Ronald L. Klueh talks with
ScienceWatch.com and answers a few questions about
this month's Fast Moving Front in the field of
Engineering. The author has also sent along images of
his work.
Article: Oxide dispersion-strengthened steels: A
comparison of some commercial and experimental
alloys
Authors: Klueh,
RL;Shingledecker, JP;Swindeman, RW;Hoelzer, DT
Journal: J NUCL MATER, 341 (2-3): 103-114 MAY 15 2005
Addresses: Oak Ridge Natl Lab, Div Met & Ceram, POB
2008,MS 6138, Oak Ridge, TN 37831 USA.
Oak Ridge Natl Lab, Div Met & Ceram, Oak Ridge, TN
37831 US.
Why do you think your paper is highly
cited?
Engineers and designers of conventional fossil-fuel-fired power-generation
plants constantly push for increased efficiency in an effort to produce
more power with less pollution. To increase efficiency, higher operating
temperatures are required. In the case of nuclear fission and nuclear
fusion power plants, not only are high temperatures required, but the
structural materials of these plants will also be subjected to high neutron
fluences and corrosive environments.
For most new designs, therefore, available materials can place limitations
on designers and engineers, and when a new material is unveiled that shows
significantly improved properties, there is a scramble to advance that
material to the point where it can be incorporated into the new designs.
In the fusion materials program at Oak Ridge National Laboratory (ORNL)
around 2000, we performed research on a Japanese oxide
dispersion-strengthened (ODS) steel that had remarkable short-term tensile
properties (yield stress and tensile strength) as a function of temperature
out to >800ºC.
It also had unusual microstructural features on the nanoscale level, as
determined by transmission electron microscopy and atom-probe tomography.
Excitement raised by observations on this ODS steel (labeled 12YWT) sent
numerous investigators throughout the world on a search for the next great
structural material for fusion. These steels were soon seen also as future
materials for improved fossil-fired and nuclear fission power plants.
In reality, ODS steels were not new. They were first investigated by
Belgian researchers for fuel cladding for nuclear reactors in the 1960s,
but problems were encountered, and the steels were not perfected for the
cladding application. International Nickel obtained a patent on an ODS
steel in the 1970s, and eventually ODS products became available
commercially with small quantities being used in niche applications. During
the 1980s, Japanese researchers again began investigating them as possible
fuel-cladding materials, and the 12YWT material ORNL investigated came from
that source.
I believe our paper (co-authors: John Shingldecker, Robert Swindeman, and
David Hoelzer) is highly cited because we were the first investigators to
conduct creep tests on 12YWT and on one commercial ODS steel and compare
the tensile and creep-rupture properties of 12YWT with other experimental
and commercial ODS steels.
The paper compared elevated-temperature creep-rupture properties for five
different steels, and an effort was made to explain the mechanical
properties behavior in terms of microstructural characteristics to explain
why some ODS steels are stronger than others.
Based on our tests, we postulated microstructures for some of the
commercial steels that were later verified by atom-probe studies. Our work
therefore provided a roadmap with which later investigators and alloy
developers could compare and analyze the microstructure and mechanical
properties of their ODS material.
Does it describe a new discovery, methodology, or
synthesis of knowledge?
I would say it describes a synthesis of knowledge in that it offers an
explanation for the variation in the elevated-temperature strength
properties of the commercial and experimental ODS steels. Based on our
observations, we suggested avenues of research that should be pursued to
verify our hypotheses on the origin of the elevated-temperature strength of
the steels.
Would you summarize the significance of your paper
in layman's terms?
To explain our work, it is necessary to provide a little background on the
nature of ODS and conventional steels. Properties of conventional steels
depend on the alloy content and heat treatment. Steels are iron-based
alloys, and conventional high-temperature steels contain 2-12% chromium and
lesser amounts of a combination of elements such as carbon, nitrogen,
tungsten, molybdenum, vanadium, niobium, etc.
Conventional steels are produced by melting and casting. The casting is
then fabricated into the desired structural component, after which it is
heat treated to optimize the desired properties.
When the steel is viewed under a light microscope up to 1,000X, it is seen
that there is a matrix phase that contains a distribution of relatively
small particles of other phases that precipitated during the heat
treatment. The precipitates are usually carbides or nitrides of chromium,
vanadium, and/or niobium, etc., depending on the elements in the steel
composition.
There are two classes of elevated-temperature steels, known as ferritic and
austenitic steels. The matrix phase of ferritic steels has a body-centered
cubic structure. That is, on the atomic scale, atoms are arranged on a
cubic lattice with atoms on each of the corners of the cube and one atom in
the cube center. Austenitic steels have a face-centered cubic lattice
structure with atoms on each corner of the cube and one in each cube face.
The ODS steels of this discussion are ferritic steels, which are preferred
for elevated temperature applications because of their better thermal
properties—higher thermal conductivity and lower thermal-expansion
coefficient. They are also more irradiation resistant for nuclear
applications in fission and fusion reactors.
We conducted two types of mechanical properties tests in our studies: a
short-time tensile test and a long-time elevated temperature creep-rupture
test. In a tensile test, a rod specimen is deformed by pulling at a
constant rate until it ruptures. A measure of strength is the yield stress,
the stress at which plastic or permanent deformation first occurs. Tensile
tests are conducted at room temperature and at intervals up to
800-900ºC to determine the temperature effect.
If a rod specimen is heated to an elevated temperature and a weight hung on
the specimen such that there is no immediate deformation, with time at
temperature, the specimen will elongate—it will "creep" which is
defined as time-dependent deformation. A creep-rupture test is a creep test
that is run until the specimen ruptures.
Deformation of a metal alloy occurs by the movement of line
defects—called "dislocations"— through the alloy matrix.
Dislocations can be imaged by transmission electron microscopy (TEM) at a
magnification of several thousand times. One way to slow the movement of
dislocations is to introduce a high number density (number per unit volume)
of small precipitate particles into the matrix. In conventional steels,
precipitates become unstable at elevated temperature, and they grow into a
smaller number of large particles, thus limiting their strengthening effect
and limiting the upper-use temperature of the steel.
A solution to this problem is to produce steel with particles that remain
stable to higher temperatures. It turns out that oxides are much more
stable than most carbides and nitrides that form in conventional steels.
Unfortunately, it is not possible to precipitate stable oxides in the
lattice, so fabrication processes other than melting and casting are
required.
Powder metallurgy techniques are used to produce ODS steels with a fine
distribution of stable oxide particles. These ODS steels are produced from
a fine powder of a steel alloy composition that serves as the matrix, such
as an iron-chromium-tungsten composition, and mixing it with an oxide
powder. Yttrium and titanium oxides are commonly used. After the powders
are mixed, they are compacted and extruded at an elevated temperature to
form a solid. Finally, the extruded product is rolled into bar, plate, or
sheet, or extruded into rods or tubes.
Figure
1:
Transmission electron micrographs of
experimental ODS steels (a) 12Y1
(Fe–12Cr–0.25Y2O3)
and (b) 12YWT
(Fe–12Cr–2.5W–0.4Ti–0.25Y2O
3).
To understand the microstructural differences between a strong and a weak
ODS steel, ORNL investigators conducted optical microscopy, TEM, and atom
probe field ion microscopy studies on two ODS steels. The first, labeled
12Y1, was fabricated with powders of an iron-12% chromium alloy and yttrium
oxide (Y2O3), and the second, labeled 12YWT, was
obtained from powders of an iron-12% chromium-tungsten-titanium alloy and
Y2O3.
The results produced very different microstructures in the two steels (Fig.
1). For 12Y1 [Fig. 1(a)], particles were estimated to be 10–40 nm in
diameter at a number density of 1020-1021
m-3.
Diffraction studies indicated the particles in 12Y1 were essentially pure,
crystalline Y2O3. For 12YWT [Fig. 1(b)], particle
size and particle number density were estimated at 3–5 nm diameter
and about 1023 m-3, respectively. For this alloy,
three-dimensional atom probe analysis revealed compositionally distinct
nano-sized clusters enriched in yttrium, titanium, and oxygen, slightly
enriched in Cr, and slightly depleted in Fe and W.
The difference in oxide particle size in the two steels is evident. The
lines between particles are dislocations. Work reported in our paper
involved the study of the mechanical properties as a function of
temperature of these two experimental steels and three commercial ODS
steels and demonstrated the difference in properties in the two types of
ODS microstructures. As expected from the microstructural differences
observed in Fig. 1, the 12YWT was much stronger than 12Y1 in a short-time
tensile test and a long-time creep-rupture test.
Comparison of 12YWT with the commercial steels gave different results. One
of them (MA 957) had comparable short-time tensile and long-time creep
properties. This steel had been shown to have the nano-sized clusters
similar to those of 12YWT.
A second steel (PM 2000) had short-time tensile properties comparable to
12Y1 at low temperatures, but contrary to 12Y1, it had elevated-temperature
creep-rupture properties comparable or better than those of 12YWT, which
was unexpected.
Finally, and also unexpected, the third commercial ODS steel (MA 956) had
inferior tensile properties relative to those of 12Y1, but nevertheless,
its long-time creep-rupture properties were comparable to those of 12YWT.
These results indicated that differences involved more that just the sizes
and number densities of the oxide particles. Among the possibilities
discussed were grain size and the amount of recrystallization that can
affect tensile behavior at low and intermediate temperatures.
How did you become involved in this research and
were any particular problems encountered along the way?
My work in recent years has involved the physical metallurgy of steel,
studies of the relationship of mechanical properties to microstructure, and
the development of new steels. It was therefore logical for us to study the
relationship of these unique microstructures to their mechanical
properties. A major difficulty was the availability of ODS material to
test.
It is ironic that, although ODS steels are at present a hot scientific
topic, there are no longer any commercial manufacturers of the product,
since the market for ODS steels is limited, and it is not cost-effective
for the manufacturers to continue to support manufacturing facilities for
limited sales.
Where do you see your research leading in the
future?
In recent years, my work has been associated more with conventional steels,
those produced by melting, casting, fabricating, and heat treating. The ODS
steels have numerous problems to overcome. That is why they have been in a
development stage since the 1960s.
Because of the powder metallurgy processes used to produce them, their
mechanical properties are anisotropic. That is the reason they were
abandoned in the 1970s as possible nuclear fuel cladding materials for fast
reactors. Their high strength is in the direction in which they are
worked—rolled or extruded—and they are relatively weak in the
direction orthogonal to the working direction (the niche applications
mentioned above exploited the anisotropy). Much effort has been expended
over the years to solve this problem.
Transmission
electron microscopy photomicrograph of a 9Cr-1Mo steel
after a thermo-mechanical treatment. The steel contains
a high number density of fine nitride precipitates that
formed on dislocations (lines in the photo) during the
treatment.
In addition, they cannot be welded by the usual fusion-welding techniques,
since melting destroys the fine oxide distribution and thus produces a weld
of inferior strength. Should those problems be solved, a possible bigger
impediment to their eventual widespread use is economics. The complicated
manufacturing process is extremely expensive relative to conventional
steels.
In recent years I have been involved in trying to develop improved
high-temperature steels using conventional steelmaking techniques to
produce a high number density of fine precipitates into the steel matrix.
We have had some success in producing steels using conventional processing
techniques that contain a high number density of small nitride precipitates
similar to those in 12YWT (Fig. 2). This was accomplished by a
thermo-mechanical treatment, wherein hot working was introduced into the
normal heat treatment sequence typically used for conventional steels.
The normal heat treatment for conventional elevated-temperature steels
produces similar nitrides as those shown in Figure 2. However, in
conventional steels they are present in a much smaller number density and
have a much larger size. Thermo-mechanical treatment (TMT) that is used to
produce the microstructure in the figure is hot working, which produces a
high number of dislocations into the matrix. These dislocations, some of
which can be seen as lines in the figure, act as heterogeneous nucleation
sites for the nitride precipitates.
Although these precipitates are not as stable as oxides in ODS steels and
the new steels could not be of used to 800ºC and beyond, as visualized
for ODS steel, they could push the use temperature well above the
600-620ºC limit of present high-temperature ferritic steels. Such an
increased service temperature could have a significant effect on the
efficiency of a power plant. It needs to be emphasized that, just as the
ODS steels are still in a development stage, the same must be said for
these thermo-mechanical treated steels.
Of course, the best approach to better elevated temperature steels would be
to improve the properties by alloying and using conventional heat treatment
processes used in the steel industry at present. Today, computational
thermodynamics programs are available that allow us to obtain insight into
how we might accomplish that. Based on thermodynamics calculations, I have
developed some ideas on how improved steels using present-day steelmaking
processes might be produced, ideas which I would like to pursue should the
opportunity arise.
Do you foresee any social or political implications
for your research?
If ODS steels or improved conventional steels discussed here can be used in
the future, they could have considerable effect on energy production for
electricity, which could, in turn, have major social and political
implications, since energy plays such a large part in our everyday lives.
Ronald L. Klueh, Ph.D.
Oak Ridge National Lab
Division of Metals & Ceramics
Oak Ridge, TN, USA