocouple with a diameter of 1.5 mm was used to monitor the temperature. The end of the thermocouple is situated in the center of the sphere (inner temperature point). The IR sensors are targeted on the surface of the sphere (outer temperature point). The IR and the LVDT sensors are placed around the surface of the sample, distributed equally by 60°. These are the spots for the temperature measurement and the displacement measurement on the sampler surface. The axes of the sensors are situated on a plane (sensor plane), which is normal to the surface of the sphere. The sensor plane is identical to the horizontal intersection of the sphere. The application of a steel tube allows for a reuse of the thermocouple due to a protection from the erosion of the melt. The application of alumina push rods allows the protection of the LVDTs. The accuracy of the measurement is assured due to the low thermal expansion coefficient of the alumina push rod, which is 6.3 [10 grC ].

2.4. Experimental methods

2.4.1. Sample preparation

The preparation of the samples started with a raw cut by a whip-saw or a rotary blade machine. The samples were mounted in a cold-setting two component epoxy resin. After the setting, the samples were manually ground on SiC paper (P80 FEPA) until a planar surface was obtained. The rest of the preparation was performed by a semi-automatic sample preparation system. The grinding was performed in four steps on P220, P500, P1200 and P2000 FEPA designation abrasive resin bonded diamond discs and by using a water lubrication. The polishing was performed in two steps by using 3 µm and 1 µm particle size diamond suspensions on polishing clothes. The samples in Supplement B were cut on the A-A plane, as is indicated in Figure 18 and Figure 19. Investigated areas are highlighted with thick frames. The areas with the related casting defects are marked with a cross.


(a) (b)

Figure 18. (a) Isometric view of the SP sample with the gating system; (b) Cross section view of the SP sample.

(a)

(b)

Figure 19. (a) Isometric view of the MEP sample; (b) Cross section view of the MEP sample.

Samples in Supplement C were cut on a section 85 mm away from the flat end of the bar. The cut plane is represented by a dashed line in Figure 20.

Figure 20. Isometric view of the sample bar from Supplement C.

A

A

10 mm

A-A

10 mm

A

A10 mm

A-A

10 mm

25 mm


2.4.2. Color etching

Color etching is a chemical film deposition process, which is used on a polished surface. The thickness of the film depends on the local silicon content on the surface of the specimen. Silicon segregates to the austenite (k>1) and become depleted in the last solidifying area. The propagation of the incident light waves through the air-film and film-specimen interfaces causes formations of colors in the reflected light by the interference phenomena. The observed tints differs by the thickness of the film, which allows for a distinction of the different phases present in a sample [72]. Different alkaline picrate solutions have been developed for the color etching of LGI [73–77]. The Motz reagent [77] was applied to reveal the microstructure on the samples investigated in this thesis. The solution consists of 10 g NaOH, 40 g KOH, 10 g picric acid and 50 ml of distilled water. The temperature of the solution was 108 °C and it was maintained with a ±3 °C accuracy by using a thermistor controlled heating system during the etching. The state of the film formation was checked periodically by using an optical light microscope (OLM). Also, the etching process was repeated until a satisfactory level of the desired coloring was achieved.

2.4.3. Evaluation of the microstructure

Micrographs were taken from the color etched surfaces with OLM. Then, the primary austenite phase was outlined manually on a Wacom Cintiq touch-sensitive display. The pre-processing of the micrographs resulted in binary images, as shown in Figure 21.

(a)

(b)

Figure 21. (a) Color etched micrograph from Supplement B. The primary austenite appears in dendritic shape with whitish, light blue color; (b) Pre-processed binary image. Black color represents the space between the primary dendrites. White color represents the dendrites and the non-evaluable areas of the eutectic colonies.

200 μm 200 μm


2.4.4. Quantitative characterization of the microstructure

The pre-process and the analysis of the binary images were performed in Adobe Photoshop CS6. The stereological parameters used in Supplement B for the description of the dendritic interspace are the surface area A , the perimeter P and the hydraulic diameter (D ). The hydraulic diameter of the interspace can be calculated as follows [44, 45]:

iihydIP P/AD (3)

The secondary dendrite arm spacing (SDAS) λ was used for the characterization of the primary dendrites.

2.4.5. Calculation of fraction solid with thermal analysis

Thermal analysis (TA) by means of cooling curve analysis (CCA) is a wide spread tool to interpret the solidification of an alloy. The processing of temperature values registered during cooling can be used to quantify solidification parameters like the latent heat, fraction solid, etc. [78–80]. Two different methods are used, the Newtonian method and the Fourier method. The Newtonian TA (NTA) method uses one cooling curve and the temperature is usually registered at the geometrical center of the specimen (one temperature point). The Fourier TA (FTA) method uses at least two cooling curves. Consequently, the temperatures are usually registered in the geometrical center T and near to the surface of the specimen T . Several authors have published results suggesting that the FTA method is more reliable than the NTA method, since it is based on the larger number of measured thermal points [81–84]. The FTA method used in the supplements has been developed at the department of Materials and Manufacturing – Casting, Jönköping University and it has been used in previous research publications [80, 82, 83, 85].

2.4.6. Calculation of pressure

The solidification of an alloy is usually described under atmospheric pressure conditions. Only a few literature contributions are known where the effect of the pressure on the solidification is invoked [86–89]. The thermodynamic equation that involves the pressure into the description of the phase transformation with coexisting phases is called the Clausius-Clapeyron relation.


The formula is written in Equation (4) and it represents the slope of the pressure-temperature boundary dP/dT between the solid and liquid phases, related to the specific enthalpy of the solidification ∆H (Jkg ), the specific volume change ∆v (kgm ), and the local temperature.

vT

H

dT

dP

(4)

If the pressure and the temperature vary along the phase transition line, an integration of Equation (4), yields the following integral

s

c

2

1

T

T

P

P

dTv

HdP (5)

which has the following solution:

c

s12 T

Tln

v

HPP

(6)

The outcome of the calculation ∆P P P is the pressure deviation from the equilibrium atmospheric pressure in the extradendritic area during solidification. It is obtained by inserting the calculated heat release during solidification along with the measured volume change. A negative value is considered to represent a tensile state, while a positive value is considered to represent a compressive state on the residual liquid within the dendritic interspace.

Chapter 3. Results and discussion

Chapter presentation In this chapter, the main results of the supplemented papers are summarized and discussed.

3.1. Columnar to equiaxed transition in LGI

As described in detail in the introduction, the CET is a key event which influences the development of the microstructure and macrostructure during solidification. The properties of the structures are crucial conditions from a defect formation point of view. Therefore, Supplement A focused on studying the process of CET in a hypoeutectic LGI. The idea behind the work presented in Supplement A, was to use known experimental techniques to investigate samples with different metallurgical parameters in order to explore CET values. Parts of the experimental setups found in the literature were reported to be used particularly to study light alloys [46, 47]. The three methods introduced in Supplement A are based on the analysis of temperature differences (∆T) between the columnar and the equiaxed zones, the released heat (q ), and the expansion force (F ) in the solidification interval. The fundamental concept of the CET determination in all methods, is the shift between the solidification sequences of the columnar and equiaxed zones. The temperature differences and the solid fractions are calculated from measured values, while the expansion force is a directly measured value. A comparison of methods to determine CET applied, on various experiments, are shown in Table 7.

Table 7. Summary of the CET determination results with the metallurgical parameters.

Charge id

Tcasting (°C)

tinoculation (min)

wtinoculation (%)

fs at CET by

∆T Fexp qs 14 1440 00 0.05 20.3 20.3 13.0 15 1380 05 0.10 16.7 35.7 07.3 16 1440 10 0.05 19.7 34.6 10.9 19 1440 00 0.15 13.8 13.8 07.9 20 1440 00 0.15 21.4 24.9 08.8 21 1320 10 0.05 16.8 31.4 08.0 23 1320 10 0.15 11.7 14.9 06.3 34 1380 05 0.10 15.7 26.7 08.1

28 | RESULTS AND DISCUSSION

It can be concluded from the results in Table 7, that the addition of inoculants decreased the calculated fraction of primary phase, at the event of CET by all conditions. Considering the growth mechanisms of the primary austenite grains, differences in reaching CET under similar thermal conditions are only possible if the number of the nucleated grains is different. The differences in the CET determination results within one charge are related to the nature of each method. For example, the ∆T method is insensitive to the mechanical influences. As soon as the crystals collide, the thermal conductivity of the system will change instantly. Therefore, this method is suitable to determine the beginning of the CET. The F method is mechanical, as the expansion force only appears in the case of a well-developed dendrite network. Therefore, this method is applicable to determine the end of the CET. The CET determination results for Charge 14 is shown in Figure 22.

Figure 22. Comparative graph of CET determined by different methods on

Charge 14.

In Figure 22 is important to note, that the expansion force is continuously increasing after the CET.

∆Tmax

qs max

Fexp start

0

20

60

80

40

ΔT [

°C]

and F

[kP

]

0

2

6

8

10

4

12

qs [M

J/m

3se

c]

Fraction of solid

0 0.80.1 0.2 0.3 0.4 0.5 0.6 0.7 0.9 1

Charge 14

RESULTS AND DISCUSSION | 29

This supports the theory of a formation of a coherent network. CET determination results for Charge 19 is shown in Figure 23.

Figure 23. Comparative graph of CET determined by different methods on

Charge 19.

In Figure 23 is important to note the break in the expansion force. The CET is indicated with F . The expansion force starts to increase until an unexpected break (F ) is reached. After an idle state, the force accumulation continues to the end of the solidification. The break can be explained by the weakness of the dendrite network. At an early coherency when a relative low fraction of solid phase (q byF 13.8) is precipitated, the compressive load developed due to the measurement setup led to the collapse of the austenite network at F . The differences between the results from the different determination methods leads to a conclusion, namely that the CET runs over a period of time. The application of different measurement techniques shows, that the CET is not represented by a single moment but rather by an interval. This occurs, when the morphological and mechanical properties of the dendrite network changes significantly.

Fb

Frs

ΔTmax

qs max

Fexp start

0

20

60

80

40

ΔT [

°C]

and F

[kP

]

0

2

6

8

10

4

12

qs [M

J/m

3se

c]

Fraction of solid

0 0.80.1 0.2 0.3 0.4 0.5 0.6 0.7 0.9 1

Charge 19


3.2. Investigation of the dendrite coarsening

The investigation of the CET by different determination methods showed that dendrite network can be formed by different volume fractions from an identic melt composition. A CET value with a low solid fraction is supposed to occur with long and thin dendrites, while a CET value with a high solid fraction is supposed to occur with short and thick dendrites. The morphology of the initial austenite network is demanding, as the further growth of the solid phase is driven by the coarsening process after the formation of CET. Also, this morphology defines the paths for the residual liquid to supply the solidifying parts through the dendritic network. The coarsening process has been investigated in compact geometries under isothermal conditions by several authors [44, 90–92]. The aim of Supplement B was to investigate the coarsening process of the austenite dendrites in the defected areas of complex shaped castings. The SP and MEP casting defects are formed during the solidification and they are usually situated in the last solidifying part of the casting (LTS). The LTS is explained as a region in the extradendritic space, where the path of the liquid transportation is open for the longest time. The space between the dendrites inside and outside of the LTS area was quantified by using image analysis. Also, the thermal field of the casting was characterized by using the simulated local solidification times.

3.2.1. Shrinkage porosity

By the SP sample, the simulated local solidification times indicate the position of the slowest cooling area (hot spot) at the bottom of the cylindrical component, which is close to the casting surface (indicated with light gray color in Figure 24). The location of hot spot can be explained by a local overload of the mold from a thermal point-of-view. This part of the mold is surrounded by the thick bottom plate, the cylinder and the vertical plate. Therefore, the heat transported from the melt towards to the atmosphere is obstructed more than in any other part of the mold. Accordingly, the solidification is slower in this region compared to the other regions in the casting.


Figure 24. Simulated local solidification time plotted in gray scale on the cross

section of the SP sample (left) and the calculated hydraulic diameter of the dendritic interspace (right).

The largest hydraulic diameters are found in the same positions where the slowest solidification times are found (Figure 24). The morphological differences can be recognized on the micrographs collected from the positions with different solidification times. Figure 25 shows a coarsened dendrite morphology in the slowest cooling area, where a SP defect was formed.

Figure 25. Micrograph representing a local solidification time of 566t sol s,

2.26DhydIP µm and SDAS = 76.8 µm in the SP sample.

Away from the defect formation area, the solidification time is shorter. Consequently, the morphology is finer than in Figure 25. The dendritic interspace decreases as it is shown in Figure 26.

22.6

22.9

15.4

18.6

21.1

17.8

17.6

18.5

18.1

27.5

20.7

19.7

26.7

26.2

24.921.3

23.224.2

24.923.1

21.7

hydDIP

[μm]

Solidification Time

[s]

590559

529

498461

436

406

375

344

314

283252

221

190160

0

10 mm

200 μm



6.17DhydIP µm and SDAS = 51.5 µm in the SP sample.

A comparison between the dynamic coarsening in a hypoeutectic LGI alloy of a compact geometry [44] and the complex geometry of the SP sample in the present work is presented in Figure 27.

Figure 27. Comparison of results of the morphologic parameters with literature

data.

200 μm

0

20

40

60

80

100

50

10

30

70

90

Dh

yd a

nd

SD

AS

[μm

]

tsol

[s1/3]

0 2 4 6 8 10 12

DIP

hyd SP Sample

SDAS[45]

DIP

hyd [45]

SDASSP Sample


The compared morphology parameters fit well with each other, except for the SDAS values. The longer the solidification time, the more distortion appears in the morphology of the dendrites. Thus, the accuracy of the SDAS measurement become less. The distortion is caused by the dendrite fragmentation phenomena, which has been reported recently [90].

3.2.2. Metal expansion penetration

Considering the MEP sample, the simulated local solidification times indicate the position of the slowest cooling area at the concave surface. This is close to the thermal center of the cylinder specimen. The location of a hot spot can be explained, by the thermally overloaded mold. The melt surrounds the sand insert, which is not able to transfer away heat of that intensity in this region compared to other parts of the mold. Consequently, in this section the solidification takes a much longer time. Therefore, there is more time for a coarsening of the dendrites to take place. Because of the coarsening, the largest hydraulic diameters were found in the same positions where the slowest solidification times were found (Figure 28).

Figure 28. Simulated local solidification time plotted in gray scale on the cross section of the MEP sample (left) and the calculated hydraulic diameter of the

dendritic interspace (right).

The morphological differences by the MEP sample can be recognized even with bare eyes on the micrographs collected from the positions with different solidification times. Figure 29 shows a coarsened dendrite morphology in the slowest cooling area, where MEP defects are formed.

16.5

17.5

18.5

17.0

20.5

22.1

20.6

19.2

18.8

17.7

19.5

19.6

20.2

27.0

18.1

22.8

23.5

29.6

25.6

33.234.7

46.0

35.1

32.9

40.2

35.3

35.8

38.3

43.3

40.4

36.3

35.0

38.5

37.0

18.0

18.5

19.7

20.0

22.6

23.3

22.0

22.7

24.1

27.1

26.6

28.5

34.7

29.4

30.9

27.6

30.0

36.9

26.5

hydDIP

[μm]

Solidification Time

[s]

590559

529

498461

436

406

375

344

314

283252

221

190160

0

10 mm



hydIPD = 35 µm and SDAS = 59.7 µm in the MEP sample.

In an area where the solidification time is shorter, the morphology is finer compared to Figure 29. Therefore, the dendritic interspace decreases as is shown in Figure 30.


hydIPD = 18.5 µm and SDAS = 30.4 µm in the MEP sample.

A comparison of the morphological parameters with the literature data [44] shows a decent fit of the dynamic coarsening between a compact and a complex geometry of the MEP sample (Figure 31).

200 μm

200 μm


Figure 31. Comparison of results of the present morphologic parameters with

literature data.

The presented results of the dendrite coarsening are in correlation with earlier observations [44, 90–92]. The morphological parameters such as SDAS and hydraulic diameter of the dendritic interspace are strictly time dependent. In both cases of the SP and MEP samples, the coarsening of the primary austenite creates a discontinuity in the columnar shell under the LTS area (Figure 32).

0

20

40

60

80

100

50

10

30

70

90

Dhyd a

nd

SD

AS

[μm

]

tsol

[s1/3]

0 2 4 6 8 10 12

DIP

hyd MEP Sample

SDAS[45]

DIP

hyd [45]

SDASMEP Sample


Figure 32. Schematics of dendrite coarsening in the LTS area covered region of the columnar zone for compact and complex casting geometries. Dashed

lines indicate the LTS area determined by the numerical simulations.

The increment of the interspace within the dendrite network in the areas where the solidification is slowest is a result of the coarsening phenomena.

3.3. Volume change measurements in LGI

Volume, as defined by thermodynamics, is an extensive parameter for the description of the thermodynamic state of a system [93]. Hence, the volume change measurements have been used for more than one hundred years to study the solidification of different alloys [94]. However, the in-situ or the real-time observation of the solidification is difficult due to the high temperature and due to the fact, that the material behaves different in a liquid state and in a solid state. Every volume change measurement methods used for in-situ investigation follows the same principle. Therefore, they can generally be identified as a dilatation analysis.

penetration

shrinkage porosity

complex shape castingsdiscontinuous columnar zone

compact shape castingcontinuous columnar zone

columnar grain equiaxed grain

extradendriticliquid

intradendriticliquid


During solidification, the movement of the surface of the specimen is registered. Thereafter, the volume change is calculated based on the displacement. The measurement of the surface movement is usually indirect, where an intermediary medium connects to the surface with a displacement sensor. This medium is usually made of a material with a low linear expansion such as sapphire or polycrystalline alumina, to avoid the disturbance of the measured results. The main component in most of the known measurement methods is the specimen holder. The material and the geometric design both vary on the specimen holder. Also, it’s shape defines the thermal field of the specimen. Although the volume change measurement has a history of 150 years, there is only one apparatus available on the commercial market namely, the dilatometer. In a dilatometer, the specimen holder is a cylindrical chamber. Therefore, the specimen has a cylindrical shape. Three different versions of the method are known to register displacements: the capacitive-, the optical- and the push rod method. The most common version is the push-rod type: a specified size of cylindrical sample (diameter 6mm, length 12mm) is placed into the ceramic chamber with ceramic pistons in the ends, where the surrounding furnace provides the requested temperature. The movement of the pistons is registered by a displacement sensor, which usually is a LVDT sensor. Whereas only the dilatometer exists as a commercial solution to study the solidification by volume change measurements, many different layouts have been used in scientific investigations. Several methods have been developed during the period of 1874-2017, 34 experimental setups were found from 28 authors / groups [61, 95–130]. Overall, the scientific versions are all self-developed methods by the authors, created individually for solidification research. Besides certain differences in the design (eg. various layout of the displacement sensors), they follow the same dilatation principle as mentioned above. Also, the sample holder is usually a sand mold and the investigated alloy is cast into the mold. The sand mold plays an important role in the solidification as it is the first medium to be in contact with the melt. Several studies have been published on the effect of the mold on the solidification [60, 126, 128, 129] together with investigation of the thermo-physical properties of different sand mixtures [85]. The indirect volume change measurement is usually referred to as a linear measurement, as the push rods and the displacement sensors are aligned along an axis.


The linear volume change measurement can be classified into basic concepts by the number of axes and direction of the displacement recording as it is represented in Figure 33.

Figure 33. Examples for the classification of linear volume change

measurements by the number of axes and the direction of the displacement registration.

The experiment in Supplement C was conducted with the three axis multidirectional measurement layout equipped with the two point cooling curve recording and expansion force measurement on a cylindrical specimen. The linear axial ∆l and the linear radial ∆l displacement together with the expansion force are presented as a function of time in Figure 34 for Heat I and in Figure 35 for Heat II.

axis

direction

uni

multi

1 2 3

Turner [99]

25 mm

Mrvar et al. [122]

25 mm

Gedeonova et al. [120]

25 mm

Nándori et al. [94]

25 mm


Figure 34. Axial ∆l and radial ∆l displacement and F registered in Heat I.

Figure 35. Axial ∆l and radial ∆l displacement and F registered in Heat II.

CET

-0.01

0.02

0.01

0.01

Δl [

%]

00

80

20

40

60

F [

kP

]

t [s]

0 100 200 300 400 500 600 700

ΔlA

ΔlR

Fexp

Heat I

CET

-0.01

0.02

0.01

0.01

Δl [

%]

00

80

20

40

60

F [

kP

]

t [s]

0 100 200 300 400 500 600 700

Fexp

ΔlA

ΔlR

Heat II


The starting point of the measurement is t 0, when the registered temperature starts to increase over the ambient temperature. A displacement starts in both the axial ∆l and radial ∆l directions at the same time right after the mold filling, in every case. Only fluctuations can be registered at the early stage of the solidification, based on the expansion force measurements. The continuous increment of the values starts later, the point is at 100s for Heat I and 139s for Heat II. This was discussed in more details in Chapter 2.3.1. Specifically, the start of the continuous increment of the expansion force is associated with the CET. This is the macrostructural development phase, from where the specimen can be considered as an aggregate. Therefore, the displacement values that were registered before the CET represent a combination of the effect of the metallostatic pressure, the plastic deformation of the columnar zone (the mantle of the cylindrical specimen), and the wall movement of the mold that originates from the thermal dilatation of the sand. Consequently, volume change calculations for cylindrical shaped specimens should implement displacement values only after the CET. The linear axial ∆l and the linear radial ∆l displacement values that have been reset to the point of CET determined by the expansion force measurements are shown in Figure 36 for Heat I and in Figure 37 for Heat II.

Figure 36. Axial ∆l and radial ∆l displacement values as a function of time, reset at

CET by F in Heat I.

-0.01

0.00

0.01

0.02

Δl [

%]

t [s]

0 1000 2000 3000 4000 5000 6000

ΔlA

ΔlR

Heat I


The displacement values for Heat I and Heat II show a difference within an alloy. The deviation between the axial and the radial surface displacement characteristics can be explained by the interaction of the movements in different directions. For example, the shrinkage in the axial direction may counteract or assist the displacement in the radial direction. The opposite may also occur, due to the low tensile strength of the columnar zone in the mantle of the cylinder.

Figure 37. Axial ∆l and radial ∆l displacement values as a function of time, reset at

CET by F in Heat II.

The comparison of the displacement values in distinct directions show the anisotropic character of the solidification, which has an impact on the evaluation of the volume change. Figure 38 and Figure 39 show the calculated volume change results for multidirectional measurements. The ∆V values are based on a single axis measurement, where the volume change is calculated based on only the axial displacement. The ∆V value represents a multiaxis measurement, where the axial and the radial displacement values have been used in the calculations. Based on the comparison of the volume change measurements, the multiaxial experiment layout clearly shows the advantage over the single axis by specimens, which solidify with an anisotropic characteristics.

-0.01

0.00

0.01

0.02

Δl [

%]

t [s]

0 1000 2000 3000 4000 5000 6000

ΔlA

ΔlR

Heat II


Figure 38. Volume change based on axial vs. axial and radial displacement values as a

function of the temperature in Heat I.

Figure 39. Volume change based on axial vs. axial and radial displacement values as a

function of the temperature in Heat II.

-0.01

0.00

0.01

0.02ΔV

[%

]

T [°C]

1200 1000 800 600 400

ΔVA+R

ΔVA

Heat I

-0.01

0.00

0.01

0.02

ΔV [

%]

T [°C]

1200 1000 800 600 400

ΔVA+R

ΔVA

Heat II


3.4. Pressure state in the extradendritic liquid area

The outcome of the work in Supplement B, is that the increased dendritic interspace facilitates the transport of the residual liquid leading to RQ 4. To answer which condition controls the direction of the liquid transport within the dendrite network, a new measurement method was introduced. The extensive literature review on the existing volume change measurements and the outcome of Supplement C, that the solidification is anisotropic on thermally unbalanced geometries, led to a specimen design to optimize the thermal field. More detailed information is presented in chapter 2.3.4.

The aim of Supplement E, is to investigate the pressure state of the extradendritic liquid during the solidification by using the newly developed spherical sample. The measured temperatures and the calculated volume change are represented as a function of the fraction solid in Figure 40.

Figure 40. Cooling curves and the calculated volume in the solidification

interval.

The contraction from the beginning of the solidification is assigned to the decrease of the specific volume as the liquid transforms into a solid.

1000

1050

1100

1150

1200

1250

Tem

per

ature

[°C

]

1.874

1.875

1.876

1.877

1.878

1.879

Volu

me

[10

-5 m

3]

Fraction of solid

0 0.80.1 0.2 0.3 0.4 0.5 0.6 0.7 0.9 1

Tc

Ts

DV


During cooling, a columnar zone is developed on the internal surface of the sampler. When the temperature in the columnar zone T reached a required undercooling value, an eutectic phase transformation started in this zone. This led to a precipitation of graphite into the intradendritic space. The increment in the specific volume is supposed to cause an expansion of the whole sample from f 0.32. At the end of the eutectic transformation in the columnar zone, the volume is expected to contract again. Parallel to this event, the temperature of the equiaxed zone T also reaches the required undercooling to start the eutectic transformation with the related increment of the specific volume. Therefore, the whole volume continues to expand until the volume increment of the graphite precipitation is larger than the contraction of the columnar shell. From f0.52 to the end of the solidification the reduction of the volume continues according to the decrease of the specific volume. With the introduction of the results of the FTA together with the volume change measurement into the Clausius-Clayperon equation, the pressure state in the extradendritic liquid P can be calculated in the solidification interval as it is presented together with the volume change derivative dV/df in Figure 41.

Figure 41. Liquid pressure and volume change during solidification.

The dV/df curve in Figure 41 illustrates the direction of the volume change.

III.

IV.

II.

I. V.

-3

3

6

0

Pre

ssure

[10

7 P

a]

-7

0

7

14V

olu

me

chan

ge

[10

-10 m

3]

Fraction of solid

0 0.80.1 0.2 0.3 0.4 0.5 0.6 0.7 0.9 1

dV/dfs

P


Negative values of the dV/df curve indicate a contraction and positive values indicate an expansion of the specimen. The deviation of the pressure of the extradendritic area from the atmospheric pressure ∆P 0 starts as soon as the solidification starts and continues until the end of solidification. The positive values of the calculated pressure are interpreted as a compressive pressure, while the negative values are considered to be an expansion pressure. The solidification interval can be divided into five periods with respect to the pressure state of the extradendritic liquid. The first period is from f 0 to f 0.32 and is indicated with I. in Figure 41. In this zone, the solidification starts with the formation of the columnar dendritic zone. The contraction of this columnar zone causes a compressive load that results in a pressure increment in the extradendritic liquid. In the solidification interval between f 0.32 to f 0.52 indicated with II. in Figure 41, the extradendritic liquid is supposed to be exposed to a tensile load due to the expansion of the columnar zone. It is important to note that a short reduction of the expansion pressure occurs at f 0.52, which is indicated with III. in Figure 41. This coincide with the start of the eutectic phase precipitation in the equiaxed zone. The precipitation of the graphite in the equiaxed zone compensates for the expanded sample volume. It is worth to note that for values from f 0.52 until f 0.72 which is indicated with IV. in Figure 41., the contraction of the columnar zone starts while the extradendritic liquid continues to be exposed to a tensile load. The fifth period is from f 0.72 to f 1 and indicated with V. in Figure 41. In this zone, the interaction between the contracting columnar zone and the graphite precipitation in the equiaxed zone expose the residual liquid to a compressive load. With a further contraction of the columnar zone the pressure on the residual liquid increases in this interval of the solidification.

3.5. Concluding discussion

The general aim of this work was to promote an understanding of the mechanism of volume change related to casting defects formations in lamellar graphite iron. The literature was reporting the shrinkage porosity and the metal expansion penetration as volume change related phenomena, where the density variation of the different phases in various combinations cause the defect formations. An important result from the literature study was the observed shrinkage cavity found in the austenite grain boundary, which indicates a metal deficit in the extradendritic area [62]. The surface of the naked austenite dendrite was oxidized, which indicated a contact with the surrounding gaseous atmosphere. In other words, the gaseous atmosphere was driven from the surroundings in


between the extradendritic area. In the case of a metal expansion penetration, a metallic phase was squeezed between the sand grains. Here, the penetrated metal has an eutectic composition which was interpreted as liquid metal formed in the extradendritic area. Consequently, a basic mechanism to promote both defect formations is the gas intrusion alternatively liquid transport in the extradendritic area. The gas intrusion respectively the liquid transport is supposed to happen in opposite directions over the casting surface. The investigations in this thesis were dedicated to understand under which conditions it is possible to have those gaseous and liquid flows in opposite directions. Furthermore, to find hindering methods in the future and to avoid those casting defects. The obtained results and how they are related to the liquid transport can be illustrated in the sketch given in Figure 42.

Figure 42. Summary of the outcomes and the relations of the supplements with the main condition of the casting defect formation in lamellar graphite iron.

structure

conditions

microstructure

formation

pressure

conditions

development

macrostructure

form

atio

n

mac

rost

ruct

ure

mic

rost

ruct

ure

deve

lopm

ent

pressuredeviation

-com

pres

sive

pre

ssur

e on th

e liquid

- exp

ansi

on p

ress

ure o

n the li

quid

- coarse dendrite morphology

- extended interspaceby diff

eren

t sol

id fr

acti

on

- d

endrit

e coh

eren

cy

- unstable thermal field

- anisotropic casting skin

Supp

lem

ent D

and E

Supplement B

Supplem

ent A

Supplement C

liquid transportwithin the dendritic

network


The extradendritic area where the gaseous and liquid flow are supposed to take place was investigated in Supplement A. Since the study of the primary dendrite network in lamellar graphite iron has previously been neglected, the study of the columnar to equiaxed transition in LGI represents a pioneering attempt to describe the related phenomena. It has been observed that the CET is an interval rather than a single moment and that it can be identified with different interpretation methods. The wide variation in the solidification kinetics of LGI was indicated by the observations where samples with similar chemical composition were cast and where the CET take place at different fractions of solidified primary austenite. The importance of the casting geometry and its influence on the dendritic network was studied in Supplement B. Complex shaped casting geometries, known to promote shrinkage porosities and metal expansion penetrations were investigated. The common properties of the samples provoking the different type of defects was the interference between the casting surface and the thermal hot spot of the studied geometry. In both cases, a coarsened dendrite morphology was observed in the metal-mold interface. Furthermore, the dendrite morphology in the entire casting was found to be related to the local solidification time. This has previously only been observed in separate cast samples with various solidification times. The presence of a directed extradendritic space, which is a result of the dendrite coarsening process, indicate the liquid and gaseous transport directions through the casting surface. Since the liquid and gaseous transport directions were determined, the focus was transferred to understand what is the driving force of the transport phenomena and why it takes place in either the one or the other direction. The macrostructure of the LGI is formed by an outer / columnar dendritic shell, which is also called a casting skin. It separates the bulk material, also called the equiaxed zone, from the surrounding. In earlier literature [94], the columnar zone was termed as a container containing the solidifying equiaxed dendrites and the adjacent liquid phase under solidification. A deformation of the container was supposed to give rise to a formation of a tensile stress in the solidifying liquid phase, which is necessary to drive the gaseous atmosphere into the extradendritic area under the formation of a shrinkage porosity. The same deformation was supposed to be the reason for the presence of a compressive stress in the extradendritic liquid. This causes a squeezing effect of the eutectic liquid to the outside surface of the casting.


Since many different methods have been reported in the literature [94] to investigate LGI with inconsistent interpretations, Supplement C is focused on investigating the problems related to volume change measurements. The main effect of the observations is the uncertainty of the volume change measurements, due to the thermal instability of the used samples and the volume change anisotropy because of the shape and curvature of the columnar zone. Based on the observed problems during the traditional volume change measurements, a novel instrument was built. The aim was to create a thermally stable field, and from a volume change point of view, an isotropic columnar zone. The instrument is described in Supplement D and some basic measurements and interpretations of the results is presented in Supplement E. Thermal analysis based on temperature differences, and volume change calculations based on dilatation measurements are introduced on a thermally stable spherical sample. The released latent heat in relation to the volume change is interpreted as a function of the local stress state in the extradendritic liquid. The temperature of the coherent columnar zone is always below the temperature of the bulk in the equiaxed zone. In the beginning of the solidification and in the later stage of the solidification interval the columnar zone is contracting faster relative to the equiaxed bulk causing a compressive stress in the extradendritic liquid. In between when the temperature and composition of the columnar zone reach the eutectic state the columnar zone start to expand. This, in turn, causes an overall volume increase. This volume increase is interpreted as the reason for the formation of a tensile tension in the extradendritic liquid. The alternation between the compressive and tensile stresses during the solidification within the same solidifying zone is considered as the possible driving force for the shrinkage and metal expansion formations, respectively. These observations are considered valid even for the case of complex shaped casting solidification. Thus, the expansion of the columnar zone causing a tensile stress in the liquid will drive the gaseous atmosphere in between the extradendritic area. However, the contraction of the columnar zone causes a compressive stress and a squeeze of the extradendritic liquid to the metal-mold interface.

Chapter 4. Conclusions

Chapter presentation In this chapter, the conclusions from the present work is drawn. The investigation of the CET in LGI was conducted by the evaluation of an experiment that used FTA and records the expansion force during the measurements. An evaluation of the microstructure was carried out to investigate the effect of the dendrite coarsening phenomena on the morphological properties of the dendritic interspace. An extensive literature work has been performed to summarize the known volume change measurement methods. Also, the investigation of the development of the macrostructure was carried out based on the evaluation of a dilatation experiment on a thermally unbalanced specimen. In addition, a new measurement method based on a thermally balanced specimen was developed. Also, pressure calculations were performed to compliment the investigations of the substantial conditions that controls the casting defect formation mechanisms by the transport of the residual liquid within the dendritic network. After the completion of these studies, the following main conclusions can be stated: Supplement A

The CET determination methods shows differences both with respect to time and the precipitated amount of solid fraction when the CET is reached.

The CET determination method based on a temperature differences is applicable for the identification of the beginning of the CET.

The CET determination method based on the expansion force

measurement is applicable for the identification of the end of the CET.

The differences in the results of the determination methods

indicate that the CET progresses over a period of time.

Supplement B

A solidification time dependent coarsening was identified within a casting in complex shaped casting geometries. This was found to locally influence the austenite morphology.

50 | CONCLUSIONS

The increment of the dendritic interspace creates a favorable condition for the transport of the residual liquid towards or away from the LTS area.

The accurate measurement of the SDAS is disturbed by the

morphological changes during the coarsening process.

Supplement C

The comparison of the axial and radial displacement values in a linear volume change measurement on a cylindrical specimen showed an anisotropic character of the solidification.

The interpretation of the solidification is more accurate when

using multidirectional and multiaxial linear dilatation measurements compared to when using uniaxial measurements. This is due to that the former considers the volume change in multiple dimensions.

The anisotropic character of the solidification is a result of the

thermally unbalanced geometry of the specimen.

Supplement D

A new measurement concept was developed, which included the implementation of FTA and a dilatation analysis on a specimen in combination with using an optimized thermal field.

Supplement E

Based on the calculations, the solidification of the LGI showed a -1.5x107 Pa expansion pressure value in zones II. and IV. Moreover, different maximum contraction pressure values were obtained in zone I. III. and V.

During the solidification, both compressive and tensile loads can

be identified in the liquid within the columnar shell. The presence of the pressure deviation from the atmospheric value

demonstrates the driving force that controls the direction of the transport of the residual liquid within the dendrite network, when SP or MEP appears in the casting.

CONCLUSIONS | 51

The research questions are answered as follows

RQ 1 How does the CET influence the development of the dendrite network?

Answer 1: The obtained results indicate a considerable influence on the formation and progress of coherency due to variation of casting parameters. The investigation of the CET by different determination methods showed that dendrite network can be formed by different volume fractions.

RQ 2 How does the development of the dendrites influence the morphology of the dendritic interspace?

Answer 2: The investigation of the morphology in the defected positions indicates a maximum space between the dendrites, which can be explained by the dynamic coarsening process of the primary dendrites.

RQ 3 Does the development of the macrostructure has any influence on the volume change during the solidification?

Answer 3: Yes. The investigation showed the anisotropic nature of the solidification process. This is also influencing the interpretation of the dilatation results.

RQ 4 Which condition controls the direction of the residual liquid transport within the dendrite network?

Answer 4: It is shown, that there is a deviation in the pressure of the residual liquid from the atmospheric pressure during solidification in LGI. The existence of the tensile respective compressive stress in the extradendritic liquid is considered as a condition for a volume change related casting defect formation.

Chapter 5. Future work

Chapter presentation In this chapter, a number of proposals for future investigations are presented. The studies presented in this thesis treat the solidification of hypoeutectic LGI where a novel measurement method was introduced. In future work a study using the spherical sampler with different diameters would be of interest to investigate the effect of the cooling rate on the pressure of the extradendritic liquid. The investigation in Supplement E involves one alloy. Different metallurgical parameters like carbon content, oxygen content and inoculation materials are also of interest for future studies. Volume change measurements could be used to compare results from dilatometric measurements. Since the molding material is supposed to play an important role in the solidification rate and the volume change in interaction with the columnar shell, the interaction between the mold and the metallic part must be studied. The use of the novel instrument with the spherical sample and the adjacent interpretation algorithms should be possible to use instantaneously in the casting process to identify if the ongoing solidification will end up or not with defect formation. Calculated thermo-physical properties by using FTA in combination with the volume change measurements are considered suitable for input data to numerical calculation of solidification. Consequently, the methods are supposed to be suitable for validation of numerical simulation.

References

[1] R. F. Tylecote, A history of metallurgy, 2nd. ed. United Kingdom: Institute of Materials, 1992.

[2] M. Goodway, “History of casting,” in ASM Handbook, vol. 15. Casting, 1988, pp. 15–23.

[3] J. C. Waldbaum, From bronze to iron: the transition from the Bronze age to the Iron age in the eastern Mediterranean. Paul Astroms Forlag, 1978.

[4] T. A. Wertime and J. D. Muhly, The Coming of the Age of Iron. Yale University Press, 1980.

[5] C. E. Cramer, “What Caused The Iron Age?” 10-Dec-1995. [6] D. B. Wagner, Iron and Steel in Ancient China. E.J. Brill, 1993. [7] D. M. Stefanescu, Science and Engineering of Casting

Solidification. Kluwer Academic / Plenum Publishers, 2009. [8] Modern Casting Staff, “50th census of world casting production,”

Modern Casting, pp. 25–29, 25-Dec-2016. [9] Thermo-Calc 2016b, database: TCFE7 - Steels/Ferrous Alloys.

Stockholm, Sweden: Thermo-Calc AB. [10] H. Berns, G. Scheibelein, and W. Theisen, Ferrous Materials:

Steel and Cast Iron. Springer Berlin Heidelberg, 2008. [11] W. Hume-Rothery, H. M. Finniston, D. W. Hopkins, and W. S.

Owen, The Structures of Alloys of Iron: An Elementary Introduction. Elsevier Science, 2016.

[12] “ISO 945 Microstructure of Cast Irons,” in Graphite classification by visual analysis, Switzerland, 2008.

[13] R. Ruxanda and D. M. Stefanescu, “Graphite shape characterisation in cast iron - from visual estimation to fractal dimension,” Int. J. Cast Met. Res., vol. 14, no. 4, pp. 207–216, 2002.

[14] J. M. Radzikowska, “Metallography and Microstructures of Cast Iron,” in ASM Handbook, vol. 9. Metallography and Microstructures, 2004, pp. 565–587.

[15] Z. Morita and T. Tanaka, “Thermodynamics of Solute Distributions between Solid and Liquid Phases in Iron-base Ternary Alloys,” Trans. Iron Steel Inst. Jpn., vol. 23, no. 10, pp. 824–833, 1983.

[16] A. Kagawa and T. Okamoto, “Coefficients for equilibrium partition of a third element between solid and liquid in iron-carbon

54 | REFERENCES

base ternary alloys and their relation to graphitization during iron-carbon eutectic solidification,” J. Mater. Sci., vol. 19, no. 7, pp. 2306–2318, 1984.

[17] M. Tisza, Physical Metallurgy for Engineers. ASM International, 2001.

[18] H. Fredriksson and U. Åkerlind, Solidification and Crystallization Processing in Metals and Alloys. John Wiley & Sons, Ltd, 2012.

[19] T. S. Piwonka, “Casting,” in ASM Metals Handbook, Asm International, 1998, pp. 728–730.

[20] H. Fredriksson, “Interpretation and Use of Cooling Curve - Thermal Analysis,” in ASM Handbook, vol. 15: Casting, 2008, pp. 353–356.

[21] J. Lacaze and G. Lesoult, “Birth, Growth and Ripening of Solidification Structures,” Acta Stereol., vol. 5, no. 2, pp. 331–336, Sep. 1986.

[22] A. Diószegi, K. Z. Liu, and I. L. Svensson, “Inoculation of primary austenite in grey cast iron,” Int. J. Cast Met. Res., vol. 20, no. 2, pp. 68–72, 2007.

[23] L. Elmquist, S. Salera, and A. Diószegi, “Inoculation and its effect on primary solidification structure of hypoeutectic grey cast iron,” Int. J. Cast Met. Res., vol. 23, no. 2, pp. 124–129, 2010.

[24] J. Campbell, Castings, Second Edition. Elsevier, 2003. [25] M. Rappaz and J. A. Dantzig, Solidification. EFPL Press, 2009. [26] S. P. Marsh and M. E. Glicksman, “Overview of Geometric

Effects on Coarsening of Mushy Zones,” Metall. Mater. Trans. A, vol. 27, no. 3, pp. 557–567, 1996.

[27] R. W. Heine and C. R. Loper, “Dendrites and Eutectic Cells in Gray Iron,” Mod. Cast., vol. 56, no. 4, pp. 185–191, 1969.

[28] G. L. Rivera, R. E. Boeri, and J. A. Sikora, “Solidification of gray cast iron,” Scr. Mater., vol. 50, no. 3, pp. 331–335, Feb. 2004.

[29] W. Xumin, Y. Xiaohua, X. Jun, and Z. Youling, “Model of Primary Austenite Dendrite Structure in Hypoeutectic Cast Iron,” J. Iron Steel Res. Int., vol. 10, no. 1, pp. 33–35, 2013.

[30] P. Beeley, Foundry Technology, 2nd ed. Oxford: Butterworth-Heinemann, 2001.

[31] W. Kurz and D. J. Fisher, Fundamentals of Solidification, 4th ed. Switzerland: Trans Tech Publications, 1998.

[32] M. C. Flemings, Solidification processing. New York,N.Y.: McGraw-Hill, 1974.

REFERENCES | 55

[33] A. Hellawell, S. Liu, and S. Z. Lu, “Dendrite fragmentation and the effects of fluid flow in castings,” JOM, vol. 49, no. 3, pp. 18–20, Mar. 1997.

[34] F. Mampaey and Z. A. Xu, “Grain impingement during solidification of cast iron,” Cast Met., vol. 7, no. 3, pp. 165–174, Nov. 1994.

[35] B. Chalmers, Principles of Solidification. John Wiley & Sons, New York, 1964.

[36] R. Morando, H. Biloni, G. S. Cole, and G. F. Bulling, “The development of macrostructure in ingots of increasing size,” Metall. Mater. Trans., vol. 1, no. 5, pp. 1407–1412, May 1970.

[37] M. H. Burden and J. D. Hunt, “Mechanism for the Columnar to Equiaxed Transition in Castings or Ingots,” Met. Trans A, vol. 6A, no. 1, pp. 240–241, 1975.

[38] J. D. Hunt, “Steady-State Columnar and Equiaxed Growth of Dendrites and Eutectic,” Mater Sci Eng, vol. 65, no. 1, pp. 75–83, 1984.

[39] H. Fredriksson and A. Olsson, “Mechanism of transition from columnar to equiaxed zone in ingots,” Mater. Sci. Technol., vol. 2, no. 5, pp. 508–516, 1986.

[40] S. C. Flood and J. D. Hunt, “Columnar to equiaxed transition,” in ASM Handbook. Vol. 15. Casting, ASM International, 1988, pp. 130–136.

[41] P. W. Voorhees, “The theory of Ostwald ripening,” J. Stat. Phys., vol. 38, no. 1–2, pp. 231–252, Jan. 1985.

[42] D. Kammer and P. W. Voorhees, “The morphological evolution of dendritic microstructures during coarsening,” Acta Mater., vol. 54, no. 6, pp. 1549–1558, Apr. 2006.

[43] M. E. Glicksman and P. W. Voorhees, “Ostwald Ripening and Relaxation in Dendritic Structures,” Metall. Trans. A, vol. 15, no. 6, pp. 995–1001, 1984.

[44] R. Lora and A. Diószegi, “Dynamic Coarsening of 3.3C – 1.9Si Gray Cast Iron,” Metall. Mater. Trans. A, vol. 43, no. 13, pp. 5165–5172, 2012.

[45] A. Diószegi, V. Fourlakidis, and R. Lora, “Austenite dendrite morphology in lamellar graphite iron,” Int. J. Cast Met. Res., vol. 28, no. 5, pp. 310–317, Sep. 2015.

[46] J. Tamminen, “Thermal analysis for investigation of solidification mechanism in metals and alloys,” Chemical Communications, University of Stockholm, 1988.

56 | REFERENCES

[47] G. Chai, “Dendrite coherency during equiaxed solidification in aluminium alloys,” Chemical Communications, University of Stockholm, 1994.

[48] J. Campbell, “Feeding Mechanisms in Castings,” AFS Cast Met. Res. J., vol. 5, no. 1, pp. 1–8, Mar. 1969.

[49] R. Boeri and F. Weinberg, “Microsegregation of Alloying Elements in Cast Iron,” Cast Met., vol. 6, no. 3, pp. 153–158, Nov. 1993.

[50] H. D. Brody, “Solute redistribution in dendritic solidification,” Massachusetts Institute of Technology. Dept. of Metallurgy, Massachusetts Institute of Technology, 1965.

[51] T. Skaland, Ø. Grong, and T. Grong, “A model for the graphite formation in Ductile Cast Iron: Part I. Inoculation Mechanisms,” Metall. Trans. A, vol. 24, no. 10, pp. 2321–2345, Oct. 1993.

[52] D. D. Double and A. Hellawell, “The nucleation and growth of graphite - the modification of cast iron,” Acta Metall. Mater., vol. 43, no. 6, pp. 2435–2442, Jun. 1995.

[53] I. Riposan, M. Chisamera, S. Stan, and T. Skaland, “Graphite nucleant (microinclusion) characterization in Ca/Sr inoculated grey irons,” Int. J. Cast Met. Res., vol. 16, no. 1–3, pp. 105–111, Aug. 2003.

[54] E. Moumeni, D. M. Stefanescu, N. S. Tiedje, P. Larrañaga, and J. H. Hattel, “Investigation on the Effect of Sulfur and Titanium on the Microstructure of Lamellar Graphite Iron,” Metall. Mater. Trans. A, vol. 44, no. 11, pp. 5134–5146, Nov. 2013.

[55] A. Hultgren, Y. Lindblom, and E. Rudberg, “Eutectic Solidification in Grey, White and Mottled Hypo-eutectic Cast Irons,” J. Iron Steel Inst., vol. 176, pp. 365–374, Apr. 1954.

[56] H. Fredriksson and S. E. Wetterfall, “A study of Transition from Undercooled to Flake Graphite in Cast Iron,” in The Metallurgy of Cast Iron, Geneva, Switzerland: Georgi Publishing Company, 1975, pp. 277–293.

[57] H. Fredriksson and U. Åkerlind, Materials processing during casting. Wiley, 2006.

[58] K. Hellström, A. Diószegi, and L. Diaconu, “A Broad Literature Review of Density Measurements of Liquid Cast Iron,” Metals, vol. 7, no. 5:165, 2017.

[59] Measurement data by Kristina Hellström, .

REFERENCES | 57

[60] S. Engler, D. Boenisch, and B. Kohler, “Metal and Mold Wall Movements During Solidification of Cast Iron,” Cast Metals Research Journal AFS, vol. 9, pp. 20–30, Mar. 1973.

[61] D. M. Stefanescu, M. Moran, and S. Boonmee, “Use of Combined Liquid Displacement and Cooling Curve Analysis in Understanding the Solidification of Cast Iron,” in Transactions of the American Foundrymen’s Society, vol. 120, 2012, pp. 365–374.

[62] L. Elmquist and A. Diószegi, “Shrinkage porosity and its relation to solidification structure of grey cast iron parts,” Int. J. Cast Met. Res., vol. 23, no. 1, pp. 44–50, 2010.

[63] A. Diószegi, I. Dugic, and I. L. Svensson, “Metal expansion penetration on concave casting surfaces of grey cast iron cylinder heads,” AFS Trans., vol. 115, pp. 609–615, 2007.

[64] L. Elmquist, S. Adolfsson, and A. Diószegi, “Characterizing shrinkage porosity in gray cast iron using microstructure investigation,” in AFS Transactions, Atlanta, GA, USA, 2008, vol. 116, pp. 692–704.

[65] L. Elmquist, K. Soivio, and A. Diószegi, “Cast iron solidification structure and how it is related to defect formation,” Mater. Sci. Forum, vol. 790–791, pp. 441–446, May 2014.

[66] H. G. Levelink and F. P. M. A. Julien, “Penetration and shrinkage by interaction of solidifying cast iron and casting mold - Part 1,” AFS Cast Met. Res. J., vol. 9, no. 2, pp. 56–63, 1973.

[67] H. G. Levelink and F. P. M. A. Julien, “Penetration and shrinkage by interaction of solidifying cast iron and casting mold - Part 2,” AFS Cast Met. Res. J., vol. 9, no. 2, pp. 105–109, 1973.

[68] I. Dugic and I. L. Svensson, “An investigation of the effect of inoculants on the metal expansion penetration in grey iron,” Int. J. Cast Met. Res., vol. 11, no. 5, pp. 333–338, 1999.

[69] A. Diószegi and I. Dugic, “The mechanisms of metal expansion penetration in grey cast iron,” in Proceedings of the Eight International Symposium on Science and Processing of Cast Iron, Beijing, China, 2006, pp. 92–97.

[70] K. Williamson, Research Methods for Students, Academics and Professionals: Information Management and Systems. Elsevier Science, 2002.

[71] S. Rumsey, How to Find Information: A Guide for Researchers. McGraw-Hill Education, 2004.

[72] H. Gahm and F. Jeglitsch, “Color methods and their application in metallography,” Met. Prog., vol. 120, no. 3, pp. 48–53, 1981.

58 | REFERENCES

[73] J. Kilpatrick, A. Benscoter, and A. Marder, “Tint Etching Improves Resolution And Contrast Microstructures,” Met. Prog., vol. 100, no. 6, pp. 79–81, 1971.

[74] E. Beraha and B. Shpigler, Color metallography. American Society for Metals, 1977.

[75] F. S. Lepera, “Improved etching technique for the determination of percent martensite in high-strength dual-phase steels,” Metallography, vol. 12, no. 3, pp. 263–268, 1979.

[76] G. F. Vander Voort, Metallography, Principles and Practice. ASM International, 1984.

[77] J. M. Motz, “Microsegregations - An easily unnoticed influencing variable in the structural description of cast materials,” Pract. Metallogr., vol. 25, pp. 285–293, 1988.

[78] S. Stan, M. Chisamera, I. Riposan, and M. Barstow, “Application of thermal analysis to monitor the quality of hypoeutectic cast irons during solidification in sand and metal moulds,” J. Therm. Anal. Calorim., vol. 110, no. 3, pp. 1185–1192, 2012.

[79] D. Emadi, L. V. Whiting, S. Nafisi, and R. Ghomashchi, “Applications of thermal analysis in quality control of solidification processes,” J. Therm. Anal. Calorim., vol. 81, no. 1, pp. 235–242, Jul. 2005.

[80] A. Diószegi and I. L. Svensson, “Interpretation of Solidification by Thermal Analysis of Cooling Rate,” Trans. Indian Inst. Met., vol. 58, no. 4, pp. 611 – 616, Aug. 2005.

[81] E. Fras, W. Kapturckiewicz, A. Burbielko, and H. F. López, “Numerical simulation and fourier thermal analysis of solidification kinetics in high-carbon Fe-C alloys,” Metall. Mater. Trans. B, vol. 28, no. 1, pp. 115–123, Feb. 1997.

[82] A. Diószegi and I. L. Svensson, “A comparison of Fourier vs. Newtonian thermal analyse and its influence on the inverse kinetic growth calculation,” in A. Diószegi: Dissertation No. 871, Linköping, Sweden, 2004, pp. 311–318.

[83] A. Diószegi and I. L. Svensson, “On the problems of thermal analysis of solidification,” Mater. Sci. Eng. A, vol. 413–414, pp. 474–479, Dec. 2005.

[84] D. Emadi, L. V. Whiting, M. Djurdjevic, W. T. Kierkus, and J. Sokolowski, “Comparison Of Newtonian And Fourier Thermal Analysis Techniques For Calculation Of Latent Heat And Solid Fraction Of Aluminum Alloys,” Metal. - J. Metall., vol. 2, no. 10, pp. 91–106, 2004.

REFERENCES | 59

[85] J. T. Svidró, A. Diószegi, and J. Tóth, “The novel application of Fourier thermal analysis in foundry technologies,” J. Therm. Anal. Calorim., vol. 115, no. 1, pp. 331–338, 2014.

[86] J. Campbell, “Shrinkage Pressure in Castings,” Trans. Metall. Soc. AIME, vol. 239, pp. 138–142, Feb. 1967.

[87] J. Campbell, “Hydrostatic Tensions in Solidifying Alloys,” Trans. Metall. Soc. AIME, vol. 242, pp. 268–271, Feb. 1968.

[88] J. Campbell, “Hydrostatic Tensions in Solidifying Metarials,” Trans. Metall. Soc. AIME, vol. 242, pp. 264–268, Feb. 1968.

[89] J. J. Sobczak, L. Drenchev, and R. Asthana, “Effect of pressure on solidification of metallic materials,” Int. J. Cast Met. Res., vol. 25, no. 1, pp. 1–14, 2012.

[90] J. C. Hernando and A. Diószegi, “An Overview of Isothermal Coarsening in Hypoeutectic Lamellar Cast Iron,” in Advances in the Science and Engineering of Casting Solidification: An MPMD Symposium Honoring Doru Michael Stefanescu, L. Nastac, B. Liu, H. Fredriksson, J. Lacaze, C.-P. Hong, A. V. Catalina, A. Buhrig-Polaczek, C. Monroe, A. S. Sabau, R. E. L. Ruxanda, A. Luo, S. Sen, and A. Diószegi, Eds. Cham: Springer International Publishing, 2016, pp. 295–302.

[91] J. C. Hernando, B. Domeij, and A. Diószegi, “Influence of Ti and Mo additions on the isothermal coarsening process of primary austenite in Lamellar Graphite Iron,” Present. 5th Decenn. Int. Conf. Solidif. Process. Old Windsor UK 25th‐28th July 2017.

[92] J. C. Hernando, E. Ghassemali, and A. Diószegi, “The Morphological Evolution of Primary Austenite During Isothermal Coarsening,” Mater. Charact., vol. 131, pp. 492–499, Sep. 2017.

[93] Y. A. Çengel and M. A. Boles, Thermodynamics: an engineering approach. McGraw-Hill Higher Education, 2006.

[94] P. Svidró and A. Diószegi, “On problems of volume change measurements in lamellar cast iron,” Int. J. Cast Met. Res., vol. 27, no. 1, pp. 26–37, Dec. 2013.

[95] R. Mallet, “On the Alleged Expansion in Volume of Various Substances in Passing by Refrigeration from the State of Liquid Fusion to That of Solidification,” Proc. R. Soc. Lond., vol. 23, pp. 209–234, Jan. 1874.

[96] F. Osmond, “Sur les phénomenes qui se produisent pendant le chauffage et le refroidissement de l’acier fondu,” Comptes Rendus, vol. ciii, p. 743 and 1135, 1886.

60 | REFERENCES

[97] F. Osmond, “On the Critical points of Iron and Steel,” Journal of the Iron and Steel Institute, vol. 1, pp. 38–80, 1890.

[98] W. J. Keep, “Cooling curves and test for cast iron,” Journal of the Iron and Steel Institute, vol. 2, pp. 227–248, 1895.

[99] T. Turner, “Volume and Temperature Changes during the Cooling of Cast Iron,” Journal of the Iron and Steel Institute, vol. 1, pp. 48–74, 1906.

[100] F. Wüst, “Die Schwindung von Metallen und Metalllegierungen,” Metallurgie, vol. 6, pp. 769–792, 1909.

[101] T. D. West, Metallurgy of cast iron. 1912. [102] A. Diefenthäler, “Die Ursachen der Lunkerung und ihr

Zusammenhang mit Schwindung und Gattierung,” Stahl und Eisen, vol. 32. II., pp. 1813–1819, Oct. 1912.

[103] F. Wüst and G. Schitzkowski, “Einfluß einiger Fremdkörper auf die Schwindung des Eisens,” Mitteilungen aus dem Kaiser-Wilhelm-Institut für Eisenforschung zu Düsseldorf, vol. 4, pp. 105–124, 1922.

[104] K. Bornemann and F. Sauerwald, “Dichtemessungen von Metallen und Legierungen,” Zeitschrift für Metallkunde, vol. 14, pp. 145–159, Apr. 1922.

[105] F. Sauerwald, “Über die Gesetzmässigkeiten der Volumengestaltung und Schwindung von Metallen und Legierungen,” Giesserei-Zeitung, vol. 20, pp. 391–393, 1923.

[106] P. Bardenheuer and C. Ebbefeld, “Beitrag zur Analyse des Schwindungsvorganges von weißem und grauem Gußeisen,” Mitteilungen aus dem Kaiser-Wilhelm-Institut für Eisenforschung zu Düsseldorf, vol. 6, pp. 45–60, 1926.

[107] K. Honda and H. Endo, “On the Volume Change in Cast Iron during Solidification,with a Criticism of the Double Diagram of the Iron-Carbon System,” Science Reports of the Tohoku Imperial University, vol. 1, pp. 9–25, 1927.

[108] P. Bardenheuer and W. Bottenberg, “Einfluß der Schmelzbehandlung auf den Gasgehalt und die Schwindung von weißem und grauem Gußeisen,” Mitteilungen aus dem Kaiser-Wilhelm-Institut für Eisenforschung zu Düsseldorf, vol. 13, pp. 149–159, 1931.

[109] W. A. Schmidt, E. Sullivan, and B. F. Brown, “Risering of Gray Iron Castings,” in Transactions of the American Foundrymen’s Society, vol. 62, 1954, pp. 70–77.

REFERENCES | 61

[110] A. Tatur, “La solidification des alliages légers, étude de la retassure,” Fonderie, vol. 116, pp. 4681–4693, Sep. 1955.

[111] M. Degois, “Solidification of Nodular Cast Iron Volume Changes - Effects of Different Parameters,” in The Metallurgy of Cast Iron, Geneva, Switzerland, 1974, pp. 741–752.

[112] J. C. Margerie, “Stages of Crystallization and Apparent Volume Variations During Solidification of Different Kinds of Cast Iron,” in The Metallurgy of Cast Iron, Geneva, Switzerland, 1974, pp. 723–739.

[113] D. M. Stefanescu, L. Dinescu, S. Crăciun, and M. Popescu, “Production of vermicular graphite cast-irons by operative control and correction of graphite shape.,” in Proceedings of the 46th International Foundry Congress, CIATF, Madrid, Spain, 1979, p. 37–37/15.

[114] A. Gazda and E. Zemlak, “Apparatus for Measurement of Dimensional Changes of Solidifying Casting Alloys,” Pr Inst Odlew, vol. 33, no. 2–3, pp. 159–173, 1983.

[115] R. Hummer, “A Study of the Shrinkage and Dilatation during Solidification of Nodular Cast Iron - its Relation to the Morphology of Crystallisation,” in MRS proceedings, 1984, vol. 34, pp. 213–222.

[116] G. Nándori and P. Jónás, “A tágulási erő mint az öntöttvasak és acélok minőségi jellemzője a kristályosodás folyamán,” Bányászati és Kohászati Lapok, Öntöde, vol. 12, pp. 169–173, Dec. 1988.

[117] Y. Yang and J. Alhainen, “Microstructure Estimation of SG Iron With a Dilatometer,” presented at the CASTCON ’91, 1991, pp. 1–14.

[118] Y. Yang and J. Alhainen, “Estimation of the Tendency Towards Shrinkage in Spheroidal-Graphite Iron Melts Using Thermal Analysis,” Cast Met. UK, vol. 5, no. 2, pp. 73–77, 1992.

[119] Y. Yang and J. Alhainen, “Functional Improvement of Dilatation Analysis on Thermal Analysis for S.G. Iron Quality Estimation,” in Proceedings of the 5th International Symposium on the Physical Metallurgy of Cast Iron, 1994.

[120] Z. Gedeonova, S. Bodi, J. Dúl, G. Nándori, and L. Vigh, “Displacement on the surface mould and metal during the solidification of nodular graphite iron castings,” Materials Science Forum, vol. Solidification and Gravity II (215-216), pp. 391–398, 1995.

62 | REFERENCES

[121] A. Kagawa, S. Kiguchi, and M. Osada, “Volumetric change in freezing cast irons,” Trans. Jpn. Foundrymens Soc. Jpn., vol. 14, pp. 18–23, 1995.

[122] P. Mrvar, M. Trbižan, and J. Medved, “Investigation of cast iron solidification with dilatation analysis,” Kovine Zlitine Tehnol. Slov., vol. 33, no. 1–2, pp. 45–49, 1999.

[123] J. Mutwil, “Device for the investigation of casting stresses,” Solidification of Metals and Alloys, vol. 2, no. 44, pp. 503–508, 2000.

[124] P. Mrvar, M. Trbižan, and J. Medved, “Examination of Dimensional Changes of Casting and Mould During the Solidification of Modified and Unmodified Aluminium Cast Alloys,” presented at the Metal 2002, Hradec nad Moravicí, 2002, pp. 1–8.

[125] J. Mutwil, “Design changes of device to investigation of alloys linear contraction and shrinkage stresses,” Archives of Foundry Engineering, vol. 9, no. 4, pp. 149–156, 2009.

[126] M. Chisamera, I. Riposan, S. Stan, P. Toboc, T. Skaland, and D. White, “Shrinkage evaluation in ductile iron as influenced by mould media and inoculant type,” Int. J. Cast Met. Res., vol. 24, no. 1, pp. 28–36, Feb. 2011.

[127] M. Chisamera, I. Riposan, S. Stan, and M. Barstow, “Simultaneous cooling and contraction / expansion curve analysis during ductile iron solidification,” in Transactions of the American Foundrymen’s Society, vol. 120, 2012, pp. 375–388.

[128] Y. Zou, Y. Hayashi, H. Kojima, K. Yamazaki, and H. Nakae, “Influence of Mold Material and CE Value on Shrinkage of Spheroidal Graphite Cast Iron,” presented at the 70th World Foundry Congress, Monterrey, Nuevo Leon, Mexico, 2012.

[129] G. Alonso, D. M. Stefanescu, R. Suarez, A. Loizaga, and G. Zarrabeitia, “Understanding graphite expansion during the eutectic solidification of cast iron through combined Linear Displacement and Thermal Analysis,” Int. Foundry Res., no. 4, pp. 18–29, 2014.

[130] G. Alonso, D. M. Stefanescu, R. Suarez, A. Loizaga, and G. Zarrabeitia, “Kinetics of graphite expansion during eutectic solidification of cast iron,” Int. J. Cast Met. Res., vol. 27, no. 2, pp. 87–100, Feb. 2014.

on the mechanisms of volume change related casting defects

Documents