Top Banner
RWTH Aachen Institute of Ferrous Metallurgy MASTER THESIS Master Student Alireza Saeed-Akbari Matr. –Nr. 268696 Subject: Determination of Steels Microstructural Components Based on Novel Characterization Techniques Supervisors: Univ. Prof. Dr.-Ing. W. Bleck M.Sc. Malek Naderi
66

Master Thesis

Nov 12, 2014

Download

Documents

irfriend

Determination of Steels Microstructural Components Based on Novel Characterization Techniques

By: Alireza Saeed-Akbari (RWTH Aachen)
Submission Date: Fall 2007
Supervisors:
Professor Wolfgang Bleck
Dr. Malek Naderi
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Master Thesis

RWTH Aachen Institute of Ferrous Metallurgy

MASTER THESIS

Master Student Alireza Saeed-Akbari

Matr. –Nr. 268696

Subject:

Determination of Steels Microstructural

Components Based on Novel Characterization

Techniques

Supervisors: Univ. Prof. Dr.-Ing. W. Bleck M.Sc. Malek Naderi

Page 2: Master Thesis

To mum and dad,

for the whole love they gave me in my life…

Page 3: Master Thesis

Acknowledgment The achievements of the current work are the results of helps

and supports of all technicians, assistants, and colleagues at the

Institute of Ferrous Metallurgy (IEHK), RWTH Aachen

University.

Among all, I should appreciate Professor Wolfgang Bleck –

head of the Institute of Ferrous Metallurgy – for his kind

considerations during the project; Mr. Malek Naderi for his

invaluable indications, instructions, and cooperation, and IEHK

technicians for their quick and accurate experiments.

I would like to give special thanks to Mrs. Aida Nonn for her

kind devotion to this investigation; without her helps, the

examination of weld structures was impossible. The

microstructural and relevant primary hardness data of welds in

the current work are parts of her assistance to broaden the

discussions of the project.

Alireza S. Akbari

Page 4: Master Thesis

Abstract In the present work, hardness mapping and dilatometry experiments were

employed as the novel methods for the quantitative and qualitative

assessments of the heterogeneous microstructures in different steels. The

results were initially compared with the optical microscopy images to

show the lack of accuracy in case of handling the heterogeneous

compressive specimens using optical microscopy. It was then

demonstrated that dilatometry and hardness mapping results are

consistent.

In another attempt, a mathematical formula was developed to predict the

fraction of martensite based on the bulk hardness of the investigated

specimens. The formula gave reliable results due to both steels studied.

Surface hardness mapping and optical microscopy methods were also

utilized to study the microstructural phases within the fusion zones and

the heat affected zones of steel welds. Both macro- and microstructures

were examined in this regard. It was concluded that the parallel

application of hardness measurements and optical microscopy was

essential when handling the weld zones.

Page 5: Master Thesis

Table of Contents

1. Introduction 1

2. Fundamentals 3 2.1. Introduction, 3

2.2. Hardness Testing, 4 2.2.1. Background, 4 2.2.2. Microstructural Aspects, 5 2.3. Dilatation Experiment, 7 2.4. Martensitic Transformation and Microstructure, 9 2.5. Bainitic Transformation and Microstructure, 11 2.6. Characterization Difficulties, 15 2.7. Microstructural Aspects of Weld Zones, 16

3. Experimental Procedure 22 3.1. Material Characterization, 22 3.1.1. General Considerations, 22 3.1.2. Chemical Composition, 22 3.1.3. Microstructure, 23 3.2. Isothermal and Non-isothermal Compression Tests, 25 3.3. Execution of Welding, 26 3.4. Hardness Mapping and Metallography, 27

4. Results and Discussion 30 4.1. Revealing Heterogeneous Microstructures, 30 4.2. Reliability of Dilatometry and Hardness Mapping Methods, 33

4.2.1. Hardness Mapping, 33 4.2.2. Dilatometry Data, 35

4.3. Prediction of Martensite Fraction Using Hardness Criterion, 38 4.3.1. Developing the Formula, 38 4.3.2. Checking the Consistency of Results, 42

4.4. Microstructural Study of Weld Zones, 45

5. Conclusions 58 References i

Page 6: Master Thesis

Chapter 1 – INTRODUCTION 1

CHAPTER

ONE

INTRODUCTION

The fast yet accurate estimation of microstructural phases in steels might

be interesting due to both scientific and industrial points of view. In

addition, developing the reliable alternative experiments for the

calculation of phase distribution and phase fraction which can replace the

conventional microscopy techniques is the matter of interest based on its

time and cost saving impacts.

It has been well established that each single microstructural phase in steel

has its own hardness range. The distribution of such hardness values in

the pre-assumed hardness intervals is interconnected with the

arrangement and amount of different phases within the microstructure.

There are many research works and reports regarding the relationship

among the hardness of each phase, its chemical composition, and its

related mechanical properties [1-3]. In some reports [4,5], hardness

profiles were used to study the homogeneity of microstructures. Some

others [6-8] used surface hardness data to correlate hardness maps and

macrostructure of the entire weld zones, including heat affected zone and

the steel base plate.

Phase transformations occur during cooling or heating. Any

transformation results in a specific phase and causes a definite contraction

or expansion. For instance, transformation of austenite to martensite

during cooling yields a certain dilatation. It should be pointed out that the

dilatation term is mainly the magnitude of plastic deformation resulting

from martensitic transformation. This dilatation is an invariant plane

Page 7: Master Thesis

Chapter 1 – INTRODUCTION 2

strain, which is the combined effect of a uniaxial dilatation and a simple

shear. The higher volume fraction of martensite results in higher

dilatation magnitudes.

It is known that the relative length changes are proportional to the relative

volume changes if the system undergoes a macroscopically isotropic

phase transformation. In the case of anisotropy that may occur in the

specimens exhibiting preferred orientation for the constituent

crystallites/grains, distinction should be made between the length and the

volumetric changes [9]

In the present work, hardness mapping and dilatometry experiments were

employed as the novel methods for the assessment of microstructural

phases in different steels. The results were initially compared with the

optical microscopy images to show the lack of accuracy in case of

handling the heterogeneous compressive specimens using optical

microscopy. It was then demonstrated that dilatometry and hardness

mapping results are consistent.

In another attempt, a mathematical formula was developed to predict the

fraction of martensite based on the bulk hardness of the investigated

specimens. The formula gave reliable results due to both steels studied.

Surface hardness mapping and optical microscopy methods were also

utilized to study the microstructural phases within the fusion zones and

the heat affected zones of steel welds. Both macro- and microstructures

were examined in this regard. It was concluded that the parallel

application of hardness measurements and optical microscopy was

essential when handling the weld zones.

Page 8: Master Thesis

Chapter 2 – FUNDAMENTALS 3

CHAPTER

TWO

FUNDAMENTALS

2.1. Introduction

Phase transformations can be well studied using quantitative and

qualitative microstructural investigations as well as consideration of

dilatation data. The first method helps to recognize present phases and

fraction of each phase within the microstructure. The latter not only aids

to specify the start and finish temperatures of the phase transformations,

but also can be used to quantify the volume fraction of the produced

phases.

Based on the above mentioned reasons, the microstructural investigations

using light optical and scanning electron microscopy techniques are

carried out. The problem which has to be taken into account is that, the

compressive deformation is heterogeneous, i.e. the effective strain and the

cooling rate along the horizontal centerline of the samples are varied.

Consequently, the distribution of the produced phases is not

homogeneous when handling the compression specimens. It means that

light optical or scanning electron microscopy investigations on few

selected points of the heterogeneously deformed surfaces are not proper

techniques for the recognition of the phase fractions inside the

experimental specimen.

In this chapter brief statements are given regarding the concept and the

application of hardness measurement in the materials investigations.

Page 9: Master Thesis

Chapter 2 – FUNDAMENTALS 4

In addition, the dilatometry technique is introduced to measure the

temperature and the amount of the possible transformations during

cooling from the austenitization temperature.

Moreover, the fundamentals of martensitic and bainitic transformations

are reviewed. It is shown that similarities between certain bainitic and

martensitic microstructures may also reduce the accuracy of the

conventional optical microscopy techniques to assess the fraction of

microstructural phases in steels.

It is then concluded that due to both mentioned problems, i.e.

microstructural heterogeneity and distinguishing between fine bainitic

and martensitic structures, hardness mapping and dilatometry can be

employed as the alternative characterization methods.

The chapter is closed by notes regarding the development of

microstructures in the weld heat-affected-zone (HAZ) and the weld fusion

zone (FZ). It is also shown that in this case, hardness mapping can be

utilized as a reasonable method to study the microstructural state.

2.2. Hardness Testing

2.2.1. Background

With regard to metallic materials, hardness has always been (and still is) a

subject of much discussion among metallurgists, and material scientists.

Attributes like wear resistance, deformation behavior, tensile strength, as

well as modulus of elasticity are connected with the term hardness.

An exact description of the method must be made if one wishes to

compare the obtained readings with each other in order to achieve a

usable hardness value.

Page 10: Master Thesis

Chapter 2 – FUNDAMENTALS 5

Hardness testing is almost nondestructive and in many cases used for

determining parameters to differentiate and describe materials. For

example, hardness values can easily provide data on the strength

properties of a material.

The term hardness is generally understood as being the resistance of a

material against the penetration of a body made of a stronger material.

Hardness is therefore not a fundamental quantity of a material but always

a response of the material to certain load or test method. A hardness value

is calculated on the basis of the response of the material to this load [10].

2.2.2. Microstructural Aspects

The concept of hardness in steels is divided into two major categories

namely hardening capacity, i.e. the maximum achievable hardness level

in the selected steel, and hardness penetration, i.e. the change in hardness

as dependent on the distance from the surface. While the first item is

dependent on the carbon content, the latter is strongly affected by the

content of alloying elements [11].

As the hardness penetration is out of the discussions of the current work,

the hardening capacity is outlined here.

Maximum hardness in steels is obtained by producing a fully martensitic

structure. This can be done by austenitizing the steel and then quenching

it. During the austenitizing treatment all of the carbides dissolve and the

ferrite transforms into austenite. Quenching this structure causes the

austenite to transform via a shear mechanism into martensite. This

transformation is so fast (martensite needles grow at close to the speed of

sound) that there is no time to the carbon to diffuse out of the martensite

Page 11: Master Thesis

Chapter 2 – FUNDAMENTALS 6

grains or to form carbide phases. The martensite, supersaturated with

carbon, is very hard and also very brittle.

Carbon, being a very effective solid solution strengthening agent,

essentially determines the hardness of martensite. Cases where a lesser

degree of hardening can be attributed to the presence of other alloying

elements, but these elements tend to also make it more difficult to obtain

a fully martensitic microstructure. So while maximum hardness in a given

steel is dependent on our ability to produce a fully martensitic

microstructure, the hardness of the martensite is largely determined by its

carbon content [12].

In some instances, especially for hot rolled steels requiring enhanced

capability to resist stretching on a blanked edge (as typically measured by

hole expansion capacity), the microstructure can also contain significant

quantities of bainite [13].

The hardness of bainite also linearly increases with carbon concentration

by approximately 190 HV per wt%. This contrasts with a change of about

950 HV per wt% in the case of carbon supersaturated martensite.

For mixed microstructures, the hardness depends on the transformation

temperature and composition. This is because the stability of the residual

austenite to martensitic transformation changes with its carbon

concentration, the limiting value of which depends on the transformation

temperature [3].

Page 12: Master Thesis

Chapter 2 – FUNDAMENTALS 7

2.3. Dilatation Experiment

Changes in physical properties can be determined through differential

thermal analysis, measurement of magnetic or electrical properties, or

through measurements using a dilatometer, which is the most prevalently

used method.

The dilatometer measures the change in length ( L∆ ), which is dependent

on temperature and time (Figure 2-1). According to SEP 1681, test

evaluation is either isothermal (T=const.) or continuous [11].

Figure 2- 1. Change in length during isothermal and continuous

transformation for the determination of TTA and TTT

diagrams, after SEP 1681 [11].

Figure 2-2 shows how to evaluate an experiment involving continuous

cooling. The beginning of a transformation range is signified by the

deviation from a straight line in the graph of change in length over time.

If the curve continues into another straight line, then the transformation in

that field is finished. If different microstructural components form

directly after one another, then a turning point in the curve signifies the

boundary between transformation ranges.

Page 13: Master Thesis

Chapter 2 – FUNDAMENTALS 8

Figure 2- 2. Example of how to evaluate experiments using continuous

cooling, after SEP 1680 [11].

Finally, the temperature is graphed over the logarithmic time axis in a

nonequilibrium diagram. This makes it possible to investigate the

transformation behavior of the steel from short to long periods of time.

The accuracy of nonequilibrium diagrams varies for temperature by +10K

and for time by +10%. Variations due to different chemical compositions

and initial states can further increase the range of distribution. All

diagrams are valid, in the strictest sense, only for the chemical

compositions observed and the conditions given for each specific case

(austenitizing temperature and austenitizing time).

Page 14: Master Thesis

Chapter 2 – FUNDAMENTALS 9

In every nonequilibrium diagram, a multitude of transformation points are

given along with the specifications for the material and experimental

procedures, in accordance with the Stahl-Eisen-Prüfblatt 1680

(SEP 1680).

2.4. Martensitic Transformation and Microstructure

Martensitic transformation in steels occurs when austenite is cooled

rapidly (quenched) below Ms temperature so that diffusional

transformation is suppressed. Even pure iron may transform

martensitically if the cooling rate is sufficiently high.

Crystallographic characteristics of martensite transformation involve

shape change, surface relief, internal defects and existence of habit

planes. Other characteristics such as athermal transformation,

stabilization and autocatalytic effect also make the mechanism of

transformation difficult to understand.

The diffusionless transformation accompanies lattice distortion and

supersaturation of carbon. Inner microstructure of martensite plates

consists of either dislocations or twins and each has differently shaped

surface, consequently named as lath martensites and plate martensites.

Lath martensite has waved surface whereas plate martensite a flat one as

schematically depicted in Figure 2-3 [14].

Page 15: Master Thesis

Chapter 2 – FUNDAMENTALS 10

Figure 2- 3. Schematics of two typical kinds of martensite

microstructures observed by transmission electron

microscopy. Plate martensite (left) has twins inside the plates

of flat boundaries, whereas lath martensite (right) contains

dislocations inside the lath of wavy boundaries [14].

The changes in volume and in configuration associated with martensitic

transformation cause strong lattice distortions, which naturally counteract

the transformation. At high carbon content values (>1.4%) and low

transformation temperatures, these distortions can be reduced only to a

small degree through gliding and recovery mechanisms, so that the

transformation from austenite to martensite does not always fully

proceed. Metallographically, martensite forms in plates as seen in

Figure 2-4 [11].

Page 16: Master Thesis

Chapter 2 – FUNDAMENTALS 11

Figure 2- 4. Different morphologies of martensite in Fe-C alloys; Left:

Plate martensite, and Right: Lath martensite [11].

At lower carbon content values (<<0.5%) and higher transformation

temperatures, the elastic lattice stresses can be removed more easily

through dislocation movements. Martensite then no longer consists of

single crystal plates but rather of lath-shaped crystals that are bunched

into blocks and oriented in various directions [11].

2.5. Bainitic Transformation and Microstructure

In contrast to the diffusional characteristics, the surface relief during

bainitic transformation is very similar to that of martensitic

transformation. Bainite also forms as isolated plate unlike lamellae of

ferrite and cementite in pearlite. The speed of the formation of plates,

however, is much lower than that of martensitic transformation.

Page 17: Master Thesis

Chapter 2 – FUNDAMENTALS 12

The morphological characteristics of bainite are divided into two kinds,

upper bainite and lower bainite according to the microstructural

characteristics of carbides. Upper bainite, on {111} habit planes of

austenite, is obtained at a temperature range of 350-550°C and the major

microstructural constituent is either needle or laths of ferrite. Carbides,

mainly Fe3C, grow at the interface of the ferrite plates. The

transformation mechanism of the upper bainite is similar to that of

Widmanstatten side plates. Lower bainite, which forms at the lower

temperature than upper bainite and on irrational habit planes of austenite,

has a similar microstructure to that of tempered martensite. Due to the

lower diffusion rate of carbon, carbides, either Fe3C or epsilon carbides

grow mostly inside the ferrite plates with a characteristic angle with

respect to the ferrite plate and with characteristic habit planes. Schematic

features of both types of bainite plates are shown in Figure 2-5 [14].

Figure 2- 5. Two carbide shapes in bainite: left: upper bainite, right:

lower bainite. Note that the carbides in the upper bainite form

at the plate boundaries and are coarser than those in the lower

bainite. Also, the carbides in the lower bainite are aligned

along a certain crystallographic direction within the plate

[14].

Page 18: Master Thesis

Chapter 2 – FUNDAMENTALS 13

Lower bainite structure that is usually obtained in the temperature range

of 250-350°C, shows higher tensile strength than that of upper bainite.

This is a consequence of the higher dislocation density in the lower

bainite plates. Also, the lower bainite, due to the fine carbide structure,

has a higher toughness than upper bainite. Therefore, the ductile

transition temperature for the lower bainite is lower than that of upper

bainite [14].

It is also possible to distinguish the bainitic structures according to the

various metallographic appearances as:

• fine acicular,

• coarse acicular,

• granular.

Figure 2-6 gives examples of bainitic microstructures based on this

classification.

Page 19: Master Thesis

Chapter 2 – FUNDAMENTALS 14

Figure 2- 6. Optical microscope photos of bainitic microstructures: a) fine

acicular bainite, b) coarse acicular bainite, and c) granular

bainite [11].

Page 20: Master Thesis

Chapter 2 – FUNDAMENTALS 15

2.6. Characterization Difficulties

The macroscopic mechanical properties of steel are highly dependent

upon the microstructure, morphology, and distribution of each phase

present [13].

As stated before, there are certain situations where the conventional

optical microscopy techniques and/or electron microscopy are not proper

or sufficiently accurate methods for the evaluation of steels'

microstructures [15].

The examples are the situations where an inhomogeneous distribution of

strain and cooling rate along the horizontal centerline of the experimental

specimen are varied, thereby influencing the microstructural distribution

of phases; also cases where a fine bainitic-martensitic phase mixture

makes it impossible to distinguish between the present phases in a precise

manner (see Figures 2-4 and 2-6).

However, the variations of hardness values when passing through

bainitic-martensitic phase boundaries (see section 2.2.2), and the position

of sharp dilatation peaks during the dilatometry experiments

(see section 2.3) make it possible to indirectly detect and quantify the

present phases, getting precise information from the whole experimental

body.

Page 21: Master Thesis

Chapter 2 – FUNDAMENTALS 16

2.7. Microstructural Aspects of Weld Zones

Austenite formation and decomposition in both the weld heat-affected

zone (HAZ) and the weld fusion zone (FZ) is a well-studied phenomenon

in the welding metallurgy of steels [16].

In the HAZ, the room temperature microstructure transforms to 100%

austenite on heating above the Ac3 temperature. This transformation may

or may not reach completion depending upon the peak temperature

(between Ac1 and Ac3) attained in the HAZ and the time over which the

material experiences this temperature. Interestingly, the temperature near

the fusion line may be high enough to form the high temperature

δ –ferrite phase.

The austenite and δ –ferrite phases that form in the FZ and HAZ during

heating will transform to several different low-temperature phases during

cooling. For example, the δ –ferrite in the HAZ may transform back to

austenite, which may then transform to different morphologies of

α –ferrite, namely, allotriomorphic ferrite, Widmanstatten ferrite, and

bainite. At rapid cooling rates, the austenite will transform to martensite.

In the fusion zone during cooling, the liquid solidifies as δ –ferrite, which

then transforms to austenite. With continued cooling, the austenite

transforms to the different α –ferrite morphologies observed in the HAZ

microstructure.

Due to steep temperature gradients and dynamic cooling conditions of

welds, the resulting material contains significant microstructural

gradients. These microstructural gradients affect the strengths, ductility,

toughness, fatigue and creep rupture properties of steel welds. The

knowledge of these gradients is well developed and is routinely

considered in the design of weldments. However, there is a need to

Page 22: Master Thesis

Chapter 2 – FUNDAMENTALS 17

develop predictive models to describe these gradients as a function of

steel composition and weld thermal cycles [17].

The microstructure obtained as the weld cools from the liquid phase to

ambient temperature is called the as-deposited or primary microstructure.

As stated before, it may consists of allotriomorphic ferrite α ,

Widmanstatten ferrite Wα , acicular ferrite aα , and the so-called

microphases, which might include small amounts of martensite, retained

austenite or degenerate pearlite (Figure 2-7).

Figure 2- 7. An illustration of the essential constituents of the primary

microstructure of a steel weld deposit. The diagram is

inaccurate in one respect, that inclusions cannot be expected

to be visible in all of the acicular ferrite plates on a planar

section of the microstructure. This is because the inclusion

size is much smaller than that of an acicular ferrite plates, so

that the chances of sectioning an inclusion and plate together

are very small indeed [18].

Page 23: Master Thesis

Chapter 2 – FUNDAMENTALS 18

Bainite is also found in some weld deposits, particularly of the type used

in the power generation industry. Allotriomorphic ferrite is sometimes

called polygonal ferrite or proeutectoid ferrite, but polygonal simply

means many sided (like all ferrite morphologies) and Widmanstatten

ferrite can also be proeutectoid. Widmanstatten ferrite (Figure 2-8) is

sometimes included under the general description 'ferrite with aligned

MAC', the abbreviation referring to martensite, austenite and carbide.

However, bainite plates can also form in a similar shape, although their

thermodynamic and kinetic characteristics are quite different.

Among others, acicular ferrite is a phase most commonly observed as

austenite transforms during the cooling of low-alloy steel weld deposits.

It is of considerable commercial importance because it provides a

relatively tough and strong microstructure. It forms in a temperature

range where reconstructive transformations become relatively sluggish

and give way to displacive reactions such as Widmanstatten ferrite,

bainite and martensite.

The term 'acicular' means shaped and pointed like a needle, but it is

generally recognized that acicular ferrite has in three-dimensions the

morphology of thin, lenticular plates (Figure 2-9).

Page 24: Master Thesis

Chapter 2 – FUNDAMENTALS 19

Figure 2- 8. (a) Schematic diagrams illustrating the development of

microstructures in weld deposits. The hexagons represent

cross-sections of columnar austenite grains whose boundaries

first become decorated with uniform, polycrystalline layers

of allotriomorphic ferrite, followed by the formation of

Widmanstatten ferrite. Depending on the relative

transformation rates Widmanstatten ferrite and acicular

ferrite, the former can grow entirely across the austenite

grains or become stifled by the intragranularly nucleated

plates of acicular ferrite. This diagram takes no account of

the influence of alloying additions on the austenite grain

structure. (b) Actual optical micrographs illustrating the

unhindered growth of Widmanstatten ferrite in a weld

deposit. (c) Optical micrograph showing how the growth of

Widmanstatten ferrite is stifled by the formation acicular

ferrite [18].

Page 25: Master Thesis

Chapter 2 – FUNDAMENTALS 20

Figure 2- 9. Replica transmission electron micrograph of acicular ferrite

plates in a steel weld deposit [18].

The experimental data to date indicate that acicular ferrite is essentially

identical to bainite. Its detailed morphology differs from that of

conventional bainite because the former nucleates intragranularly at

inclusions within large austenite grains whereas in wrought steels which

are relatively free of nonmetallic inclusions, bainite nucleates initially at

austenite-austenite grain surfaces and continues growth by the repeated

formation of subunits, to generate the classical sheaf morphology.

Acicular ferrite does not normally grow in sheaves because the

development of sheaves is stifled by hard impingement between plates

nucleated independently at adjacent sites. Indeed, conventional bainite or

acicular ferrite can be obtained under identical isothermal transformation

conditions in the same (inclusion rich) steel. In the former case, the

austenite grain size has to be small in order that nucleation from grain

surfaces dominates and subsequent growth then swaps the interiors of the

austenite grains. For a larger austenite grain size, intragranular nucleation

on inclusions dominates, so that acicular ferrite is obtained (Figure 2-10).

Page 26: Master Thesis

Chapter 2 – FUNDAMENTALS 21

Figure 2- 10. An illustration of the effect of austenite grain size in

determining whether the microstructure is predominantly

acicular ferrite or bainite. A small grain sized sample has a

relatively large number density of grain boundary nucleation

sites and hence bainite dominates, whereas a large number

density of intragranular nucleation sites leads to a

microstructure consisting mainly of acicular ferrite [18].

Hence, the reason why acicular ferrite is not usually obtained in wrought

steels is because they are relatively free of inclusions and because most

commercial heat treatments aim at a small austenite grain size. It is ironic

that bainite when it was first discovered was referred to as acicular ferrite,

and that the terms acicular ferrite and bainite were often used

interchangeably for many year after 1930 [18].

Page 27: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 22

CHAPTER

THREE

EXPERIMENTAL PROCEDURE

3.1. Materials Characterization

3.1.1. General Considerations

Five different steels in shapes of compression specimens and welded

sheets were employed. The compression specimens were made of

22MnB5 and MSW1200 steels which were received in form of plates,

while the welded sheets were composed of S355, EH36 and RQT701

steels. The S355 sheet had a thickness of 12mm, while EH36 and

RQT701 sheets had the thickness of 20mm.

3.1.2. Chemical Composition

Chemical compositions of the investigated steels are given in tables 3-1

and 3-2 in case of compression and welded specimens.

Table 3-1. Chemical composition of the compression specimens (wt-%).

C Si Mn Cr Ni Al Ti B

22MnB5 0.23 0.22 1.18 0.16 0.12 0.03 0.040 0.002

MSW1200 0.14 0.12 1.71 0.55 0.06 0.02 0.002 -

Page 28: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 23

Table 3-2. Chemical composition of the welded sheets (wt-%).

S355-12 EH36-20 RQT701-20 C 0.054 0.121 0.144 Si 0.018 0.389 0.436

Mn 1.085 1.290 1.360 P 0.011 0.012 0.012 S <0.001 0.002 <0.001

Cr 0.015 0.015 0.016 Mo <0.005 <0.005 <0.005 Ni 0.019 0.024 0.021 Al 0.021 0.037 0.036 Nb 0.023 0.032 0.031 Cu 0.010 0.016 0.014 Ti 0.009 0.001 0.026 V 0.003 0.004 0.068 B <0.0005 <0.0005 0.002

3.1.3. Microstructure

The 22MnB5 steel contains ferrite and pearlite phases (together with

carbide) in as-received condition. Figure 3-1 shows the microstructure of

the 22MnB5 steel in the rolling direction.

The image analysis data shows that the microstructure contains around

78% ferrite besides 22%, combination of pearlite and carbide. Ferrite

grain size was measured to be comparable with 11 ASTM grain size

standard.

Page 29: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 24

Figure 3-1. Microstructure of the as-received 22MnB5 sheets in the

rolling direction a) 500X, and b) 1000X.

The microstructures of the base metal of the investigated welded sheets

are given in Figure 3-2. As seen in this illustration, in case of S355 and

EH36 specimens, the microstructures are predominantly ferritic-pearlitic,

while RQT701 has a ferritic-bainitic microstructure.

Figure 3-2. Microstructure of the base metals: a) S355, b) EH36, and

c) RQT701.

Page 30: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 25

3.2. Isothermal and Non-isothermal Compression Tests

The isothermal and non-isothermal compressive deformations were

exerted on the 22MnB5 and MSW1200 compression specimens.

The whole deformations and dilatometry experiments were carried out

using a Baehr 805 dilatometer (Figure 3-3) with a deformation unit

designed to perform a simple uniaxial deformation. In all cases, the

geometry of the sample used was cylindrical with a length of 10mm and a

5mm diameter (Figure 3-4). In all of the tests, the temperature was

measured using a surface mounted Pt/Pt-Rh10% thermocouple located at

the mid-length relative to the specimen. Non-isothermal deformations

were established through several simultaneous forming and quenching

tests at the temperatures between 600°C–850°C and by the strain rates of

0.05s-1–1.0s-1. Regarding the duration of the experiments, the samples

were deformed up to the strains of 0.1-0.5 in a single step. All samples

were austenitized at 900°C for five minutes and quenched to the

compression temperatures at a cooling rate of 50K/s.

In case of isothermal experiments, deformations were carried out at the

temperature range of 600°C – 850°C by the strain rates of

0.1 s-1 – 10.0 s-1. Due to the technical limitations in the Baehr deformation

dilatometer, higher strain rates could not be applied in the simultaneous

forming and quenching tests.

Page 31: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 26

Figure 3-3. Baehr DIL 805 deformation dilatometer, and the sample

set up.

Figure 3-4. Schematic illustration of the cylindrical specimen used

during the dilatation experiments.

3.3. Execution of Welding

The aforementioned steel sheets (see 3.1), were joint together using arc

welding and laser welding methods. Different parts of a welded structure

Page 32: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 27

are depicted in Figure 3-5. The caps and roots of fusion zones (FZ) and

heat-affected-zones (HAZ) are plotted as well.

In the current work, the joining sheets were put together before welding

using two methods: either with a simple touching of the two 'V' notched

neighboring sheets, or making a gap between them for further

improvement of the final attachment.

Figure 3-5. Schematic illustration of the welded structure.

3.4. Hardness Mapping and Metallography

In the present study, an innovative surface hardness mapping technique

was developed by the Department of Ferrous Metallurgy (IEHK), RWTH

Aachen University. In this technique, the surface of the sample is scanned

using an indenter which exerts a 0.8g force on the surface of the sample

Page 33: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 28

and records hardness of the points in Vickers or Rockwell hardness

scales. Figure 3-6 shows the setup of the mentioned surface hardness

mapping device and relevant specimen insertion.

Figure 3-6. Surface hardness mapping relevant instruments at IEHK,

RWTH Aachen University.

Surface hardness measurements were performed on the previously

deformed and welded samples. To do this, in case of the deformed

samples, they were cut lengthwise. The samples were mounted after

cutting. The Vickers hardness of the whole deformed surfaces was

measured in 0.3mm steps using an exerted force of 0.8g. In case of

welded sheets 0.5mm steps were used. Afterwards, the surface hardness

maps of the samples were plotted.

Page 34: Master Thesis

Chapter 3 – EXPERIMENTAL PROCEDURE 29

Surface hardness maps of the deformed, cut and mounted samples were

started one millimeter out of the sample, i.e., from the polymeric

mounting material. The hardness of mounting material was recorded as

999 and was ignored. Due to the boundary conditions between the

mounting material and the sample, hardness values of the sample's edge

must not be considered. Accordingly, the quantitative measurements

using surface hardness mapping data were performed by means of the

reliable hardness data which were taken from inside the sample. To

evaluate the consistency of the hardness mapping results regarding the

determination of the present phases, optical microscopy investigation

were also executed on the whole deformed specimens.

The welded sheets were also scanned using hardness mapping method

throughout the fusion zones and heat-affected-zones. Four sets of optical

microscopy images were taken from caps and roots of FZ and HAZ in

every single specimen. In addition, the macrostructures of the welds were

revealed using lower magnification microscopy.

Page 35: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 30

CHAPTER

FOUR

RESULTS AND DISCUSSION

4.1. Revealing Heterogeneous Microstructures

As stated in the previous sections, the compressive deformation results in

an inhomogeneous distribution of microstructural components. To clarify

this issue, and by consideration of various hardness levels due to different

phases in steels, a complete surface hardness map a compressed 22MnB5

specimen deformed up to the strain of 0.1 is plotted in Figure 4-1.

Figure 4-1. The deformed 22MnB5 sample with maxε =0.1, ε& =0.05s-1,

TiD =750°C; the complete surface hardness map with the

scale of 25HV0.8 representing heterogeneity.

Page 36: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 31

In the present study, the hardness values of more than 400HV0.8g were

assumed as martensite. The hardness between 200 and 400HV0.8g were

assumed as bainite and below 200HV0.8g as ferrite. These assumptions

were calculated using the CCT diagram, microscopic investigations and

some reports [1-2,19]. Pearlite was not considered because no pearlite

was seen in the final microstructure using LOM and SEM images.

For further observation of the microstructural state, the whole surface of

the investigated specimen was scanned using optical microscopy. The

scanning sequence followed a pattern as depicted in Figure 4-2. Red spots

demonstrate the approximate positions where an optical image was taken.

Figure 4-2. The pattern corresponding to scanning the deformed surface

using optical microscopy.

Figure 4-3 shows the relevant microstructures of points 'X' and 'Y'

illustrated in Figure 4-1 using this scanning method.

Page 37: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 32

Figure 4-3. Optical microscopy images representing martensite and/or

bainite in zones 'X' and 'Y' as mentioned in Figure 4-1.

As seen in Figure 4-3, the microstructure in zone 'X' is fully martensitic

while bainite is the dominant phase in zone 'Y'. This proves that the

pre-assumed hardness intervals regarding different phases are reliable. In

addition, it is clearly seen that one may make a colossal mistake when

reporting the phase fractions within the microstructure of the

heterogeneously deformed specimens using single selected optical

microscopy images.

Page 38: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 33

4.2. Reliability of Dilatometry and Hardness Mapping Methods

4.2.1. Hardness Mapping

The complete surface hardness maps of two samples which were

deformed by the strain rate of 0.05s-1=ε& are plotted in Figure 4-4. These

samples were austenitized at 900°C for 5min. The initial deformation

temperature, TiD, was 850°C and the cooling rate,T& , was 50K/s. The first

map, Figure 4-4a, exhibits the sample which was deformed up to 10

percent strain (ε ), while the second, Figure 4-4b, deformed to the

maximum strain of 15 percent. As seen in Figure 4-4a, hardness values

are greater than 400HV0.8 indicating a fully martensitic microstructure. In

the second map, Figure 4-4b, concerning with the hardness values, bainite

and martensite phases were distributed almost equally, i.e., 50%

martensite and 50% bainite. Besides, the heterogeneity of hardness values

in the range of each phase can be well studied. For instance, the hardness

values within the fully martensitic microstructure vary between 400 and

550HV0.8. It might be due to the heterogeneous distribution of carbon

content during austenization. This fact can be seen even in the bainitic

zone.

Page 39: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 34

Figure 4-4. Surface hardness maps of two 22MnB5 samples deformed

non-isothermally, CTiD °= 850 , austenization at 900°C for

5min, 105.0 −= sε& and sKT /50=& ; a) 1.0=ε and b)

15.0=ε (scale = 25HV0.8).

Another conclusion is that, higher hardness values, i.e. martensite phase,

can be detected in the inner parts of the compressed specimen. On the

contrary, the phase with lower hardness levels, i.e. bainite, is formed in

the outer parts. It is seen that, this technique provides one of the best tools

for the physical understanding of the phases' heterogeneity using hardness

criterion.

Page 40: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 35

4.2.2. Dilatometry Data

It is known that the dilatation observed during martensitic transformation

becomes anisotropic if there is variant selection due to stress or strain

[20,21]. This makes it difficult to asses the fraction of transformation

using one-dimensional dilatometry measurements. However, due to sharp

dilatation curves regarding martensitic transformation in all of the

experiments, martensite phase percent within the deformed samples were

also calculated using the dilatation data. The calculation was based on a

simple mathematical method. In this method, the amount of dilatation due

to martensitic transformation in one of the samples in the CCT diagram

was assumed as a reference. In this case, the sample was cooled down at a

cooling rate of approximately 50K/s. A fully martensitic microstructure

caused a dilatation of around 0.26 percent. All other calculations –

regarding the martensite phase percent resulting from the dilatation data

in the rest of the specimens – were completed using this reference value.

By using this calculation technique, the amount of martensite in the 10

percent deformed sample is about 92 percent and in the sample which

was deformed 15 percent is about 46 percent, Figure 4-5.

Page 41: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 36

0 200 400 600 800 1000-1.8

-1.5

-1.2

-0.9

-0.6

-0.3

0.0

0.3

Martensite (%)

εmax Ms(°C) Mf(°C) Dil% (Dilatation Data)____________________________________________________________ 0.0 410 230 0.26 100% 0.1 395 292 0.24 92% 0.15 379 258 0.12 46%

0.15

0.1

0.0

TiD = 850°C

dε/dt = 0.05 s-1

Dila

tatio

n (%

)

Temperature (°C)

Figure 4-5. Dilatation curves of the previously mentioned samples in

Figure 4-4, as well as a fully martensitic sample produced

after cooling and without applying any force.

These values differ from the previously estimated values using surface

hardness maps up to eight and four percent, respectively. These

discrepancies might be due to the anisotropy of crystal and/or grains. In

Figure 4-6, the results of measurements using these two techniques

corresponding to 26 different specimens are compared.

These samples were deformed non-isothermally by the strain rates

ranging from 0.05s-1 to 1.0s-1, initial deformation temperatures between

600°C to 850°C, and the final strains of 0.05 to 0.5.

Page 42: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 37

0 20 40 60 80 100

0

20

40

60

80

100

% M

hard

ness

map

% M dilatation

%Mh = %Md + 3

Figure 4-6. Comparison of the estimated martensite content in the

non-isothermally deformed samples using surface hardness

mapping and dilatation data.

It is seen that the mentioned techniques are consistently reliable methods

for the quantitative and qualitative assessment of the steels'

microstructure.

Page 43: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 38

4.3. Prediction of Martensite Fraction Using Hardness Criterion

4.3.1. Developing the Formula

A dimensionless parameter so-called 'relative hardness' is defined as

follows:

100.)(

m

m

HHH −

=κ (1)

where κ is the relative hardness, Hm is the average hardness of a fully

martensitic microstructure of the same steel in HV10 scale (measured by

macrohardness indentation on the known specimen containing almost

100% martensite), and H is the bulk hardness of an unknown specimen in

HV10 scale. The Hm value in the case of 22MnB5 steel is amounted to be

473. In this stage, the κ factors of different 22MnB5 steel specimens are

calculated. There are reports in which the maximum achievable hardness

of steel can be measured using its chemical composition. These methods

can be employed as an alternative for the mentioned calculation of the

hardness of a fully martensitic specimen. However, in the current work, a

direct measurement is used to achieve a higher accuracy in case of the

developed formula.

The mentioned values are plotted against Mp which is the precise

martensite fraction coming from surface hardness mapping data

(Figure 4-7).

Page 44: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 39

0 20 40 60 80 10035

30

25

20

15

10

5

0

50%M|22MnB5 ~ κ15

κ ~ -0.3Mp + 30κ

% Mp

Figure 4-7. κ factor against precise martensite fraction of 22MnB5.

As is seen in Figure 4-7, the equation of the resultant solid trend line is:

κ = 0.3Mp + 30 (2)

The point which should be noted here is that, because of the definition of

κ equation at the beginning, the effect of the hardness of a fully

martensitic microstructure has been almost removed. Therefore, it is

simply possible to compare two specimens (one with higher and one with

lower carbon contents) with exactly the same H value (hardness of the

unknown bulk) and different Hm values (one higher for the higher carbon

content, and one lower, for the lower carbon content).

If one assumes that Hm1>Hm2 and H1=H2, then 1κ > 2κ . This can be proved

by putting different numbers within the κ equation.

If 1κ and 2κ values are put inside the previous diagram, it is found out

that 1κ gives a lower martensite fraction. This is true, because the higher

carbon content will increase the hardness of bainite and martensite phases

simultaneously. As the bulk hardness of both specimens is the same, and

it is known that the martensite hardness increases by increasing the

Page 45: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 40

carbon content, then the only possibility to keep the bulk hardness

constant, is to reduce the fraction of harder martensite and increase the

fraction of less harder bainite.

In another attempt, κ was plotted against different H (unknown bulk

hardness) values. The plot is given in Figure 4-8. The trend line equation

is:

κ = -0.2 H + 94 (3)

250 300 350 400 450 50050

40

30

20

10

0

κ

κ ~ -0.2H + 94

H (HV0.8

)

Figure 4-8. κ factor against bulk hardness values of 22MnB5 specimens.

Putting equations (2) and (3) together results in:

3205.1 += κMH (4)

where Mκ is the calculated martensite fraction using mathematical

method.

To show the accuracy of this method in case of 22MnB5 steel, Mp and

Mκ values were plotted against each other (Figure 4-9).

Page 46: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 41

0 10 20 30 40 50 60 70 80 90 1000

10

20

30

40

50

60

70

80

90

100

MRH ~ Mp + 6

% M

RH

% Mp

Figure 4-9. Accuracy test of MRH against MP in case of 22MnB5 steel.

As seen in Figure 4-9, there is a +6% tolerance factor regarding the

mathematical estimation in comparison with the precise martensite

fraction.

To expand the application of the presented formula, upper and lower

limits of 320 and 470HV10 were set as the states of no martensite and

fully martensitic microstructures respectively. These numbers were

chosen based on consideration of a large number of differently deformed

specimens of the investigated steels.

Finally, the whole equation package for the calculation of martensite

fraction using hardness of an unknown bulk is as follows:

H = 1.5 Mκ + 320

H ≤ 320 no martensite (5)

H ≥ 470 fully martensitic

Page 47: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 42

4.3.2. Checking the Consistency of Results

To check the applicability of this formula in case of other steels, another

steel called MSW1200 was employed. The point which must be

mentioned here is that, the carbon content of MSW1200 is 0.14%wt

(almost half of 0.23%wt carbon content in case of 22MnB5 steel) and

there is no boron available in its composition (in comparison with

0.002%wt of boron within the composition of 22MnB5 specimens).

Figures 4-10 and 4-11 gives the hardness maps of differently deformed

MSW1200 specimens. In each case, the fraction of microstructural phases

is shown. In addition the bulk hardness values of specimens are inserted

in the hardness mapping diagrams.

Using equation 5, Mκ values are calculated. As is easily seen and based

on the suggested equation, presented specimens in Figure 4-10 are

martensite-free. This is confirmed through both hardness mapping and

mathematical estimation methods. In case of the specimens in

Figure 4-11, mathematical estimation gives 71% and 60% martensite

contents against the actual martensite contents of 76% and 71%,

respectively. By this, the equation seems to be reliable.

It is concluded that the developed formula is consistently applied for

MSW1200 steel with different chemical composition to that of 22MnB5.

Page 48: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 43

Figure 4-10. Hardness maps of ferritic-bainitic MSW1200 specimens;

a) 64% bainite and 36% ferrite , and b) 54% bainite, 46%

ferrite, and almost no martensite.

Page 49: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 44

Figure 4-11. Hardness maps of bainitic-martensitic MSW1200 specimens;

a) 24% bainite and 76% martensite , and b) 29% bainite,

71% martensite.

Page 50: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 45

4.4. Microstructural Study of Weld Zones

As mentioned in chapter 3, three different steel sheets were chosen for the

study of weld zones microstructure.

The S355 sheets (the lowest carbon content among others) were chosen

with a thickness of 12mm, while EH36 and RQT701 sheets had the

thickness of 20mm. In addition two sets of welding were exerted on

RQT701 sheets one under the condition of making a gap between the 'V'

notched joining parts (called as 'F' specimen) and one without making a

gap (called 'I' specimen).

In each case, the whole surface (including 'HAZ' and 'FZ') was scanned

using hardness mapping method. Moreover, optical microscopy images

were taken from four different locations: a) cap of FZ, b) root of FZ,

c) cap of HAZ, and d) root of HAZ.

Figure 4-12 shows the macrostructure and surface hardness map of the

entire weld zone for S355-12I welded sheets.

Page 51: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 46

Figure 4-12. Macrostructure (a) and hardness map (b) of S355-12I welded

sheet; the numbers indicate the hardness values in HV0.8.

Page 52: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 47

Figure 4-13. Microstructure of S355-12I welded sheet; a) cap of FZ, and

b) root of FZ.

Based on the pre-assumed hardness intervals as mentioned in section

4.2.1 and by the evaluation of hardness map of Figure 4-12, it is seen that

the FZ cap mostly consists of bainite and ferrite while in the FZ root,

ferrite is the dominant phase.

Figure 4-13 confirms the hardness mapping data using optical microscopy

images. Due to the presence of inclusions on the FZ cap, the conditions

lead to the formation of acicular ferrite beside bainite (see chapter 2 for

details), while the root mostly shows a ferritic microstructure.

It is seen that both methods are consistent when working on a

bainitic-ferritic microstructure. Hardness mapping can easily distinguish

between ferritic and bainitic zones in an accurate manner.

Although not mentioned here, the HAZ cap consists of ferrite (based on

hardness data), while its root is mostly occupied by the mixtures of

bainite and ferrite.

Figure 4-14 illustrates EH36-20F specimen based on the same method.

Page 53: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 48

Figure 4-14. Macrostructure (a) and hardness map (b) of EH36-20F

welded sheet; the numbers indicate the hardness values.

Page 54: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 49

Figure 4-15. Microstructure of EH36-20F welded sheet; a) cap of FZ, and

b) root of FZ.

Figure 4-16. Microstructure of EH36-20F welded sheet; a) cap of HAZ,

and b) root of HAZ.

Figure 4-15 and 4-16 gives the microstructures of FZ and HAZ of the

same steel in cap and root zones. As depicted by the mentioned hardness

map, the entire FZ consists of bainite phase with distribution of ferritic

islands in it. This is confirmed by consideration of Figure 4-16. On the

other hands, the whole HAZ (both cap and root) contains ferrite and no

bainite. Looking at the microstructures gives the same information,

besides the fact that the microstructure of HAZ in cap also contains large

fraction of acicular ferrite. It is seen that, although hardness mapping

method can easily distinguish between different phases, it is almost

impossible to find out which morphology is dominant in each case using

Page 55: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 50

single hardness criterion. This is a major deficiency when working on the

weld zones which having an idea of the presented morphologies is vital

due to their impact on mechanical properties like toughness. Therefore,

the parallel utilization of both hardness mapping and optical microscopy

methods is suggested in such cases.

Figure 4-17. Macrostructure (a) and hardness map (b) of RQT701-20F,

welded sheets; the numbers indicate the hardness values in

HV0.8.

Page 56: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 51

Figure 4-18. Macrostructure (a) and hardness map (b) of RQT701-20I,

welded sheets; the numbers indicate the hardness values in

HV0.8.

Figures 4-17 and 4-18 show RQT701-20F and RQT701-20I specimens

respectively. It is seen that bainitic-martensitic mixed microstructures are

expected in the cap and root of FZ in RQT701-20F sheet. The root has a

higher fraction of martensite than cap. The HAZ instead is fully bainitic.

Page 57: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 52

In addition, the RQT701-20I specimen is expected to show a fully

bainitic microstructure in both FZ and HAZ parts.

Figure 4-19. Microstructure of RQT701-20F welded sheet; a) cap of FZ,

and b) root of FZ.

Figure 4-20. Microstructure of RQT701-20F welded sheet; a) cap of HAZ,

and b) root of HAZ.

Looking at the microstructures in Figures 4-19 to 4-22 gives

contradictory results. Although the cap and root of FZ in RQT701-20F

specimens are martensitic-bainitic as expected before, the HAZ seems to

be fully martensitic in cap and dominantly martensitic – although

partially bainitic - in root. It is in contrast with the hardness mapping

results which predict a fully bainitic structure in HAZ. This can be

justified by consideration of left hand side of Figure 4-20b. By

comparison of this figure and what seen in the relevant hardness map in

Page 58: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 53

Figure 4-17b, it can be assumed that the optical microscopy image was

taken in the boundary of FZ and HAZ; therefore, left hand side of this

image is strictly bainitic while the rest is martensitic. Based on the same

justification, a wrong capturing zone is the reason for a fully martensitic

microstructure as reported by metallography.

Figure 4-21. Microstructure of RQT701-20I welded sheet; a) cap of FZ,

and b) root of FZ.

Figure 4-22. Microstructure of RQT701-20I welded sheet; a) cap of HAZ,

and b) root of HAZ.

As stated before, hardness mapping results show a fully bainitic

microstructure for RQT701-20I specimen in both FZ and HAZ sites.

However, microstructural investigations (Figures 4-21 and 4-22) give a

ferritic-bainitic microstructure. The observed ferritic microstructure is

completely acicular particularly in case of FZ parts of the specimen.

Page 59: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 54

Considering the hardness map of RQT701-20I specimen (Figure 4-18b),

it is found that in spite of no prediction due to the presence of ferrites,

lower hardness values in the center of cap and root of FZ and the increase

of hardness till the FZ/HAZ boundary might be an indirect indication of

the presence of a softer phase in that areas.

In some reports [8,22], the equations were suggested to relate the

hardness of each phase to the chemical composition of steel. By this and

consideration of thermal profiles from different temperature regimes in

the weld zone, also having the hardness map of the specimen, it is

possible to have an estimation of the present phases without optical

microscopy [8]. However, as described in the current work, due to the

fact that the knowledge of phases' morphology in the weld zones is as

essential as the distribution of microstructural components based on their

influence on mechanical properties, utilization of both hardness mapping

and optical microscopy methods is suggested.

The linear hardness profiles of RQT701-20F and RQT701-20I specimens

are demonstrated in Figure 4-23 and Figure 4-24, respectively. In each

case, the profiles are given in three different zones: a) cap, b) middle, and

c) root. The black and red colored lines in the diagrams are related to the

examination of two closely selected paths within each specific zone.

Looking through the linear hardness profiles, it is seen that there is a

certain drop in the hardness values when passing across the FZ/HAZ

boundary. The reason is however different when talking about

RQT701-20F and RQT701-20I specimens.

In the case of RQT701-20F samples (Figure 4-23), the FZ consists of

bainite and martensite phases as mentioned before by means of optical

microscopy. On the contrary, the HAZ is completely bainitic. Passing

Page 60: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 55

through such a boundary reduces the hardness substantially from

400HV0.8 to 280HV0.8 (Figure 4-23).

Figure 4-23. Linear hardness profiles of RQT701-20F specimens: a) cap;

b) middle; and c) root.

Page 61: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 56

Figure 4-24. Linear hardness profiles of RQT701-20I specimens: a) cap;

b) middle; and c) root.

Page 62: Master Thesis

Chapter 4 – RESULTS AND DISCUSSION 57

In addition, the hardness values throughout the FZ area is almost stable at

around 400HV0.8. Such an average is expected to indicate an almost

50%-50% bainitic-martensitic microstructure using the mathematical

estimation method mentioned in 4.3; although an exact estimation is

possible only after improvement of the formula based on the chemical

composition of the investigated steel welds.

Different trends are depicted in Figure 4-24 in case of RQT701-20I

specimens. Despite of a sharp drop of hardness values in the borderline of

FZ and HAZ (when coming from FZ into HAZ), it is seen that the

mentioned drop in hardness values is available on both sides of the

boundary. The microstructural images of these zones show a

bainitic-ferritic structure in both FZ and HAZ. The average hardness

values of 250-350HV0.8 on both sides of the FZ/HAZ boundary can also

confirm the presence of a ferrite-bainite mixture. Therefore, the presence

of the hardness peak on the borderline of FZ and HAZ – despite of no

change in the available microstructural phases - can be justified due to the

developed equiaxed structure in the vicinity of FZ. This fact plus the later

development of rather coarser structures when going farther from

FZ/HAZ boundary, decreases the hardness dramatically.

Hence, the finer structure is the major reason for the increase of hardness

in the FZ/HAZ contact zone. Furthermore, it is possible that the change in

grain structure has been also affected the mentioned drop in hardness

when entering HAZ in case of RQT701-20F. However, as said before, the

stable high hardness in the FZ in comparison with the hardness of

FZ/HAZ borderline shows that the present martensite phase has a

dominant effect on the grain structure.

Page 63: Master Thesis

Chapter 1 – INTRODUCTION 58

CHAPTER

FIVE

CONCLUSIONS

- Non-isothermal compressive deformation of cylindrical steel

samples is heterogeneous, i.e., the effective strain and cooling rate

along the horizontal centerline of the samples vary. Consequently,

distribution of the produced phases is not homogeneous. This

means that light optical or scanning electron microscopy

investigations on only a few selected points of the heterogeneously

deformed surfaces are not the proper techniques to identify the

phase fractions within the experimental specimen. Hence, surface

hardness mapping technique as well as dilatation data were used to

identify the present phases. It is concluded that these techniques are

accurate, applicable, and comparable.

- Hardness mapping technique is a powerful method to distinguish

between the phases with close microstructural appearances under

the optical microscopy. It means that the usual problems regarding

discrimination between martensite and bainite can be easily

removed. In addition, further works on the field can enhance the

possibility for the determination of different morphologies of a

single phase, e.g. different ferritic and/or bainitic structures, based

on specific hardness intervals.

- Despite of various factors affecting the resultant dilatation values,

the suggested comparative calculation method in the current work,

seems to be reliable due to its consistency with other

characterization methods.

Page 64: Master Thesis

Chapter 1 – INTRODUCTION 59

- The presented mathematical formula in the current study has been

developed based on simple physical assumptions. The results are

applicable on two different steels. However, simple improvements

might be required – i.e. due to the hardness of a fully martensitic

microstructure – when handling the steels with compositions far

different from the investigated steels here.

- Because of the current deficiencies for the determination of

different morphologies of ferrite and bainite using a single

hardness criterion, and the importance of those information for the

mechanical properties of the weld zone, optical microscopy is still

required beside the hardness mapping data.

Page 65: Master Thesis

REFRENCES

i

REFRENCES [1] K. Hasegawa, K. Kawamura, T. Urabe and Y. Hosoya. ISIJ Int. 2004; 44,

No.3: 603-609.

[2] R.W. K. Honeycombe. Steels-microstructure and properties. Edvard Arnold;

1981.

[3] H.K. D. H. Bhadeshia. Bainite in steels. University of Cambridge: Institute

of Materials; 2001.

[4] G. Schiessl, T. Possehn, T. Heller, S. Sikora. IDDRG 2004: 158-166.

[5] W. Bleck, V. Uthaisungsuk, M. Naderi, U. Prahl. Proc. Moderne

thermomechanische Prozessstrategien in der Stahlumformung 2007: 127-

141.

[6] E. A. Metzbower, D. W. Moon. Sci. Technol. Weld. Joi. 2007; 12 (2): 189-

196.

[7] D. W. Moon, S. G. Lambrakos, R. J. Wong, E. A. Metzbower, Sci. Technol.

Weld. Joi. 2003; 8 (5): 334-339.

[8] D. W. Moon, S. G. Lambrakos, R. J. Wong, E. A. Metzbower, Sci. Technol.

Weld. Joi. 2003; 8 (2): 95-101.

[9] E. J. Mittemeijer, A. Van gent, and P. J. van der schaaf. Metall. Trans. A

1986; 17A: 1441-1445.

[10] S. Frank. Mobile hardness testing. Application Guide for Hardness Testers.

General Electric Inspection Technologies 2005.

[11] W. Bleck. Material Science of Steel. Textbook for RWTH Students.

RWTH Aachen University: Department of Ferrous Metallurgy; 2007.

[12] M. Meier. The Hardenability of Steel. UC Davis: Department of Chemical

Engineering and Materials Science; 2004.

[13] http://www.azonano.com.

[14] S. K. Hwang. Martensitic Steels and Bainitic Steels. Inha Univ: 2001.

[15] M. Naderi, A. Saeed-Akbari, W. Bleck. Mater. Lett. In Press 2007.

Page 66: Master Thesis

REFRENCES

ii

[16] K. E. Easterling. Introduction to the Physical Metallurgy of Welding.

Butterworths, London; 1983.

[17] S. S. Babu, J. M. Vitek, S. A. David, T. Palmer, J. W. Elmer. The

Minerals, Metals and Materials Society (TMS); 2003.

[18] H.K. D. H. Bhadeshia, L. E. Svensson. Modelling the Evolution of

Microstructure in Steel Weld Metal. Institute of Materials; 1993, 109-182.

[19] D. N. Hanlon, J. Sietsma, S. Zwaag. ISIJ International 2001; 41 (9):

1028-1036.

[20] H.K. D. H. Bhadeshia, S. A. David, J. M. Vitek and R. W. Reed. Mater.

Sci. Technol. 1991; 7: 686-98.

[21] A. Matsuzaki, H.K. D. H. Bhadeshia and H. Harada. Acta Mater. 1994; 42:

1081-90.

[22] Ito Y., Besseyo K. Weldability formula of high strength steels related to

heat affected zone cracking. Technical Report Doc IX-567-68, IIW, 1968.