EFFECT OF MICROSTRUCTURE ON THE COLD

HEADABILITY OF A MEDIUM CARBON STEEL

by

Xiaoyu Ma

Department of Mining, Metals and Materials Engineering

McGill University
Montreal, Canada

A Thesis Submitted to the Faculty of Graduate Studies and Research

in Partial Fulfillment of the Requirements for the Degree of
Master of Engineering

© Xiaoyu Ma 2003
 ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my project supervisor Prof. John J.

Jonas for his continual guidance, valuable advice and supervision throughout the
entire project. His knowledge and assistance were of great value. I additionally
thank Prof. James Nemes for his support and advice. The help of Dr. Alan
Humphreys with this thesis was also greatly valuable.

I wish to thank IVACO Rolling Mills for their support of this work, particularly Dr.
Michel Hone and Dr. Nick Nickoletopoulos. Thanks are also due to Dr. Baohong
Cao for his organization of the tested materials.

I am grateful to Lorraine Mello for her administrative help and Edwin Fernandez
for his laboratory expertise.

On a personal level, I thank my parents, brother and sisters for their

understanding and unconditional love. Finally I thank my wife and daughter,
whose love, patience and support are very much appreciated.
 ABSTRACT

Cold headability is the ability of a cylindrical metallic specimen to be shaped at

high strain rate into the head of a bolt, screw or other cold-formed part without
cracking. This property is material dependent and can be influenced by many
factors such as chemical composition, surface condition, and microstructure. This
project focuses on the effect of microstructure upon the cold headability of a
medium carbon steel (1036M).

Six different microstructures were produced by various heat treatment conditions.

Drop Weight Tower (DWT) tests, developed at McGill in 2000 by Dr. N.
Nickoletopoulos, were used on samples of these materials. Visual inspection,
metallographic and SEM analysis were performed to identify cracks on the
surface of tested samples, and trace their cause. The axial and circumferential
strains of tested samples were measured and the strains at which cracks first
occurred were used to assess the cold headability.

This research further indicates that DWT testing is a valid method for evaluating
the cold headability of metallic materials and that the cold headability is
particularly sensitive to the microstructure of a material.
 RESUME

La forgeabilite a froid est I'habilite d'un specimen metallique cylindrique d'etre

forme a haut taux de deformation dans la tete d'un boulon, vis ou autre pieces
produite par formage a froid sans fissuration. Cette propriete est dependante du
materiau et peut etre influenced par plusieurs facteurs comme la composition
chimique, I'etat de surface et la microstructure. L'emphase de ce projet est
I'etude de I'effet de la microstructure sur le refoulage a froid d'un acier a teneur
moyenne en carbone (1036M).

Six microstructures differentes ont ete produites par divers traitements

thermiques. L'essai de frappe developpe a I'Universite McGill en 2000 par le Dr.
N. Nickoletopoulos, a ete utilise sur les divers echantillons. Des inspections
visuelles, metallographiques et analyses au microscope a balayage ont ete
utilises pour identifier les fissures a la surface des echantillons testes et
d'identifier leurs causes. Les deformations axiales et circonferentielles des
echantillons testes ont ete mesurees et les deformations ou les fissures debutent
ont ete utilisees pour decrire le refoulage a froid.

Cette recherche indique que l'essai de frappe est une methode valide pour
evaluer le refoulage a froid des materiaux metalliques, et que le refoulage a froid
est particulierement sensible a la microstructure du materiau.
 TABLE OF CONTENTS

CHAPTER 1 INTRODUCTION 1

CHAPTER 2 LITERATURE REVIEW 4

2.1 The Cold Heading Process 4

2.2 CHQ Steel and Cold Headability 6

2.2.1 CHQ Steel 6

2.2.2 Cold Headability 7

2.3 Typical Fastener Defects 10

2.4 Tests of Cold Headability 11

2.4.1 Drop Weight Impact Test 12

2.4.2 Split Hopkinson Pressure Bar Test 13

2.4.3 Drop Weight Test 14

2.5 Ductility Criteria 20

2.6 Ductile Fracture 25

2.6.1 Microvoid Nucleation 26

2.6.2 Growth of Voids 27

2.6.3 Void Coalescence 28

2.7 Improvement of Microstructure for Cold Heading 28

2.7.1 Intercritical Process 29

2.7.2 Subcritical Process 29

2.8 Development of Non-heat-treated CHQ Steels 30

CHAPTER 3 EXPERIMENTS 32

3.1 Materials 32

3.1.1 Chemical Composition 32

3.1.2 Microstructures 33

3.2 Heat Treatment 35

3.3 Drop Weight Test (DWT) 36

3.3.1 Drop Weight Tower 36

3.3.2 Specimen Preparation 37

3.3.3 Drop Weight Test 37

3.4 Fracture Evaluation 38

CHAPTER 4 RESULTS 40

4.1 Heat Treatment 40

4.2 Drop Weight Test Results 43

CHAPTERS DISCUSSION 47

5.1 Drop Weight Test Results 47

5.2 Effect of Microstructure 52

5.3 Crack Analysis 61

CHAPTER 6 CONCLUSIONS 75

REFERENCES 76
 LIST OF FIGURES

Figure 2.1 Typical sequence of operations performed in cold heading 5

Figure 2.2 Brief flowchart of the cold heading process 6

Figure 2.3 Strain measurement on compressed specimens 8

Figure 2.4 Longitudinal crack on the head of a bolt 10

Figure 2.5 Oblique crack on the head of a bolt 10

Figure 2.6 Drop weight impact test device 13

Figure 2.7 Schematic diagram of the Split Hopkinson Pressure Bar Test 14

Figure 2.8 Structure of DWT machine 15

Figure 2.9 Exploded view of the die-set assembly for the DWT 16

Figure 2.10 Load-Time curve of 1038 steel in the as-rolled state 18

Figure 2.11 Load-Time curve of 1038 steel in the as-spheroidized state 18

Figure 2.12 Load-displacement curve of as-hot rolled 1038 steel 19

Figure 2.13 Forming limit diagram for upsetting test 21

Figure 2.14 Surface strain states in various deformation processes 22

Figure 2.15 Distribution of fracture points on the fracture line at different

deformations of 1045 steel 23

Figure 2.16 Fracture locus consisting of three segments for AIS11045 steel 24

Figure 2.17 Fracture lines for steel containing different carbon and alloy contents

24
Figure 3.1 (a) Microstructure of 1036M steel, as-hot rolled state 34

Figure 3.1 (b) Microstructure of 1036M steel, heat treated state 34

Figure 3.2 Die set assembly of DWT 36

Figure 3.3 Measurement of surface strain 38

Figure 3.4 illustration of three crack classes in DWT specimens 39

Figure 4.1 Microstructure of HT-1 41

Figure 4.2 Microstructure of HT-II 41

Figure 4.3 Microstructure of HT-III 42

Figure 4.4 Microstructure of HT-IV 42

Figure 4.5 Fractures on DWT samples 46

Figure 5.1 DWT fracture results for all microstructures at an aspect ratio of 1.24 51

Figure 5.2 DWT fracture results for the HR and HT microstructures at aspect ratios

of 1.0,1.24 and 1.6 51

Figure 5.3 Microstructure of annealed and quenched base steel, consisting of

ferrite and martensite, X 1000 53

Figure 5.4 Fracture strain vs. volume fraction of pearlite 55

Figure 5.5 Differences between undeformed pearlite and ferrite microhardnesses

57

Figure 5.6 Relationship between microhardness difference and fracture strain....57

Figure 5.7 SEM micrographs 60

Figure 5.8 Cracks on DWT samples 62

Figure 5.9 Widely opened fracture on HR sample, aspect ratio 1.24, axial strain 1.8

circumferential strain 0.9 64

Figure 5.10 Widely opened crack on HR sample, aspect ratio 1.24, axial strain 1.68,

circumferential strain 0.84 64

Figure 5.11 Widely opened crack on HT sample, aspect ratio 1.24, axial strain 1.69,

circumferential strain 0.86 65

Figure 5.12 Widely opened fracture on HT-I sample, aspect ratio 1.24, axial strain

1.89, circumferential strain 0.93 65

Figure 5.13 Dimples on fracture surface of an HR sample 67

Figure 5.14 A spheroidal carbide particle on the fracture surface of an HR sample

67

Figure 5.15 Dimples and inclusions on fracture surface of an HR material 68

Figure 5.16 Carbide particle on the fracture surface of an HT-I sample 68

Figure 5.17 X-ray spectrum analysis of arrowed particle in Figure 5.13 69

Figure 5.18 X-ray spectrum analysis of the spheroidal particle in Figure 5.14 69

Figure 5.19 X-ray spectrum analysis of the particle in Figure 5.16 70

Figure 5.20 Carbide particles and a thin crack (arrowed) on the fracture surface of

an HR sample 71

Figure 5.21 X-ray spectrum analysis of the large particle in Figure 5.20 71

Figure 5.22 Small crack (circled) on an HT sample 72

Figure 5.23 Higher magnification of the crack in Figure 5.22 73

Figure 5.24 Root area of the widely opened fracture in Figure 5.9 73
 LIST OF TABLES

Table 3.1 Chemical composition of 1036M CHQ steel 33

Table 3.2 Heat treatments to produce different microstructures 35

Table 4.1 DWT fracture strains for various microstructures 44

Table 4.2 DWT fracture strains for different aspect ratios 44

Table 5.1 Minimum impact energy to cause cracking 49

Table 5.2 Volume fraction of pearlite 54

Chapter 1

Introduction

Cold heading is a forging operation that is widely employed in the fastener

industry to manufacture bolts, screws, and various fasteners. This operation
involves an impact force applied to an end of a metal workpiece, to deform it into
a designed contour. The most commonly used materials in the fastener industry
are cold heading quality (CHQ) steels, including low carbon, medium carbon, and
alloy steels.

However, some workpieces fail during the cold heading operation. These
failures can be attributed to external reasons, such as die design and lubrication,
or to properties of the material itself, known as cold headability.

The cold headability of steels can be affected by chemical composition,

surface condition and microstructure. In general, steels are delivered to the
fastener industry in the as-hot rolled state, of which the microstructure consists of
ferrite and laminar pearlite. Laminar pearlite has a higher deformation resistance
and less ductility, due to laminar cementite dividing a continuous matrix of ferrite.
For medium carbon and alloy steels, a heat treatment is conducted before cold
heading to optimise the microstructure for cold headability.

Conventional test methods to determine the workability of cold heading

materials mainly involve traditional tensile and compression testing. However,
although these methods are widely employed in the fastener industry, they are
Chapter 1 Introduction 2

not able to successfully assess the cold headability of materials. Their primary
weakness is that the strain rate of the test does not represent that of the cold
heading operation.

Previous workers [e.g. Nickoletopoulos 2001] have developed a Drop

Weight Test (DWT) method for cold headability assessment. The DWT simulates
the deformation occurring in the cold heading operation by dropping a weight to
impact the sample being tested. The strain rate of the DWT can be as high as
800s"1 and is varied by changing the dropping height and specimen aspect ratio.
The cold headability is assessed by measuring the strain at which a crack
initiates.

In the present work, as a further development of this test method, a

medium carbon CHQ steel, grade 1036M, was investigated. This research
focused on the effect of microstructure on the cold headability. Two initial
microstructures were used: as-hot rolled and heat treated, and four additional
microstructures were produced through further heat treatment in the lab. DWTs
were performed on samples of all six microstructures. The fracture strain at
which a crack initiated was measured for the assessment of cold headability.
Optical microscope (OM) and scanning electron microscope (SEM) observations
were conducted to evaluate the effect of microstructure on the cold headability.

The specific objectives of the present project were:

• To verify the validity of the DWT on the different microstructures of a

medium carbon steel; and

• To determine the effect of microstructure on the cold headability of a

medium carbon steel.
Chapter 1 Introduction 3

This thesis is divided into the following chapters:

Chapter 2 presents a literature review, including a description of the

definition of cold headability and its methods of assessment. Fracture
mechanisms and the effect of microstructure on fracture are also reviewed
in this chapter.

Chapter 3 explains the details of the experiments conducted.

Chapter 4 presents the experimental results including those pertaining to

heat treatment and the DWT.

These results are discussed in Chapter 5.

Conclusions are presented in Chapter 6.

Chapter 2

Literature Review

2.1 The Cold Heading Process

Cold heading is a forging operation that is performed without an external

heat source to preheat the material. In the course of cold heading, a force is
applied to the free end of a metal workpiece, which is held between a die and a
punch, to form the metal into a predetermined contour. It is widely employed in
the fastener industry, where the end or a head of a metal blank is plastically
deformed to produce fasteners such as bolts, screws, and so on [Nickoletopoulos,
2001; Wick, 1960]. A typical sequence of operations performed in the cold
heading operation is indicated in Figure 2.1.

The production of fasteners involves quite sophisticated processing steps.

Hot rolled CHQ steels are heat-treated, depending on their chemical composition
and hardness, before the cold heading process. The main purpose of this
treatment is to decrease the hardness and increase the workability of the steels
[Bickford and Nassar, 1998].

These heat-treated CHQ steels are then descaled by either a mechanical

or chemical method to remove oxide from their surface. Normally, pickling is
performed on coiled material and hydrochloric, sulphuric, or hydrofluoric acid is
used in the fastener industry. Then the materials are uncoiled, straightened, lime-
coated, drawn and then cold headed. The cold heading operation can consist of
Chapter 2 Literature Review

one or more strokes. This series of operations is performed continuously in a

production line.

V
Figure 2.1 Typical sequence of operations performed in cold
heading

After cold heading, the workpieces sometimes still need to be heat-treated,

depending on the steel quality and technical and engineering requirements. A
brief flowchart of a typical cold heading process is shown in Figure 2.2 [Sarruf,
2000].
Chapter 2 Literature Review 6

2.2 CHQ Steel and Cold Headability

2.2.1 CHQ Steel

Although many kinds of metallic materials can be employed for the

manufacture of fasteners, no other material offers the high strength at low cost
that can be achieved by steel. For this reason, steel has always been the major
material for manufacturing fasteners worldwide. Steels whose characteristics are
specially developed for the purpose of cold heading are known as cold heading
quality (CHQ) steels [Bickford and Nassar, 1998].

CHQ Steel
ir
Hot Rolling
+
Spheroidizing/Annealing
i
Pickling
*

Coating

Cold Drawing
v
Cold Heading

I 4 r

Quenching +Tempering Annealing

Figure 2.2 Brief flowchart of the cold heading process [Sarruf, 2000]
Chapter 2 Literature Review 7

According to the strength grade of the fasteners produced, the steels used
for cold heading can be either plain carbon or alloy types. Low carbon and
medium carbon steels are often used for cold heading. Low carbon steel is
generally used in the as-hot rolled state, whilst a heat treatment is normally
applied to medium carbon steel before cold heading. In general, steels containing
less than 0.08 or more than 0.45 wt.% carbon are seldom used for cold heading.
When the diameter of fasteners becomes too large for carbon steels to provide
homogeneous hardening through heat treatment, alloy steels are used. The
selection of alloy steels depends on their hardening ability. However, high alloy
steels are not chosen because of their greater strength and higher cost [Wick,
1960; Bickford and Nassar, 1998].

The ideal hardness for the raw material generally varies from 72 Rockwell
B for low carbon steels to about 80 Rockwell B for medium carbon and alloy
steels. In any case, the hardness cannot exceed 90 Rockwell B and material
over this hardness may require annealing prior to cold heading. A spheroidized
structure is generally preferred in medium carbon and alloy steels [Wick, 1960]

2.2.2 Cold Headability

Cold headability is the capability of a cylindrical piece of material to be

shaped into the head of a bolt, screw, or other cold formed enlarged part, without
cracking under the high strain rate imposed. It is a material-dependent property
and is affected by the chemical composition and microstructure of the steels.
Additionally, cold headability depends on the surface quality of the material and
working conditions [Sarruf, 2000; Yoo, 1997].

Cold headability is expressed by the maximum strain achieved without

generating a crack. The critical location for crack initiation in a cylindrical sample
is at the barrel surface [Thibau, 1999]. Axial and circumferential strains are
Chapter 2 Literature Review 8

usually used to measure cold headability. The measurement, as shown in Figure

2.3, can be conducted by measuring the space between grids on the surface of a
sample, using the following relationship [Lee and Kuhn, 1973]:

Axial strain: . h . H
e, = l n — = ln- (1)
Hr

w D
Circumferential strain: , ,
ea - In — (2)
wn = In D—
n

Figure 2.3 Strain measurement on compressed specimens [Lee and Kuhn, 1973]

The chemical composition is the strongest factor to influence cold

headability. Cold headability is directly proportional to ductility. In general, raising
the carbon content increases the yield and tensile strength, but decreases the
ductility, impact and toughness properties of the steels. Therefore, high carbon
content adversely affects the cold headability of a material [Nickoletopoulos,
2001].
Chapter 2 Literature Review 9

Besides carbon content, alloy additions also affect the cold headability in a
similar way to that of carbon. The investigation made by Muzak et al. [1995]
showed a reduction of cold headability when the equivalent carbon content in the
steels was increased. The equivalent carbon content (Ceq) is calculated by the
equation:

r -r\Mn \ (Cr + Mo + V
^ (Ni + Cu) , ,
{
6 5 15 '

The presence of surface defects can influence a material's workability. It is

believed that surface defects act as sources of circumferential stress
concentration, which cause defects to open up during uniaxial compression. This
results in ductile fracture at an earlier stage than would otherwise occur for
specimens without defects. [Thomason 1969B; Jenner and Dodd, 1981].
Thomason reported that the defect depth is the primary factor to decrease the
apparent ductility.

Microstructure is another factor that influences cold headability. Most CHQ

steels are hypoeutectoid, with microstructures in the as-rolled state consisting of
a ferrite matrix with varying amounts of lamellar pearlite. The most desirable
microstructure for cold heading is a spheroidized microstructure, because this
leads to higher ductility during processing [Matsunaga and Shiwaku, 1980].

Low carbon steels are soft enough to be cold-headed in the as-hot rolled
state without prior heat treatment, whilst medium carbon and alloy steels have to
be heat-treated before cold heading in order to decrease their hardness and
hence improve their cold headability [Lehnert, 1995; Bickford and Nassar, 1998].
Chapter 2 Literature Review 10

2.3 Typical Fastener Defects

The main defects that occur during cold heading operations are surface
defects. Two types are commonly found on a fastener surface: longitudinal and
oblique. Longitudinal defects are commonly found on the lateral surface of
enlarged heading parts and are parallel to the longitudinal axis of fastener
products. Oblique defects are shear cracks on the upset portion of a piece,
typically at 45° to the longitudinal axis. Figures 2.4 and 2.5, respectively, show a
longitudinal crack and an oblique crack present on the surfaces of the enlarged
parts of two bolts.

Figure 2.4 Longitudinal crack on the head of Figure 2.5 Oblique crack on the head of a
a bolt bolt

The cause of these surface defects is complicated. Nickoletopoulos [2001]

believes that longitudinal cracks are due to the exhaustion of matrix ductility,
whilst shear cracks are the result of localization induced by local softening and
plastic instability.
 11
Chapter 2 Literature Review

Muzak et al. [1995] showed that a crack parallel to the longitudinal axis
indicated the presence of a surface defect. A 45° shear crack signified that the
plastic limit of the material had been reached.

Surface defects in cold heading materials can be initiated during previous

processing, such as casting, and then elongated during subsequent rolling and
drawing. If surface defects are present on a cold heading wire, ductile failure will
occur at an early stage of the compression process. This is the result of the
tensile surface stress, which has a strong influence on the ductility of cold
heading wire, because the tensile stress acts normal to the longitudinal surface
defects and mechanical fibreing. The effective stress concentration at the root of
these defects will cause a large hydrostatic stress and severe localized straining,
which can open up the defects and possibly produce cracks. If the cylindrical
surface of cold heading wire is completely free from defects, ductile failure would
eventually occur by shear fracture [Thomason, 1968].

Longitudinal cracking caused by surface defects occurs at a very early

stage of compression, at low strain and prior to the presence of a shear crack. If
a cylindrical specimen were compressed in the axial direction with very good
lubrication to eliminate barreling, longitudinal defects would not be a source of
rupture or fracture [Thomason, 1968].

2.4 Tests of Cold Headability

As explained above, cold headability is a function of many variables; thus,

it is hard to assess cold headability by means of only one test. Conventional
assessment involves tensile and simple compression testing to gauge yield and
tensile strength, hardness, reduction of area and so on. These tests are easy to
perform and represent the basic properties of the materials. However, there are
some drawbacks in these conventional assessments. Firstly, these methods test
 12
Chapter 2 Literature Review ~~~

the samples at significantly lower strain rates than those involved in actual cold
heading [Sarruf, 2000]. In fact, deformation at high strain rate is one of the most
important features of the cold heading process.

Secondly, some researchers have pointed out that the tensile test is not
suitable for the prediction of cold headability, because fracture in the tensile test
does not initiate at the surface of the sample, as that in the cold heading
operation. This is due to the change in the stress state once necking begins
[Olsson, 1986].

Furthermore, the friction conditions associated with a conventional upset

test are not similar to those that apply to cold heading. Also, the verification of
failures appearing on the surface of the sample are visual and therefore
subjective to the analyser [Muzak et al., 1995].

In order to overcome these problems, several new assessment methods

have been developed. These have a common factor of simulating the
deformation at a high strain rate.

2.4.1 Drop Weight Impact Test

One of the newly developed assessment methods is the drop weight

impact test. The philosophy of this test is to deform the material to failure by an
impact force and analyze the data acquired from the impact test. Figure 2.6
shows the principle of the test device. During a test, a wire specimen is held in an
instrumented impact machine. The specimen diameter used on this device is 3.5
mm. The high speed instrumented impact tester is able to initiate up to 500 joules
impact energy at a velocity of up to 7m/s. An impact force-displacement curve is
recorded and the failure point can be identified, therefore the cold headability can
be calculated. [Muzak et al., 1995].
Chapter 2 Literature Review 13

TOT
Before Upsetting After Upsetting

DIE CASE

WIRE BLANK-
SPACER—

SET
SCREW

Figure 2.6 Drop weight impact test device [Muzak et al., 1995]

2.4.2 Split Hopkinson Pressure Bar Test

Another development to improve conventional assessment is the Split

Hopkinson Pressure Bar Test, as shown in Figure 2.7.

This test consists of two pressure bars that are specially processed. One
is known as the incident bar, the other is the transmitted bar. Strain gauges are
attached to these and a lubricated specimen is mounted between them. The
impact is provided by a striker bar, which is propelled by an air pressure gun.
Upon impact, the elastic deformation of both the incident bar and the transmitted
bar are recorded by the strain gauges. The deformation features of the sample,
including the dynamic stress-strain curve, can be obtained via a computer
program. By this method, an order of magnitude higher strain rates (as high as
103 s"1) can be reached [Yoo et al., 1997].
Chapter 2 Literature Review 14

Striker bar loci dent bar Tnmsmiurd hi-

Specimen

Sarin gage I Strain fajc

_[jW
V J\ Sioppo

Strain fag* Strain (age fjvjra

bridge box -ggg] bridge box

Nicolct
JMWWOS
distal
oscillojcose

Impact Telocity Socxs-ttrain curve

Figure 2.7 Schematic diagram of the Split Hopkinson Pressure Bar Test [Yoo et al., 1997].

2.4.3 Drop Weight Test

The drop weight test (DWT) machine, developed by Nickoletopoulos,

simulates the deformation that occurs in the case of cold heading, by dropping a
weight to impact the sample that is being tested. In general, it is similar to the
design employed for the drop weight impact tester discussed in 2.4.1, but the
present DWT machine can test a larger specimen diameter of 5.2mm.

The DWT machine consists of a tower enabling a set of weight plates to drop
downward from a height of up to 2.5 meters onto a die set and stop blocks.
Figure 2.8 shows the structure of the DWT machine and Figure 2.9 the details of
the die set. The dropped weight impacts the die set, which transfers the impact
 Chapter 2 Literature Review 15

load from the crosshead to the sample. The whole die set configuration is
mounted on a central column that is fixed to the base of the DWT machine
[Nickoletopoulos, 2001].

'«*®Kfe'3sm^>:*»ss3K!@g

Figure 2.8 Structure of DWT machine (a) weight plates, (b) load cell, (c) die set
configuration, (d) pneumatic shock absorbers
Chapter 2 Literature Review 16

I c

40 m m

Figure 2.9 Exploded view of the die-set assembly for the DWT
including: (a) die protection cap, (b) upper die, (c) test
specimen, (d) die sleeve, (e) lower die, and (f) load cell.

A load cell, located between the die set and the central column, measures
real time load and displacement data at the moment of impact. From these, a
load vs. displacement curve can be obtained. The area below the curve is the
energy that the sample absorbs during the impact period.

The impact load and strain rate are adjustable by varying the dropping
weight and height. A sample can be deformed until failure occurs by cracking.
The axial and circumferential failure strains of the sample are then measured.
The minimum strain at which a crack starts is taken as the fracture strain, which
is an indication of cold headability.
 17
Chapter 2 Literature Review

Nickoletopoulos [2001] has used the DWT to examine 7 heats of 1038

steel, with different nitrogen, copper and residual element contents, in both the
as-hot rolled and spheroidized states. The diameter of the sample was 5.2 mm
and an aspect ratio of 1.3 was chosen for all 7 heats, although one of the heats
was additionally tested using ratios of 1.0 and 1.6. The samples of all 7 heats
were cracked using a constant dropping height of 1.5 m at various weights from
11.8-26.3kg.

The tests demonstrated that the DWT simulated the cold heading process
by generating an ideal (constant high strain rate) impact load. The load applied
to the sample reached a maximum value within a very short time. Figures 2.10
and 2.11 show the load-time curves of the DWT for the as-hot rolled and
spheroidized states, respectively. The load increased from zero to a maximum in
around 0.002s, resulting in an approximate strain rate of 630s"1.
Chapter 2 Literature Review 18

0.0005 0.0010 0.0015 0.0020 0.0025 0.0030

Time [s]

Figure 2.10 Load-Time curve of 1038 steel in the as-rolled state [Nickoletopoulos, 2001]

10O0O

0.0000 0.0005 0.0010 0.0015 0.0020 0 0025 0 0030

Time [s]

Figure 2.11 Load-Time curve of 1038 steel in the as-spheroidized state

[Nickoletopoulos, 2001]
Chapter 2 Literature Review 19

The approach to the maximum load has a steeper gradient for the
spheroidized material than for the as-hot rolled state. This is due to the higher
work hardening coefficient for the spheroidized material. This demonstrates that
the DWT is able to identify feature differences associated with different
microstructures.

Figure 2.12 shows a typical load-displacement curve of a DWT. Load-

displacement curves are useful for analysis of the power impacted during loading.

0.000 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008

Displacement [m J

Figure 2.12 Load-displacement curve of as-hot rolled 1038 steel [Nickoletopoulos, 2001]

The transformation band, which is caused by rapid deformation heating

and cooling, is a potential initiation site of internal cracks. Generation of this band
requires a high strain rate and large deformation; therefore it is not easy to
simulate this in the laboratory. However, Nickoletopoulos has successfully
produced transformation bands in the DWT.
Chapter 2 Literature Review 20

He concluded that the DWT is sensitive to the effect of surface integrity

and microstructural differences between as-hot rolled and spheroidized materials
on the fracture behavior during cold heading.

Therefore, the DWT is employed in the present work to evaluate the cold
headability of the materials investigated. However, the load cell system was not
used and the mechanical characteristics of the DWT were not researched in the
present work.

2.5 Ductility Criteria

Ductility is defined as the ability of a material to withstand deformation

without fracture. The apparent ductility of a workpiece depends on the stress and
strain at the surface; once a crack appears at any position on the surface, the
workability limit has been reached [Jenner and Dodd, 1981]. Cockroft and
Latham [1968] proposed that when the work done by the maximum tensile stress
reaches the critical value "C", ductile fracture will occur. This critical value is
constant for a given material under the same conditions of heat treatment, test
temperature and strain rate [Thomason, 1969]. The ductile criteria can be written:

\°,d£eq=C (4)
0

a, - maximum tensile stress

ef - fracture strain

de - effective plastic strain increment

C - critical value
Chapter 2 Literature Review 21

During tensile testing, the maximum tensile stress acts along the
centerline, whilst during upset deformation, the maximum tensile stress is located
at the barrel surface in the circumferential direction [Nickoletopoulos, 2001].

The concept of forming limit diagrams, as shown in Figure 2.13, was

developed to determine the criteria for fracture. A fracture line or fracture locus,
that is a line joining each fracture point along different strain paths, is employed
to predict fracture. The fracture line is parallel to the homogeneous compression
line and has a gradient of -1/2. A material does not fracture at strains below the
fracture line limit and fractures at strains above this line. This concept has been
applied to various deformation processes such as upsetting, rolling and bending.
Figure 2.14 shows the strain states in various deformation processes and Figure
2.15 the distribution of fracture points on the fracture line at different
deformations. [Lee and Kuhn, 1973; Kuhn, 1977; Shah and Kuhn, 1986].

Fracture line

Homogeneous
deformation
Slope=-1/2

Axial Strain

Figure 2.13 Forming limit diagram for upsetting test

Chapter 2 Literature Review 22

The slope of the strain path has the greatest influence on the stress state
at the free surface during upset deformation. When the slope equals -0.5
(homogenous deformation), the circumferential stress is zero. Under this
situation, longitudinal defects will not be a source of fracture. For inhomogeneous
deformation, the strain path gradient is greater than 0.5; thus the circumferential
stress on the free surface is positive. The circumferential tensile stress greatly
affects the workability, because the stress is normal to the longitudinal defects
and mechanical fibers. This may therefore open up the defects and produce
longitudinal cracking. [El-Domiaty, 1999].

Upset Rolling Bending Plane

strain

Surface
strain

Figure 2.14 Surface strain states in various deformation processes

Chapter 2 Literature Review 23

Upsetting

0.4 0.6 0.8

Compression Strain

Figure 2.15 Distribution of fracture points on the fracture line at different deformations of
1045 steel [Lee and Kuhn, 1973]

El-Domiaty [1999] reported that the fracture line of AISI 1045 steel
consisted of three line segments. The first segment was almost parallel to that of
homogenous deformation (slope = -0.5), the second segment had a smaller
gradient than the first, and the slope of the third segment was almost zero, as
shown in Figure 2.16.
Chapter 2 Literature Review 24

09,

01

Fracture
•—•- Region
CD

W
04 CD"
en
Safe —\
Region OJ

I
•Oil -07 4t -0.5 •04 -03 -0J -0.1

Compression strain

Figure 2.16 Fracture locus consisting of three segments for AISI 1045 steel [El
Domiaty, 1999]

H
CD
=5
U)
CD"
U)
—I

&
13'

at -o 7 -os -oi
Compression strain

Figure 2.17 Fracture lines for steel containing different carbon and alloy contents [El-
Domiaty, 1999]
Chapter 2 Literature Review 25

The effect of carbon content on these fracture lines has been reported [El-
Domiaty, 1999] for several steels containing 0.18-0.78 wt.% carbon. The results,
as shown in Figure 2.17, indicate that the locus for the low carbon steel (0.18
wt.% C) is at the highest level in the forming limit diagram, whilst the high carbon
steel (0.78 wt.% C) is located at the lowest level. This indicates that the low
carbon steel has higher workability, together with a higher fracture strain, than
the high carbon steel.

For some materials, fractures cannot be achieved in the lab, because the
required fracture strains exceed the capacity of the test equipment. Thibau et al.
[1999] developed a solution to this by introducing a notch into the surface of the
sample. The fracture strain at a crack occurring at the root of this notch was then
measured.

Thomason also introduced longitudinal grooves of various depths to the

free surface of uniaxial compression specimens. During heading testing, cracks
were initiated at the roots of these grooves. The critical depth of groove, which is
defined as the minimum depth of the groove at which cracking occurs, was
employed to measure the fracture criterion [Thomason, 1969/1970].

2.6 Ductile Fracture

Ductile fracture occurs in three stages: microvoid nucleation, growth of the

microvoids with continued deformation and, finally, coalescence of the voids to
produce complete fracture. [Magnusen et al., 1988; Maheshwari et al., 1978].
The total fracture strain, ef, is the sum of the strains of nucleation, growth and

coalescence [Poruks et al., 1998]:

£
f=£n+£g+£c (5)
Chapter 2 Literature Review 26

where en is the strain at which microvoid nucleation occurs, eg is the strain

during which void growth developed, and e(. is the strain during final

coalescence [Poruks et al., 1998].

2.6.1 Microvoid Nucleation

Microvoid nucleation may occur on second phase particles by particle-

matrix interface separation or by particle cracking. The second phase is usually
less ductile than the matrix, so that the highest stress concentration is in this
region. The voids nucleate preferentially at large particles, particularly at the tip of
elongated precipitates [Gurland, 1972]. Fisher and Gurland, [1981] revealed that
in low and medium carbon steels, voids occurred preferentially at particles
located on grain boundaries. Interfacial decohesion generally occurred at
particles of larger than average size. This is due to the fact that the critical stress
for void nucleation at large particles is lower than that at small particles [Van
Stone etal., 1985].

In low alloy steels, manganese sulphide inclusions are void nucleation

sites at which interfacial decohesion is most frequently observed [Cox and Low,
1974]. Carbides are other commonly occurring void nucleation sites. The large
carbides fail by cracking, whilst smaller carbides fail by interfacial decohesion
[Cox and Low, 1974; Gurland, 1972].

The distribution of second phase particles affects void nucleation.

Microvoid nucleation occurs more easily in microstructures with clusters of
particles than those with homogeneous particle distributions. This is because the
interparticle spacings in particle clusters are smaller than those in homogeneous
structures. [Poruks, 1998]. Voids are often formed between closely spaced
particles [Fisher and Gurland, 1981].
Chapter 2 Literature Review 27

Void nucleation is a function of true plastic strain. Cox and Low [1974]
have researched this relation in AISI 4340 steel and revealed that when the
plastic strain reached a certain degree, voids were found at all inclusion particles.
It is evident that void nucleation also depends upon the volume fraction of
second phase particles. More second phase particles result in the formation of
more microvoids.

Fisher and Gurland [1981] revealed that void nucleation in low and
medium carbon spheroidized steels occurred preferentially by interfacial
decohesion, whilst in high carbon spheroidized steel, the primary mode of void
formation was particle fracture. The cracks were perpendicular to the applied
stress direction in tensile samples and parallel to the loading direction in
compression. At a given strain, void density increased with carbon content and
decreasing cementite particle size [Gurland, 1972].

2.6.2 Growth of Voids

Void growth is the process of cavity enlargement by highly localized

plastic deformation [Van Stone et al., 1985]. It has been estimated that more than
87% of the energy expended during the ductile failure of a high strength steel is
absorbed during void growth [Shockey et al., 1977].

Cavity growth is controlled by the plastic flow of matrix material around the
nucleation site and by the decohesion of smaller second phase particles. These
voids tend to preserve a globular or ellipsoidal shape during growth if it is
controlled by plastic flow of the matrix material, in order to minimize surface and
local plastic energy [Floreen and Hayden, 1970]. Alternatively they can grow by
repeated void nucleation and tearing along a path of neighboring particles. Voids
exhibit irregular shapes if their growth is controlled by the decohesion of smaller
second phase particles [Cox and Low, 1974]. It was revealed that voids grew
Chapter 2 Literature Review 28

more rapidly in the high strength condition of a material than in the equivalent low
strength condition [Floreen and Hayden, 1970].

2.6.3 Void Coalescence

Void coalescence is the last stage of ductile fracture. With continuing

deformation, nucleated voids grow until they impinge on each other. Void
impingement is one of the most common mechanisms of coalescence.

Secondary void formation may also occur in the bands of intense localized
shear between large particle-nucleated voids. Sheets of these voids are
produced in the bands connecting larger voids. These void sheets generally
exhibit angles of 30-40° to the tensile direction [Van Stone et al., 1985].

Cox and Low [1974] revealed that in AISI 4340 steel, large voids may be
formed by decohesion of the interface between the matrix and MnS inclusion
particles. The growth of these voids was interrupted before impingement by the
formation of void sheets at 45° to the tensile direction.

2.7 Improvement of Microstructure for Cold Heading

The as-hot rolled state of CHQ steels consists of ferrite and pearlite.
Lamellar cementite subdivides the ferrite, causing deformation resistance and
stress concentration in the structure. Spheroidized microstructures, which
represent the most thermodynamically stable structures in steels, are comprised
of spherical carbide particles uniformly dispersed throughout the ferrite matrix.
This continuous matrix provides the softest and most ductile condition; thus it is
the most desirable structure for cold heading operations [Muzak et al., 1995].
Chapter 2 Literature Review 29

The purpose of a spheroidizing treatment is to obtain a microstructure of

globular cementite distributed in a ferrite matrix. In general, spheroidizing
annealing is performed closely above or just below the Ac1 temperature for a
certain time after which the steel is cooled slowly to 600°C [Matsunaga et al.,
1980]. According to whether the temperature is above or below the Ac1, two
types of spheroidization may occur: these are known as the intercritical and
subcritical processes.

2.7.1 Intercritical Process

The intercritical process is where the steel is initially heated to above the
Ac1 temperature, then cooled down to below the Ac1 and held for a longer time
[Muzak et al., 1995]. When a hypoeutectoid steel is held just above the A d , the
microstructure partially changes from ferrite plus pearlite to ferrite plus austenite.
Some fine cementite particles are retained in the austenite phase. During
subsequent slow cooling and holding, these fine cementite particles function as
nuclei for spheroidized cementite, mostly on the grain boundaries between the
ferrite and austenite. This higher number of retained cementite particles leads to
a higher ratio of spheroidized to lamellar cementite [Ochi and Koyasu, 1980]
Therefore, retaining the fine cementite particles in the austenite during heating is
of prime importance for obtaining spheroidized cementite [Muzak et al., 1995].
Some researchers have further pointed out that the complete dissolution of
carbon would generate pearlite below the Ac1 temperature rather than
spheroidite [O'Brien and Hosford, 2000].

2.7.2 Subcritical Process

Subcritical annealing occurs when a steel is annealed just below the Ac1
temperature. O'Brien and Hosford [2000] suggested that the fine pearlite
produced by extremely rapid cooling after hot rolling should rapidly spheroidized
Chapter 2 Literature Review 30

at the lower critical temperature. The driving force for subcritical spheroidization
is reduction of the ferrite-carbide interface area by the transformation of carbide
plates to spheres. The mechanism of the transformation is the diffusion of carbon
from high-(surface) energy to low-energy sites.

O'Brien and Hosford [2000] revealed that the ductility (reduction of area)
of a medium carbon steel was close to its maximum value after one hour of such
annealing. Spheroidization was nearly completed within the first two hours, and
further time increased the particle size with an associated reduction in their
number.

In the present work, subcritical annealing was performed to produce a

spheroidized microstructure.

2.8 Development of Non-heat-treated CHQ Steels

Heat treatment, especially the spheroidizing treatment that is often

performed on medium carbon and alloy steels prior to cold heading, is a
procedure that consumes large amounts of time and energy. Many attempts
have been made to eliminate this treatment, including the development of new
non-heat-treated steels.

One approach to non-heat-treated CHQ steel is to decrease the carbon

content. Koyama et al. [1995] developed SUC80 steels to take the place of the
JIS-S45C grade; this was for the manufacture of high strength bolts. Compared
with JIS-S45C, the new steels involve a reduction of carbon content from 0.45
wt.% to 0.18-0.22 wt.% and higher contents of Mn and Cr together with additions
of V, Ti and B. Research has indicated that these new low carbon steels,
combined with a controlled rolling process, have equivalent mechanical
Chapter 2 Literature Review 31

properties and better cold workability than JIS-S45C; plus they eliminate the
necessity of a spheroidizing treatment before cold deformation.

Similar research was conducted by Lee et al. [1992]. Low C and high Mn
steels were developed for the manufacture of high strength (70kg/mm2) bolts.
These were conventionally made from S45C grade steel, with a heat treatment
performed before the cold heading procedure. The composition of the newly
developed steels included 0.15-0.20 wt.% C, 1.39-1.60 wt.% Mn, and the
addition of Nb or V. Through controlled rolling, low reheating temperature, low
rolling temperature, and rapid cooling, a combination of high strength and good
ductility was obtained in the as-hot rolled condition.

Microalloyed extra low carbon steels were developed by Tsai et al. [2000].
These normally have very good formability, hence are suitable for cold heading
applications. In order to further improve cold headability, the new steels were
microalloyed with titanium, niobium and/or boron. Hot-deformation parameters
were also investigated. This revealed that cold headability is sensitive to
microalloying additions, degree of hot deformation, and cooling rate. The addition
of Ti+Nb has a positive influence on the cold headability. When the hot strain is
over 20%, the average grain size of the ferrite is significantly refined; therefore
the cold headability is increased.

Dual-phase steels are other newly developed CHQ steels that avoid the
need for heat treatment. These dual-phase microstructures consist mostly of
ferrite and a second phase of bainite, martensite and/or retained austenite. They
potentially offer good strength and ductility combinations. Two dual phase steels,
DP2 and DP6, with carbon contents of 0.09 wt.% and 0.11 wt.% respectively,
have been developed for CHQ. Cold headed bolts made from as-hot rolled DP2
met the strength and hardness requirements for high strength fasteners without
any heat treatment before or after cold heading [Wanjara et al., 1999].
Chapter 3

Experiments

The experiments performed in the project included:

• Heat treatment

• Drop weight testing

• Visual inspection

• Optical microscopy (OM) and scanning electron microscopy (SEM)

inspection

3.1 Materials

3.1.1 Chemical Composition

As mentioned in Chapter 2, cold headability is affected by the chemical

composition of a material. Any deviation in chemical composition may cause a
difference in cold headability, thus a single chemical composition was used in
this project. The material selected was a medium-carbon grade 1036M CHQ
steel, containing 0.34 wt.% carbon, and the chemical composition is listed in
Table 3.1. This steel is produced by IVACO for the manufacture of fasteners.
Chapter 3 Experiments 33

Table 3.1 Chemical composition of 1036M CHQ steel (wt.%)

c Mn P S Si Cu

0.34 0.94 0.009 0.015 0.24 0.2

Ni Cr Mo Sn Al N

0.09 0.25 0.30 0.012 0.004 0.0065

3.1.2 Microstructures

In the as-received state, two microstructures were investigated. One was

the as-hot rolled state (HR), whose microstructure consisted of lamellar pearlite
and ferrite. The other was a basic heat treated state (HT). The heat treatment
cycle involved annealing at 745°C for 4 hours, then cooling to 500°C in 5 hours.
The microstructure of the HT condition consisted of heat treated pearlite and
ferrite. Figure 3.1 shows the microstructures of 1036M steel in the two as-
received states.
Chapter 3 Experiments 34

SEI _15.0_kV x2000 5pm

Figure 3.1 (a) Microstructure of 1036M steel, as-hot rolled state, x2000

SEI 15.0kV x2000 5pm

Figure 3.1 (b) Microstructure of 1036M steel, heat treated state (HT), x2000
Chapter 3 Experiments 35

3.2 Heat Treatment

In addition to the above mentioned conditions, four other microstructures

were produced by means of various heat treatments in the lab, so as to
determine the effect of microstructure on cold headability. The four
microstructures, referred to as Heat Treated I - IV, (HT-I, II, III, IV), were
produced as follows:

HT-I: samples were heated to 745°C, held at temperature for 24 hours,

then furnace cooled to room temperature.

HT-II, III, IV: samples was heated to 715°C and held at temperature for 24
hours, 4 hours and 1 hour, respectively, then furnace cooled.

The aim of HT-I was to compare the effect on cold headability of

microstructures containing lamellar pearlite with that containing completely
spheroidized cementite. HT-II, III and IV were used to compare the effect of
different degrees of spheroidizing.

These heat treatment cycles are summarized in Table 3.2.

Table 3.2 Heat treatments to produce different microstructures

Heat Treated I (HT-I) Annealed at 745°C for 24 hours

Heat Treated II (HT-II) Annealed at 715°C for 24 hours

Heat Treated III (HT-III) Annealed at 715°C for 4 hours

Heat Treated IV (HT-IV) Annealed at 715°C for 1 hour

The furnace used was an electric muffle type, with no protective

atmosphere applied during the heat treatment.
Chapter 3 Experiments 36

3.3 Drop Weight Test (DWT)

Drop weight tests were performed to measure the cold headability of the
tested materials.

3.3.1 Drop Weight Tower

The DWT machine consisted of a tower that guided weight plates

dropping down onto a die set and stop blocks. The weights were changeable to
generate various impact loads. A cylindrical sample was held vertically in the die
set on a central column that was fixed to the base of the machine. The die was
set so that the impact force was forwarded to workpiece. Two pneumatic
cylinders, fixed to the base, worked as shock absorbers to avoid secondary
impact caused by rebounding of the weight. The die set assembly is shown in
Figure 3.2.

ma&.
Figure 3.2 Die set assembly of DWT; (a) die sleeve; (b) lower die, (c) upper die
Chapter 3 Experiments 37

3.3.2 Specimen Preparation

As explained in Chapter 2, cold headability is affected by the surface

condition of the material. In order to avoid any influence of surface condition in
the current investigation, all DWT samples were smoothed by machining.

The DWT samples were of a cylindrical shape with a diameter of 5.2mm.

An aspect ratio of 1.24 was chosen for all microstructures, whilst 1.0 and 1.6
were also tested for the HR and HT materials. The resulting sample heights
were 8.24, 7 and 10 mm respectively.

3.3.3 Drop Weight Test

The DWT can be controlled by changing either the drop weight or the drop
height. Varying the drop height causes a change in strain rate; therefore, in this
work, the height was kept constant and varying weights were used in ascending
order. In the tower, the maximum drop weight was limited by the length of the two
thread bars, which penetrated through the weight plates and tied them together.
Thus when this limit was reached, a new drop height was taken and kept
constant, whilst the drop weight was varied again in an ascending manner.

Previous work revealed that a drop height of 1.5 m is suitable for fracturing
a similar medium carbon (1038) steel [Nickoletopoulos, 2001]. Therefore, the
primary drop height was set at 1.5 m and the drop weight was varied in
increments of 2.3kg from 11.35kg until the sample cracked. Once this rough
weight limit was established, the weight was decreased at intervals of 0.6 kg to
determine the minimum weight at which a crack was generated. Three samples
were tested for each condition.
Chapter 3 Experiments 38

3.4 Fracture Evaluation

Cold headability is measured in terms of the surface strain at which the

first crack is initiated; thus observing and measuring cracks is the key for its
determination. However, this is a subjective measurement; there is no industry
standard available to evaluate the surface quality of cold headed parts. In this
work, the surface observations on the DWT's samples were divided into three
classes: no crack, ghost line crack, and open crack, as illustrated in Figure 3.4. A
stereomicroscope was employed to inspect these surface cracks and the
presence of a ghost line crack at a magnification of x 25 was used to classify the
failure of a tested sample. The surface strain of a deformed sample was
measured as shown in Figure 3.3.

Axial strain:
£. - In
H,

Ho Circumferential strain:
w
ee =in- Dn
w,

Df

Figure 3.3 Measurement of surface strain

A JEOL JSM-840A scanning electron microscope was used to observe

the cracks and fracture surfaces of tested samples. Energy Dispersive Spectrum
(EDS) analysis was employed to qualitatively determine the chemical
composition of particles observed on these fracture surfaces.
Chapter 3 Experiments 39

(a) Surface with no cracks

ik <~]9 W (b) Ghost line cracks

•KB^^^T^fetilr::'
IHfiSiftl ^- - -1
HHHHHlfPs*^ I
-i
•1ftj:
1 HHIHiP'
1i f f l i f f i
(c) Open crack

Figure 3.4 Illustration of three crack classes in DWT specimens

Chapter 4

Results

4.1 Heat Treatment

Heat treatments were performed according to the cycles listed in Table 3.1
to produce four additional microstructures, as shown in Figures 4.1 - 4.4.

HT-I, annealed at 745°C for 24 hours, then furnace cooled, consisted of

equiaxed pearlite and ferrite. This is because the hot rolled pearlite was
transformed into austenite above the A d temperature (approximately 720 °C), and
transformed to pearlite again during cooling.

HT-II, annealed at 715°C for 24 hours, then furnace cooled, consisted of

spheroidized cementite distributed in a ferrite matrix. Because this temperature
was below the A d , no phase transformation occurred during heating, holding and
cooling. However, due to the elevated temperature, lamellar cementite in the
pearlite was converted into spheroidal cementite.

HT-III and HT-IV, annealed at 715°C for four hours and one hour
respectively, had microstructures similar to that of HT-II, but were characterized by
different degrees of spheroidization.
Chapter 4 Results 41

SEI 15.0kV X2000 5um ^

Figure 4.1 Microstructure of HT-I, x2000

SEI 15.0kV x2000 5u_m i i

Figure 4.2 Microstructure of HT-II, x2000

Chapter 4 Results 42

SEI 15.0kV x2000 5u_m

Figure 4.3 Microstructure of HT-III, x2000

SEI 15.0kV X2000 5ym

Figure 4.4 Microstructure of HT-IV, x2000
 4
Chapter 4 Results 3

4.2 Drop Weight Test Results

DWT testing was performed on the 6 materials with specimen aspect ratios
of 1.24, so that all the microstructures could be compared at an identical aspect
ratio. HR and HT were additionally tested at aspect ratios of 1.0 and 1.6.

The dimensions of the samples were measured before and after the DWTs,
so that the axial and circumferential strains could be calculated. The surfaces of
the tested samples were carefully inspected under a magnification of x25 and the
minimum strain at which the first crack initiated was taken as the fracture strain.
Table 4.1 lists the DWT fracture strains for all six microstructures at an aspect ratio
of 1.24. These are expressed using lower and upper limits; the lower limit is
defined as the strain at which the first crack occurred, whereas the upper limit is
the maximum strain at which no crack was found. No cracks were found below the
lower limit and all samples were cracked above the upper limit.

HR, the as-hot rolled material, had the lowest axial and circumferential
fracture strains. Compared with HR, there was no significant improvement with the
HT microstructure. However, HT-II had the best cold headability, with the largest
axial and circumferential strains among the six microstructures. HT-I and HT-III
had similar intermediate values. Comparing HR and HT-II, the lower limits of the
axial and circumferential strains of HT-II were 2 1 % and 17% higher than those of
HR. The differences between the fracture strains are due to the different
microstructures of the tested materials.

Table 4.2 indicates the fracture strains of HR and HT for aspect ratios of 1,
1.24 and 1.6. The fracture strains increased with increasing aspect ratio. This is in
agreement with the theory of fracture limit diagrams.
Chapter 4 Results 44

Table 4.1 DWT fracture strains for various microstructures

Aspect Axial Circumferential

Ratio Strain Strain

HR 1.24 1.52 - 1.70 0.77-0.88

HT 1.24 1.59-1.63 0.82 - 0.84

HT-I 1.24 1.75-1.89 0.89-0.93

HT-II 1.24 1.84-1.95 0.9-0.99

HT-III 1.24 1.76-1.81 0.88 - 0.93

HT-IV 1.24 1.58-1.84 0.8 - 0.94

Table 4.2 DWT fracture strains for different aspect ratios

Aspect Axial Circumferential

Ratio Strain Strain

HR 1 1.4-1.59 0.71-0.79

HR 1.24 1.52-1.70 0.77-0.88

HR 1.6 1.76-1.85 0.85-0.92

HT 1 1.5-1.58 0.75-0.83

HT 1.24 1.59-1.63 0.82 - 0.84

HT 1.6 1.78-1.8 0.88-0.91

Chapter 4 Results 45

The samples cracked by DWT were carefully examined. Most cracks were
of a longitudinal type, whilst some equiaxed small fractures and widely opened
fractures were also found. Figure 4.5 shows typical cracks and fractures that were
observed in the DWT samples. Figure 4.5 (a) depicts a longitudinal crack that is
narrow and extends along the length of the sample. Figure 4.5 (b) is a small
"equiaxed" fracture on the barrel of the sample; whilst Figure 4.5 (c) is a widely
opened fracture, which penetrates through the entire length of the deformed
sample. The equiaxed fracture can be seen to be made up of two perpendicular
shear fractures.
Chapter 4 Results 46

«v~_-':- yxsz- -r

(a) longitudinal crack

(b) small equiaxed fracture

(c) widely opened fracture. Mag. x25

Figure 4.5 Fractures on DWT samples

Chapter 5

Discussion

5.1 Drop Weight Test Results

The objective of the present work was to evaluate the cold headability
associated with different microstructures by means of measuring the axial and
circumferential strains at which a crack is initiated. As discussed in Chapter 2, it is
widely agreed that commonly used test methods in the fastener industry, such as
tensile and upset testing, are not able to assess cold headability satisfactorily,
[Olsson et al., 1986; Muzak et al., 1995]. The main problem with these traditional
techniques is that they do not correspond to the conditions of cold heading
deformation in terms of strain rate, stress state and friction. In the present work,
the DWT was employed to simulate the cold heading process.

The DWT is designed to deform and crack a cylindrical sample, held in a

pair of sleeve-guided dies, under an impact load caused by a dropping weight.
The impact energy (E) is a function of the mass (M) and velocity (V),
:
1
E = -MV (6)
2
 48
Chapter 5 Discussion

The impact velocity is a function of the drop height (h),

V = j2g~h (7)

Therefore, the impact energy is adjustable by varying the drop weight

and/or drop height.

The DWT is capable of applying deformations at very high strain rates.

During impact, the strain rate decreases from the initial strain rate, corresponding
to the impact velocity, to zero. Using a 1.5 m drop height, the initial strain rate can
be as high as 630s"1 for an aspect ratio of 1.6 and 820s"1 for an aspect ratio of 1.3
[Nickoletopoulos, 2001].

The die set of the present DWT is designed in the form of a sleeve guide,
with a pocket on each die to stabilize the specimen. The upper die moves along
the die sleeve, so that the specimen is deformed exactly along its axis.

When a cylindrical specimen is deformed in uniaxial compression, the strain

is initially uniformly distributed along the axis. Nevertheless, stress concentrations
develop between the matrix and inclusions or second phase particles. When the
circumferential stress on the equatorial free surface of the cylindrical specimen
becomes increasingly tensile, due to the occurrence of barreling, a state of tensile
plastic instability can be reached. The high stress concentrations at inclusions or
second phase particles within this highly strained region rupture the bonds with the
matrix or alternatively crack particles, thus nucleating voids [Thomason, 1969].

Many researchers have employed "naked eye" observations to detect the

presence of cracks on the barrel surface. However, Nickoletopoulos [2001] used a
stereomicroscope at x25 magnification for such determinations. This latter method
is more stringent than the "naked eye" technique, which may result in an
Chapter 5 Discussion 49

overestimate of the fracture strain. In the present work, the stereo microscopy
method was again employed. The observation of a ghost line at x25 magnification
was used to define the failure of a sample.

DWT was performed on all six microstructures; all samples had the same
chemical composition. The test results, which were listed in Chapter 4, are
presented in a different format in Table 5.1 and Figures 5.1 and 5.2.

Table 5.1 Minimum impact energy to cause cracking (aspect ratio: 1.24)

Drop Weight Drop Height Impact Energy Axial Strain

(kg) (m) (J)

HR 28.8 1.5 423 1.52

HT 24.1 1.5 354 1.59

HT-I 29.9 1.9 567 1.75

HT-II 28.8 2.1 592 1.84

HT-III 25.8 1.9 480 1.76

HT-IV 29.9 1.5 440 1.58

Table 5.1 lists the values of impact energy that caused initial (ghost line)
cracking in the tested materials. From the table, it can be seen that the HT
microstructure has the lowest impact fracture energy, but the fracture strain is not
significantly greater than that for the HR material. This indicates that the HT heat
treatment only softened the material, without significantly improving its formability.
HT-II had the highest impact energy among the six tested materials, corresponding
Chapter 5 Discussion 50

to an axial strain of 1.84. This indicates that the HT-II heat treatment best improved
the cold headability. HT-I and HT-III led to similar fracture strains, but the HT-I
impact energy was 16% higher than that of HT-III. This indicates that a
spheroidized microstructure has a lower deformation resistance than a
microstructure containing lamellar pearlite. The HT-IV fracture strain was close to
that of HT, whose impact energy was a little higher than that of HR.

As explained in Chapter 4, lower and upper limits were used to define the
range of strains attainable without fracture. The strains below the lower limit were
deemed to be well within the safe zone, whilst strains above the upper limit were
definitely in the fracture zone. Strains between the lower and upper limits were
deemed to be in the danger zone. No cracks were observed in the safe zone,
cracks were always found in the fracture zone, whilst cracks may or may not have
developed in the danger zone. It is evident that the larger the safe zone, the better
the cold headability of the material.

Figure 5.1 shows that the as-hot rolled microstructure, which contains large
amounts of lamellar pearlite, has the lowest amount of allowable strain, whereas
HT-II has the largest safe zone amongst the six tested materials. HT and HT-IV
were similar; both were more formable than HR. HT-I and HT-III were also similar
over the extent of the safe zone.

Figure 5.2 displays the DWT results for the HR and HT samples tested
using different aspect ratios. In general, the fracture strain increased with
increasing aspect ratio. The greater the aspect ratio in a heading operation, the
lower the stress concentration effect at the free surface [Thomason, 1969B] and
hence the greater the cold headability. This is in agreement with the theory of
workability limit diagrams. Therefore, comparisons of the cold headability should
be conducted on specimens of similar aspect ratio.
Chapter 5 Discussion 51

HR HT HT-I HT-II HT-III HT-IV

El safe zone • danger zone • fracture zone

Figure 5.1 DWT fracture results for all microstructures at an aspect ratio of 1.24

1.8

1.6
9
1.4

1.2
re
*-*
w 1
"re

<
0.8
&, ••'••

0.6 : '
0.4

0.2 : •--,,..
:"

0 ;!-•:•
HR-1.0 HR-1.25 HR-1.6 HT-1.0 HT-1.25 HT-1.6

IB safe zone • danger zone • fracture zone

Figure 5.2 DWT fracture results for the HR and HT microstructures at aspect ratios
of 1.0,1.24 and 1.6
 52
Chapter 5 Discussion

5.2 Effect of Microstructure

As mentioned in previous chapters, cold headability depends on chemical

composition, surface condition and microstructure. The present work was focused
on the effect of microstructure. All the material used in this work was of the same
chemical composition, containing 0.34 wt.% carbon. The surfaces of the samples
were smoothed by machining. Hence it can be concluded that all the variations
between test results were due to microstructural differences.

The as-hot rolled (HR) microstructure consisted of ferrite and pearlite.

These constituents formed during phase transformation on the STELMOR
conveyor. Due to the fast cooling on the conveyor, relatively high pearlite fractions
were present in this microstructure.

The HT material had been prepared by heating to and holding at 745° C for
4 hours, then cooling to 500° C in 5 hours. The microstructure contained ferrite and
pearlite; the latter was mostly of lamellar form and was spheroidized in a few areas.
The pearlite fraction was lower in the HT microstructure than in the HR material.
As the annealing temperature was in the ferrite-austenite dual phase region, phase
transformation from pearlite to austenite occurred at 745° C and the long holding
time at this temperature also resulted in the formation of pro-eutectoid ferrite.
During subsequent cooling, the remaining austenite transformed into pearlite. Due
to the relatively slow cooling rate, some of the pearlite was spheroidized during this
cooling period.

In order to confirm the above analysis, a heat treatment test was conducted.
A sample was heated to 745° C and held at this temperature for 4 hours, then
quenched to room temperature. Ferrite and martensite were found in the
microstructure of this sample, as shown in Figure 5.3.
Chapter 5 Discussion 53

Figure 5.3 Microstructure of annealed and quenched base steel, consisting of ferrite and
martensite, X 1000

The HT-I microstructure was developed by annealing at 745° C for 24 hours.

The microstructure consisted of equiaxed ferrite and lamellar pearlite, formed
during cooling. The pearlite fraction in HT-I was much lower than in HR.

HT-II, HT-III and HT-IV were developed by annealing at 715°C for 24 hours,
4 hours and 1 hour, respectively. Because of the long holding time close to the Aci
temperature, the pearlite in HT-II was completely spheroidized. Partial
spheroidization occurred in the HT-III sample, whereas only small regions of the
HT-IV microstructure were spheroidized.

The tested materials consisted of a two-phase mixture of ferrite and pearlite

or spheroidized cementite. The volume fraction of each phase in the mixture
influences the strength and formability of the material. In general, the hard phase
of such a mixture contributes to the material's strength and hardness, whereas the
soft phase contributes ductility and toughness [Gurland, 1988].
Chapter 5 Discussion 54

The volume fraction of pearlite in each material was measured by means of

an optical microscope using image analysis (Clemex) software. HR had a volume
fraction of 78% pearlite, HT-IV was close to HR with 75%, and HT had 63%. HT-II
had the lowest pearlite fraction with an average of 43%, while HT-I and HT-III
contained 48% and 52%, respectively. The measured results are listed in Table
5.2.

Table 5.2 Volume fraction of pearlite (%)

HR HT HT-I HT-II HT-III HT-IV

78 63 48 43 52 75

Pearlite in carbon steel is the hard phase and contributes to its hardness.
Spheroidized cementite also works as a second phase. The effect of the second
phase is to decrease ductility and impact toughness [Pickering, 1980], Therefore,
decreasing the pearlite fraction will improve the cold headability.

Goods and Brown [1979] have given the relationship between the critical
strain to nucleate a void and the volume fraction of second phase particles:

i V
(Ti-<7H - (70 V
e =1.7- 1 + 3/,,+ (8)
b M 1.8

ECri - critical strain to nucleate a void oi - interfacial strength

b - Burgers vector o0 - Orowan stress

fv - particle volume fraction OH - hydrostatic tension

r - particle radius p - shear modulus

Chapter 5 Discussion 55

It is evident from this relationship that increasing the second phase volume
fraction (fv) results in a decrease in the critical strain required to nucleate a void
(Ecri).

Figure 5.4 shows the relationship between the fracture strain and the
volume fraction of pearlite. With decreasing pearlite fraction, the fracture strains
(axial and circumferential) increase. This is because the interfacial areas of the two
phases are the regions of high stress concentration. Void nucleation occurs at the
sites of high stress concentration [Van Stone at al., 1985]. Raising the volume
fraction of pearlite, therefore, results in an increase in the density of void
nucleation sites; hence it deteriorates the cold headability.

2.5

c
(0
1.5
*->
<fi
0)
1-
3
•¥••
o<D 1

0.5

0
35 45 55 65 75 85
Fraction of Pearlite (%)
• Circumferential strain • Axial strain

Figure 5.4 Fracture strain vs. volume fraction of pearlite

Chapter 5 Discussion 56

The stress concentration at the interface results from the difference in

mechanical properties between the two phases. As ferrite is the soft phase, it
contributes to the material's ductility; pearlite is the hard phase and contributes to
the hardness. The difference in ductility between the two phases leads to high
stress concentrations at their interfaces during deformation. When the
concentration exceeds a critical value, voids nucleate on these interfaces or at
second phase particles. The smaller the difference in ductility between the two
phases, the better the cold headability.

In order to determine the difference in strength and hence ductility between

the two phases, microhardness analysis was used. The microhardness (Vickers
microhardness at a load of 50 grams) of ferrite and pearlite colonies in each
material was measured at multiple points on each phase for an undeformed
sample, so that the average microhardness and in this way the disparity between
the ferrite and pearlite colonies was obtained.

Figure 5.5 shows the average microhardnesses and the differences

between the phases for all the tested materials. HR, whose microstructure was
formed on the STELMOR conveyor, had the highest values in the pearlite and
ferrite, as well as the largest disparity between them amongst all six tested
materials. After heat treatment, the microhardnesses of both the pearlite colonies
and the ferrite decreased to different degrees, as did the disparity values.
Chapter 5 Discussion 57

400

HR HT HT-I HT-II HT-III HT-IV

FJ Pearlite • Ferrite • Difference

Figure 5.5 Differences between undeformed pearlite and ferrite microhardnesses

2.5

20 40 60 80 100 120
Microhardness difference before DWT
• Hoop strain • Axial strain

Figure 5.6 Relationship between microhardness difference and fracture strain

 58
Chapter 5 Discussion

The HT-II material had the smallest microhardness difference between

phases. Note, however, that some uncertainty may be involved in the
microhardness values, because the dimensions of some spheroidal cementite
particles were so small that the test indenter probably covered some adjacent
ferrite in certain cases.

Figure 5.6 displays the relationship between the microhardness difference

before DWT and the fracture strain. The fracture strain decreased with an increase
in the microhardness disparity. It is evident that higher disparity values result in
higher stress concentrations at the phase interfaces and therefore to an increase
in the probability of nucleating microvoids and fractures.

The sizes and shapes of second phase particles are also factors that affect
the characteristics of fracturing. Fisher and Gurland [1981] reported that voids in
spheroidized carbon steel were generally associated with particles whose
dimensions were of greater than average size and that they rarely formed at very
small particles. Non-equiaxed or irregularly shaped cementite particles were often
subject to internal fractures. Particles situated at ferrite grain boundaries were
favored sites for the nucleation of voids.

The fracture stress of a second phase particle may be calculated using an

energy balance approach. It has been assumed that when the elastic energy
stored in a second phase particle exceeds the surface energy for a newly formed
void, fracture can occur at the particle. The relationship between the critical stress
for particle cracking (oc) and particle size (d) was provided by Van Stone [1985].

<J, = (9)
Kq-dj

E - particle Young's modulus y - surface energy of particle

q - particle stress concentration factor

Chapter 5 Discussion 59

Trufiakov [1992] revealed that the fracture toughness of two-phase steels

was determined by the volume fraction and particle size of the second phase
component. He gave the relationship between the crack resistance and the second
phase fraction and size as the following:

S =^ = (10)

5 - crack resistance

Vsp - volume fraction of second phase

h sp - size of second phase

A - coefficient

From the above discussion, it is evident that increasing the dimensions of

the second phase will decrease the critical stress required to cause particle
cracking, as well as the crack resistance and so deteriorate the formability.

The HT-II, III and IV materials, which were annealed at the same
temperature but for different holding times, differed in the distribution of the
carbides present; this is shown in Figure 5.7. The pearlite was spheroidized to
different degrees in these microstructures. Here the "spheroidization ratio" is
defined as the ratio of the number of spheroidal cementite particles to the total
number of cementite particles including those of lamellar form [Das et al., 2002].
Chapter 5 Discussion 60

SEI 15.0kV x200d'5pm

Figure 5.7 SEM micrographs of (a) HT-II, (b) HT-III, (c) HT-IV, x2000

Several researchers [Chattopadhyay and Sellars, 1977; Das et al., 2002]

have employed the "aspect ratio" of the cementite to evaluate the degree of
spheroidization. The aspect ratio is defined as the length ratio of the long axis to
the short axis of a cementite particle. When this ratio is less than a critical value,
the particle is deemed to be spheroidized. Chattopadhyay and Sellars chose 5 as
the criterion, whilst Das chose 2.

Such a quantitative analysis of spheroidization was not carried out in the

present work. Nevertheless, it is qualitatively evident that the aspect ratios of
Chapter 5 Discussion — 61

almost all the cementite particles in the HT-II and III materials are less than 5, as
shown in Figures 5.7 (a) and (b), which indicates complete or nearly complete
spheroidization. The aspect ratios of most of the cementite particles are greater
than 5 in the HT-IV material, Figure 5.7 (c). Because the second phase aspect
ratio was the highest, the headability of the HT-IV material was the lowest of the
four steels heat treated in the laboratory.

5.3 Crack Analysis

The purpose of DWT is to ascertain the crack resistance by deforming a

cylindrical sample to cracking. In addition to the fracture strain, the crack profile
and orientation were also observed. From the profile and orientation of a crack, the
fracture mechanism can be determined qualitatively.

Nickoletopoulos [2001] concluded that longitudinal cracks are due to

exhaustion of the matrix ductility, whilst shear cracks are the result of localization
induced by local softening and plastic instability. Other research has revealed that
longitudinal cracks can be caused by the presence of surface defects on cold
heading rod, so that the plastic limit of the material is reached more quickly.
[Thomason, 1968; Muzak etal., 1995].

In the present work, all the samples tested by DWT exhibited longitudinal
cracking at the point of initiation. Figure 5.8 shows the first crack on one sample of
each of the tested materials. Most of the DWT samples exhibited a narrow
longitudinal crack on the barrel surface, even when the strain was moderately high.
Some of the "longitudinal" cracks were seen to consist of segments of shear
cracks, e.g. Figure 5.8 e.
Chapter 5 Discussion 62

Figure 5.8 Cracks on DWT samples, (a) HR, (b) HT, (c) HT-I, (d) HT-II, (e) HT-III, (f) HT-IV, X25
 63
Chapter 5 Discussion

However, widely opened fractures occurred on some samples of the HR,

HT and HT-IV materials, even at moderate strains. These all contained non-
equiaxed pearlite. No widely opened cracks were found in the HT-II and III grades.
Figure 5.9 illustrates a widely opened fracture on an HR sample with an aspect
ratio of 1.24, an axial strain of 1.8 and a circumferential strain of 0.9. The fracture
edges are widely separated on the barrel surface and several internal cracks of a
shear nature extend from the root of the fracture along the different directions.
Figure 5.10 shows a widely opened fracture on another HR sample with an aspect
ratio of 1.24; here the axial and circumferential strains were 1.68 and 0.84,
respectively. The morphology of the fracture is similar to that of the previous HR
sample and again the shear character of the internal cracks is evident.

A widely opened crack of the HT material is shown in Figure 5.11. The

aspect ratio of the sample was again 1.24 and the measured axial and
circumferential strains were 1.69 and 0.86, respectively. Compared with the two
previous HR samples, the HT fracture is narrow and short. This fracture does not
completely traverse the barrel surface. Figure 5.12 displays a widely opened
fracture on the surface of an HT-I sample whose strains were 1.89 and 0.93,
respectively. This fracture was widely opened on the barrel surface and a zigzag
(shear) crack can be seen to characterize the root of the fracture.
Chapter 5 Discussion 64

Figure 5.9 Widely opened fracture on HR sample, aspect ratio 1.24, axial strain 1.8
circumferential strain 0.9, x25

SEI 15.0kV X25

Figure 5.10 Widely opened crack on HR sample, aspect ratio 1.24, axial strain 1.68,
circumferential strain 0.84, x25
 Chapter 5 Discussion 65

Figure 5.11 Widely opened crack on HT sample, aspect ratio 1.24, axial strain 1.69,
circumferential strain 0.86, x25

Figure 5.12 Widely opened fracture on HT-I sample, aspect ratio 1.24, axial strain 1.89,
circumferential strain 0.93, x25
Chapter 5 Discussion 66

SEM observations were also made. The fracture surfaces were seen to
exhibit large amounts of voidal depressions separated by ridges that are remnants
of ligaments. The dimples, which were distributed randomly, were mainly equiaxed
and of different sizes. Figure 5.13 illustrates some of the dimples on the fracture
surface of an HR sample. Some particles are observed to be lodged in the bases
of the dimples and most of these are elongated.

The presence of the dimples on the fracture surface indicates that the
failures of the DWT samples took place by ductile fracture controlled by a
microvoid coalescence mechanism. The microvoids were initiated at the second
phase particles or at interfaces between the matrix and the particles as a result of
the stress concentrations discussed above [Van Stone et al., 1985]. These second
phase particles can be inclusions, particles of spheroidal cementite, or an entire
pearlite colony, where the latter consists of alternate cementite and ferrite platelets
[Metal Handbook, MCIC, 1975]. The dimples were formed as a result of the growth
and coalescence of the microvoids.

Second phase particles play important roles in fracture development.

Particles found at the bottom of dimples, such as those in Figure 5.13, are
responsible for the initiation of voids. Figure 5.14 shows a spheroidal carbide
particle on the fracture surface of an HR sample. Figure 5.15 displays the particles
on another area of the fractured sample of Figure 5.13. Most of the particles are
elongated and some are broken into pieces. Figure 5.16 presents a particle on the
fracture surface of an HT-I sample.

Energy Dispersive Spectroscopy (EDS) was employed to determine the

particle compositions qualitatively. Most of the particles shown in Figure 5.13 were
carbides, whilst a few particles were manganese sulphides. Figure 5.17 is the X-
ray spectrum of the particle identified by an arrow in Figure 5.13. The spectrum
revealed that the particle, which is located in a dimple, was an MnS particle. MnS
Chapter 5 Discussion 67

inclusions were also found on the fracture surface (particle A in Figure 5.15). This
particle was elongated and broke during deformation.

Figure 5.13 Dimples on fracture surface of an HR sample, x1500

Figure 5.14 A spheroidal carbide particle on the fracture surface of an HR sample, x1300
Chapter 5 Discussion 68

Figure 5.15 Dimples and inclusions on fracture surface of an HR material, x1600

Figure 5.16 Carbide particle on the fracture surface of an HT-I sample, x1300
Chapter 5 Discussion 69

au_ 1000™

Mn

b
Fe

.^.iL^jJ^JLJ^L.^
0.70 1.40 2.10 2 . SO 3.50 4.20 4.90 5.60 6.30 7.00 7.70

Figure 5.17 X-ray spectrum analysis of arrowed particle in Figure 5.13

Figure 5.18 X-ray spectrum analysis of the spheroidal particle in Figure 5.14
Chapter 5 Discussion 70

Cl
K
UuilkiJljH.* '•+••- - A ^
0.70 1.40 2.10 2.90 3.50 4.20 4.90 5.60 6 30 7.00 7.70

Figure 5.19 X-ray spectrum analysis of the particle in Figure 5.16

Figure 5.18 illustrates that the composition of the spheroidal particle

presented in Figure 5.14 included C, O and Ca. Figure 5.19 is the X-ray spectrum
of the particle in Figure 5.16, which is in a dimple on the fracture surface of an HT-I
sample. The spectrum indicates that the particle is a carbide. The Na, CI and K
peaks are due to contamination of the fracture surface.

As discussed above, second phase particles lead to the initiation of fracture

because of the stress concentrations caused by the differences in ductility between
the particles and the matrix material. Particle B in Figure 5.15 is a carbide platelet
that had been broken into pieces. A crack was initiated at one end and extended
from it. Another example is shown in Figure 5.20. Two particles were located on
the fracture surface of an HR sample. EDS analysis revealed that both particles
are iron carbides, whilst the X-ray spectrum of the larger particle reveals the
presence of a small S peak (Figure 5.21). A crack originated on the particles and
then connected them up.
Chapter 5 Discussion 71

Figure 5.20 Carbide particles and a thin crack (arrowed) on the fracture surface of an
HR sample, x600

Fe
J! F e

1.00
*
2 00 3.00 4.00 5.00 6.00 7.00 .00 9.00 10.00

Figure 5.21 X-ray spectrum analysis of the large particle in Figure 5.20
Chapter 5 Discussion 72

Additional evidence was obtained from an HT sample. A small crack was

present on the free surface of the sample at a magnification of x25 (Figure 5.22)
and the crack was quite shallow. It is evident that the crack was initiated just below
the surface. Under higher magnification, a particle was found on the crack and the
fracture occurred at the interface between the particle and the matrix (Figure 5.23).
EDS analysis indicated that this was a carbide. Thus a particle on or near the
surface of a sample can act as a surface defect and then initiate a surface crack in
this way.

A pearlite colony can also act as a stress concentration site and cause
fracture during DWT. Figure 5.24 displays a pearlite colony on the fracture surface
of an HR sample. The pearlite colony, located at the root of the fracture, was
broken by the crack that penetrates through the colony.

Figure 5.22 Small crack (circled) on an HT sample

Chapter 5 Discussion 73

Figure 5.23 Higher magnification of the crack in Figure 5.22, carbide particle is arrowed, x550

Figure 5.24 Root area of the widely opened fracture in Figure 5.9, x300
Chapter 5 Discussion 74

During plastic deformation, voids nucleate at harder particles, then grow

and coalesce during further flow of the matrix metal [Thomason, 1969; Narasimhan
and Kamat, 1994]. The large particles fail by cracking while the smaller ones fail by
interfacial decohesion [Van Stone et al., 1985]. That is why unbroken small
particles were present at the fracture surface, while the pearlite colonies were
cracked.

The growth and coalescence of voids depend on the flow instability of the
matrix metal. During stable plastic flow, there is little potential for ductile fracture.
Only when the tensile or shear instability condition of the matrix is reached does
ductile fracture progress rapidly. The strain in the matrix from the moment of
instability to fracture depends primarily on the proportion of second phase particles
[Thomason, 1969]. The greater the second phase fraction, the lower the fracture
strain. The HR microstructure had the highest proportion of the second phase;
therefore it displayed the lowest fracture strains. By contrast, the HT-II material
had the highest fracture strain and the lowest amount of second phase.

Although a few sulphide particles were found on some of the fracture

surfaces, EDS analysis revealed that most of the second phase particles
responsible for fracture in the DWT tests were carbides. In spheroidized
microstructures, such as the HT-II and HT-III, the dimensions of the carbide
particles were small as were their aspect ratios. This resulted in lower stress
concentrations and higher ductility of the matrix. In relatively unspheroidized
microstructures, such as those of the HR, HT and HT-IV samples, large amounts
of the cementite were lamellar and of uneven dimensions. The platelet particles
caused high stress concentrations during DWT; this resulted in early fracture. This
is also the reason why microstructures containing spheroidal cementite possess
better cold headability than those containing pearlite.
Chapter 6

Conclusions

1. The tests indicate that the drop weight test is able to differentiate between
the cold headabilities of samples containing different microstructures.
Therefore it is a valid method for assessing the suitability of materials for
cold heading.

2. The cold headability of medium carbon steel is sensitive to the difference

between the mechanical properties of the ferrite and the pearlite colonies.
Therefore, it can be improved by reducing this disparity by means of heat
treatment.

3. Cold headability is also sensitive to the inclusions and second phase

particles contained in the microstructure. When such particles are present
near the external surface of CHQ rod, they act as surface defects and lead
to the nucleation of cracks.

4. The present results indicate that a totally spheroidized microstructure (the

HT-II) has the best cold headability amongst the six tested microstructures.
 References

Aihara, K. and Kanbara, S., 1990, "Influence of Prior Structure upon

Spheroidization Rate and Cold Forge Ability of Annealed Wire", Sumitomo
Search, No. 42. pp. 1-8.

Aihara, K., 1991, "A New Thermomechanical Processing for Spheroidizing

Carbide Directly in a Rolling Line", 33rd Mechanical Working and Steel
Processing Conference Proceedings, The Iron and Steel Society, Inc. (USA), pp.
285-290.

Baquet, I.G., Kaspar, R., Richter, J., Nussbaum, G. and Koethe, A., 1997,
"Influence of Microalloying on the Mechanical Properties of Medium Carbon
Forging Steels after Newly Designed Post Forging Treatment", Steel Research,
No. 12, pp. 534-540.

Bickford, J. and Nassar, S., 1998, "Handbook of Bolts and Bolted Joints",
M. Dekker, pp. 69-74.

Bruns, H. and Kaspar, R., 1997, "Pearlite Spheroidization by

Thermomechanical Treatment of Directly Charged Thin Slabs", Steel Research,
Vol.68, No. 4, pp. 158-162.
References 77

Chattopadhyay, S. and Sellars, CM., 1977, "Quantitative Measurements

of Pearlite Spheroidization", Metallography, Vol. 10, No. 1, pp. 89-105.

Cockcroft, M.G. and Latham, D.J., 1968, "Ductility and Workability of

Metal", Journal of the Institute of Metals, Vol. 96, pp. 33-39.

Cox, T.B. and Low, J.R. Jr, 1974, "Investigation of the Plastic Fracture of
AISI 4340 and 18 Ni-200 Grade Maraging Steels", Metall. Trans., Vol. 5, pp.
1457-1470.

Das, T , Li, J.Y., Painter, M. and Summerville, E., 2002, "Evaluation of

Two AISI 4037 Cold Heading Quality Steel Wires for Improved Tool Life and
Product Quality", Journal of Materials Engineering and Performance (USA), Vol.
11, pp. 86-91.

El-Domiaty, A., 1999, "Cold Workability Limit for Carbon and Alloy Steels",
Journal of Materials Engineering and Performance, Vol. 8, No. 2, pp. 171-183.

Fisher, J.R. and Gurland, J., 1981, "Void Nucleation in Spheroidized

Carbon Steels", Metal Science, No. 5, pp. 185-192.

Floreen, S. and Hayden, H.W., 1970, "Some Observations of Void Growth

During the Tensile Deformation of a High Strength Steel", SCR Met. Vol. 4, pp.
87-94.

Goods S.H. and Brown, L.M., 1979, "The Nucleation of Cavities by Plastic
Deformation", Acta Metall. Vol. 27, No. 1, pp. 1-15.

Gurland, J., 1972, "Observation of the Fracture of Cementite Particles in a

Spheroidized 1.05%C Steel Deformed at Room Temperature", Acta Metall., Vol.
20, pp. 735-741.
 78
References —

Gurland, J., 1988, "A Review of the Application of the Law of Mixtures to
the Plastic Deformation of Two Phase Alloys", Israel Journal of Technology, Vol.
24, pp. 243-253.

Jenner, A. and Dodd, B., 1981, "Cold Upsetting and Free Surface
Ductility", J. Mech. Work. Technol, Vol. 5, pp. 31-43.

Kohlmann, R., Kruse, M., Meyer, M. and Plociennik, U., 2000,

"Optimization of Material Properties for Bars and Wire Rod", Metallurgical Plant
and Technology Int., Vol. 23, No. 4, pp. 56-62.

Koyama, R., Higurashi, M., Fukusima, S. and Hagita, H., 1995, "Non-
Heat-Treated Steel for Cold Forging Application", Recent Development of Rolling
and Following Process Technology and/or Application of Steel Products, Penang,
Malaysia, Session 10, pp. 4/1-4/15.

Kruse, M., Mauk, P.J. and Plociennik, U., 1996, "Temperature Control
Conception for Thermomechanical Rolling in Bar and Wire Rod Mills-Plant
Technology, Computer Simulation and Results", Metallurgical and Foundry
Engineering, Vol.22, No. 3, pp. 183-198.

Kuhn, H.A., 1977, "Workability Testing and Analysis for Bulk Forming
Process", Formability Topics - Metallic Materials, Toronto, Canada, American
Society for Testing and Materials, pp. 206-219.

Lee, D.L, Lee, J.B., Kim, J.K. and Choi, J.H., 1992, "Steel Wire Rod for
Cold Heading without Heat Treatment", SEAISI Quarterly (Malaysia), Oct., pp.
63-74.

Lee, P.W. and Kuhn, H.A., 1973, "Fracture in Cold Upset Forging - a
Criterion and Model", Metallurgical Transactions Vol. 4, pp. 960-974.
References 79

Lehnert, W., 1995, "Production and Use of High Strength Cold Heading
Steels", Wire (Germany) Vol. 45, No. 5, pp. 275-280.

Magnusen, P.E., Dubensky, E.M. and Koss, D.A., 1988, "The Effect of
Void Arrays on Void Linking During Ductile Fracture", Acta Metall., Vol. 36, p.p.
1503-1509.

Maheshwari, M.D., Dutta, B. and Mukherjee, T., 1978, "Quality

Requirement for Cold Heading Grades of Steel", Tool Alloy Steel, Vol. 12, p.p.
247-251.

Matsunaga, T. and Shiwaku, K., 1980, "Manufacturing of Cold Heading

Quality Wire Rods and Wires", SEAISI, Quarterly, Vol. 9, No. 1, pp. 45-55.

Mazanec, K. and Mazancova, E., 1997, "Physical Metallurgy of

Thermomechanical Treatment of Structural Steels", Cambridge International
Science.

ASM Metals Handbook, 1992, "Interpretation of Scanning-Electron

Microscope Fractographs", 8th ed., Vol. 9, pp. 64-78.

Muzak, N., Naidu, K. and Osborne, C , 1995, " New Methods for
Assessing Cold Heading Quality", Wire Association International 65th Annual
Convention, pp. 104-111.

Narasimhan, R. and Kamat, S.V., 1994, "A Numerical Investigation of

Ductile Fracture Initiation in a High Strength Low Alloy Steel", Bulletin of
Materials Science, Vol. 17, No. 3, pp. 259-282.
 80
References

Nickoletopoulos, N., 2001, PhD Thesis, "Physical and Numerical Modeling

of Steel Wire Rod Fracture During Upsetting for Cold Heading Operations",
McGill University.

Nickoletopoulos, N., Hone, M., Wanjara, P., Verreman, Y., Nemes, J.A.,
Jonas J.J., and Yue, S., 2001, "Development of Fracture Criterion for Cold
Heading", Wire Journal International, Vol. 34, No. 9, pp. 90-98.

O'Brien, J.M. and Hosford, W.F., 2000, "Spheroidizing of Medium Carbon

Steels", Industrial Heating, Vol. 67, No. 9, pp. 79-82.

Ochi, T. and Koyasu, Y., 1992, "A Study of Spheroidizing Mechanism of

Cementite in Annealing of Medium Carbon Steel", 33rd Mechanical Working and
Steel Processing Conference Proceedings, The Iron and Steel Society, (USA),
pp. 303-309.

Olsson, K., Karlsson, S. and Melander A., 1986, "The Influence of Notches,
Testing, Geometry, Friction Conditions and Microstructure on the Cold Forge
Ability of Low Carbon Steel", Scandinavian Journal of Metallurgy, Vol. 15, No. 5,
pp. 235-256.

Peisker, D., Doktorowski, A., and Dittrich, D., 2001, "Contribution on the
Influence of Transformation Structure Under Consideration of Austenite
Deformation", Steel Research, Vol. 72, No. 3, pp. 111-114.

Pickering, F.B., 1980, "Some Aspects of the Relationships between the

Mechanical Properties of Steels and Their Microstructure", TISCO, Vol. 27,
pp.105-132.

Pittinato, G.F. (editor), 1975, "SEM/TEM Fractography Handbook",

McDonnell Douglas Astronautics Company.
References 81

Poruks, P., Wilkinson, D.S. and Embury, J.D., 1998, "Role of Particle
Distribution on Damage Leading to Ductile Fracture", ASM International, Analysis
of In-Service Failures and Advances in Microstructural Characterization, Vol. 26
(USA), pp. 491-496,

Ray, A., Paul, S.K. and Jha, S., 1996, "Effect of Inclusions and
Microstructural Characteristics on the Mechanical Properties and Fracture
Behavior of a High-Strength Low-Alloy Steel", Journal of Materials Engineering
and Performance (USA), Vol. 4, No. 6, pp. 679-688.

Ren, C. and Varo, R., 1999, "Controlled Cooling Process in the Tianjin
High Speed Wire Rod Mill", Metallurgical Plant and Technology Int., Vol. 22, No.
4, pp. 88-93.

Sarruf Y., 2000, "Criteria and Tests for Cold Headability", Master's thesis,
McGill University.

Sarruf, Y., Cao, B.H., Jonas, J.J. and Nickoletopoulos, N., 1998, "Criteria
and Tests for Cold Headability", 68th Annual Convention of the Wire Association
International, Wire Association International Inc., pp. 141-148.

Shah, J.J. and Kuhn, H.A., 1986, "An Empirical Formula for Workability
Limits in Cold Upsetting and Bolt Heading", J. Appl. Metalwork., Vol. 4, No. 3, pp.
255-261.

Shockey, D.A., Curran, D.R. and Seamen, L., 1977, "Computer Modeling
of Microscopic Failure Processes Under Dynamic Loads", Proc. Conf. of High
Velocity Deformation of Solids, Tokyo, Japan, pp. 149-162.
 82
References

Shore, S.M., Rixham, B. and Keyzer, PL., 1998, "Rod Mill Improvement for
Increased Profitability", 40 th Mechanical Working and Steel Processing
Conference, Iron and Steel Society, Vol. 36, pp. 653-661.

Sun, Z.Q., Yang, W.Y., Hu, A.M., and Yang, P., 2001, "Deformation
Enhanced Ferrite Transformation in Plain Carbon Steel", Acta Metall., Vol. 14, No.
2, pp. 115-121.

Thibau, R., Trudel, A., Baragar, D. and Hastings, P.R., 1999,

"Development of a Test to Evaluate the Formability of CHQ Wire", Wire
Association Int. 69th Annual Convention, Wire Association Int., pp. 232-241.

Thomason, P.F., 1968, "The Use of Pure Aluminium as an Analogue for

History of Plastic Flow in Studies of Ductile Fracture Criteria in Steel
Compression Specimens", Int. J. Mech. Sci. 1968, Vol. 10, pp. 501-518.

Thomason, P.F., 1969, "Tensile Plastic Instability and Ductile Fracture

Criteria in Uniaxial Compression Test", Int. J. Mech. Sci., Vol. 11, pp. 187-198.

Thomason, P.F., 1969/1970, "The Effect of Heat Treatment on the

Ductility of Alloy Steel Wires in a Cold Heading Process", Proc. Inst. Mech. Engrs,
Vol. 184, Pt. 1, No. 47, pp. 875-884.

Thomason, P.F., 1969B, "The Free Surface Ductility of Uniaxial

Compression Specimens with Longitudinal Surface Defects", Int. J. Mech. Sci.,
Jan., pp. 65-73

Trufiakov, V.I., Kirjan, V.I. and Novikova, D.P., 1992, "Relationship

between the Structure and Ductile Fracture Resistance of Ferritic-Pearlitic
Structure Steel", Soviet Material Science, Vol. 28, No. 3, pp. 289-294.
References 83

Tsai, D.H., Jou, D.S., Hwang, I.S., Lin, H.R. and Guo, Y.T., 2000,
"Development of Microalloyed Extra Low Carbon Steel Bars for Severe Cold
Heading Quality", SEAISI Quarterly, Jul., pp. 19-34.

Van Stone, R.H., Cox, T.B., Low, J.R. and Psoda, J.A., 1985,
"Microstructure Aspects of Fracture by Dimpled Rupture", Int. Met. Rev., vol. 30,
No. 4, pp. 157-179.

Wanjara, P., Donaberger, R., Hone, M., Nemes, J.A., Nickoletopoulos, N,

Tendler, A. and Yue, S., 1999, "Development of Application of Dual Phase Steels
for Cold Heading High Strength Steel Fasteners", AIME 41 s t Mechanical Working
and Steel Processing Conference, pp. 151-157.

Wick, C.H., 1960, "Chipless Machining", New York Industrial Press, pp.
34-71.

Wu, F., Liu, J., and Hu, J., 1999, "Rapid and Direct Spheroidization of
Carbide in Hypoeutectoid Steels", Journal of National University of Defense
Technology, Vol. 21, No. 6, pp. 27-31.

Yoo, S.J., Lee, D.L. and Chung, D.T, 1997, "Prediction of Cold Heading
Quality with High Strain Rate Compression Test", Wire Journal Int., Vol. 30, No. 9,
pp. 84-89.