Ferry M. Direct Stripcasting of Metals and Alloys: Processing, Microstructure and Properties
Подождите немного. Документ загружается.
As-cast
microstructure,
texture
and
properties
193
5.4.4 Other non-ferrous alloys
Titanium
alloys
The processing of titanium foils
by
conventional ingot metallurgy involves
casting ingots,
hot
forging into billets followed
by
several
hot
rolling, heat
treatment and surface grinding sequences to produce plate or sheet that is
suitable for cold rolling into thin-gauge sheet (Collings
1984). Gapsar
et
al.
(1991, 1994) and Weaver and Garmestani (1998) have successfully cast
commercial purity titanium, titanium alloys
and
Ti-base intermetallics into strip
form using a single-roll casting technique termed melt overflow rapid
solidification technology
(MORST).
These workers noted several potential
advantages of
DSC
over conventional ingot metallurgy routes such as
improved purity, increased chemical homogeneity and a reduction in
processing losses resulting in reduced overall costs.
Figure 5.40. Solidification texture of commercial purity titanium strip produced
by
single roll casting, after Weaver
and
Garmestani (1998)
(with kind permission of Elsevier Limited).
As-cast strip of 0.5
mm
thickness and 100
mm
width has been produced
by
the
MORST
process. The as-cast foils were found to exhibit microstructures,
textures and mechanical properties comparable to those produced
by
conventional routes. The as-cast microstructure of a commercial purity
titanium alloy consists largely of large columnar grains originating from the
surface of the roll with grain diameters of
35 to
200
)lm
and
some equiaxed
grains forming near the free surface of the strip. Due to the allotropic
transformation
in
titanium (§1.3.3), the structure transforms
on
cooling to a
mixture of acicular
and
serrated a phases. The solidification texture produced
on the wheel surface of the caster is shown
in
Figure 5.40 as ¢ = 0
0
and
¢ = 30
0
sections
in
Euler space
(Roe
notation). The final texture is very weak and
consists mainly of a TO-rotated
{1120}
/ /NO texture
but
there is evidence of a
tube of orientations from
(1010)[1210] to (1018)[0110]. This weak texture is
194
Direct
strip
casting
of
metals
and
alloys
likely to be a result of the
fJ
~
a transformation that occurs
in
commercial
purity titanium during cooling.
Nickel
alloys
The limited work
on
DSC
of nickel alloys indicates that the most popular
process is
TRC
(Yukumoto and Yamane
1995;
DaIle
et
al.
2003).
For example,
Yukumoto
and
Yamane
(1995)
twin roll cast Ni-8%Fe-17%Cr (Inconel
606)
to
produce strip of
0.2
to
0.8
mm
thickness and
250
to
500
mm
width for the
production of overlay welding hoops. In order to suppress precipitation and
other unwanted structural changes after casting, the strip was subjected to
water spray cooling
at
300-350°C/s on leaving the caster and coiled at -400°C.
The resulting microstructure is similar to that of other strip-cast alloys that
do
not
undergo any allotropic transformations
and
consists of a fine columnar
grain structure almost parallel to ND. These grains exhibit a characteristic as-
cast dendritic structure with
SDAS
varying from 2 to
3.5
!lm at the surface to
the core of the strip.
Palladium
alloys
Korzekwa
and
co-workers
(1992)
carried
out
melt spinning experiments
on
Pd-
Ni alloy using the apparatus shown in Figure 3.16a. The results are similar to
those found for
DSC
of other cubic metals that
do
not undergo further solid-
state transformations. The melt-spun ribbons exhibited a near-random texture
at the strip surface that intensified
due
to the orientation selection mechanism
described
in
§2.2.2.6 to develop the classic TO-rotated
<001>
fibre texture with
dendrites inclined
-20° from ND (Figure
5.41).
4000
~
III
jj 3000
.5
i
.!!l
~
2000
o
I::
1
1000
o
U
20
40
60
80 .
Angle from ribbon normal (0)
Figure 5.41. Intensity distribution of grains rotated about TD towards CD
in
a melt-
spun
Pd-Ni alloy, after Rollett
and
Wright
(1998)
(with kind permission
of Cambridge University Press, UK).
6.1
Introduction
Chapter
6
Secondary Processing
and
Fabrication
Depending
on
the nature of the process, strip casting produces material with a
gauge thickness of over
10
mm
to less
than
0.8
mm
(§3.2.3). In
twin
roll casting,
thin gauge strip « 2
mm
for steel
and
2-12
mm
for aluminium) is produced
at
high solidification rates to generate a range of non-equilibrium microstructures
(chapter 5). In moving belt/block casting processes such as the Hazelett (Figure
3.9)
and
Alusuisse caster II (Figure 3.10), the as-cast strip is
up
to
40
mm
thick
and
the solidification rates are lower than TRC. Nevertheless, a common
feature of all materials produced
by
high speed casting is the presence, to some
degree, of micro-
and
macrosegregation, porosity
and
undissolved second-
phase constituents. The as-cast strip therefore has a particular microstructure
and
surface condition which
mayor
may
not
be
suitable for immediate
fabrication into the desired component.
It
may
be
necessary to further modify
the strip to some degree
by
hot
and/or cold rolling
and/or
thermal processing in
order
to alter the strip dimensions, internal structure
and
surface condition.
This chapter provides
an
overview of these secondary processing routes
and
discusses other possible options for structural/surface modification to enhance
the final properties of the material.
6.2
Scope
of
secondary processing
One of the critical challenges for
near
net
shape casting (NNSC) processes such
as DSC is the restricted secondary processing
window
available for modifying
195
196
Direct
strip
casting
of
metals
and
alloys
the as-cast microstructure
and
surface condition of the strip. For example,
conventional continuous casting produces thick slab that enables the alloy to be
homogenised,
hot
and
cold rolled
and
annealed
at
various stages of the process
to generate the desired final microstructure
and
properties of thin-gauge strip.
While NNSC processes such as
TSC
are restricted to a few
hot
rolling passes,
the as-cast material is sufficiently thick to allow substantial structural
modification. In the extreme case of direct casting of thin-gauge strip, both
economic considerations
and
strip dimensions tend to restrict the
number
and
type of secondary processes that are achievable.
Nevertheless,
it
is possible to modify the microstructure
by
one
or
more
relatively simple thermal treatments that
may
allow the material to be
fabricated directly
or
prepared for further downstream processing if a more
homogeneous microstructure is required. More sophisticated processes such as
in-line rolling are useful for adjusting the strip thickness, surface condition and
profile as well as modifying the as-cast microstructure. The narrow processing
window
becomes evident for DSC of low carbon steel since solidification
generates a coarse columnar austenite grain structure (Table 5.2). Complete
recrystallization of the austenite during in-line rolling is difficult to achieve
and
an
understanding of the effect of processing parameters, composition
and
austenite grain morphology on the recrystallization kinetics is a pre-requisite if
the limited downstream processing opportunity
in
DSC is to be optimised.
Another important secondary processing route to produce final-gauge sheet is
cold rolling
and
annealing. Depending
on
the type of strip casting process
and
the final application of the material, one or more rolling
and
annealing
processes may be required to generate strip products
with
mechanical
properties comparable to those obtained via conventional production routes.
6.3
Control
of
thermal path after strip casting
6.3.1 Direct cooling and coiling
There are a few reports of structure control of strip-cast low carbon steels
by
direct cooling
and
coiling (Kopp
et
al.
1998; Blejde
et
al.
2000a;
Mukunthan
et
al.
2000).
With reference to the transformation behaviour of low carbon steel,
austenite decomposition is governed
by
the cooling
path
through the critical
transformation range
and
structures ranging from ferritic to martensitic are
possible. Figure
6.1
is a transformation diagram of strip-cast 0.36% C steel
showing the possible phases generated after cooling from the austenite phase
field. The type of decomposition product
and
its volume fraction
in
the final
microstructure is governed
by
the composition of the steel, morphology of
austenite
and
cooling
path
(§1.2.3.1). Superimposed
on
the CCT diagram in
Figure
6.1
are several cooling paths which illustrate the effect of coiling
Secondary
processing
and
fabrication
197
temperature on phase formation. Low cooling rates
and
high coiling
temperatures (paths 1
and
2)
generate polygonal ferrite ( a )
and
pearlite
(P)
that
become increasingly refined as either the cooling rate is increased or coiling
temperature is decreased (path
2).
High cooling rates
and
low coiling
temperatures (path
3)
may generate a microstructure containing ferrite, pearlite,
bainite
(B)
and
martensite (M). In the extreme case (path
4),
rapid cooling to
below
400°C will generate a martensitic structure.
1000
,....----.:--------------.....,
-t
•••.
~
".
:~
".
Fe-O.36C-O.64Mn-O.27Si
~
\
.....
.
800F-----~·~~.---~~~~~------------------~
~
\
.............
.
600
400
Ms
M
200
t
Pinch roller
0
0.1
.
....
'.
'.
.............
.
.•..
......
'.
......
.
.....
'.
'.
e.
••••
e,
....
~
.............................
..........
........
'.
\.!)
'.
"'®
.....
'®
t
Coiler
10
100
1000
Time (8)
10000
Figure
6.1.
CCT diagram of 0.36 wt.% C steel showing the influence of cooling rate
and
coiling temperature
on
final microstructure, after Kopp
et
al.
(1998)
(with kind
permission of Elsevier Limited).
The general trends shown in Figure
6.1
have been confirmed by the strip-
casting/cooling
path
experiments
by
Mukunthan and co-workers
(2000).
Table
6.1
is a summary of plant trials
on
twin roll cast
0.06%
C silicon-killed steel in
which cooling rates and coiling temperatures were controlled to generate a
wide range of microstructures and properties.
It
is clear that controlled
cooling/coiling has the capability of transforming the coarse columnar austenite
grain structure into a wide range of useful microstructures. The final
microstructure is altered further by:
(i)
in-line
hot
roIfing to deform or
recrystallize the austenite
(§6.4.2);
(ii) alloying additions that tend to modify the
CCT diagram
(§5.3.1.3);
(iii) directly refining the austenite during solidification
using special alloying additions such as
S or
Te,
and (iv) texturing the mould
surfaces. Finally, cold rolling
and
annealing may be used to further extend the
range of properties.
198
Direct
strip
casting
of
metals
and
alloys
Table
6.1.
Effect of cooling rate
and
coiling temperature
on
microstructures
and
properties of
IRC
low carbon steel strip, after Mukunthan
et
al.
(2000)
Cooling
rate
(K/s)
0.1
13
25
100
Coiling
Microstructure
Yield
strength
temperature
(0C)
constituents
>800
670
580
<400
O
Residuals
or
fast cooling
Polygonal ferrite, pearlite
Polygonal ferrite,
Widmanstatten ferrite
Polygonal ferrite, bainite
Polygonal ferrite, bainite,
martensite
O
MOderate
cooling
~ow
temperature
~
hot rolling
20
~Aircooling
~
(typical)
Slow
~
cooling Hot rolling +
~
cold rolling and
o annealing
Cold rolling
~
and
annealing
"-----'"
30
Elongation
(%)
40
o
C = 0.08/0.1%
Mn =
1.411.5%
Ti
= 0.0110.03%
Nb = 0.04/0.05%
Nb = 0.03/0.04% C = 0.14/0.17%
Mn = 0.710.8%
C=0.08/0.11%
9
Mn = 0.5/0.6%
Drawing quality
~
semi-killed steel
C=0.09/0.17%
~
20
Mn
= 0.5/0.75%
~
30
Elongation
(%)
40
(MPa)
210
320
390
490
50
50
Figure
6.2.
(a)
Influence of processing parameters
on
the strength
and
ductility of strip-
cast low carbon steel produced from a single steel chemistry. (b) Alloying
requirements to generate the same range of properties as that
given in (a), adapted from Mukunthan
et
al.
(2000).
Secondary
processing
and
fabrication
199
The effect of processing parameters on the yield strength
and
elongation of
strip-cast low carbon steel is illustrated
in
Figure 6.2a which shows the
remarkable range of properties achievable using a single base alloy
composition. This figure may be compared with a similar gauge
hot
band
material produced
by
conventional slab casting
and
secondary processing
(Figure 6.2b), which shows that a wide the range of alloying additions are
needed to generate the same combination of mechanical properties.
6.3.2 Direct heat treatment processes
6.3.2.1
Austenitisation treatments
A simple method of altering the microstructure in strip-cast iron (and titanium)
alloys is to make use of the allotropic transformations that occur during
reheating and subsequent cooling. This heat treatment, termed austenitising
in
the ferrous industry, is a simple and inexpensive mode of structural alteration
with a range of final microstructures achievable
by
controlling the chemistry,
austenitising temperature and cooling rate through the critical transformation
temperature range.
Other heat treatment such as sub-critical
and
inter-critical
annealing can also alter the microstructure to some degree
(Shiang
and
Wray
1989;
Sasiain
1998).
1200
1100
IT
'L...1000
~
:::J
T§
900
~
E
800
~
700
600
0.1
FERRITE
MORPHOLOGY
W =
Widmansatten
IS
=
Irregularty
shaped
FG
=
Fine
grained
CG
=
Coarse
grained
•
10
100
1000
10000
Time
(8)
Figure
6.3.
Processing
map
showing the final microstructure
in
strip-cast AI-free low
carbon steel, after
Shiang
and
Wray
(1989)
(with kind permission of
The Metals, Minerals
and
Materials Society, USA).
Shiang
and
Wray
(1989)
carried
out
a study of the effect of reheating on
microstructural development of strip-cast low carbon steels. The effect of post-
cast heat treatment
on
microstructural development of AI-free low carbon steel
is summarised in Figure
6.3.
Sub-critical annealing followed
by
air cooling
alters the microstructure slightly with the gradual formation of large grains of
200
Direct
strip
casting
of
metals
and
alloys
polygonal ferrite (not shown in Figure
6.3)
dispersed in a Widmanstatten ferrite
microstructure. Inter-critical annealing followed
by
air cooling generates a
combination of fine- and coarse-grained ferrite together with Widmanstatten
ferrite. However, water quenching generates some fine-grained martensite
due
to the partial transformation of ferrite to austenite during heating. During the
full austenitisation treatment, time, temperature and cooling rate have a
substantial influence on the final ferrite microstructure (Sasiain
1998).
An
increase in the austenitising temperature generates a coarser austenite grain size
which,
on
subsequent cooling, transforms to produce a microstructure ranging
from fine polygonal to Widmanstatten ferrite.
Figure
6.4.
Final microstructures in heat treated twin roll cast low carbon steel after
austenitising for 1h at: (a)
927°C
and
(b) 1093°C followed
by
air cooling, after Sasiain
(1998).
Figure
6.4
shows some microstructures produced in strip-cast C-Mn steel by
simply controlling the austenitising temperature. A low austenitising
temperature generates fine-grained polygonal ferrite
but
both Widmanstatten
and coarse-grained polygonal ferrite are generated at higher temperatures, see
e.g.
Figure
6.3.
Overall, a simple low-temperature austenitising treatment
followed by slow cooling creates a similar fine polygonal ferrite microstructure
to that produced by hot direct rolling.
Such a microstructure is desirable for
increasing the strength of the
{lll}<uvw> texture after cold rolling and
annealing
and
will tend to improve formability
(§6.7.1.2).
6.3.
2.2
Hornogenising
and
other treatments
A significant problem with most casting processes is the microsegregation that
develops during solidification
(§2.4.1.2).
While the dendrite spacing of an alloy
is considerably finer after
DSC
than in conventional casting (see
e.g.
Figure
5.3),
microsegregation has a strong influence on the recrystallization behaviour after
cold rolling (see
e.g.
§6.6.3.3 for aluminium alloys) and may be detrimental to
Secondary
processing
and
fabrication
201
the formability of the material
(§6.7.1).
A substantial improvement in strip
properties can be achieved
by
a simple homogenising treatment which alters
the microstructure appreciably
by
redistributing solute elements, precipitating
out
supersaturated elements, dissolving unstable phases
and
precipitates and
coarsening intermetallic particles.
1000
-
E
:::I..
-
"#.
100
0
.....
II
r3
-
(,)
10
.£:
0
:c
~
-
CII
~
c:::
~
is
0.1
10-
18
10-
17
10-
16
10-
15
10-
14
Diffusivity,
Ds
(m2Js)
Figure
6.5.
The
effect
of
diffusivity on the times required
to
reduce segregation
to
10%
of
the original value
(C~
in
Eq.
2.45).
The
shaded zones are values
for
half
SDAS,
as
expected
for
Dee
and
TRe
(taken
from
Figure
5.2)
showing the bulk diffusivity
of
various alloying elements in aluminium at
soooe.
A major factor affecting the rate of homogenisation is the group of variables:
D.t!
A}
where D. is the diffusivity of a given solute element
in
the matrix, t the
homogenising time
and
A the dendrite
arm
spacing (see
Eq.
2.45).
It
is often
assumed that homogenisation is essentially complete
when
the initial
concentration profile between dendrite arms is reduced to
10% of the original
amplitude, as shown in Figure
2.21.
Hence, Eq.
2.45
can
be
rewritten as the time
to achieve a uniform dispersion of a particular alloying element:
t
o
.
l
=
0.058..1,7
/ D.
(6.1)
An
example of the use of
Eq.
6.1
is given in Figure
6.5
for aluminium
and
shows
the time to reach
90% of homogenisation as a function of SDAS
and
diffusivity.
For a given diffusivity, homogenisation of a particular alloying element is
achieved more rapidly
by
reducing the spacing between segregations via a
202
Direct
strip
casting
of
metals
and
alloys
small
dendrite
arm
spacing; a favourable consequence
of
DSC. The trends
shown
in
Figure 6.5
are
in
general accord
with
data
on
a strip-cast Cu-Sn alloy
where
segregation associated
with
small SDAS homogenises
at
a faster rate
than
DCC-produced material (Mahulikar 1988).
An
interesting aspect of Eq.
6.1
is the influence of the variation
of
diffusivity,
shown
by
the different alloying element,
on
the rate of homogenisation.
Superimposed
on
Figure 6.5 are the expected zones (shaded) for as-cast
aluminium
produced
by
DCC
(~=50-100
)lm)
and
DSC
(~=5-1O
)lm)
showing the time to reduce segregation
of
Mn, Cr, Cu,
Mg
and
Zn
at
500°C.
Since the diffusivity
of
Mn
is considerably lower
than
the
other
elements, AI-
Mn
alloys require long homogenising times
or
high temperatures to reduce
segregation to
an
acceptable level. The foregoing discussion illustrates the
importance
of
controlling both the type of alloying elements
and
dendrite
arm
spacing if homogenisation is
an
essential step
in
the
production
of
strip-cast Al
alloys.
(b)
4<f
•
\
,..
i:
~i;:>,
~
"':"'~:'
......
Q
o 1050·C
e1100·C
l:J.
1150-C
A 1200·C
.!!l
...
,:
......
,
·····
....
0
~
1 •
····t:.::::::::::::
..
~~~~:::::::::::::::::::::::·
......
.
......
c.
...........•
..............
c.
.....
.
o~~--~--_~I--~··_·-L--~I--~
o
50
100
150
200
Time
(5)
250
300 350
Figure 6.6. Twin roll cast AISI 304 showing: (a) skeletal
morphology
of delta-ferrite,
and
(b) the effect of temperature
on
the
dissolution kinetics of this phase,
after Kim
et
al.
(2003) (with kind permission of Elsevier Limited).
In
strip-cast austenitic stainless steels, the presence of delta-ferrite can influence
the recrystallization behaviour after cold rolling (§6.6.3.2). This
phase
also
has
a
deleterious effect
on
high
temperature workability
and
corrosion resistance
(Kim
et
al.
2003). The production of 2-4
mm
gauge strip
by
TRC provides little
scope for dissolving delta-ferrite
by
processes
such
as
hot
rolling
but
continuous annealing before cold rolling
appears
to
be
a feasible route. Kim
et
al.
(2003) carried
out
a detailed
study
of the rate
of
dissolution
of
delta-ferrite
in