THE GRAPHITISATION
PROCESS IN MEDIUMCARBON STEEL
David V Edmonds and Kejian He
Institute for Materials Research,
University of Leeds, Leeds, UK
SolidSolid Phase Transformations
Research Programme Objectives
To study graphitisation in steels, with the
overall aim of identifying a new route to the
development of a plain carbon cold-forging
steel with good machinability.
To accelerate the kinetics of graphitisation.
To examine the mechanism of graphite
formation in steels by high-resolution
microanalytical electron microscopy.
SolidSolid Phase Transformations
Outline of Presentation
Background to overall research programme
objectives.
Choice of steel alloying - thermodynamic
modelling.
Examination of graphite nodule formation
using transmission electron microscopy
(TEM), electron energy loss spectroscopy
(EELS) and energy-filtered transmission
electron microscopy (EFTEM).
SolidSolid Phase Transformations
Machining Operation
(after Metal Cutting, EM Trent, Butterworths, 1977)
SolidSolid Phase Transformations
Traditional Free-Machining
(Free-Cutting) Steels
The world market is very large from threaded
screws and bolts to accurately machined
components e.g. in the automotive industry.
Plain carbon steels are alloyed with elements such
as Pb, S, Te, Bi and P to act as a lubricant at the
tool/workpiece interface, and additionally, to assist
with chip break-up.
Disadvantages can be toxicity, impairment of cold
forgeability, and, at least for Te and Bi, steel that is
more expensive and difficult to recycle.
SolidSolid Phase Transformations
Free-Machining Plain Carbon Steels
An Alternative
Why not anneal plain carbon steels to exchange cementite for
graphite?
The presence of graphite (rather than cementite) in the
microstructure of a plain carbon cold-forging steel would act as
an internal lubricant during machining, and also assist chip
break-up, as in the case of grey cast irons which have
customarily exhibited excellent machinability. Forgeability
should also be improved.
However, the annealing time required to convert cementite to
graphite in steels has traditionally been too long (of the order of
100+ hours) to integrate successfully into a production heat
treatment schedule.
Thus, consider whether graphitisation can be influenced by
alloying, to reduce these heat treatment times.
SolidSolid Phase Transformations
Thermodynamic Assessment of Alloying
The effect of various alloying elements on graphitisation were
evaluated in terms of the driving force for the precipitation of
graphite in an Fe-0.54wt%C system at 680C using Thermo-Calc.
Si and Al have
equivalent effects.
strong
and
roughly
Ni and Co are similar but less strong.
Cu has a positive effect but is much weaker.
Mo is mildly suppressive, Mn more so.
Cr is strongly suppressive.
SolidSolid Phase Transformations
Experimental Steels
Alloying Philosophy
Increase Si and Al alloying
Minimise Mn alloying
(Avoid alloying with Cr)
C
Si
Steel 1 0.47 0.19
Al
Mn
Ni
0.32 1.5
Time for
graphite to
appear in
microstructu
re at 680C
(hours)
Approx.
time for
completion
of
graphitisati
on at 680C
(hours)
18
115
Note faster graphitisation
in0.5
Si-Al steel.
-
Steel 2 0.38 1.82 1.44 0.07
3.5
SolidSolid Phase Transformations
Light Microscopy
Steel 2 : Si-Al steel
Steel 1
Annealed 115 hours
Annealed 3.5 hours
Light micrographs of steels 1 & 2, austenitised at 1150oC,
quenched to martensite, and annealed at 680oC.
SolidSolid Phase Transformations
Experimental further examination of
graphitising Si-Al steel (Steel 2)
As-quenched (martensite)
Quenched and annealed 0.5 hours
50 m
50 m
Graphitisation has started
SolidSolid Phase Transformations
Si-Al steel (Steel 2)
0.5 hours
3.5 hours
50 m
50 m
55 hours
Progress of
graphitisation at
680C.
50 m
SolidSolid Phase Transformations
Transmission Electron Microscopy
0.5m
0.5m
5 m
Bright field TEM image
showing coarsened cementite
particles located mainly at the
interfaces of ferrite laths after
0.5 hours at 680.
Bright field TEM image showing
only graphite nodules present in
a more equiaxed ferrite structure
after 1.5 hours at 680.
SolidSolid Phase Transformations
Further Electron Microscopy
Observations Particle Nucleation
G
Nucleation of graphite (G)
on an aluminium oxide
inclusion (O) after 0.5
hours. Note also the
carbide particle dispersion
and remnants of the
martensite structure.
Nucleation on an AlN particle. Note the
irregular
graphite
morphology
and
structure.
The inner ring of the diffraction pattern is
(002) graphite, and the single crystal
reflections are from the [111] zone of AlN.
SolidSolid Phase Transformations
Graphite Nodule Morphology
5 m
Small, ~4m diameter, regular
spheroidal graphite nodules,
apparently without a coring
oxide or nitride particle, after 1.5
hours at 680C.
____
0.5m
SolidSolid Phase Transformations
Graphite (002) Lattice Fringes Within
Conical Segments
SolidSolid Phase Transformations
Structure of Spheroidal Nodules
BF image
DF image: non-diametric
section
DF image: diametric section
Schematic diagram
illustrating cone-helix
growth model for
graphite nodules in
cast irons [after DD
Double and A
Hellawell, Acta
Metall., 22(1974)481].
SolidSolid Phase Transformations
(002) Lattice Fringes in a Spheroidal
Graphite Nodule
Central region of nodule (BF image)
Away from centre
Near centre
SolidSolid Phase Transformations
High-Resolution Images - Pitch
Graphitization Series
200oC
All images
show (002)
fringes.
600oC
2730oC
1200oC
2000oC
Howard
Daniels,
IMR,
Leeds
University
SolidSolid Phase Transformations
EELS- Shifting of the + Plasmon as a
Function of Graphitisation
lowloss2
Plasmon Position / HTT
70
26.5
26
25.5
25
24.5
24
23.5
23
60
50
Plasmon
Energy (eV)
40
30
1000
2000
3000
HTT (oC)
20
10
0
-0
10
20
30
Energy Loss (eV)
40
50
Plasmon peak energy closely
follows the change in density.
Not a perfect match due to the
effect of crystallite size.
SolidSolid Phase Transformations
EFTEM Plasmon Mapping
By taking the ratio of the intensities in the windows shown for a
large number of different plasmon positions, it is possible to
calibrate the intensity in the image.
27/22ev using 3ev windows
lowloss2
70
60
1.5
50
40
30
0.5
20
0
Ratio
27/22ev
10
0
-0
10
20
30
Energy Loss (eV)
40
50
21
23
25
27
Plasmon energy (eV)
SolidSolid Phase Transformations
29
EFTEM Plasmon Imaging of a Graphite
Nodule
23 eV
24 eV
25 eV
26 eV
0.5m
27 eV
Bright Field ( montage of 25 images)
Plasmon (27eV/22eV) Ratio Map
(montage of 25 images)
Plasmon ratio map suggests a more amorphous core.
SolidSolid Phase Transformations
Coarsened Carbide Particles
0.5m
100nm
SolidSolid Phase Transformations
BF TEM Images of Surviving Carbides
100nm
Annealed 50 min.
Annealed 58 min.
Annealed 58 min.
Are these carbide particles?
SolidSolid Phase Transformations
Carbon K-edge ELNES Pitch
Graphitization Series
As order within the carbon
increases, the electronic structure
follows suit, resulting in higher
definition of the unoccupied states.
750oC
1500oC
Howard Daniels,
IMR, Leeds University
2730oC
SolidSolid Phase Transformations
Carbon K edge EELS Spectra
Fe3C
1.7
centre
1.2
centre1
Counts
0.7
edge 1
edge
0.2
680 graphite
-0.3
280
285
290
295
300
305
310
315
320
325
330
Energy Loss (eV)
EELS spectra collected from coarse particles, and cementite and
graphite for comparison.
Carbon content - 30 atom% in crystalline cementite part, 70 atom%
in amorphous part.
SolidSolid Phase Transformations
EFTEM Jump Ratio Images of a Particle
C
Fe
Mn
The particle is not simple
cementite it consists of
crystalline cementite and a
more amorphous part.
Remaining cementite
TEM BF image; C K- jump ratio image; O K- jump ratio image;
Fe L2,3- jump ratio image and Mn L2,3 - jump ratio image.
SolidSolid Phase Transformations
Jominy Bar Analysis after Annealing for
6 hours at 680C
600
550
500
450
400
350
300
250
Hardness
(HV30)
200
150
100
0
5
Martensite
Martensite
Lower bainite
Lower
bainite
Upper bainite / acicular ferrite
Upper
bainite
Si-Al
/ acicular ferrite
Si-Al-B
Ferrite + pearlite /
Widmanstatten ferrite
Ferrite +
pearlite
Widmanstatten
10
15
Si-Al 6hr
ferrite
20
25
30
Distance from quenched end (mm)
Different graphite nucleation kinetics
and
microstructural
dispersions
result
from
different
starting
microstructures, possibly related to
the different routes for carbide
formation
between
martensite,
bainite and pearlite in the Si-Al
experimental steels.
160
1200
140
1000
120
800
Si-Al
600
Si-Al-B
400
100
Si-Al
80
Si-Al-B
60
40
200
20
0
0
0
10
15
20
10
15
SolidSolid Phase Transformations
20
Conclusions
Graphitisation of carbon steels in the tempered martensitic condition can
be achieved after short annealing times (2-3 hours) by alloying.
Regular spheroidal nodules appear to consist of cone-shaped segments
radiating from a central core.
Within the cones the circumferential stacking of the graphite layers during
growth is very regular, equivalent to that which can be achieved in the
graphitisation of carbonaceous materials only at temperatures around two
thousand degrees higher.
Microanalysis by high-resolution TEM suggests that, in the experimental
steels, either cementite dissolution is accompanied by loss of crystallinity
and the formation of amorphous regions, or these regions form on the
decomposing cementite.
Observations of a more amorphous centre to the small spheroidal
nodules (lacking an obvious nucleating particle), suggests that the
amorphous carbon regions associated with the decomposing cementite
may be the nuclei for these graphite nodules an intermediate stage in
graphite nodule formation.
SolidSolid Phase Transformations
