KTH Industrial Engineering and Management

Microstructure Evaluation and Wear-Resistant

Properties of Ti-alloyed Hypereutectic High

Chromium Cast Iron

Qiang Liu

Doctoral Dissertation

Stockholm 2013

Division of Applied Process Metallurgy

Department of Materials Science and Engineering

Royal Institute of Technology

SE-100 44 Stockholm

Sweden

Akademisk avhandling som med tillstånd av Kungliga Tekniska Högskolan i

Stockholm, framlägges för offentlig granskning för avläggande av Teknologie

Doktorsexamen, fredagen den 27 september, kl. 13.00 i sal B1, Materialvetenskap,

Brinellvägen 23, Kungliga Tekniska Högskolan, Stockholm

ISBN 978-91-7501-842-3

Qiang Liu Microstructure Evaluation and Wear-Resistant Properties of Ti-alloyed

Hypereutectic High Chromium Cast Iron

Division of Applied Process Metallurgy

Department of Materials Science and Engineering

School of Industrial Engineering and Management

Royal Institute of Technology

SE-100 44 Stockholm

Sweden

ISBN 978-91-7501-842-3

© Qiang Liu (刘强), July 2013

Tryck: Universitetsservice US AB, Stockholm, Sweden

To my beloved parents

献给我的父亲母亲

Abstract High chromium cast iron (HCCI) is considered as one of the most useful wear

resistance materials and their usage are widely spread in industry. The mechanical

properties of HCCI mainly depend on type, size, number, morphology of hard

carbides and the matrix structure (γ or α). The hypereutectic HCCI with large volume

fractions of hard carbides is preferred to apply in wear applications. However, the

coarser and larger primary M7C3 carbides will be precipitated during the solidification

of the hypereutectic alloy and these will have a negative influence on the wear

resistance.

In this thesis, the Ti-alloyed hypereutectic HCCI with a main composition of

Fe-17mass%Cr-4mass%C is studied based on the experimental results and calculation

results. The type, size distribution, composition and morphology of hard carbides and

martensite units are discussed quantitatively. For a as-cast condition, a 11.2μm border

size is suggested to classify the primary M7C3 carbides and eutectic M7C3 carbides.

Thereafter, the change of the solidification structure and especially the refinement of

carbides (M7C3 and TiC) size by changing the cooling rates and Ti addition is

determined and discussed. Furthermore, the mechanical properties of hypereutectic

HCCI related to the solidification structure are discussed.

Mechanical properties of HCCI can normally be improved by a heat treatment process.

The size distribution and the volume fraction of carbides (M7C3 and TiC) as well as

the matrix structure (martensite) were examined by means of scanning electron

microscopy (SEM), in-situ observation by using Confocal Laser Scanning Microscope

(CLSM), Transmission electron microscopy (TEM) and electron backscattered

diffraction (EBSD). Especially for the matrix structure and secondary M7C3 carbides,

EBSD and CLSM are useful tools to classify the fcc (γ) and bcc (α) phases and to

study the dynamic behavior of secondary M7C3 carbides. In conclusion, low holding

temperatures close to the eutectic temperature and long holding times are the best heat

treatment strategies in order to improve wear resistance and hardness of Ti-alloyed

hypereutectic HCCI.

Finally, the maximum carbides size is estimated by using statistics of extreme values

(SEV) method in order to complete the size distribution results. Meanwhile, the

characteristic of different carbides types will be summarized and classified based on

the shape factor.

Keywords: High Chromium Cast Iron, cooling rate, Ti addition, M7C3, TiC, carbides

size distributions, volume fraction, heat treatment, microstructure, mechanical

properties, wear resistance, statistics of extreme values (SEV), maximum carbides

size.

Acknowledgements

I would like to take this opportunity to thank the people who have contributed and

supplied the help to this work.

All of first, I would like to express my sincere gratitude to two great supervisors

Professor Keiji Nakajima and Professor Pär Jönsson. Your endless support,

enthusiasm, knowledge and scientific discussion are very helpful for me during my

study and growth at KTH, Sweden. Your guidance makes benefit greatly on my

research life.

I would like to thank Professor Qiang Wang, associate Professor Hongwei Zhang,

associate Professor Kai Wang and Mr. Xiangkui Zhou from Northeastern University,

China. Many thanks for yours simulation support and experiment help.

I would like to say many thanks to Professor Hiroyuki Shibata at Tohoku University,

Japan. Thank you very much for taking care of me all the time when I was in Japan.

Your knowledge and experiences are really helpful to my work. I also want to say

thanks to assistant Professor Junichi Takahashi (Tohoku University) for helpful

discussion and nice suggestions on the statistical analysis. Thanks also to Akifumi

Harada for the help of SEM measurement and teaching me many Japanese words at

Tohoku University.

I also appreciate the help and support from Dr. Peter Hedström, KTH. Thank you very

much for your experiment support, very nice discussion and many good suggestions.

Thanks also to Ms. Wenli Long, who help me a lot on the technical problem at KTH

lab. The same thanks to associate Prof. Lyubov Belova (KTH) for the help with TEM

sample preparation by using Focused Ion Beam (FIB) technique. Mr. Oskar Karlsson

from Swerea KIMAB is acknowledged, who has been very helpful on EBSD

measurements.

Many thanks for the help from the colleagues at the Department of Material Science

and Engineering (MSE), KTH. I have a lot of fun from you during sharing Chinese

food, ice-cream time, basketball time and innebandy time. Special thanks to Erik

Roos for teaching me Swedish and your friendly help and care, especially when I was

in the hospital. Special appreciations are also to my Chinese friends: Rongzhen, Jun,

Zhili, Zhi, Yanyan, Ni, Xiaolei, Mei, Mu, Ying, Changji, Xiaobin, Bu, many thanks

for your kindly help and happy moments during living in Stockholm.

All colleagues in the division of TPM are thanked for your help and friendship, many

thanks to Dr. Mikael Ersson for introducing the innebandy and Mjöd to me. It is

very nice experiences for me to feel the Swedish culture.

I would like to acknowledge to China Scholarship Council for the financial support.

Finally, I wish to thank my family: my sister, Yuanyuan, thank you very much for

taking care of mother and father when I was abroad, I owe you a lot; Mama and Baba,

thank you very much for your endless love and support.

Qiang Liu, Stockholm, July 2013.

Preface

This is a systematic study from melting and solidification to heat treatment of a wear

resistant high chromium cast iron. The results on Ti-alloyed hypereutectic high

chromium cast iron can be widely be applied to improvements of wear resistant parts

in mineral industry plants, power plants, steel making plants, and others. Moreover,

for the development of new superior wear resistant materials. I wish you can get some

`hint` when you read this thesis, which is intend to contribute with new knowledge in

the production of wear parts.

The present thesis is based on the following supplements:

Supplement I:

“Effect of Cooling Rate and Ti Addition on the Microstructure and Mechanical

Properties in As-Cast Condition of Hypereutectic High Chromium Cast Irons”

Qiang Liu, Hongwei Zhang, Qiang Wang, Xiangkui Zhou, Pär G. Jönsson and Keiji

Nakajima

ISIJ International, Vol. 52(2012), No. 12, pp. 2210-2219.

Supplement Ⅱ:

“Solidification and Phase Transformations of Ti-added Hypereutectic High

Chromium Cast Irons”

Qiang Liu, Hongwei Zhang, Qiang Wang, Xiangkui Zhou, Pär G Jönsson and Keiji

Nakajima

ICRF 2012, 1st International Conference on Ingot Casting, Rolling and Forging,

05.06.2012, Aachen. Düsseldorf: Stahl-Zentrum 2012, ICRF-31.pdf (CD-ROM).

Supplement Ⅲ:

“Effect of Heat Treatment on Microstructure and Mechanical Properties of

Ti-alloyed Hypereutectic High Chromium Cast Iron”

Qiang Liu, Peter Hedström, Hongwei Zhang, Qiang Wang, Pär G. Jönsson and Keiji

Nakajima

ISIJ International, Vol. 52(2012), No. 12, pp. 2288-2294.

Supplement Ⅳ:

“Dynamic Precipitation Behavior of Secondary M7C3 Carbides in Ti-alloyed

High Chromium Cast Iron”

Qiang Liu, Hiroyuki Shibata, Peter Hedström, Pär Göran Jönsson and Keiji Nakajima

ISIJ International, Vol. 53(2013), No. 7, pp. 1239-1246.

Supplement Ⅴ:

“Estimation of the Maximum Carbide Size in a Hypereutectic High Chromium Cast

Iron Alloyed with Titanium”

Qiang Liu, Pär Göran Jönsson and Keiji Nakajima

Accepted by ISIJ International for publication, Vol. 53(2013), No. 12.

The contributions by the author to the different supplements of the dissertation:

Supplement I: The literature survey, a major part of the experimental work, and a

major part of the writing.

Supplement Ⅱ: The literature survey, a major part of the experimental work, and a

major part of the writing.

Supplement Ⅲ: The literature survey, a major part of the experimental work, and a

major part of the writing.

Supplement Ⅳ: The literature survey, a major part of the experimental work, and a

major part of the writing.

Supplement Ⅴ: The literature survey, a major part of the experimental work, and a

major part of the writing.

Parts of the work have been presented at the following conference:

“Estimation of the Maximum Carbide Size in a Hypereutectic High Chromium Cast

Iron Alloyed with Titanium”

Qiang Liu, Pär Göran Jönsson and Keiji Nakajima.

3rd

International Symposium on Cutting Edge of Computer Simulation of

Solidification, Casting and Refining (CSSCR2013), May 20-23, 2013, Stockholm,

Sweden and Helsinki, Finland.

Contents

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

1.1 Background ................................................................................................................................... 1

1.2 The purpose and overview of thesis .............................................................................................. 2

Chapter 2 Methodology .............................................................................................................. 5

2.1 Experiment methods ...................................................................................................................... 5

2.1.1 Melting and solidification process .......................................................................................... 5

2.1.2 Heat treatment process ........................................................................................................... 8

2.1.3 Microstructure observations and image analysis .................................................................. 10

2.1.4 Mechanical properties measurement .................................................................................... 14

2.2 Theoretical works ........................................................................................................................ 14

Chapter 3 Results and Discussion ........................................................................................... 17

3.1 Phase diagram and solidification path of Ti-alloyed hypereutectic HCCI .................................. 17

3.1.1 Phase diagram ....................................................................................................................... 17

3.1.2 Solidification path ................................................................................................................ 19

3.2 Microstructure of Ti-alloyed hypereutectic HCCI ...................................................................... 20

3.3 Effect of cooling rate on as-cast structure ................................................................................... 21

3.4 Effect of Ti addition on as-cast structure..................................................................................... 23

3.5 Effect of Ti addition and cooling rate on the volume fraction in as-cast condition of the

hypereutectic HCCI ........................................................................................................................... 27

3.6 Classification of M7C3 carbides for as-cast condition ................................................................. 29

3.7 Effect of cooling rate and Ti addition on the mechanical properties in as-cast condition

of the hypereutectic HCCI ................................................................................................................. 33

3.8 Evolution of matrix after heat treatment...................................................................................... 35

3.9 Evolution of carbides after heat treatment ................................................................................... 39

3.9.1 Evolution of primary and eutectic M7C3 carbides after heat treatment ................................ 39

3.9.2 Evolution of TiC carbides after heat treatment..................................................................... 42

3.9.3 Evolution of secondary carbides during heat treatment ....................................................... 44

3.10 Mechanical properties improved by heat treatment .................................................................. 52

3.11 Estimation of the maximum carbide size by using SEV method .............................................. 54

3.11.1 Estimation of maximum primary M7C3 carbides size ........................................................ 54

3.11.2 Estimation of maximum TiC carbides size ......................................................................... 57

3.11.3 Estimation of maximum secondary M7C3 carbides size ..................................................... 58

Chapter 4 Concluding Discussion ........................................................................................... 63

Chapter 5 Conclusions and Future Work ............................................................................... 69

References ................................................................................................................................ 73

-1-

Chapter 1

Introduction

1.1 Background

The Fe-Cr-C system alloys with high carbon have widely been applied to wear

resistant parts in steel making plants, power plants, mineral industry plants, and others

[1, 2]. In all industrial fields, however, the improvement of the abrasion

wear-resistance is strongly required because of its short service life.

In the first half of the 1980s, High Chromium Cast Iron (HCCI) with a composition

range of 2.0-4.3%C [2] and 12-30%Cr [3] were started to be applied to roll materials

for hot rolling. The wear performance level of hypereutectic HCCI is mainly

controlled by a high volume fraction (30 to 60%) of M7C3 type carbides (1200~1600

HV) [4], M23C6 type carbides (1600-2520HV) [5], MC type carbides (2400-4000HV)

[5], etc. These carbides will precipitate during the solidification and heat treatment,

which lead to an improvement of the final material properties.

Compared to the hypoeutectic HCCI, the hardness of hypereutectic HCCI is obviously

improved via a higher carbon and chromium content. This is due to that carbonization

inside the hypereutectic HCCI highly increased. Moreover, these higher additions will

also cause an increase of the volume fraction of carbides. However, the coarser and

larger primary carbides will be precipitated during the solidification of the

hypereutectic HCCI. These coarser primary carbides have a negative effect on the

improvement of the hardness and wear resistance of all alloys. Therefore, many

investigations have been executed with the aim of refining the solidification structure

and improving the properties of hypereutectic HCCI. Here, additions of alloy

-2-

elements, such as Ti, Nb, V, Zr etc. [6-13], have been used to form the NaCl type

carbides. Those elements, which are strong carbide forming elements, will react with

carbon and form NaCl type carbides, such as TiC, NbC, VC etc. Also, these can act as

heterogeneous nuclei for the precipitation of M7C3 type carbides during the cooling

and solidification of the melt. [6-8] This, in turn, results in significant refinements of

the final carbides size.

Furthermore, for many applications, a higher wear resistance is required to improve

the service life of the wear components. In comparison to the as-cast alloy, the

heat-treated alloy with higher wear resistance is prior to service. The reason is that the

matrix changes from an austenitic structure (fcc, γ) into a martensitic structure (bct, α’)

and new carbide precipitates (M7C3 [3, 14] or M23C6 [15-17] form within the matrix.

Therefore, many studies [4, 14-15, 18-22] focusing on improving the mechanical

properties of cast iron, by alternative heat treatment processes, have been reported.

Most of the previous studies are focused on identifying the new carbides by using

transmission electron microscopy (TEM). [14, 16-20, 22-26] They have furthermore

studied how the new carbides influence the final properties.

At present, few reports discuss how the carbide size distribution influences their final

solidification structure and properties. However, they only show the effects of the

average diameter and volume fraction of the total amount of carbides on the material

properties. [6-9, 11] In addition, for heat treated HCCI, only a few studies

quantitatively discuss how the size distribution of carbides and martensite units affect

the final properties of hypereutectic HCCI. However, we must point out that the size,

number, volume fraction and distribution of these carbides and martensite units will

also affect the final properties of the hypereutectic HCCIs significantly. Hence, a

quantitative study on the carbides size distribution and how these carbides influence

on mechanical properties is desired.

1.2 The purpose and overview of thesis

In this thesis, we quantitatively study the size distribution, composition and

-3-

morphology of carbides and martensite units for as-cast condition and heat treatment

condition of Ti-alloyed hypereutectic HCCI with main composition of

Fe-17mass%Cr-4mass%C based on following five supplements. An outline of this

thesis is given and organized as follows:

Fig. 1.1. The scope and focus of different supplements in the present thesis.

As shown in Fig. 1.1, the content of this thesis is mainly divided into three parts to

study the relationship between the microstructure and the final mechanical properties

of a hypereutectic HCCI alloyed with titanium. The first one is the as-cast part, which

is related to the melting and solidification of hypereutectic HCCI alloys during the

casting process. It corresponds to Supplement I [27] and Supplement Ⅱ [28]. The

solidification path, the microstructure refined by different cooling rates and titanium

additions, and the mechanical properties of hypereutectic HCCI were studied both

experimentally and theoretically.

The second part of the thesis is focus on heat treatment, which is related to the matrix

phase transformation and the secondary carbides precipitation during the heat

treatment process. It corresponds to Supplement Ⅲ [29] and Supplement Ⅳ [30].

-4-

The relationship between carbide precipitation, martensite units and mechanical

properties are discussed based on the volume fraction and the size distribution of

carbide precipitates (M7C3 carbides and TiC carbides) as well as the martensite units.

A possible mechanism is finally proposed to explain the improvement of mechanical

properties through the heat treatment process. In addition, the best heat treatment

strategy is suggested.

In the final part of thesis, which is related to Supplement Ⅴ [31], more accurate

carbides size distribution results were obtained by supplementing the maximum

carbides size information to the normal size distribution. Moreover, the combined

method between a size distribution analysis and a maximum analysis is proposed.

Then, the relationships between the carbides size distribution including the maximum

carbides size and the mechanical properties can be fully explained by synthetically

consider evaluation of carbides and matrix.

-5-

Chapter 2

Methodology

In the present thesis, the solidification process, the carbides precipitation behavior, the

phase transformations as well as these phenomena influence on the mechanical

properties of a Ti-alloyed hypereutectic HCCI were studied both experimentally and

theoretically. However, the experimental studies were taken as the major research

tools in this thesis. The detail information about experiment methods and theoretical

methods will be introduced as following sections.

2.1 Experiment methods

2.1.1 Melting and solidification process

Based on the requirements during actual production, melts were prepared in a

laboratory furnace. high-carbon ferro-chromium (Fe-8mass%C-60mass%Cr),

ferro-molybdenum (Fe-0.05mass%C-58.4mass%Mo), pure iron, pig iron

(Fe-3.8mass%C-0.34mass%Mn-1.8mass%Si), ferro-titanium(Fe-30mass%Ti),

ferro-manganese(Fe-75.9mass%Mn), and pure nickel were melted in a 8 kg capacity

graphite crucible. This was done using a medium frequency induction furnace (45Kw,

7k Hz) and using an air atmosphere. The alloy compositions are shown in Table 2.1.

Pure iron, pig iron and high carbon ferro-chromium were first melted. Then,

ferro-manganese and ferro-molybdenum were added. Finally, pure nickel and

ferro-titanium were added into this mother alloy.

The holding time over the TiC precipitation temperature was kept as a value of

-6-

Table 2.1 Chemical compositions of HCCI used in the experiments. (mass.%)

C Cr Ti Mn Mo Ni Si N S O P Fe

Sand

4.01 16.2 - 2.04 1.01 1.24 0.775 0.020 0.018 0.0086 0.026 Bal.

3.79 17.0 1.53 2.08 1.05 1.09 0.416 0.016 0.017 0.014 0.025 Bal.

Graphite

4.9 16.4 - 1.59 0.75 1.09 0.845 0.019 0.022 0.0046 0.025 Bal.

4.03 15.6 0.74 1.72 0.93 1.08 0.992 0.014 0.020 0.0047 0.023 Bal.

4.01 17.4 1.48 1.91 0.89 1.11 0.951 0.015 0.017 0.011 0.024 Bal.

3.91 17.5 3.36 1.95 1.01 1.08 0.966 0.025 0.025 0.031 0.022 Bal.

Metal

3.92 14.7 - 1.950 0.95 1.05 0.529 0.015 0.020 0.0037 0.024 Bal.

3.90 16.8 1.06 1.95 0.99 1.15 0.708 0.014 0.014 0.0094 0.024 Bal.

3min±1min, in order to prevent the agglomeration of TiC carbide precipitation.

Finally, the molten hypereutectic HCCI were poured into the mold at a temperature

value of 1450±10°C.

Fig. 2.1. Schematic diagram of melting process of hypereutectic HCCI with 1.5mass% Ti addition and

casting in the graphite mold.

Composition

s Mold

-7-

The following three mold types were used: sand mold (silica sand), graphite mold,

metal mold (ASTM: 1045). A schematic diagram of the melting process is shown in

Fig. 2.1.

The dimension of the ingot is 120mm (L) ×84mm (W) ×55mm (H). A schematic view

of the ingot and the thermocouple (Type-B) position is shown in Fig. 2.2.

Fig. 2.2. Schematic view of ingot and cutting position of sample for microstructure observation.

0 100 200 300 400 500 600 7001000

1100

1200

1300

1400

1500

Sand mold Graphite mold Metal mold

25oC/min 80

oC/min 110

oC/min

Time, (s)

Tem

pera

ture

, (o

C)

Sand mold

Metal mold

Graphite mold

Fig. 2.3.Typical cooling curves measured in three kinds of molds for hypereutectic HCCI with

1.5mass% Ti addition.

-8-

In addition, the cooling curves obtained from the three different molds are shown in

Fig. 2.3. From this information the following different cooling rates were obtained:

25oC/min, 80

oC/min and 110

oC/min within 10% errors for the sand mold, graphite

mold and metal mold, respectively.

2.1.2 Heat treatment process

In this thesis, the heat treatment process was carried out in two ways: one using a

conventional heat treatment, which is mainly focused on an evaluation of the matrix

and carbides after heat treatment. Another way is by using the in-situ observations,

which mainly focused on studying the dynamic precipitation behavior of secondary

carbides during the whole heat treatment process from heating and holding to cooling.

2.1.2.1 Heat treatment for matrix and carbides evaluation

The specimens used for the heat treatments were selected from the ingots, which were

casted in a graphite mold with a 1.5mass% titanium addition. Samples of the

dimensions 10mm (L) ×10mm (W) ×10mm (H) were cut and then heat-treated at a

furnace temperature of 900oC, 1000

oC and 1050

oC for 2hr and 6hr respectively. The

0

200

400

600

800

1000

1200

200000

10 oCmin10

oCmin

906oChr

1008oChr

1058oChr

air cooling

1005oChr

903oChr

T

em

pe

ratu

re,

(oC

)

Time, (s)

1057oChr

air cooling

10000 200000 10000

Fig. 2.4. Histories of sample surface temperatures.

-9-

samples were air cooled from the elevated temperature to room temperature. The

surface temperature of the sample during heat treatment was measured and calibrated

by using a R-type thermocouple. The temperature histories of sample surface are

shown in Fig. 2.4. It was found that the difference between the furnace temperature

and the actual temperature of the sample surface was less than 10 oC.

2.1.2.2 In-situ observation for secondary carbides evaluation

The specimens, which were used for in-situ observations, were selected from the

ingots that were casted in the graphite mold for HCCI grades with 1.5mass% titanium

additions. It is the same as the above heat treatment process. The cast ingots were cut

into small pieces (3mm (L) × 2mm (W) × 2mm (H)). Each specimen was carefully

polished following the preparation procedure as is described in Table 2.2.

Table 2.2 Sample preparation for in-situ observation and EBSD measurement

Step

Disk name and type No.

(Company: BUEHLER)

Abrasive Size and

suspension

Load(N)

Speed(rpm) and

Direction

Time

(minutes)

Step 1

ApexHercules® S rigid

No.41-27400-412-001

9μm

diamond suspension

30

150

(Comply)

5

Step 2

MicroFloc

No.40-8322

3μm

diamond suspension

30

150

(Comply)

5

Step 3

ChemoMet®

No.40-7922

0.05μm

colloidal silica

30

150

(Contrary)

2

The polished specimens were set into a high purity alumina crucible (Φ5.5mm O.D. ×

Φ4.5mm I.D. ×5mm height) and study in a Confocal Laser Scanning Microscope

(CLSM) equipped with an infrared heating furnace. The wavelength of 1.5mW He-Ne

laser, which is used in CLSM setup, is 632.8nm. A detailed description of the CLSM

setup can be found in previous works. [32, 33] The specimens were heated to 890°C

-10-

and 1026°C using a 10°C/min heating rate and thereafter they were held for 2 and 6

hours, respectively. The atmosphere in the CLSM was pure Ar; moreover, a titanium

foil was wound around the upper part of the crucible to avoid oxidation of the

specimen surface during the experiments. Once the desired temperature and holding

time was reached, the specimen was cooled at a cooling rate of 90°C/min. The furnace

temperature was measured by a PtRh30%–PtRh6% thermocouple (type B), which was

inserted at the bottom of the crucible. In this case, the actual specimen temperature

was calibrated by using a Pt–PtRh13% thermocouple (type R) through welding of a R

type thermocouple wire on the surface of the specimen. It was found that the

temperature difference between the specimen surface and the furnace was 13°C and

31°C at 890°C and 1026°C, respectively. The temperature history of the specimen

surface and furnace is illustrated in Fig. 2.5.

0

100

200

300

400

500

600

700

800

900

1000

1100

(Specimen:890oC)

(Specimen:1026oC)

100500

10oC/min

Difference:31oC Furnace

Specimen surface

Difference:13oC

Tem

pera

ture

, (o

C)

Time, (min)

90oC/min

150 100500 150

90oC/min

10oC/min

Fig. 2.5. Temperature history of specimen surface and furnace during the heat treatment process.

2.1.3 Microstructure observations and image analysis

2.1.3.1 Carbide analysis by LOM and SEM

The as-cast samples were cut at a position close to the thermocouple measurement

position, as shown in Fig. 2.2. Thereafter, the as-cast samples were polished and

etched by using a solution of 5g FeCl3+10ml HCl+100ml ethanol. Then, the size

-11-

distribution of carbides (primary M7C3, eutectic M7C3, secondary M7C3 and TiC) was

measured on photographs obtained by using a light optical microscope (LOM) (Leica

DMR) at a ×10 magnification. This is done both for primary M7C3 carbides and

eutectic M7C3 carbides, where one pixel represented 0.3356μm. Back scattered

electron (BSE) setting was used in the scanning electron microscopy (SEM) (Hitachi:

S-3700N) together with a ×600 magnifications when studding TiC carbides, where

one pixel represented 0.1656μm. Moreover, the BSE mode was used in SEM

(FE-SEM, HITACHI, S-4800 type and FEG, JEOL JSM-7401F) along with a ×10000

magnification to study secondary M7C3 carbides, where one pixel represented

0.0093μm. The composition of carbides including Fe, Cr, Ti, Mo, Mn and other

elements were determined with the Energy Dispersive X-ray spectrometer (EDX)

method. The number of carbides per unit area, NA, was calculated using Equation

(2-1).

NA (i) = n (i) /Sobs (2-1)

where n(i) is the counted number of carbides for each selected step (i) of the size

distribution. Moreover, Sobs is the total area which was observed on the photographs,

which were taken by LOM (Sobs=0.3152mm2 for primary M7C3 carbides and eutectic

M7C3 carbides) and SEM (Sobs=0.0314 mm2 for TiC carbides and Sobs=107.78μm

2 for

secondary M7C3 carbides) respectively. The size of carbides, d, and the volume

fraction of carbides were calculated as the equivalent diameter of a circle by using an

image analyzer, which was the commercial software WinROOF. The size

distributions for carbides and martensite unit were plotted using a logarithmic scale.

The value of size d(j), for the j-th step, ∆(j), of a log-normal distribution can be

determined as follows:

d(j) or D(j) = 10∆(j) (2-2)

where ∆(j) is the width of the j-th step for log-normal distributions (∆(j)= ∆(j-1)+0.15).

This range is from -0.15 to 2.4 for primary M7C3 carbides and eutectic M7C3 carbides

(d), from -1.05 to 2.1 for TiC carbides (d) and from -2.25 to 2.4 for secondary M7C3

-12-

carbides and martensite unit (D). In this case, the mean value for the j-th step of the

size distribution, d(j) or D(j) varies from 0.85 to 214.50μm, from 0.11 to 107.51μm,

and from 0.0068 to 214.5038μm respectively.

2.1.3.2 Carbide and matrix analysis by EBSD

An electron backscattered diffraction (EBSD) analysis was carried out to identify the

phases in the specimens. [34, 36] Austenite (fcc, γ), martensite (bct, α’) and carbides

(M23C6, M7C3 and TiC) were included in the analysis. The flatness and surface quality

of specimens is crucial for EBSD analysis [35] in order to properly index phases such

as e.g. martensite. [35] In this study, the specimens used for EBSD analysis were

polished; first by manual polishing using SiC abrasive papers (P240-P320) together

with water, secondly with 9μm and 3μm diamond suspensions. Finally, specimens

were polished with a 0.05μm colloidal SiO2 by automatic polishing. The detailed

preparation procedure for the hypereutectic HCCI samples is shown in Table 2.2.

The EBSD measurements were performed using a LEO 1530 FEG (Field Emission

Gun)-SEM equipped with a HKL system and the HKL Channel 5 software. The

specimens were tilted 70o towards the detector and the microscope was operated at an

acceleration voltage of 15kV, in high current mode, and using an aperture of 120μm.

Two different step sizes of 0.1μm and 0.03μm were used. The area scanned with a

0.1μm step size was 200μm×150μm and the area scanned with a step size 0.03μm was

10μm×10μm. The reason for using two step sizes was the rather broad size

distribution. The size and volume fraction of carbide and martensite units were

determined by using image analysis made on EBSD maps.

The EBSD measurements were further complemented by energy dispersive X-ray

spectroscopy (EDS) mapping using the INCA software. The EDS analysis was

conducted using a 20kV acceleration voltage and a 30μm aperture during operation in

the high current mode. The EDS analysis enabled a more sound identification of the

phases since both crystal symmetry (EBSD) as well as chemical composition (EDS)

could be considered for the observed microstructural features.

-13-

The number of carbide precipitates and martensite units per unit area (NA) was

calculated by Equation (2-1). In this case, the parameter Sobs is the total area of the

micrographs, which was 0.0314 mm2

for the SEM determinations. For EBSD the

following values were used: Sobs=0.0001mm2 (0.03μm step size) and Sobs=0.03mm

2

(0.1μm step size). The size of martensite units (D) was calculated as the equivalent

diameter of a circle, which has the same area as the measured martensite unit.

2.1.3.3 Secondary carbide identification by TEM

Thin-foil specimens (~40nm) after heat treatment, which were prepared by focused

ion beam (FIB)-SEM (FEI, Nova 600) technique, were used to identify the type of

secondary carbides. Transmission electron microscopy was conducted using a TEM,

JEOL JEM-2100F, at an operating voltage of 200 kV. Moreover, the secondary

carbides type was evaluated as a M7C3 type in the Ti-alloyed hypereutectic HCCI

grades.

2.1.3.4 Estimation of the maximum carbides size by SEV method

In present thesis, the statistics of extreme values (SEV) method was used to predict

the maximum carbides size in a Fe-17mass% Cr-4mass% C hypereutectic HCCI. The

analysis procedure of the SEV was performed according to Murakami s method

[37-40] and following the ASTM E2283-03 standard. [41] The as-cast samples with

different titanium additions, casting in the different molds and heat treated at different

conditions were used to determine the maximum carbides size. The size of a unit area

for primary M7C3 carbides and TiC carbides is 0.505mm2 (LOM) and 0.017mm

2

(SEM), respectively. This corresponds to a ×5 magnification micrograph (one pixel

represented 0.555μm) in a LOM and a ×400 magnification micrograph (one pixel

represented 0.124μm) in a SEM, respectively. In addition, the size of a unit area for

secondary M7C3 carbides is 107.8μm2

at a magnification of ×10000 and 420.3μm2

at

magnification of ×5000. The maximum carbide size was measured in each unit area.

Totally, 40 unit areas were examined for the carbide extreme value analysis according

to the suggestions in the Murakami’s method for the inclusion. [42] The total

observed area for primary M7C3 carbides TiC carbides and secondary M7C3 carbides

-14-

was 20.20mm2, 0.692mm

2, and 0.106mm

2, respectively. The maximum size of

carbides, dAmax, was calculated as the equivalent diameter of a circle by using the

WinROOF image analyzer, which is commercial software.

2.1.4 Mechanical properties measurement

The mechanical properties: bulk hardness and wear loss of the hypereutectic HCCI

were investigated in the as-cast condition and in the heat treatment condition. The

bulk hardness was measured by using a Rockwell type hardness tester, HR-150A with

the Rockwell C scale (HRC). Overall, eight measurements were done for each sample.

The wear resistance was evaluated by using a single platform rotary abraser (model:

ML-100 type, Zhangjiakou Taihua Machine Factory). The wear tests were performed

at a platform speed of 60rpm under a load of 100N. The size of the wear test samples

was 4.1mm (φ) × l0mm (H). The tests were performed on a 80 mesh Al2O3 abrasive

paper at room temperature. The total sliding distance and the testing time of the

samples on the abrasive paper were 9m and 25s, respectively. The mass loss of the

specimens due to abrasion was calculated by measuring the weight of the specimens

before and after tests with a 0.001mg accuracy. Three measurements were carried out

per sample.

2.2 Theoretical works

In this thesis, the phase diagram and volume fraction of carbides in HCCI were

predicted numerically by coupling a micro-segregation model with thermodynamic

equilibrium calculations at solid/liquid interfaces.

The Scheil model, which is based on an assumption of no diffusion in solids, indicates

that the new phases always precipitate from the residual liquid in a solidification

sequence. Furthermore, the already formed solids remain frozen.

The Partial Equilibrium (PE) model distinguishes the diffusivities of substitutionals as

negligible and interstitials as complete in solids. In the present HCCI system, only C

behaves as an interstitial. Then the compositions and the amounts of the phases are

-15-

adjusted based on equalization of the chemical potentials of the interstitials，following

the equations below:

1SL

C Cµ µ (2-3)

The mass conservation of the interstitial component C before and after partial

equilibrium:

1 1

1

1

( ) ( )0

1 ( ) 1 ( )

S SL LSLC C C C

SL

C C

w w w wf f

w w

(2-4)

The unchanged u-fraction of substitutional components such as Fe element before and

after partial equilibrium in each phase:

1 ( )( )

1

LL LCFe FeL

C

ww w

w

(2-5)

1

1 1

1

1 ( )( )

1

SS SCFe FeS

C

ww w

w

(2-6)

The mass conservation of the substitutional components before and after partial

equilibrium:

1( )

1 ( )

LL LC

L

C

wf f

w

(2-7)

1

1 1

1

SS S

S

1( )

1 ( )

C

C

wf f

w

(2-8)

The chemical potential of C in a phase is determined through phase equilibrium

calculations, based on data of the phase composition, temperature and pressure. In the

above equations,L

Cw , ( )L

Cw are the mass compositions of carbon in the liquid before

and after partial equilibrium, 1S

Cw , 1( )

S

Cw are the corresponding averaged mass

composition of C in solid 1S .

1 1 1 1* ( ) /S S S Sk k

C C

k k

w w f f (2-9)

-16-

where 1* S

Cw is the newly formed composition in S1 at the k th step. The

parametersLf , 1Sf ( 1S

kk

f ) are the mass fractions of the liquid and the solid phases

1S at k th step, respectively.

The partial equilibrium among several phases (i.e. liquid and solid phase j =1, m) is

regarded as equivalent to several partial equilibriums between two phases (e.g. liquid

and one of the solids,1S ). Therefore, the mass conservation of the interstitial

component among all phases follows equation (2-10) instead of equation (2-4),

( ) ( )0

1 ( ) 1 ( )

j j

j

j

S SL LSLC C C C

SLj

C C

w w w wf f

w w

(2-10)

The calculations are based on the ideas of Chen and Sundman. [43, 44] They are

performed in Thermo-Calc/TQ interface [45] by a coupling of thermodynamic

equilibrium calculations with the alloy database TCFE6. [46] A temperature step

of -1oC was used in the calculations. The calculations by the PE and Scheil models

were stopped when the liquid fraction in the mass was less than 10-4

.

-17-

Chapter 3

Results and Discussion

The results in the present thesis include both experimental studies and theoretical

studies. These will be shown in the following sequence in this chapter: firstly, the

phase diagram and solidification path of Ti-alloyed hypereutectic HCCI will be

presented, it corresponds to section 3.1; secondly, the solidification (as-cast) structure

evolution and its influence on the mechanical properties of as-cast Ti-alloyed

hypereutectic HCCI will be concerned, it corresponds from section 3.2 to section 3.7;

then, the relationship between heat treatment structure evolution and final mechanical

properties was discussed from section 3.8 to section 3.10; finally, in the section 3.11,

the maximum carbides size will be discussed in order to compete the size distribution

results. Meanwhile, the characteristic of different carbides types will be summarized

and classified based on the shape factor.

3.1 Phase diagram and solidification path of Ti-alloyed hypereutectic

HCCI

3.1.1 Phase diagram

The phase diagram of Ti-alloyed hypereutectic HCCI, which is based on the sample

composition given in Table 2.1, is created by thermodynamic equilibrium calculations

by using the Thermo-Calc©

software. [45, 46] The phase diagrams of the

hypereutectic HCCI with different Ti content are shown in Fig. 3.1. It was found that

the solidification was initiated with a formation of primary M7C3 carbides, which

began to precipitate at 1337°C, Thereafter, the eutectic reaction L γ+M7C3 took

-18-

Fig. 3.1. Phase diagram of the hypereutectic HCCI with (a): 0mass%, (b): 0.75mass%, (c):1.5mass%

and (d): 3.0mass% Ti addition by using Thermo-Calc calculation. A zone: LIQUID; B zone: LIQUID+

TiC (FCC_A1#2); C zone: LIQUID+M7C3; D zone: LIQUID+ γ (FCC_A1)+TiC (FCC_A1#2)+M7C3;

E zone: γ (FCC_A1)+M7C3.

place during the cooling from 1256°C to 1215°C. This is valid for the case without Ti

additions, as illustrated in Fig. 3.1(a). On the other hand, in the case with a Ti addition,

the precipitation sequence is changed as shown in Fig. 3.1(b), 3.1(c) and 3.1(d). In the

case of a 1.5mass% Ti addition, FCC_A1#2 (corresponds to the NaCl type carbides

such as TiC carbide) first precipitates at 1549°C. Then, the primary M7C3 carbides

precipitate at 1317°C. Finally, the eutectic reaction L γ+M7C3 occur from 1258°C.

Moreover, the precipitation temperature of the TiC carbide increases from 1459°C to

-19-

1734°C with an increased Ti content from 0.75mass% to 3.0mass%.

3.1.2 Solidification path

The predicted solidification path and composition evolution in the residual liquid of

hypereutectic HCCI are shown in Fig. 3.2 and Fig. 3.3.

1000 1100 1200 1300 1400 1500 16000.0

0.2

0.4

0.6

0.8

1.0

Primary M7C

3

M3C

Eutectic M7C

3

MC

FCC

Vo

lum

e f

rac

tio

n o

f p

ha

se

s,

-

Temperature, (oC)

C : 4wt%

Cr: 20wt%

Ti: 1.5wt%

PE

LIQ

Fig. 3.2. Predicted solidification path for normal composition of HCCI by PE prediction.

The solid phases precipitate in the order of MC→M7C3→FCC→M3C, as shown in

Fig. 3.2. MC carbide precipitates firstly and acts as the nuclei of the M7C3 carbide

[27]. Afterwards it forms the eutectic FCC phase and eutectic M7C3 carbide, which act

as the matrix of HCCI. Also, a M3C carbide appears at the final stage of solidification.

Its amount is so small that its effect is omitted here. Thus, the carbides including MC

and M7C3 account for approximately 33% in volume fraction of the hypereutectic

HCCI. The composition evolution in the residual liquid is closely related to the phase

-20-

0.0 0.2 0.4 0.6 0.8 1.0

0

5

10

15

20

25

Primary

Co

mp

osit

ion

in

liq

uid

, (w

t%)

Volume fraction of solid, (-)

T

C

Cr

Ni

Mo

Mn

Si

Ti

precipitation

M7C

3

precipitationMC

Eutectic

1000

1100

1200

1300

1400

1500

1600

Tem

pera

ture

, ( o

C)

C : 4wt%

Cr: 20wt%

Ti: 1.5wt%

Fig. 3.3. Variation of liquid composition during the solidification sequence for normal composition of

HCCI by PE prediction.

precipitation. As shown in Fig. 3.3, the composition of Ti in the liquid firstly

decreases largely corresponding to the first steep decrease in temperature curve. It

indicates the formation of a MC (i.e. TiC) carbide. Secondly, the Cr composition in

the residual liquid decreases greatly, relating to the second steep decrease in the

temperature curve. It is due to the formation of primary M7C3 carbide. Afterwards, the

Cr composition decreases smoothly. It indicates the formation of a FCC phase and

eutectic M7C3 carbides.

3.2 Microstructure of Ti-alloyed hypereutectic HCCI

Fig. 3.4 shows the microstructure of Ti-alloyed hypereutectic HCCI as the conditions

of as-cast and heat treatment. The solidification (as-cast) microstructure consists of

primary M7C3 carbides with a large hexagonal columnar structure, the eutectic M7C3

carbides with an irregular shape, TiC carbides with a cubic structure and a

-21-

Fig. 3.4. Microstructure of Ti-alloyed hypereutectic HCCI, (a) LOM and (b) (c) (d) SEM for carbides,

(e) and (f) EBSD for matrix

matrix containing mainly of an austenite phase (γ), as shown in Fig. 3.4(a)-(c) and (e).

After heat treatment, the secondary carbides will precipitate from austenite matrix, as

shown in Fig. 3.4(d). Meanwhile, the matrix phase will change from austenite (γ) into

martensite (α´) after heat treatment, as shown in Fig. 3.4(f).

3.3 Effect of cooling rate on as-cast structure

As mentioned in chapter 2, three kinds of cooling rates, which were examined in the

three kinds of mold, were obtained. The idea for refining the microstructure by

increasing the cooling rates is based on: that the nucleation probability of primary

M7C3 carbides increases due to the increase of the precipitation undercooling degree.

It is apparent from the photographs in Fig. 3.5(a) that in the case without Ti addition,

the size of primary M7C3 carbides decreases as the cooling rate increase. It is also

apparent from the comparison between Fig. 3.5(a) and 3.5(b) that the size of primary

M7C3 carbides in the case of a 1.5mass% Ti addition is significantly smaller than for

those in the case without a Ti addition. This is due to a precipitation of TiC carbides in

the melt. The size distribution of M7C3 carbides for different conditions was shown in

-22-

Fig. 3.5. Optical microscope observations for M7C3 carbide under different cooling rate,

(a) Ti=0mass%, (b) Ti=1.5mass%

Fig. 3.6. As shown in Fig. 3.6(a) representing a case without Ti additions, the

maximum size of primary M7C3 carbides (>11.2μm) decreases from 151.9 to 53.9μm

with an increased cooling rate (from a sand mold to a metal mold). In addition, the

number of eutectic M7C3 carbides, which in general has a size less than 11.2μm,

increases from 0.81×103 mm

-2 to 4.80×10

3 mm

-2 with an increased cooling rate.

10-1

100

101

102

103

10-1

100

101

102

103

104

4.80103

mm-2

0.81103

mm-2

53.9 um

151.9 um

Eutectic M7C

3

Ti=0%

Sand mold

Graphite mold

Metal mold

N

A, (m

m-2)

d, (m)

Primary M7C

3

Peak value

(a)

-23-

10-1

100

101

102

103

10-1

100

101

102

103

104

(b)

Peak value

53.9 um

Eutectic M7C

3

Ti=1.5%

Sand mold

Graphite mold

Metal mold

N

A,

(mm

-2)

d, (m)

Primary M7C

3

2.49103

mm-2

1.29103

mm-2

Fig. 3.6. Effect of cooling rate on M7C3 carbide size distribution of the hypereutectic HCCI with (a):

without Ti addition; (b): 1.5mass% Ti addition by using optical microscope photos.

Moreover, as shown in Fig 3.6(b) representing a case with a 1.5mass% Ti addition,

the maximum size of primary M7C3 carbides (>11.2μm) is not changed to a large

degree. However, the number of eutectic M7C3 carbides (<11.2μm) increases from

1.29×103 mm

-2 to 2.49×10

3 mm

-2 with an increased cooling rate, when a 1.5mass% Ti

addition is used.

If Fig. 3.6(a) with Fig. 3.6(b) are compared, it can be seen that the size of primary

M7C3 carbides in the case of a 1.5mass%Ti addition is significantly smaller than that

for a case without a Ti addition. Furthermore, that the maximum sizes are 53.9μm and

151.9μm respectively. Making the story short, the size of primary M7C3 carbides of

the hypereutectic HCCI can be refined both by Ti additions and by increasing the

cooling rate.

3.4 Effect of Ti addition on as-cast structure

As mentioned in chapter 1, strong carbide forming elements, such as Ti, Nb, V, Zr etc.,

will react with carbon to form NaCl structure (MC type) carbides. Those MC type

carbides can act as heterogeneous nuclei for the precipitation of M7C3 type

-24-

Fig. 3.7. SEM and EDX analysis for TiC carbide with 1.5mass% Ti addition in the graphite mold.

(Point A and C: Ti(C,N); Point B and D: (Ti, Mo)C)

carbides during the cooling and solidification of the melt. Therefore, based on this

idea, Ti was selected to study the influence on the solidification structure of HCCI. As

shown in Fig. 3.7, it is confirmed that the TiC carbides can act as a nuclei of M7C3

carbides. More specifically, it was found that TiC carbides can not only act as nuclei

Fig.3.8. Light optical microscope observations for M7C3 carbide precipitate found in the graphite mold

(a). Ti=0%, (b). Ti=0.75%, (c). Ti=1.5%, (d). Ti=3.0%.

-25-

of M7C3 carbides, but they also contain a Ti(C, N) core. This is illustrated by the black

point A, C in Fig. 3.7. Furthermore, EDX results show that the main composition

particle in B and D is Ti, Mo and C, it should be (Ti, Mo)C. However, the chemical

composition of the black points A and C shows that Ti, C and N are dominant, and

that the TiC carbides should be Ti(C, N).

Fig. 3.8 shows the micrographs solidification structure for different Ti additions. The

microstructure of M7C3 carbides is refined significantly with increasing Ti contents.

As shown in Fig. 3.9, the size distributions of M7C3 carbides also resemble a

log-normal distribution.

10-1

100

101

102

103

10-1

100

101

102

103

104

27.0 um

151.9 umEutectic M

7C

3

Graphite mold

Ti=0%

Ti=0.75%

Ti=1.5%

Ti=3.0%

N

A, (m

m-2)

d, (m)

Primary M7C

3

2.84103

mm-2

2.00103

mm-2

Peak value

Fig. 3.9. Effect of Ti addition on M7C3 carbides size distribution of the hypereutectic HCCI in the

graphite mold obtained by using optical microscope photos.

It can be seen that the maximum size of the primary M7C3 carbides decreases from

151.9 to 27.0μm with an increased Ti addition in the graphite mold. However, the size

and number of eutectic M7C3 precipitates do not change to a large degree (from

2.00×103 mm

-2 to 2.84×10

3 mm

-2) with an increased Ti content.

The back scattered electron (BSE) observations for TiC carbides are shown in Fig.

3.10. TiC carbides are mostly located inside or at the boundary of M7C3 carbides, as

shown in Fig. 3.10. Moreover, TiC carbides starts to grow and agglomerate when the

-26-

Fig. 3.10. SEM observation for TiC carbide, (a): Ti=0.75mass%, (b): Ti=1.5mass%, (c): Ti=3.0mass%.

10-1

100

101

102

103

10-1

100

101

102

103

104

Graphite mold

Ti=0.75%

Ti=1.5%

Ti=3.0%

N

A,

(mm

-2)

d, (m)

Fig. 3.11. Effect of Ti addition on TiC carbide size distribution of the hypereutectic HCCI in the

graphite obtained by using SEM photos

Ti content is 1.5mass%, as shown in Fig. 3.10(b). It is also apparent from the

photographs in Fig. 3.10(c) that larger and coarser clusters of TiC carbides exist. The

size distribution of TiC carbides in the graphite mold for different Ti additions is

-27-

shown in Fig. 3.11. These carbides generally have a size less than 10μm. Moreover, it

can be seen that both the size and number of TiC carbides increase with an increased

Ti addition.

10-1

100

101

102

103

10-1

100

101

102

103

104

Ti=1.5%

Sand mold

Graphite mold

Metal mold

N

A,

(mm

-2)

d, (m)

Fig. 3.12. Effect of cooling rate on TiC carbide size distribution of the hypereutectic HCCIs with

1.5mass% Ti addition obtained by using SEM photos

Also, Fig. 3.12 shows the size distribution of TiC carbides with 1.5mass% Ti addition

in a sand mold, a graphite mold and a metal mold. The average size of TiC carbides

with a 1.5mass% Ti addition does not practically change with an increased cooling

rate. Also, the number of TiC carbides does not change due to the increased cooling

rate. Consequently, it is clearly concluded that the precipitation of TiC carbides was

not seemingly affected by the cooling rate. Thus, it is confirmed that the cooling rate

will not affect the precipitation behavior of TiC carbides.

3.5 Effect of Ti addition and cooling rate on the volume fraction in

as-cast condition of the hypereutectic HCCI

The experiment results and the calculation results of the volume fractions of M7C3

carbides and TiC carbides are shown in Fig. 3.13.

-28-

Fig.3.13. Calculation results and experiment results on volume fraction of M7C3 carbides and TiC

carbide

It can be seen that the volume fraction of M7C3 carbides decreases with an increased

cooling rate. The condition in the low cooling rate seems to be more close to the

condition in the equilibrium state. However, the 5-10 vol.% of M7C3 carbides did not

fully precipitate in the high cooling rate conditions. Moreover, the volume fraction of

M7C3 carbides decreases with a Ti addition due to the reaction between Ti and C.

Furthermore, the volume fraction of Ti carbides increases with increased Ti additions

in the graphite mold. However, the volume fraction of TiC carbides does not change

although the cooling rate changed. If the experimental results are compared to the

-29-

calculation results, it can be seen that the tendency is the same. Moreover, that the

experimental results are closer to the PE calculation results. The agreement between

experiment and calculation results is within 5%, which is encouraging. This is due to

that the PE calculation considers the carbon back diffusion and consequently the

results will be closer to the real experimental situation.

3.6 Classification of M7C3 carbides for as-cast condition

In this study, it is suggested that the M7C3 type carbides in the HCCI can be classified

into primary M7C3 carbides and eutectic M7C3 carbides depending on the time of

precipitation. The primary carbides precipitate in the melt before solidification (that is

during cooling of the melt until the solidification temperature is reached). The eutectic

M7C3 carbides precipitate during the eutectic reaction. The precipitation possibility of

M7C3 carbides before and during solidification is estimated for the sample

composition given in Table 2.1. This was done based on PE model calculations. The

volume fraction of precipitated primary M7C3 carbides, eutectic M7C3 carbides and

1000 1100 1200 1300 1400 1500 1600 1700 18000.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

M7C

3:

Ti=0%

Ti=0.74%

Ti=1.48%

Ti=3.36%

TiC:

Ti=0.74%

Ti=1.48%

Ti=3.36%

Vo

lum

e F

racti

on

, (%

)

Temperature, (C)

Eutectic M7C

3

Primary M7C

3

TiC

Fig. 3.14. Volume fraction of M7C3 carbide and TiC carbide of the hypereutectic HCCI with different Ti

addition obtained by using PE model calculation for Fe-3.98mass%C-16.8 mass%Cr-1.86

mass%Mn-0.943 mass%Mo-1.09 mass%Ni-0.97mass% Si. "∙": transformation points of M7C3 carbide

-30-

TiC carbides are plotted against the temperature in Fig. 3.14. It can be seen that the

precipitation temperature of primary M7C3 carbides decreases with an increased Ti

content. However, the precipitation temperature of TiC carbides increases slightly

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50

5

10

15

20

25

30

35

40

(a) border size: 7.9m

Total M7

C3

PE Scheil Experiment

Total M7C

3:

Primary M7C

3:

Eutectic M7C

3:

Vo

lum

e F

rac

tio

n,

(%)

Ti, (mass%)

4.9mass%C

4.9mass%C

4.0mass%C

4.9mass%C

4.0mass%C Eutectic M7

C3

Primary M7

C3

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50

5

10

15

20

25

30

35

40

PE Scheil Experiment

Total M7C3:

Primary M7C3:

Eutectic M7C3:

Vo

lum

e F

rac

tio

n,

(%)

Ti, (mass%)

4.9mass%C

(b) border size: 11.2m

4.9mass%C

4.0mass%C

4.9mass%C

4.0mass%C

Total M7

C3

Eutectic M7

C3

Primary M7

C3

Fig. 3.15. PE model prediction and experimental results of total, primary and eutectic M7C3 carbides in

the hypereutectic HCCI with different Ti additions. M7C3 carbide’s border size: (a) d=7.9μm, (b)

d=11.2μm

-31-

with an increased Ti content.

Fig. 3.15 shows PE model prediction and experiment results of total, primary and

eutectic M7C3 carbides in the hypereutectic HCCI for different Ti additions. Here, (a)

d=7.9μm and (b) d=11.2μm were adopted as the M7C3 carbide’s border size, d.

Comparing Fig. 3.15(a) with 3.15(b), it can be seen that experimental results with a

11.2μm border size are closer to the calculation results from the viewpoint of the

primary M7C3carbides volume fraction (see ○ mark). Regarding eutectic M7C3

carbides (see ∆ mark), the experiment data are 5% lower than the calculation curve.

This might be due to the difficulty of deciding the threshold from the gray images of

the optical microscope photos. Moreover, the experimental results are close to the

Partial Equilibrium (PE) predictions. The difference between the experimental results

and the calculation results is lower than 5%. This can be explained by that the carbon

back diffusion is taken into consideration in the PE calculation. Therefore, the results

will be closer to experiment results. It is also found that the C content has a large

influence on the total volume fraction of M7C3 carbides. More specifically, the total

volume fraction increases with an increased C content, as shown in Fig. 3.15.

-32-

0 2 4 6 8 10 30 40 50 60 70 80 90 1000

1

2

3

4

30

40

50

60

70

Co

mp

os

itio

n o

f c

arb

ide

, (a

t%)

(b):

Graphite mold

d, (um)

Fe

Cr

Ti

Mo

Mn

Primary M7C

3Eutectic M7C

3

0 2 4 6 8 10 30 40 50 60 70 80 90 1000

1

2

3

4

30

40

50

60

70

Co

mp

os

itio

n o

f c

arb

ide

, (a

t%)

(c):

Sand mold

Fe

Cr

Ti

Mo

Mn

d, (um)

Primary M7C

3Eutectic M

7C

3

Fig. 3.16. Relationship between composition and size of primary M7C3 carbides and eutectic M7C3

carbides of the hypereutectic HCCI with 1.5mass% Ti addition under different cooling rate: (a) Metal

mold, (b) Graphite mold and Sand mold.

In summary, Fig. 3.15 shows that the total volume fraction of M7C3 carbides

decreases as the Ti content increases; the volume fraction of primary M7C3 carbides

decreases with an increased Ti content. However, the volume fraction of eutectic

M7C3 carbides increases with an increased Ti content. The contents of Fe, Cr, Ti, Mo,

-33-

Mn in primary M7C3 carbides and eutectic M7C3 carbides are plotted against the

carbide sizes in Fig. 3.16 for the metal mold, graphite mold and sand mold for a

1.5mass% Ti addition. In general, it is seem that the content of Cr and Ti increases

with an increased M7C3 carbide size, d. However, the content of Fe and Mo decreases

with an increased M7C3 carbide size; Moreover, the content of Mn does not change

with an increased M7C3 carbide size for the sand, graphite and metal molds. It is

concluded that the primary M7C3 carbides are dominant at higher Cr contents, and that

the eutectic M7C3 carbides have lower Cr contents.

In summary, based on the analysis of size distribution and chemical composition of

M7C3 carbides, the M7C3 carbides were roughly classified into primary M7C3 carbides,

which in general have size larger than 11.2μm. Moreover, into eutectic M7C3 carbides

which in general have a size smaller than 11.2μm.

3.7 Effect of cooling rate and Ti addition on the mechanical properties

in as-cast condition of the hypereutectic HCCI

In this section, the effect of cooling rate and Ti addition on the mechanical properties

of the hypereutectic HCCI is discussed, but only for as-cast conditions. The bulk

hardness of the hypereutectic HCCI is determined by the hardness of carbides, such as

TiC and M7C3 [21] and the matrix. The matrix of the hypereutectic HCCI will

obviously not change by a Ti addition. This is due to that it will mainly exist as a

single austenite phase (γ-phase) together with small amounts of martensite (α´-phase)

under as-cast conditions. Therefore, the bulk hardness of the hypereutectic HCCI is

mainly affected by the volume fraction of carbides. Fig. 3.17 shows the bulk hardness

of the hypereutectic HCCI with different Ti additions for different cooling rates. The

tendency is that the bulk hardness increases slightly with increased cooling rates and

titanium contents. As shown in Fig. 3.13, when Ti is added, the volume fraction of

TiC carbides increase and the volume fraction of M7C3 carbides decrease. TiC

carbides might make a bigger contribution to the bulk hardness than M7C3 carbides,

because the hardness of TiC carbides is in the range of 2000-3200HV. [4] However,

-34-

Fig 3.17. Mechanical properties of the hypereutectic HCCI with different Ti additions and different

cooling rates.

the bulk hardness of the hypereutectic HCCI in the graphite mold shows a big scatter

because the TiC carbides are not uniformly distributed. The wear loss of the

hypereutectic HCCI is determined by the amount of matrix and M7C3 carbides,

although the hardness of TiC carbides is larger than that of the M7C3 carbides and the

matrix. However, the presence of a small amount of TiC carbides will not contribute

to the wear loss. It can be seen that there is a tendency of an increased wear loss with

an increased Ti content in the graphite mold experiments, as shown in Fig. 3.17. This

is due to the decrease of the volume fraction of M7C3 carbides. It seems that the

austenite phase (γ-phase) which surrounds the agglomerated TiC might be easily

peeled off during the wear application process. This is leads to an increased Ti content.

In summary, the hardness and wear loss of the hypereutectic HCCI do not change

significantly with increased Ti contents and cooling rates under as-cast conditions.

As a summary for as-cast results, the mainly focus point on the carbides (M7C3 and

TiC) evolution is quantitative discussed by changing the cooling rates and titanium

contents. Base on this work, the heat treatment process can be performed in order to

improve the mechanical properties of the hypereutectic HCCI.

-35-

3.8 Evolution of matrix after heat treatment

From this section, we will set our focus point on how the matrix and the carbides

(M7C3 and TiC) evolution and their influence on the mechanical properties though

heat treatment process. As mentioned in the chapter 2, the specimens, which were

casted in a graphite mold with a 1.5mass% titanium addition, were used for heat

treatments. The reason for the casting process and selected chemical composition

(Table 2.1) is explained in detail in supplement Ⅲ. Firstly, the matrix evolution after

heat treatment will be introduced. The matrix of hypereutectic HCCI includes both

austenite and martensite before and after heat treatment, as shown in Fig. 3.18 and

Fig. 3.19. However, the austenite is dominating in the as-cast condition and the

martensite dominates after heat treatment. EBSD phase maps with 0.1μm and 0.03μm

step sizes are shown in Fig. 3.18 and Fig. 3.19, respectively. It is clear that the matrix

essentially consists of austenite with a small amount of martensite in the as-cast

condition. The inverse pol figure (IPF) map in Fig. 3.19 shows the orientation of the

martensite units, and it is clear that the martensite unit size differs between the

different heat treatment conditions.

Fig. 3.18. EBSD phase colored maps using 0.1μm step size. (Color in phase maps: Blue: M7C3; TiC:

Yellow; Green: Austenite; Red: Martensite; Greyscale: band contrast)

-36-

Fig. 3.19. EBSD phase colored and inverse pole figure (IPF for martensite) colored maps using

0.03μm step size. (Color in phase maps: Blue: M7C3; TiC: Yellow; Green: Austenite; Red: Martensite;

Greyscale: band contrast)

The size distribution of martensite units in as-cast condition and after the different

heat treatments is shown in Fig. 3.20. It shows the results when combining data from

the 0.03μm and 0.1μm step sizes. These are also plotted with a logarithmic scale.

From Fig. 3.20, it is seen that large martensite units above 4μm is present under the

conditions of high temperatures or long holding time: 900oC×6hr, 1000

oC×6hr and

1050oC×6hr. Compared to the higher temperatures, the size of the martensite unit

-37-

10-2

10-1

100

101

102

103

100

101

102

103

104

105

106

107

NA

, (m

m-2

)

D, (m)

Martensite:

As cast

900oC2hr

1000oC2hr

1050oC2hr

900oC6hr

1000oC6hr

1050oC6hr

0.1m step0.03m step

Fig.3.20. Size distribution of martensite unit.

decreased at lower heat treatment temperatures. The volume fraction of martensite is

shown in Fig. 3.21. The EBSD maps, which were used for calculating the volume

fraction of martensite includes the non-indexed regions. The volume fraction of

martensite calculated by using the 0.1μm and the 0.03μm step size EBSD maps give

0 2 4 6 80

10

20

30

40

50

60

70

Minimum

Martensite

0.1m step:

As cast

900oC

1000oC

1050oC

Maximum

Vo

lum

e F

racti

on

, (%

)

Time, (hr)

0.03m step:

As cast

900oC

1000oC

1050oC

Fig. 3.21. Volume fraction of martensite

-38-

quite different results, as shown in Fig. 3.21. Thus, from the viewpoint of size

distribution and volume fraction of martensite, the formation behavior of martensite is

summarized as follows:

(1) At longer holding times during heat treatment, such as 6hr, the volume fraction of

martensite increases, compared to shorter holding time.

(2) At long holding times, such as 6hr, a lower holding temperature results in a more

refined martensite.

(3) Thus, from the viewpoint of the transformation promotion of martensite (volume

fraction) and the refinement of martensite unit (size distribution), longer holding time

such as 6hr and lower holding temperatures such as 900oC are recommended.

Fig. 3.22. SEM observations before heat treatment (a, c and e) and after heat treatment (b, d and f) at

the same position.

-39-

3.9 Evolution of carbides after heat treatment

Fig. 3.22 shows SEM observations from the same position before and after heat

treatment. The size of primary M7C3 carbides and eutectic M7C3 carbides increases at

higher temperatures and longer holding times, as shown in Fig. 3.22(c)-(f). The

primary and eutectic M7C3 carbides grow and finally merge with each other, as

marked by A and A in Fig. 3.22 (c) and (d). This is also indicated by the length of

mark B and B in Fig. 3.22 (e) and (f), which shows that the distance decreases from

9.0μm to 5.3μm. Compared to the microstructure before heat treatment, it can be seen

from Fig. 3.22(b) that the secondary carbides precipitate from the matrix during heat

treatment. The following section 3.9.3 will discuss the secondary carbides in detail

after heat treatment.

3.9.1 Evolution of primary and eutectic M7C3 carbides after heat

treatment

The carbides mainly include primary M7C3 carbides and eutectic M7C3 carbides both

before and after heat treatment, as shown in Fig. 3.18 and Fig. 3.22. It is also noticed

Fig.3.23. SEM observations of M7C3 and TiC after different heat treatments.

-40-

based on SEM and EBSD observations that fine secondary carbides precipitate in the

matrix after heat treatment, as shown in Fig. 3.4(d) and Fig. 3.23. A detail information

regarding secondary carbides will be described in section 3.9.3. The size distribution

of M7C3 carbides from SEM and EBSD are shown in Fig. 3.24(a) and (b). From Fig.

3.24(a) and (b), it is possible to clearly recognize the precipitation of secondary

carbides (<1μm) after heat treatment. In order to understand the precipitation behavior

of M7C3 carbides, besides the size distribution, the total carbide volume fraction has

to be taken into consideration.

10-1

100

101

102

103

100

101

102

103

104

105

M7C

3:

As cast

900oC2hr

1000oC2hr

1050oC2hr

900oC6hr

1000oC6hr

1050oC6hr

Eutectic M7C

3

NA

, (m

m-2

)

d, (m)

Secondary M7C

3Primary M

7C

3 (a)

10-1

100

101

102

103

100

101

102

103

104

105

M7C

3:

As cast

900oC2hr

1000oC2hr

1050oC2hr

900oC6hr

1000oC6hr

1050oC6hr

Eutectic M7C

3

NA

,(m

m-2

)

d, (m)

Secondary M7C

3Primary M

7C

3(b)

Fig. 3.24. M7C3 carbide size distribution in the hypereutectic HCCI by using (a): SEM images, (b):

EBSD maps

-41-

The total volume fraction of M7C3 carbides (primary, eutectic and secondary)

obtained from SEM images and EBSD maps using image analysis is shown in Fig.

3.25.

Fig. 3.25. Volume fraction of M7C3 and TiC carbides

From the size distribution in Fig. 3.24 and the total carbide volume fraction in Fig.

3.25, the precipitation behavior of M7C3 carbides is summarized as follows:

(1) For lower holding temperatures and shorter holding times during the heat

treatment, such as 900oC×2hr, the number of fine secondary M7C3 carbides (less than

1μm) and eutectic M7C3 carbides (less than 11.2μm) increases. On the contrary, at

higher holding temperatures and longer holding times of heat treatment, such as

1050oC×6hr, the size of coarse primary M7C3 carbides larger than 11.2μm increases.

(2) This behavior is due to the following reasons. The precipitation of secondary

M7C3 occurs actively at 900oC close to the eutectic temperature and finishes

completely within 2hr. The coarsening of all sizes of M7C3 carbides occur actively at

higher holding temperatures and longer holding times during heat treatment.

-42-

(3) Thus, from the viewpoint of the precipitation of M7C3 (volume fraction) and the

growth control/refinement of M7C3 (size distribution), a lower holding temperature

such as 900oC is preferred. This is regardless of the holding time during heat

treatment.

(4) According to the reports of Wiengmoon, et al. [16, 22, 23] and Pearce, [47] in cast

irons with higher Cr contents, such as 30mass% Cr, the secondary carbides were of

M7C3 type or M23C6 type. These precipitate during conventional heat treatment, as

shown in Table 3.1, however, only the M7C3 type carbides were examined based on

the EBSD results in the present work.

Table 3.1 Secondary carbide type under different C and Cr content

Type C (mass%) Cr (mass%) References

M7C3 type

2.8 18 S K. Hann, et al

2.4 30 J.T.H. Pearce,

M23C6 type

2 26 E. Pagounis, et al

2.4 30 A. Wiengmoon, et al

1.09 10.48 Y. Wang, et al

3.9.2 Evolution of TiC carbides after heat treatment

The TiC carbides with a high precipitation temperature are uniformly distributed

within the M7C3 carbides and the matrix, as shown in Fig. 3.18 and Fig. 3.23.

The size distributions of TiC in as-cast condition and after various heat treatments are

shown in Fig. 3.26(a) and (b). Based on the volume fraction in Fig. 3.25 and the size

distribution in Fig. 3.26, the precipitation behavior of TiC can be summarized as

follows:

(1) At longer holding times of heat treatment, such as 6hr, the number of all sizes of

TiC carbides increases.

-43-

10-1

100

101

102

103

100

101

102

103

104

105

TiC:

As cast

900oC2hr

1000oC2hr

1050oC2hr

900oC6hr

1000oC6hr

1050oC6hr

N

A, (m

m-2

)

d, (m)

(a)

10-1

100

101

102

103

100

101

102

103

104

105

(b)

NA

, (m

m-2

)

d, (m)

TiC:

As cast

900oC2hr

1000oC2hr

1050oC2hr

900oC6hr

1000oC6hr

1050oC6hr

Fig. 3.26. TiC size distribution of hypereutectic HCCI by using (a): SEM images and (b): EBSD maps

(2) This behavior is caused due to the following reasons. The precipitation of TiC

occurs mainly in the melt and finishes completely at higher temperatures above

1100oC. Moreover, the coarsening of all sizes of TiC particles occurs actively under

longer holding times during heat treatment.

(3) Thus, from the viewpoint of the precipitation of TiC (volume fraction, size and

number), longer holding time such as 6hr is preferred, regardless of the holding

temperature level, as the heat treatment condition.

-44-

3.9.3 Evolution of secondary carbides during heat treatment

3.9.3.1 Identification of secondary carbides

Fig. 3.27 shows a TEM micrograph, the selected area diffraction pattern (SADP) and

the EDS results for a secondary carbide in the condition of 890oC×6hr. It shows that

the secondary carbides have a size of ~250nm in diameter. The SADP in Fig. 3.27(b)

form the particle, marked in Fig. 3.27(a), is indexed to be a M7C3 carbide type in the

Fe-4 mass%C-17 mass%Cr-1.5 mass% Ti hypereutectic HCCI alloy. It is also confirm

the EBSD identification regarding the secondary carbides. In addition, the EDS result

from the secondary M7C3 carbide also shows that the composition of this particle is

rich in Cr, Fe and C, as is shown in Fig. 3.27(c). It is different from the matrix

(Martensite, α´), which is only Fe rich, as is shown in Fig. 3.27(d).

Fig. 3.27. TEM micrograph of secondary carbide in a condition of (890°C×6h): (a) a bright–field TEM

micrograph and (b) a selected area diffraction pattern (SADP) of secondary M7C3 carbide and (c) an

EDS analysis of a secondary M7C3 carbide and (d) an EDS analysis of matrix .

-45-

3.9.3.2 In-situ observations of secondary M7C3 carbide precipitation

The images were extracted from a series of CLSM video sequences at a rate of 30

frames per second during the whole heat treatment. Fig. 3.28 illustrates the in-situ

observations of the microstructure evolution in the Ti-alloyed HCCI heated to 890°C

Fig. 3.28. In-situ observations of secondary carbide precipitations during the heat treatment process by

using a CLSM to study sample for an 890°C×2hr condition.

-46-

and held there for 2hr with a heating rate 10°C/min. The corresponding temperature

and time during the heat treatment are also displayed in the images. It was found that

the ferrite (bcc, α) start to form from the matrix at approximately 575°C during the

heating process, as is shown in Fig. 3.28(b). The ferrite (bcc, α) phase should be

stable at low and intermediate temperatures as determined from the phase diagram for

a Fe-17 mass% Cr- 4 mass% C- 1.5 mass% Ti alloy, which was calculated in a

supplement Ⅲ. It is the retained austenite (fcc, γ) that starts to transforms into ferrite

(bcc, α) at 575°C. The ferrite will disappear again when approaching the holding

temperature where austenite (fcc, γ) is the stable phase. The temperature where the

ferrite austenite transformation takes place is approximate 703°C, which is

estimated by the phase diagram of this alloy. In addition, the volume of the ferrite

increased with an increased temperature, as shown in Fig. 3.28(c). Moreover, at

666°C, it seems that some precipitates start to appear from surface of primary M7C3

carbides, as shown in Fig. 3.28(c). These precipitates increased with an increased

temperature. However, the precipitates inside the samples were polished and studied

carefully by using SEM after the heat treatment in Fig. 3.23 and none of precipitates

were found on the surface of primary M7C3 carbides. Hence, this indicates that the

observed feature on the primary M7C3 carbides is a surface phenomenon. Furthermore,

at 840°C, the precipitates start to appear from the matrix as shown in Fig. 3.28(d) and

the matrix phase is austenite (fcc, γ) according to the phase diagram. From Fig.

3.28(e)-(h), it can be seen that the number of precipitates increased and the sizes

became larger with an increased time and temperature during the holding and cooling

process. The increasing number of precipitates during the heat treatment is illustrated

in Fig. 3.29. The total area for the counted precipitate number is 4856.0μm2 (73.8μm

(L) × 65.8μm (W)). It was found that the precipitation starts at approximately 840°C

and 847°C for the 890°C×2hr and 1026°C×2hr conditions, respectively. Moreover, the

number of precipitates increased sharply at these temperatures. It was also found that

the precipitate number per unit area did not change too much when the temperature

reached the holding temperatures. Comparing the two different temperatures, the

-47-

0 20 40 60 80 100 120 140 160 180 200 220 240 260

104

0.8

1.6

2.4

3.2

4.0

4.8

890C2hr

52C1026C1026C

847C840C

876C

Time, (min)

NA

, (m

m-2)

HoldingHeating

55C

700C

890C

890C890C

849C

890C890C

Cooling

6.4

0

5.6 1026C2hr

1026C1026C1026C

Fig. 3.29. Change of number of secondary carbides with time and temperature during heat treatment

for samples with 890°C×2hr and 1026°C×2hr condition based on in-situ observation results.

number of secondary carbides is higher for 890°C than for 1026°C.

Fig. 3.30 shows the evolution of the size distribution of secondary carbides during

heat treatment with holding temperatures of 890°C and 1026°C. In this section, only

the secondary carbides, which are larger than 1μm, are discussed because of a limited

resolution of the CLSM. The smaller secondary carbides (<1μm) will be discussed in

detail in section 3.9.3.3. From Fig. 3.30(a), it was found that the number of secondary

carbides increased with an increased heating temperature. Moreover, the number of

carbides did not change to a large extent during the holding and cooling stage.

However, the number of secondary carbides decreased with increased holding and

cooling times. It can also be seen that the size of the secondary carbides increases

from the heating stage to the cooling stage. Compared to the 890°C×2hr condition, the

same tendency was obtained for a 1026°C×2hr condition, as seen in Fig. 3.30(b).

However, it was found that the number of secondary carbides is lower for a

1026°C×2hr condition compared to a 890°C×2hr condition.

-48-

10-2

10-1

100

101

101

102

103

104

105

106

Resolution limited zone

890C2hr:

Cooling

Holding

d, (m)

NA, (m

m-2)

79min 840C

80min 849C

85min 890C

205min 890C

220min 55CCooling:

Holding:

Heating:

Heating

(a)

10-2

10-1

100

101

101

102

103

104

105

106

d, (m)

NA, (m

m-2)

80min 847C

85min 1026C

205min 1026C

220min 52CCooling:

Holding:

Heating:

Cooling

Holding

Heating

1026C2hr: (b)

Resolution limited zone

Fig. 3.30. Size distribution of secondary carbides determined in-situ observation by using CLSM for

the following conditions: (a) 890°C×2hr and (b) 1026°C×2hr.

Fig. 3.31 shows the evolution of volume fraction of secondary carbides, which was

obtained from the in-situ observation. The volume fractions of secondary carbides

increase gradually with an increased heat treatment time. In addition, it was found that

the volume fraction of secondary carbides at lower temperatures is a little bit higher

-49-

0 20 40 60 80 100 120 140 160 180 200 220 240 2600

5

10

15

20

25

30

35

40

1026C1026C

1026C

890C2hr

840C

55C

700C890C890C890C

849C

890C890C

CoolingHolding

Vo

lum

e F

rac

tio

n,

(%)

Time, (min)

Heating

52C1026C1026C

1026C2hr

Fig. 3.31. Volume fraction of secondary carbides as a function of time determined from in-situ

observation by using CLSM.

0 2 4 6 8 100

5

10

15

20

25

30

35

40

Vo

lum

e f

racti

on

, (%

)

Time, (hr)

M23

C6

M7C

3

M23

C6

M23

C6

M7C

3

M7C

3

2.91 26.96 SEM

Wiengmoon,

et al[24]

Carbides

type

dataCLSM

References Content Method Temperature

C(%) Cr(%) 900oC 1050

oC

Present work 4 17 SEM, CLSM

2.3 30 SEM

Wiengmoon,

et al[16] [18]

Wiengmoon, 2.72 26.6 SEM

et al[25]

3 20

Fig. 3.32. Comparison of the volume fraction of secondary carbides between data from present work

and data from Wiengmoon, et al [16, 18, 24, 25].

than that at higher temperatures. However, the volume fraction will not change to a

large extent (5%±2%) when the holding time and holding temperature change after

holding for 2hr, as shown in Fig. 3.32.

-50-

As a confirmation of the results in the present study, data from Wiengmoon, et al. [16,

22, 24, 25,] were used as a comparison. The same tendency between the present

results and Wiengmoon, et al. [16, 22] results can be found. However, compared to

the present study, the volume fraction of secondary carbide is much higher in

Wiengmoon, et al. studies. [16, 22] One possible reason for this difference is that the

Cr content is higher in the Wiengmoon, et al. study. [16, 22] Furthermore, the volume

fraction values (~5%) in this study agree with the Wiengmoon, et al. results [23, 25].

3.9.3.3 Size distribution of secondary carbides

Fig. 3.33 shows the secondary carbides for different heat treatment conditions

observed in SEM. It can be seen that the number of secondary carbides at lower

Fig. 3.33. SEM observations of secondary M7C3 carbides for the following heat treatment conditions:

(a) 890°C×2hr, (b) 1026°C×2hr, (c) 890°C×6hr and (d) 1026°C×6hr.

temperatures such as 890°C (Fig. 3.33(a) and (c)) is higher than that at higher

temperatures such as 1026°C (Fig. 3.33 (b) and (d)). In addition, the size of secondary

carbides seems to increase with an increased holding temperature and time.

-51-

10-2

10-1

100

101

101

102

103

104

105

106

890C2hr

1026C2hr

890C6hr

1026C6hr

SEM CLSM

d, (m)

NA, (m

m-2)

CLSMSEM

Present work(a)

10-2

10-1

100

101

101

102

103

104

105

106

900C 1025CRef. :2.3% C-30% Cr

d, (m)

NA, (m

m-2)

Wiengmoon, et al [22]

SEM

950C 1025C

Wiengmoon, et al [16]

(b) References

Fig. 3.34. Comparison of the size distribution of secondary M7C3 carbides. (a) data from present work

and (b) data analyzed based on SEM images from Wiengmoon, et al. [16, 22].

Fig. 3.34(a) shows the combination size distribution of secondary carbides results

from the SEM study after heat treatment and in-situ observation by using CLSM in

this study. It was found that the number of secondary M7C3 carbides per unit area

increases with a decreased temperature (as 890°C×2hr). Meanwhile, compared to a

lower holding temperature and a shorter holding time (as 890°C×2hr), the size of

-52-

secondary M7C3 carbides increases with increased holding times and holding

temperatures (as 1026°C×6hr). Furthermore, SEM micrographs from the literatures

[16, 22] were studied by using an image analyzer and the size distribution results are

shown in Fig. 3.34(b). The total area, Sobs, which were taken based on the SEM

images from literatures, [16, 22] were 181.9μm2 and 81.2μm

2 respectively. Compared

to the data in Fig. 3.34(a), it can be seen that these previous results (order of size and

number) agree well with the results of the present study. The number of secondary

carbides is high at a lower temperature (as 900°C). Moreover, the size of secondary

carbides is large at higher temperature (as 1025°C).

In a short summary of secondary M7C3 carbides, the results from this section show

that a lower temperature and a shorter holding time, as 890°C×2hr, is preferred to

precipitate secondary M7C3 carbides.

3.10 Mechanical properties improved by heat treatment

Previous studies on the effect of a matrix structure on the wear resistance suggest that

a martensitic matrix offers a better wear resistance than an austenitic matrix. [48, 49]

It is also reported [1, 2] that two competing effects will influence the final properties.

These effects are (a) a high carbon martensite phase, which forms at higher

temperatures tends to increase the hardness, and (b) the increasing retained austenite

content at higher temperatures, tends to reduce the hardness. Fig. 3.35 gives the

correlation between hardness and wear resistance in as-cast condition and after heat

treatment. It is seen that both the hardness and the wear resistance is improved

markedly after heat treatment. In addition, the wear resistance increases with an

increased holding time from 2hr to 6hr. However, both the hardness and wear

resistance decrease with an increased holding temperature from 900°C to 1050°C. In

-53-

48 50 52 54 56 58 60 62 64 660.30

0.32

0.34

0.36

0.38

0.40

0.42

0.44

0.46

0.48

0.50

0.52

As cast

900C,1000C,10502hr

900C,1000C,10506hr

We

ar

los

s (

g/m

in)

Hardness (HRC)

900C6hr

1000C6hr

1050C6hr 900C2hr

1050C2hr

1000C2hr

As cast

Fig. 3.35. Mechanical properties of the hypereutectic HCCI.

the as-cast condition, the matrix almost consists solely of austenite. Clearly the

hardness of austenite is lower than that of martensite. By considering the results of

volume fraction and size distribution of M7C3, TiC and martensite units in sections

3.8 and 3.9 on the observed properties in Fig. 3.35, the mechanism of improved

mechanical properties was deduced as follows:

(1). Longer holding time leads to the transformation promotion of martensite (volume

fraction increase) as well as the refinement of martensite units (size distribution).

Consequently, longer holding times will increase the wear resistance, due to the

altered matrix toughness and the decreased hardness, which is due to the coarsening

of M7C3.

(2). Meanwhile, longer holding time leads to the growth promotion of TiC (volume

fractions) and the increase number of TiC carbides. Therefore, longer holding times

will also increase the wear resistance from the view point of the precipitation of TiC

carbides (volume fraction, size and number).

-54-

(3). In addition, a lower holding temperature leads to precipitation promotion of M7C3

(volume fraction) and the growth control/ refinement of martensite unit (size

distribution). Consequently, lower holding temperatures will generate an increased

wear resistance and hardness due to the fine dispersions of M7C3 carbides.

As a summary, the relationship between carbide precipitation, martensite units and

mechanical properties are discussed base on the volume fraction and the size

distribution of carbide precipitates (M7C3 and TiC) as well as martensite units.

Possible mechanism is proposed to explain the improvement of mechanical properties

through the heat treatment process. And the best heat treatment strategy with a lower

holding temperature close to eutectic temperature and a longer holding time, such as

900oC×6hr, is the most suitable heat treatment conditions for improving the wear

resistance properties of Ti-alloyed hypereutectic HCCI.

3.11 Estimation of the maximum carbide size by using SEV method

In this section, in order to obtain complete size distribution results, the maximum

carbide size in a hypereutectic HCCI produced with different cooling conditions,

titanium additions and heat treatment conditions was estimated by using the statistics

of extreme values (SEV) method. The experiments conditions and prediction results

for all carbides are summarized as table in detail in supplement Ⅴ.

3.11.1 Estimation of maximum primary M7C3 carbides size

Fig. 3.36 and Fig. 3.37 are plots of the size (dAmax) versus the reduced variate (yi)

from six groups samples. These extreme value distribution (EVD) figures confirm

that a linear relationship holds between the size of the maximum carbides and the

reduced variate. The reduced variate of each size data, yi, was calculated by using the

following equation:

ln ln( )1

i

iy

n

(1≤i≤n) (3-1)

n is the number of a unit volume or a unit area. It was found that the sand mold had

-55-

0 50 100 150 200 250 300-2

-1

0

1

2

3

4

5

6

Primary M7C

3 carbides:

SEV method

Ti=0%

Re

du

ce

d v

ari

ate

, y

i in

Eq

. (3

-1)

Sand mold

Graphite mold

Metal mold

dAmax, (m)

X fit

x high

x low

Fig. 3.36. Comparisons of the EVD for primary M7C3 carbides under different casting molds without

titanium addition.

0 50 100 150 200 250 300-2

-1

0

1

2

3

4

5

6

Primary M7C

3 carbides:

SEV method

Ti=0%

Ti=0.75%

Ti=1.5%

Ti=3.0%

Re

du

ce

d v

ari

ate

, y

i in

Eq

. (3

-1)

dAmax, (m)

Graphite mold

X fit

x low

x high

Fig. 3.37. Comparisons of the EVD for primary M7C3 carbides with different titanium additions in the

graphite mold

the slowest cooling rate, as expected. This means that the hypereutectic HCCI alloyed

has the largest primary M7C3 carbides and the smallest slope of the EVD regression

-56-

lines among the studied molds. The latter is especially true for carbide sizes larger

than 150μm, as seen in Fig.3.36. In addition, for the smaller size carbides, such as for

1.5mass% and 3.0mass% titanium additions have a similar slope of the EVD

regression lines, as seen in Fig. 3.37. It s likely this is due to the different growing

mechanisms. In present study, the shape factor, Circularity, was calculated as follows

in order to classify the type of carbides:

2

4 ACircularity

p

(3-2)

where A is the area of the shape and p is the perimeter. Fig. 3.38 shows a typical size

distribution of primary M7C3 carbides circularity versus the dAmax value for different

molds and titanium additions.

Fig. 3.38. Typical size distribution of primary M7C3 carbides circularity versus the dAmax value for

different molds and titanium additions

It was found that the circularity value of primary M7C3 carbides was in the range of

-57-

0.2-0.7. Moreover, that the circularity value is decreased slightly with an increased

cooling rate. However, the circularity values are not change too much in case of

different titanium additions.

In summary, the circularity value of maximum primary M7C3 carbides, which have

the typical hexagonal shape, is approximately 0.43, as shown in Fig. 3.38.

3.11.2 Estimation of maximum TiC carbides size

Fig. 3.39 gives an extreme value distribution (EVD) for experiments using the

graphite mold and different titanium contents. It confirms that a linear relationship

holds between the size of the maximum TiC carbide and the reduced variate.

Moreover, it was found that the tendency plots for a 3.0% titanium condition was

suddenly changed in the region of a reduced variate value of yi >2.5 and for a dAmax

0 10 20 30 40-2

-1

0

1

2

3

4

5

6

SEV method

Graphite mold

TiC carbides:

Ti=0.75%

Ti=1.5%

Ti=3.0%

Re

du

ce

d v

ari

ate

, y

i in

Eq

. (3

-1)

dAmax, m)

x fit

X low

X high

Fig. 3.39. Comparisons of EVD for TiC carbides with different titanium additions for graphite Mold

value >15μm, as shown in Fig. 3.39. It was believed that is due to the different shapes

and sizes of TiC carbides. It can be seen in Fig. 3.40 that the circularity value of TiC

carbides decrease from 0.45 to 0.16 as the titanium content is increased from 0.75% to

3.0%. In addition, the difference circularity value of TiC carbides is due to the shape

-58-

change of TiC carbides. Fig. 3.40 also shows that the typical single TiC carbides have

a cubic shape, which circularity value is about 0.45 for the case with a 0.75% titanium

addition. However, the single TiC carbides seem to be adhered to each other and start

to form cluster for the experiments with a 1.5% titanium addition. The circularity

value is around 0.36. Furthermore, many TiC carbides aggregate with each other and

form big cluster for the experiments with an addition of 3.0% titanium. Then, the

circularity value is reduced to 0.16.

Fig. 3.40. Typical size distribution of TiC carbides circularity versus the dAmax value for different

titanium additions.

3.11.3 Estimation of maximum secondary M7C3 carbides size

Fig. 3.41 shows the extreme value distribution (EVD) for different heat treatment

conditions. It confirm that a linear relationship holds between the size of the

maximum secondary M7C3 carbide and the reduce variate. In general, it was found

that the slope of the EVD regression lines is not changed by changing the heat

-59-

0 1 2 3 4 5 6-2

-1

0

1

2

3

4

5

6

Ti=1.5%

SEV method

Secondary carbides:

Graphite mold

890oC2hr

1026oC2hr

890oC6hr

1026oC6hr

Re

du

ce

d v

ari

ate

, yi

in E

q.

(3-1

)

dAmax, (m)

X fit

X low

X high

Fig. 3.41. Extreme value distribution (EVD) of secondary M7C3 carbides for different heat

treatment conditions.

Fig. 3.42. Typical size distribution of secondary M7C3 carbides circularity versus the dAmax values for

different heat treatment conditions.

treatment conditions, as shown in Fig. 3.41. However, in the region of a reduced

variate value yi>1.5 and for a dAmax>2μm, it was found that the tendency plots were

-60-

0 20 40 60 80 100-2

-1

0

1

2

3

4

5

6

SEV method Graphite mold

Ti=1.5%

R

ed

uc

ed

va

ria

te, y

i in

Eq

. (3

-1)

dAmax, m)

Secondary M7C

3 carbides

TiC carbides

Primary M7C

3 carbides

X fit

X low

X high

Fig. 3.43. Comparisons of the EVD of primary M7C3 carbides, TiC carbides and secondary M7C3

carbides for graphite mold with 1.5% titanium addition.

a little bit changed for the 1026oC×2hr/6hr conditions. In addition, the shape of the

secondary M7C3 carbide is rod-like and irregular and the mean circularity value is in

range of 0.17-0.35, as shown in Fig. 3.42.

A summary of the extreme value distribution (EVD) among primary M7C3 carbides,

TiC carbides and secondary M7C3 carbides for a case with a 1.5% titanium addition

and casting in a graphite mold is shown in Fig. 3.43. The slope of the regression line

for these three carbide types is different. The reasons for this discrepancy in the EVD

value depend on their characteristic, such as the shape factor. It was found that the

typical morphology of primary M7C3 carbides, TiC carbides and secondary M7C3

carbides is of a hexagonal shape, a cubic shape and an irregular shape, respectively.

Moreover, the corresponding circularity value for three carbide types is quite different.

Compared to the circularity value of TiC carbides and secondary M7C3 carbides, the

circularity value of primary M7C3 carbides is larger (about 0.43) than the one for TiC

carbides and secondary M7C3 carbides. A comparison of the estimated and observed

-61-

0 50 100 150 200 250 3000

50

100

150

200

250

300

Ti=1.5%moldGraphite

Ti=0%

Graphite mold

Metal mold

1/1

Primary M7C

3 carbides

Es

tim

ate

d m

ax

imu

m s

ize

, (

m)

Observed maximum size, (m)

Ti=0%Sand mold

Ti=0%

Graphite moldTi=0.75%

Graphite

Ti=1.5%

Ti=3.0%

mold

TiC carbides

Secondary M7C

3 carbides

Fig. 3.44. Comparisons of the estimated and observed maximum size of primary M7C3 carbides, TiC

carbides and secondary M7C3 carbides. Note: The error bar is standard error.

maximum size of primary M7C3 carbides, TiC carbides and secondary M7C3 carbides

for the different experimental conditions is shown in Fig. 3.44. It can be seen that the

estimated maximum size of primary M7C3 carbides are larger than the observed ones.

This is especially true for the larger primary M7C3 carbides, such as in the sand and

graphite mold experiments without titanium additions. Moreover, the estimated

maximum size shows a large error bar (18.9μm in sand mold). However, the smaller

carbides, such as primary M7C3 carbides in the graphite mold with 1.5% and 3.0% Ti

additions, TiC carbides and secondary M7C3 carbides show a smaller error. The errors

are 0.92μm-2.13μm for TiC carbides and 0.17μm-0.5μm for the secondary M7C3

carbide, respectively. The predicted smaller carbides with small errors result in agree

well with the observed one, as shown in Fig. 3.44.

-62-

In summary, the complete size distribution results can be obtained base on the

maximum carbides size analysis. In addition, the characteristic of the different

maximum carbide types are summarized and classified based on the shape factor.

Then, based on a combination of a size distribution analysis and a maximum size

analysis, the relationship between the carbide size distribution including the

maximum carbides size and the mechanical properties can be discussed.

-63-

Chapter 4

Concluding Discussion

This work has had the aim to develop superior abrasion resistant castings. Thus, a

systematic research has been carried out to study the solidification and heat treatment

of hypereutectic High Chromium Cast Iron (HCCI) containing mainly

Fe-17mass%Cr-4mass%C. The work, which is presented in Supplement I, started

with the aim to refine the carbides size and improve the mechanical properties of

hypereutectic HCCI alloyed with Ti. In Supplement I, the primary M7C3 carbides size

was successfully refined with an increased cooling rate and titanium addition. The

mechanical properties, such as hardness, were improved with an increased cooling

rate and titanium addition. However, the mechanical properties, such as the wear

resistance, were found to become worse when too much titanium was added to the

as-cast condition.

The prediction of the fraction of a solidification structure and a phase diagram of

hypereutectic HCCI (Fe-17mass%Cr-4mass%C multicomponent alloys) was also

done in Supplement I. This was performed by using a multiphase micro-segregation

model (Gulliver-Scheil and Partial Equilibrium), which was coupled with the

CALPHAD method and using the Thermo-Calc software database. Furthermore, in

Supplement Ⅱ , the solidification path and effect of C, Cr, Ti content on the

compositions and amounts of Fe-17mass%Cr-4mass%C multicomponent alloys were

studied both experimentally and theoretically. Here, the focus was to understand the

solidification mechanism of hypereutectic HCCI more deeply.

-64-

Finally, an as-cast sample with a 1.5mass% titanium addition and being casted in a

graphite mold was selected as a good candidate for the following heat treatment

studies.

Supplement Ⅲ focus on different heat treatment processes that were carried out on

as-cast hypereutectic HCCI to study how improved mechanical properties could be

obtained. The relationship between carbide precipitation, martensite units and

mechanical properties were discussed after heat treatment. This was done based on

the volume fraction and the size distribution of carbide precipitates (M7C3 and TiC) as

well as the martensite units. The best heat treatment strategy, which is 900oC×6hr, is

suggested in Supplement Ⅲ. However, the main focus of Supplement Ⅲ is set on

the martensite units. The results show that the fine martensite units make a big

contribution for wear resistance properties of hypereutectic HCCI. However, in

Supplement Ⅲ, it did not go into a lot of detail about the behavior of the secondary

carbides, which precipitated from austenite matrix during the heat treatment process.

Consequently, in Supplement Ⅳ , in-situ observation on secondary carbide was

performed by using a Confocal Laser Scanning Microscope (CLSM) in order to

understand the dynamic precipitation behavior of secondary carbides. The type of

secondary carbide is identified as a M7C3 type by using transmission electron

microscopy (TEM). It was found that the precipitation of secondary M7C3 carbides

started during the heating process (~840°C). Furthermore, the precipitation of

secondary carbides will almost be finished within 2hr after reaching the holding

temperature. This is due to that the volume fraction of secondary carbides appeared to

be unaffected by the holding time and the holding temperature after a holding time of

2hr.

Finally, the focus in Supplement Ⅴ was to get more exactly size distribution results

to fully explain the relationship between the size distribution and the mechanical

properties. More specifically, the maximum carbide size, such as primary M7C3

carbides, TiC carbides and secondary M7C3 carbides in Fe-17mass% Cr-4mass% C

hypereutectic HCCI alloyed with titanium, was studied by using the statistics of

-65-

extreme values (SEV) methods. Moreover, the combined method between a size

distribution analysis and a maximum analysis was proposed in Supplement Ⅴ. A

more accurate carbides size distribution result was obtained by supplementing the

maximum carbides size information to normal size distribution, as shown in Fig. 4.1.

Fig. 4.1. Combination of size distribution analysis and maximum size analysis of primary M7C3

carbides, TiC carbides and secondary M7C3 carbides.

-66-

Then, based on a combination of a size distribution analysis (Supplement I, Ⅲ，Ⅳ)

and a maximum size analysis (Supplement Ⅴ), the relationship between the carbides

size distribution including the maximum carbides size and the mechanical properties

could be discussed.

Fig. 4.1(a) shows that the size of the primary and eutectic M7C3 carbides decreased

and the number of primary and eutectic M7C3 carbide increased with an increased

cooling rate ranging from a sand mold to a metal mold. As a result, the hardness of the

hypereutectic HCCI in an as-cast condition increased with a decreased primary and

eutectic M7C3 carbide size, as shown in Fig. 4.1 and Fig. 4.2.

48 50 52 54 56 58 60 62 64 660.30

0.32

0.34

0.36

0.38

0.40

0.42

0.44

0.46

0.48

0.50

0.52

Primary M7C

3:

Sand mold

max. 259.1m

TiC: max. 26.7mTi=3.0%

Ti Hardness Wear lossMold

Hardness, HRC

Sand : 0%

Graphite: 0%

0.75%

1.5%

3.0%

Metal: 0%

W

ea

r lo

ss

, g

/min

Primary M7C

3:

Graphite mold

max. 196.6m

Primary M7C

3:

Metal mold

max. 132.9m

TiC: max. 19.8mTi=0%-1.5%

Fig. 4.2. Mechanical properties of the hypereutectic HCCI in as-cast conditions.

Moreover, this tendency was also confirmed by the predicted maximum primary

M7C3 carbide sizes for different molds. In short, the fine primary and eutectic will

improve the hardness of hypereutectic HCCI. In addition, compared to 0%-1.5%

titanium additions, it was found that the predicted maximum TiC carbide size, which

was found for a 3.0% titanium addition, is 26.7μm, as shown in Fig. 4.1(b). Moreover,

higher titanium additions of 3.0% were found to be worse to improve the wear

resistance of hypereutectic HCCI in as-cast conditions, as shown in Fig. 4.2. The

reason might be that the agglomerated TiC carbides for 3.0% titanium additions are

easily peeled off during the wear application. As discussed in chapter 3, the wear

-67-

48 50 52 54 56 58 60 62 64 660.30

0.32

0.34

0.36

0.38

0.40

0.42

0.44

0.46

0.48

0.50

0.52

Graphite mold

Ti=1.5%

max. 3.79m

Secondary M7C

3:

As-cast

900C,10502hr

900C,10506hr

W

ear

loss

, g/m

in

Hardness, HRC

900C6hr

1050C6hr900C2hr

1050C2hrAs-cast

max. 2.58m

Secondary M7C

3:

max. 2.16m

Secondary M7C

3:

max. 3.15mSecondary M

7C

3:

Fig. 4.3. Mechanical properties of the hypereutectic HCCI with different heat treatment conditions.

resistance properties of hypereutectic HCCI are mainly affected by size and number

of martensite units after heat treatment. More specifically, the presence of fine

martensite units for a 900oC×2hr/6hr conditions make a big contribution to improve

the wear resistance properties of hypereutectic HCCI, as shown in Fig. 4.3. However,

the secondary M7C3 carbide, which precipitated from the matrix during the heat

treatment process, might also have an influence on the mechanical properties of a

hypereutectic HCCI. This is particularly true for the hardness of hypereutectic HCCI.

As shown in Fig. 4.1(c), it was found that the size of the secondary M7C3 carbide

increased as the heat treatment conditions changed from 900oC×2hr/6hr to

1050oC×2hr/6hr. This result led to a reduced hardness of hypereutectic HCCI, as

shown in Fig. 4.3.

-68-

-69-

Chapter 5

Conclusions and Future Work

5.1 Conclusions

This thesis focuses on refining the microstructure and improving the mechanical

properties of Ti-alloyed hypereutectic high chromium cast iron (HCCI). The size

distribution, composition and morphology of these carbides and martensite units in

the Ti-alloyed hypereutectic HCCI with main composition of Fe-17mass%

Cr-4mass% C were quantitatively studied by using both experimental methods and

theoretical methods. The most important conclusions reached in this study can be

summarized as follows:

General conclusions of supplement I:

1). The microstructures of as-cast Ti-alloyed hypereutectic HCCI were refined by an

increased cooling rates and titanium addition;

2). The M7C3 carbides were roughly classified into “primary M7C3 carbides” and

“eutectic M7C3 carbides” with a 11.2μm border size based on size distribution results

and volume fraction results; The TiC carbides with high formation temperatures can

not only act as nuclei of M7C3 carbides, but they also contain a core, Ti(C, N). It was

also found that TiC carbides precipitation was not obviously affected by the cooling

rate;

3). The hardness of the hypereutectic HCCI increased slight with an increased the

cooling rats and titanium addition. The wear resistance were found become worse

-70-

when too much titanium addition. However, the mechanical properties (hardness and

wear resistance) do not change significantly under the as-cast conditions.

General conclusions of supplement Ⅱ

1). The precipitation sequence of MC and M7C3 carbides is greatly affected by the

nominal contents of C, Cr and Ti elements in the HCCI. The composition of Cr in the

M7C3 phase increases with both the Cr and Ti contents, but decreases with the C

content. Both the increase of C and Cr lead to an increased volume fraction of M7C3

carbide. However, they have no influence on MC carbide;

2). The PE predictions shows that a potential exists to predict the solidification path of

HCCI at cooling rates between 0.4°C

/s to 1.5

°C

/s. This conclusion is based on the

agreement between the predictions and the experimental thermal analysis data from

the ingot.

General conclusions of supplement Ⅲ:

1). From the viewpoint of the precipitation promotion of M7C3 (volume fraction) and

the growth control/ refinement of M7C3, the growth promotion of TiC (volume

fractions) and the transformation promotion of martensite (volume fraction) and the

refinement of martensite unit (size distribution), the best heat treatment process for

improving the wear resistance properties is suggested as 900oC×6hr.

General conclusions of supplement Ⅳ

1). The secondary carbides were identified as a M7C3 type found in Fe-4 mass%C-17

mass%Cr-1.5 mass% Ti hypereutectic HCCI alloys.

(2). Precipitation of secondary M7C3 carbides started during the heating process

(~840°C) and the number of these secondary M7C3 carbides increased sharply at

temperatures higher than 840°C. Moreover, the precipitation of secondary M7C3

-71-

carbides will almost finish within 2hr after reaching the holding temperature since the

volume fraction of secondary carbides appeared unaffected by the holding time and

holding temperature after holding for 2hr. In summary, a lower temperature and a

shorter holding time, such as 890°C×2hr, are preferred to precipitate secondary M7C3

carbides.

General conclusions of supplement Ⅴ

1). The shape factor (circularity value) of maximum primary M7C3 carbides, TiC

carbides and secondary M7C3 carbides was found to be different. It leads to different

slope of the extreme value distribution (EVD) regression lines for these carbides. The

large sized carbides such as primary M7C3 carbides have a lower slope than the

smaller carbides such as TiC carbides and secondary M7C3 carbides.

2). The combined method between a size distribution analysis and a maximum

analysis is proposed in this study. Moreover, a more accurate carbides size

distribution result was obtained by supplementing the maximum carbides size

information to normal size distribution.

3). The combination of a size distribution analysis and a maximum analysis results

showed that the primary M7C3 carbides and secondary M7C3 carbides in as-cast

condition and heat treatment condition are related to the hardness of hypereutectic

HCCI. More specifically, it will decrease with an increased maximum size of primary

M7C3 carbides and secondary M7C3 carbides. Moreover, it is confirm the wear

resistance property will become worse due to increasing the size of TiC carbides

(titanium addition) from viewpoint of maximum size analysis.

5.2 Suggestions for future work

In the present thesis, Ti as a carbide promoter was selected to refine the

microstructure of hypereutectic HCCI. Actually, other strong carbides forming

-72-

elements, such as Zr, Hf, V, Nb and B can also be used to refine the microstructure of

hypereutectic HCCI according to the wear applications. Moreover, the complex

addition of those elements in hypereutectic HCCI may also improve the

microstructure due to the nucleation theory.

The three-dimensional 3D X-ray computed tomography technic or electric extraction

experiment for hypereutectic HCCI might be used to reconstruct or extract 3D

carbides information. Thus, it will be possible to get more clearly (3D) experimental

results on the carbides size distribution. In addition, statistical analysis methods, such

as Schwartz-Saltykov (ss) method, is also very important to estimate the 3

dimensional (3D) carbides size distribution base on 2 dimensional (2D) carbides size

distribution by using Schwartz-Saltykov (ss) method. Furthermore, these results may

thereafter be used to verify the prediction and calculation.

In order to obtain very accurate carbides size distribution, not only maximum carbides

size, the small size and volume fraction, which is omitted due to the resolution of

characterization equipment, also should be estimated by statistical methods.

-73-

References

[1] F. Maratray and A. Poulation: Transactions of the American Foundry Society,

90(1982) 795.

[2] F. Maratray: Transactions of the American Foundry Society, 79(1971) 121.

[3] J.T.H. Pearce: Journal of Materials Science Letters, 2(1983), 428.

[4] Ö.N.Dogan, J.A.Hawk and J.H. Tylczak: Wear, 250(2001), 462

[5] H. Miyahara, S. V. Bravo, K. Yamamoto and K. Ogi: ISIJ International, 49(2009),

1075.

[6] H. G. Fu, X. J. Wu, X. Y. Li, J. D. Xing, Y. P. Lei and X. H. Zhi: Journal of

Materials Engineering and Performance, 18(2009), 1109.

[7] X. J. Wu, J. D. Xing, H. G. Fu and X. H. Zhi: Materials Science and Engineering

A, 457(2007), 180.

[8] X. H. Zhi, J. D. Xing, H. G. Fu and Y. M. Gao: Materials Characterization,

59(2008), 1221.

[9] R.J.Chung, X. Tang, D.Y. Li, B. Hinckley and K. Dolman: Wear, 267(2009), 356.

[10] M.M. Arikan, H. Çimenoglu and E.S. Kayali: Wear, 247(2001), 231.

[11] A. Bedolla-Jacuinde, R. Correa, J.G. Quezada and C. Maldonado: Materials

Science and Engineering A, 398(2005), 297.

[12] A. Bedolla-Jacuinde, R. Correa, I. Mejıa, J.G. Quezada and W.M. Rainforth:

Wear, 263(2007), 808.

[13] A. Bedolla-Jacuinde, S. Aguilar and C. Maldonado: Journal of Materials

Engineering and Performance, 14(2005), 301.

[14] S K. Hann, J.D. Gates and J.V. Bee: Journal of Materials Science, 32(1997),

3443.

-74-

[15] E. Pagounis, M. Talvitie and V.K. Lindroos: Metallurgical and Materials

Transactions A, 27A(1996), 4171.

[16] A. Wiengmoon, T. Chairuangsri, A. Brown, R. Brydson, D.V. Edmonds and

J.T.H. Pearce: Acta Materialia, 53(2005), 4143.

[17] Y. Wang, M.Y. Li, B. Han, T. Han and Y.Y. Cheng: Journal of Materials

Science, 45(2010), 3442.

[18] J. Wang, C. Li, H. H. Liu, H.S. Yang, B. L. Shen, S. J. Gao and S. J. Huang:

Materials Characterization, 56(2006), 73.

[19] Z. P. Sun, R. Zuo, C. Li, B. L. Shen, J. Yan and S. J. Huang: Materials

Characterization, 53(2004), 403.

[20] G. L. F. Powell: Journal of Materials Science, 31(1996), 707.

[21] X. Zhi, J. Xing and Y. Gao: Materials Science and Engineering A, 487(2008),

171.

[22] A. Wiengmoon, T. Chairuangsri and J.T.H. Pearce: ISIJ International, 44(2004),

396.

[23] A.Wiengmoon, T. Chairuangsri and N. Chomsang: Journal of Materials Science

and Technology, 24(2008), 330.

[24] A. Wiengmoon and T. Chairuangsri: Materials Science and Engineering A,

480(2008), 333.

[25] A. Wiengmoon, J. T. H. Pearce and T. Chairuangsri: Materials Chemistry and

Physics, 125 (2011) 739.

[26] J. T. H. Pearce and D. W. L. Elwell: Journal of Materials Science Letters, 5

(1986), 1063.

-75-

[27] Q. Liu, H.W. Zhang, Q. Wang, X.K. Zhou, P. G. Jönsson and K. Nakajima: ISIJ

International, 52(2012), 2210.

[28] Q. Liu, H.W. Zhang, Q. Wang, X.K. Zhou, P. G. Jönsson and K. Nakajima:

ICRF 2012, 1st International Conference on Ingot Casting, Rolling and Forging,

05.06.2012, Aachen. Düsseldorf: Stahl-Zentrum 2012, ICRF-31.pdf (CD-ROM).

[29] Q. Liu, H.W. Zhang, Q. Wang, X.K. Zhou, P. G. Jönsson and K. Nakajima: ISIJ

International, 52(2012), 2288.

[30] Q. Liu, H. Shibata, P. Hedström, P. G. Jönsson, K. Nakajima: ISIJ International,

53(2013), 1239.

[31] Q. Liu, P. G. Jönsson, K. Nakajima: accepted by ISIJ International for

publication, No. 12, 2013.

[32] H. Chikama, H. Shibata, T. Emi and M. Suzuki: Materials Transactions, JIM,

37(1996), 620.

[33] H. Shibata, H.B. Yin, S. Yoshinaga, T. Emi and M. Suzuki: ISIJ International, 38

(1998), 149.

[34] J. A. Small and J. R. Michael: Journal of Microscopy, 201(2001), 59.

[35] M. Godec, B.Š. Batič, D. Mandrino, A. Nagode, V. Leskovšek, S.D. Škapin and

M. Jenko: Materials Characterization, 61(2010 ), 452.

[36] F.J. Humphreys: Journal of Materials Science, 36(2001), 3833.

[37] Y. Murakami: Metal Fatigue: Effects of Small Defects and Nonmetallic

Inclusions, (Tokyo, Japan: Yokendo, 1993).

[38] Y. Murakami: Journal of Research of NIST, 99(1994), 345.

[39] Y. Murakami: Transactions of the Japan Society of Mechanical Engineers,

56(1990), 162.

-76-

[40] Y, Murakami , T. Toriyama and E. M. Coudert: Journal of Testing and Evaluation,

22(1994), 318.

[41] ASTM E2283-03, West Conshohocken, PA, American Society for Testing and

Materials, (2003).

[42] M. Tsuda: Tetsu-to-Hagané, 80 (1994), 74.

[43] Q. Chen and B. Sundman: Materials Transactions, JIM, 2002, No. 3, P. 551.

[44] H.W. Zhang, Ch. –A. Gandin, J.C. He and K. Nakajima: IOP Conf. Ser.:

Materials Science and Engineering, 33(2012), 012061.

[45] B. Sundman and Q. Chen: Thermodynamic Calculation Interface TQ-Interface

V7.0, 2008, Thermo-Calc software AB (Stockholm, SE).

[46] P. Shi: TCS steels/Fe-alloys database V6.0, 2008, Thermo-Calc Software AB

(Stockholm, SE).

[47] J.T.H. Pearce, T. Chairuangsri, A. Wiengmoon, N. Poolthong and H. Nomura:

China Foundry, 4(2007), 38.

[48] I.R. Sare and B.K. Arnold: Metallurgical and Materials Transactions, 26A(1995),

357.

[49] A. Bedolla-Jacuinde, R. Correa, I. Mejia, J.G. Quezada and W.M. Rainforth:

Wear, 263(2007), 808.