Niobium in modern steels
A. J. DeArdo
It is well known that niobium is added to a wide
range of steels for improving processing,
microstructure, properties and performance. Over
the past 20 years, the use of niobium has also
permitted new steels with attractive properties to
be developed. Furthermore, the addition of
niobium to existing steels such as ferritic stainless
steels has also led to improvement. The goal of
this paper is to review the basic behaviour of
niobium in a wide range of steels, including not
only the traditional steels but also some of the
newer versions. Particular emphasis has been
placed on the basic metallurgical principles that
apply to these steels, for it is the application of
these principles that allows the composition–
processing–microstructure–mechanical property
relationships to be rationalised and exploited. The
application of basic metallurgical principles has
resulted in a predictive capability that has led to
alterations in composition and processing for the
purpose of producing steels with superior
mechanical properties and improved overall
performance. IMR/401
The author is in the Basic Metals Processing Research
Institute, Department of Materials Science and
Engineering, University of Pittsburgh, Pittsburgh, PA
15261, USA (deardo@engr.pitt.edu).
© 2003 IoM Communications Ltd and ASM International.
Published by Maney for the Institute of Materials, Minerals
and Mining and ASM International.
Introduction
As we celebrate the 200th anniversary of the discovery
of niobium by Charles Hatchett, we will be reminded
repeatedly of how this metal has changed the face of
science and technology in the materials industry.1,2
Nowhere is this more noticeable than in the case of
steels. Although niobium is classiŽ ed as a transition
metal, perhaps the most important transition made
by niobium was from being little more than a labora-
tory curiosity prior to 1960 to being a viable, commer-
cially produced ferroalloy suitable for addition to
steel in 1965. While we celebrate 1801 as the date of
the discovery of niobium, in fact, the Ž rst success-
ful commercial production of ferroniobium by the
Companhia Brasileira de Metalurgia e Mineração
(CBMM) mine in Araxa, Brazil took place in 1965.3
After 1965, ferroniobium was abundantly available to
the steel industry for the Ž rst time as a microalloying
element. Prior to this date, only vanadium and
titanium were available on a commercial scale for
microalloying steel.
The 41st element in the periodic table is called
columbium within the industrial sector of the USA
and niobium by the rest of the world. The story
behind the dual names is quite interesting. John
Winthrop (1606–1676), the Ž rst governor of
DOI 10.1179/095066003225008833
Connecticut and a scientist and rock collector, dis-
covered a new mineral around 1734. He named it
columbite, possibly in honour of Christopher
Columbus. Winthrop’s grandson sent it to the British
Museum for display and analysis in 1753. Decades
later, in 1801, Charles Hatchett (1765–1847), a British
chemist working at the British Museum, attempted
to analyse the constitution of the columbite ore.
Hatchett could tell that there was an element
unknown at that time present in the columbite, which
he called columbium. In fact, unbeknownst to
Hatchett, there were two then unknown elements in
the columbite: columbium and tantalum. In 1802, the
Swedish chemist A. G. Ekeberg discovered a new
element in Finnish minerals similar to columbite and
named it tantalum after the Greek god Tantalus. The
situation became confused when, in 1809, the British
chemist and physicist, William Hyde Wollaston, com-
pared the minerals tantalite and columbite and
declared that the element columbium was actually
the element tantalum. This confusion was not unex-
pected, since the elements are so similar, are always
found together in natural deposits, and are very
diYcult to separate. In 1844, Heinrich Rose, a German
chemist, produced two apparently new and diVerent
acids from columbite and tantalite. One came to be
known as niobic acid, while the other was called
tantalic or pelopic acid. Rose found that the element
contained within the tantalic acid was very similar to
the tantalum described by Ekeberg, hence the name
tantalic acid. The undeŽ ned element found in the
other acid was called niobium, hence the name niobic
acid, after Niobe, mythical daughter of Tantalus and
goddess of tears. About 20 years later, the Swiss
chemist Jean Charles Galissard de Marignac proved
that these two acids were both distinct and produced
by two diVerent elements. Finally, the Swedish chem-
ist Christian Wilhelm Blomstrand isolated metallic
niobium in 1864. The name niobium was oYcially
adopted by the International Union of Pure and
Applied Chemistry in 1950, although the metal is still
referred to as columbium, especially in the industrial
metallurgical sector of the USA.
Although the beneŽ ts of niobium additions to low
carbon steel had been known since the late 1930s,4 it
was not until 1958 that the Ž rst heat of niobium
microalloyed steel in the form of hot strip was com-
mercially produced by National Steel in the USA.5
The international symposium, ‘Niobium’, held in San
Francisco in 1981, reviewed the science and tech-
nology of using niobium in a broad range of materials
as practised at that time.6 By 1981, the concepts of
controlled rolling, austenite conditioning, and proper
alloy design for ferrite–pearlite microstructures with
adequate properties in a broad range of products
were understood and practised. Perhaps the plate and
linepipe microalloyed steels were the best understood
and optimised at that time. On the other hand, the
beneŽ ts of niobium microalloying to product lines
International Materials Reviews 2003 Vol. 48 No. 6 371
372 DeArdo Niobium in modern steels
1 Abbreviated periodic table showing some of elements that form solid solution with iron13
such as strip, sheet, bars, shapes and castings were
just starting to be exploited. Most of the work pub-
lished in the steel papers in the ‘Niobium’ proceedings
concerned ferrite–pearlite microstructures.6
Over the past 20 years, there have been at least
two major beneŽ cial changes in niobium bearing steel
products. First, since polygonal ferrite–pearlite micro-
structures cannot readily exceed yield strength levels
of about 400 MPa in reasonable section sizes or with
the low carbon contents found in niobium steels, the
applications-driven requirement for higher strength
levels has led to the development of ferritic micro-
structures in virtually every product class that are
based on low temperature transformation products
such as acicular ferrite and bainite or to multiphase
microstructures. Sometimes these microstructures are
achieved through the use of accelerated cooling, some-
times by using hardenability approaches, or some-
times additions of niobium improve both. Regardless
of the approach, niobium continues to play a major
positive role in optimising the manufacturability, Ž nal
properties and performance of these steels.
The second major change since 1981 has been in
the wide acceptance of niobium as a stabilising
element, often in combination with titanium, whereby
carbides are formed thus reducing the amount of free
carbon in solid solution, in both ultralow carbon
(ULC) sheet steel and ferritic stainless steel. Since the
niobium levels employed in the ULC steels frequently
fall in the range 100–300 ppm, they can be considered
as a microalloying addition. However, the 1500–
5000 ppm niobium sometimes used in the ferritic
stainless steels would be considered an alloying
addition. The so called dual stabilised Ti+Nb con-
taining ULC steels and ferritic stainless steels are
viable, successful, and growing members of their
respective families.
In the proceedings of ‘Niobium’, the forerunner of
the present review appeared.7 In that paper, the
authors attempted to present the fundamental aspects
of niobium and its behaviour in steel, at least as
understood at that time. These fundamental aspects
have remained valid over the course of time. However,
much has been learned in the intervening years and
the goal of the present review is to provide the reader
with the background necessary to understand modern
niobium microalloyed steels. The paper is organised
into sections devoted to the following aspects of the
International Materials Reviews 2003 Vol. 48 No. 6
role of niobium:
F fundamental considerations
F thermomechanical processing and austenite
conditioning
F niobium and transformation
F niobium and strengthening
F niobium and stabilisation.
It is hoped that the basic principles presented here
will permit the reader to design better steels and
processing routes to further improve future steels, all
through the intelligent use of niobium. The appli-
cations side is becoming even more demanding:
weight reduction, safety, lower cost, etc. These increas-
ing demands can only be met through a thorough
understanding of the literature and of the basic prin-
ciples of microalloying with niobium upon which
much progress depends There exists a large and
impressive literature which summarises the rich his-
tory of R&D in this Ž eld. The reader is encouraged
to consult this work.8–12
Fundamental considerations
Electronic structure
Iron, manganese, and niobium belong to the family
of metallic elements known as transition metals.
Transition metals are characterised by the electronic
structure of the atoms where the outer shell (4s energy
level for period 4 elements and 5s for period 5
elements) contains electrons, while the inner shell (3d
energy level for period 4 elements and 4d for period
5 elements) is not completely Ž lled. The transition
metals are the only elements that have unŽ lled inner
shells. Many of the elements that form substitutional
solid solutions with iron in HSLA ( high strength low
alloy) steels are shown in Fig. 1.13
Fe–Nb phase diagram
As might be expected from the crystallographic data,
niobium and iron are completely miscible at high
temperatures. The equilibrium diagram for the Fe–Nb
system is well established: details may be found in
standard texts.14 The most interesting feature of the
diagram, in relation to the use of niobium in steels,
is the existence of a c loop which limits the existence
of the c phase to alloys containing less than
0·83 wt-%Nb (0·50 at.-%). Clearly, niobium is a ferrite
stabiliser. However, additions of up to 0·10–0·20%
 DeArdo Niobium in modern steels 373

2 Possible diffusion distance of Nb in ferrite and

austenite based on diffusion coefficients16

lower the A temperature, i.e. the c loop apparently

3
displays a minimum similar to that observed in the
Fe–Cr and Fe–V systems; thus, in small quantities,
niobium acts as a c stabiliser.
In the ternary Fe–C–Nb system, the major feature
of the equilibrium diagram relevant to steels is the
marked reduction of the solubility of carbon in
austenite and ferrite as a result of the ready formation
of niobium carbide.

Diffusion 3 Tendency of certain metals to form oxides,

sulphides, carbides, and nitrides and their
The early literature contains several estimates of the precipitation strengthening potential (arranged
interdiVusion coeYcients in dilute solutions of similar to periodic table) (after Meyer et al.20 )
niobium in c-Fe.7 The data show considerable scatter
and are considered unreliable.
in the bcc crystal structure. The diVusion coeYcient
Kurokawa et al.15 have studied the diVusion of
is usually expressed as an Arrhenius law of the form
niobium in high purity iron containing no detectable
interstitial levels. Iron, Fe–0·6Si and Fe–0·6Si–1·5Mn D=D exp(­ Q/RT ) . . . . . . . . . (1)
alloys were investigated at temperatures in excess of 0
where D is a constant depending on the nature of
1080°C. The experiments were conducted using the 0 and the bulk composition and called the
the element
radioactive tracers Nb95 and Nb96, and the concen-
frequency factor, R is the ideal gas constant
tration proŽ les were determined by direct sectioning.
(8·314 J molÕ 1 KÕ 1), Q is the activation energy, and
Two important observations emerge:
T is the temperature in kelvin. The constants for the
(i) the interdiVusion coeYcient of niobium in
diVusion of niobium in iron can be compared to the
austenite is not strongly in uenced by the
coeYcients for the self-diVusion of iron in Table 1.17
composition of the solid solution matrix
(ii) the coeYcient for niobium is somewhat higher
than the coeYcient for the self-diVusion of iron. Compound forming tendencies
The diVusion coeYcient in bcc ( body centred cubic) Transition metals are known to form a series of
a-Fe is about 100 times higher than that in c-Fe, simple and solid solution compounds of oxides, sul-
resulting in much greater diVusion distances in ferrite phides, carbides, and nitrides. The compound forming
(see Fig. 2) at lower temperatures.16 This is thought tendencies of several of the transition metals in steel
to be partially caused by the looser atomic packing have been reviewed18,19 and have been summarised
by Meyer et al.20 (see Fig. 3). Niobium shows a strong
tendency to form carbonitrides, but relatively little
Table 1 Values of Q and D0 for diffusion of Nb in tendency to form oxides, sulphides, or solid solutions
Fe and for self-diffusion of Fe of these compounds. In this regard it behaves similarly
Element Bulk Q, kJ molÕ 1
D0 , m2 sÕ 1
×104 to vanadium. This characteristic distinguishes it from
titanium which does not act as a carbide former until
Diffusion of Nb in Fe all oxygen, nitrogen, and sulphur have been consumed
Nb c-Fe 266·5 (±1·8) 0·83 (±0·69)
by initial additions of titanium.
Nb c-Fe 264 0·75
Nb c-Fe 209 to 334 400 to 530
Nb a-Fe 252 (±2·5) 50·2 (±3·0) Precipitation of niobium compounds:
Nb a-Fe 289 100D0 (c) Nb–NbC system
Self-diffusion of Fe The Nb–NbC system has been studied extensively
Fe c-Fe 284 0·49 over the past 25 years,21,22 the most deŽ nitive study
Fe a-Fe 241 2·01 being that of Storms and Krikorian.22 The Nb–NbC
Q is activation energy and D0 is frequency factor. phase diagram is shown in Fig. 4. There are three
International Materials Reviews 2003 Vol. 48 No. 6
374 DeArdo Niobium in modern steels

4 Nb–NbC phase diagram (after Storms and Krikorian) 21,22

solid solution, single phase regions: a, b, and c. To per unit cell (or 4 lattice points, each consisting of one
avoid confusion with the ferrite phases, these will be niobium and one carbon site). The crystal structure of
referred to as NbC-a, NbC-b, and NbC-c. The NbC- NbC can be represented by two interpenetrating fcc
a phase is an interstitial solid solution of carbon in x with niobium atoms (or niobium atom vacanc-
lattices
niobium, with a maximum solid solubility of ies) residing on one set of fcc lattice points and carbon
0·2 at.-%C near 2300°C. A vanishingly small carbon atoms (or carbon atom vacancies) residing on the other
solid solubility would be expected at temperatures of set. Therefore, NbC can be considered to consist of
interest in steelmaking. The NbC-a phase has an A2 niobium atoms, carbon x atoms, and vacancies. If N0 is
(bcc) crystal structure with a lattice parameter that deŽ ned as the number of atom positions on either fcc
will vary with the carbon content and is 3·294 A Ã for lattice containing niobium or carbon atoms, then
pure niobium.21 The NbC-b phase (Nb C) also
2 N0=N +NV =N +NV . . . . . . (2)
appears to have a very limited range of carbon Nb Nb C C
solubility around the stoichiometric composition where N is the number of niobium (or carbon)
(33·3 at.-%C) below 1500°C. This phase has an hcp Nb (or
atom sites perC) unit volume containing an niobium
(hexagonal close packed ) crystal structure with lattice (or carbon) atom and NV is the number of
parameters a=3·12 A Ã and c=4·95 A Ã .21 The phase of Nb (or C)
vacant niobium (or carbon) atom sites per unit
the Nb–NbC system that is of most interest in steels volume. Subtracting NV from both sides gives
is the NbC-c phase ‘NbC’, but it should be noted Nb
that Nb C is observed in steels having high Nb/C N =N +(NV ­ NV ) . . . . . . . (3)
Nb C C Nb
ratios.23 2The NbC-c phase has a range of carbon
Since, by convention, the subscript of the niobium is
solubility which, for example, at 1100°C, appears to
taken as one, the terms on both sides of equation (2)
extend from near NbC to NbC. This range of
possible carbon contents0·72
should be divided by N . This gives
has led to the adoption of Nb
the symbol NbC to describe the NbC-c phase, with N NV ­ NV
x being equal to xthe mole ratio of carbon to niobium 1= C + C
N N
Nb . . . . . . . . (4)
(at 1100°C, therefore, 0·72<x<1·0). The NbC-c Nb Nb
phase has a B1 (NaCl ) type fcc (face centred cubic) where
crystal structure24 whose symmetry is described by
N
the Schoen ies point group O5 and Hermann–
h C =x (in NbC ) . . . . . . . . (4a)
Mauguin space group (shortened ) Fm3m with 8 atoms N x
Nb
International Materials Reviews 2003 Vol. 48 No. 6

5 Variation of NbC lattice parameter with
composition (after Storms and Krikorian 22 )
and
NV ­ NV
C Nb =1­ x . . . . . . . . . (4b)
N trations; both nitrogen and vacancies act to reduce
Nb
Equations (4) indicate that if the number of vacancies
on niobium and carbon sites is equal, then N =N
and x in NbC would be equal to one and the Nb carbideC
x
would have the stoichiometric composition. If, how-
ever, there are more vacancies in the carbon lattice,
then N >N and x<1·0; this, of course, is the
Nb It can
usual case. C be concluded, therefore, that there
is a direct relationship between the deviation in x
from x=1 and the vacancy concentration in NbC .
The lattice parameter of NbC has been shown xto
x assessed.
be a strong function of x (see Fig. 5).22 Pure, stoichio-
metric NbC has a lattice parameter of 4·470 A Ã and
the lattice parameter decreases with decreasing x to
a value of about 4·43 A Ã at NbC . Clearly, the
vacancy concentration in the NbC-c 0·7 phase has a very
strong in uence on the lattice parameter.
Microalloyed steels contain both carbon and nitro-
gen and when niobium precipitates it does so as
niobium carbonitride.25–29 The crystallography of
‘NbC’ has been discussed above and, in fact, is very
similar to ‘NbN’.24 Since ‘NbC’ and ‘NbN’ are very
similar compounds, it is quite reasonable to expect
the two to have complete solid solubility, i.e. to form
a carbonitride. This is, in fact, correct.30
The NbC–NbN system has been studied31,32 and
reviewed.22 The NbC–NbN system is actually a ter-
nary system of NbC, NbN, and vacancies. This ter-
nary system and associated lattice parameters are
shown in Fig. 6. Two very important points emerge
from an inspection of Fig. 6. First, the composition
of niobium carbonitride can be represented by
NbC N , where x is the C/Nb mole ratio, y is the
N/Nbx moley ratio, and 1­ (x +y) is the mole ratio of
vacancies. Figure 6 also indicates that the quantities
x, y, and (x +y) can all be variables. Second, the
DeArdo Niobium in modern steels 375
6 Lattice parameter versus composition in
NbN–NbC system (after Storms and
Krikorian) 22,31,32
lattice parameters of the carbonitride will be strongly
in uenced by the nitrogen and the vacancy concen-
the lattice parameter of pure NbC. Figure 6 further
illustrates that the extent to which the lattice para-
meter is lowered with increased nitrogen will depend
on the vacancy concentration of the carbonitride. The
higher the vacancy concentration, the larger will be
the lowering of lattice parameter per unit increase in
nitrogen content in the carbonitride. Storms and
Krikorian22 point out that the interpretation of lattice
parameter measurements will be diYcult unless the
vacancy concentration in the carbonitride can be
Composition of NbCx N y formed in niobium
steels
The composition of precipitates found in commercial
steels has been extensively studied and the results
obtained prior to 1973 have been reviewed by Gray.33
Brie y, it can be stated that the cubic forms of
carbonitride are most frequently found at normal
niobium levels (<0·04%). The reported34,35 range of
stability from NbC to NbC may allow substi-
0·72
tution of molybdenum36,37 1·0 as the nitrogen/
as well
vacancy contents discussed above without causing
changes in crystal structure.
Other studies of the composition of the carbonitride
which forms in niobium bearing microalloyed
steels26–29,38–40 have shown a direct relationship
between the composition of the carbonitride and the
composition of the steel: the larger the N/C ratio in
the steel, the more nitrogen rich the carbonitride. A
typical example of the relationship is shown in
Fig. 7.28
These and other studies33,41–44 have also shown
that the composition of the carbonitrides in any given
steel can depend upon the thermal conditions under
International Materials Reviews 2003 Vol. 48 No. 6
376 DeArdo Niobium in modern steels

9 Change in NbCx Ny composition with Nb content

and precipitation temperature44
7 Influence of C/N ratio in 0·1%Nb steel on C/N
ratio in carbonitride precipitate (after Meyer
et al.28 )
approximated by the Kolomojorov–Johnson–Mehl–
Avrami (KJMA) equation47

A B
which they form. The data indicate that the precipi- p
X =1­ exp ­ bG
N b 3t4 . . . . . . (5)
tates contain more nitrogen when formed at higher p 3
temperatures,27,29,44,45 depending on carbon or
where X is the volume fraction of precipitates formed,
niobium content (Figs. 8 and 9),44 which is similar to
Nb is thepnucleation rate (usually assumed constant ),
results for precipitation of vanadium carbonitride in b
G is the growth rate (usually assumed constant), and
vanadium strengthened microalloyed steels.45,46 It is
t is the reaction time. The equation shows that a
to be expected that the presence of other elements
precipitation reaction will attain a given level of
such as titanium and aluminium, that have strong b, Gb , and t have suYciently
completion only when N
nitride-forming tendencies, will aVect the amount of
large values. The nucleation rate is controlled by
nitrogen in NbC N . This has been observed by
Ouchi and Izumi,40 x especially
y in steels austenitised at
solute supersaturation, while the growth rate is con-
trolled by both the diVusion coeYcient and the solute
high temperatures (Fig. 10).
supersaturation.48 Since solute supersaturation con-
Long austenitising treatments (" 150 h) and higher b and Gb , it is very important to understand
trols both N
niobium (" 0·10%) and nitrogen (" 0·012%) contents
the factors that determine it.
give rise to the formation of non-cubic compounds,
The extent to which elements can be maintained in
often of the hexagonal NbC-dê or NbC-e carbonitride
solid solution in austenite is governed by the appro-
type.41–43 Also, as stated above, Nb C (NbC-b) type
precipitates are observed at high 2 Nb/C mole
priate solubility product. If, for example, the interest
is in the extent of solid solubility of niobium and
ratios.23,33
carbon in austenite which is in equilibrium with pure,

Solubility of NbCx Ny in austenite

The formation (and dissolution) of precipitates nor-
mally exhibits sigmoidal kinetic curves which can be

10 Relationship between nitrogen content

8 Change in NbCx Ny composition as function of uncombined with Al or Ti and composition of
C content and precipitation temperature 44 niobium carbonitride (after Ouchi and Izumi40 )

International Materials Reviews 2003 Vol. 48 No. 6

 DeArdo Niobium in modern steels 377

stoichiometric NbC, then the following reaction must coordinate system. The use of this standard state
be considered49 eliminates the necessity of converting weight percent-
NbC =[Nb] +[C] . . . . . . . . (6) ages, obtained via chemical analysis, to mole fractions
c c for the purpose of thermodynamic calculations. This
and, at equilibrium, the reaction can be written standard state is particularly convenient to use in
as follows metallurgical systems containing dilute solutes.50,51
NbC (s)=Nb (s)+C(gr) . . . . . . . (7a) This standard state can formally be deŽ ned as
DG° =130 122­ 1·67T . . . . . . . . (7b) f ! 1 as %Nb! 0 . . . . . . . . (12)
1 Nb
where DG° represents the positive change in free Hence, with respect to the 1 wt-% standard state
1
energy (J molÕ 1) for the reaction. It is important to having unit activity, the activity of niobium is given by
realise that in this reaction, NbC, Nb(s), and C(gr)
h = f %Nb . . . . . . . . . . . (13)
are taken as the pure components in a stable state of Nb Nb
existence at some given temperature. This is referred This expression for the activity of niobium is often
to as the Raoultian standard state.50,51 For cases simpliŽ ed by two further assumptions which reduce
where the pure component may exist in a physical f to unity. These assumptions are often made in the
Nb
state which is diVerent from that of the solution, the calculation of the equilibrium constant K, also
Henrian 1 wt-% standard state may be more referred to as the solubility product.38 The Ž rst of
convenient. these assumes that niobium is a dilute solute in
The Henrian standard state is obtained from austenite. The second assumption neglects any inter-
Henry’s law, which, strictly being a limiting law action between solutes in the system. The fact that
obeyed by the solute niobium (or carbon) at inŽ nite f goes to unity under these assumptions can be
Nb
dilution in austenite, is expressed as51 realised given its dependence on the Ž rst order free
a energy interaction coeYcient for dilute multicompon-
Nb ! c° as X ! 0 . . . . . . . . (8) ent solutions.52,53 This interaction coeYcient was Ž rst
X Nb Nb introduced by Wagner52 and later was extensively
Nb
where a is the activity of niobium in austenite with used by Chipman53 and colleagues in the study of
respect Nbto the Raoultian standard state, X is the molten alloy steels. It is related to the activity
mole fraction of niobium in solution, and c° Nb is the coeYcient such that
Nb
Henry law constant, i.e. the activity coeYcient which n
quantiŽ es the diVerence between Raoultian solution log f = æ ej (wt-% j ) . . . . . . . . (14)
i i
behaviour of niobium and Henrian solution behav- j= 1
iour of niobium.51 If the solute (niobium or carbon) where ej (read ‘e j on i’) is the Wagner interaction
parameter i on a weight per cent composition coordi-
obeys Henry’s law over a Ž nite composition range,
then nate and n would represent the total number of solute
elements or compounds in the system.52 It is apparent
a =c° X . . . . . . . . . . . (9)
Nb Nb Nb from equation (14) that if either a solute is dilute in
The Henrian standard state is obtained by extrapolat- a given system (wt-% j! 0) or interactions between
ing the Henry law line deviating from ideal solution solutes are negligible (ej ! 0), the activity coeYcient
behaviour to X =1. This state represents pure for the given solute will ibe unity.
niobium in the Nb hypothetical, non-physical state in Returning to the dissolution of NbC precipitates,
which it would exist as a pure component if it obeyed the change in free energy (J molÕ 1) as carbon and
Henry’s law over the entire composition range (i.e. as niobium are dissolved in austenite is given by38
it does for a dilute solution).50,51
Having deŽ ned the Henrian standard state, the C(gr)=[C] . . . . . . . . . . . (15a)
activity of niobium in solution in austenite h with DG° =35 062­ 33·22T . . . . . . . (15b)
respect to the Henrian standard state having Nb unit 2
activity is given by Nb(s)=[Nb] . . . . . . . . . . (16a)
h =f X . . . . . . . . . . . (10) DG° =­ 27 949­ 29·54T . . . . . . (16b)
Nb Nb Nb 3
where f is the Henrian activity coeYcient. where DG° and DG° represent the free energies associ-
Nb be noted that the mole fraction of niobium
It should ated with2 the change
3 between the Raoultian and
in solution in austenite can be related to its concen- Henrian standard states. The overall reaction is
tration in weight per cent (%Nb) by given as
%Nb/MW NbC =[Nb] +[C] . . . . . . . . . (17)
X = Nb
Nb [%Nb/MW ]+[(100­ %Nb)/MW ] The corresponding change in free energy is given by
Nb Fe
. . . . . . . . (11) the addition of equations (7 ), (15), and (16)
where MW and MW are the respective atomic h h
weights of Nb
niobium and Fe iron. This conversion is DG° =137 235­ 64·43T =­ 2·303RT log Nb C
a
necessary in order that a hybrid Henrian standard NbC
. . . . . . . . (18)
state can be introduced. This is referred to as the
1 wt-% standard state. This standard state diVers where a is taken as unity for the pure component
from the previous one in that a weight per cent NbC inNbCits standard state. Furthermore, assuming
coordinate system is used rather than a mole fraction that niobium and carbon are dilute in austenite, h
Nb
International Materials Reviews 2003 Vol. 48 No. 6
378 DeArdo Niobium in modern steels

and h may be represented by their weight per cents. Problems arising from the separation technique are
Hence,C the solubility product under this theoretical that very Ž ne precipitates may not be included in the
treatment is found to be analysis61 and that discrepancies may exist as to the
exact composition of the precipitate.38 These prob-
7167
log [Nb][C ]=3·36­ . . . . . . (19) lems also plague the equilibration methods, as carbon
T contents are often analysed assuming that a uniform
There are three factors that must be taken into stoichiometric or non-stoichiometric compound is
account when applying a solubility product of the present.38,61 Additionally, both of these methods,
form shown above that has been derived from thermo- including the thermodynamic methods, neglect the
dynamic considerations: eVect of precipitate size on solubility. Thermo-
(i) it applies strictly to equilibrium conditions dynamics indicate that small particles are more sol-
that rarely exist in practice uble than large particles.69 Hence, solubility products
(ii) it might be strongly altered by the presence of obtained via these methods may predict a more stable
other solutes through their eVect on the inter- precipitate than would be expected. A possible excep-
action coeYcients presented above tion to this may be found in the work by Simoneau
(iii) it will surely be altered by the Gibbs– et al.70 who used electrical resistivity measurements
Thompson or capillarity eVect that accounts to determine the solubility of Nb(CN) in austenite.
for the in uence of particle curvature on solu- This technique is based upon the measured reduction
bility. (The solubility of a given particle varies in bulk resistivity as niobium, carbon and nitrogen
inversely with its curvature.54,55 Following in solid solution precipitate to form Nb(CN). The
Lupis, the solubility of particles of radius 1 nm reduction in resistivity is caused by the loss of solid
would be nearly twice as high as those of solution atoms which act as scattering centres. The
radius 100 nm.55) results of Simoneau et al.70 show that the thermo-
dynamic stability of Nb (CN) is less than that pre-
dicted using equilibrium methods.
Nb(CN) solubility products in austenite Hardness techniques are questionable, since they
The importance of niobium as a microalloying are based upon the assumption that an increase in
element in steels is apparent from the numerous
studies over the past 30 years. Much of this work is
concerned with the solubility of niobium mono- Table 2 Solubility products for Nb–C, Nb–N, and
carbides,7,18,38,56–62 mononitrides,7,18,38,58,62,63 and Nb–C–N systems in austenite
carbonitrides2,7,29,64–67 in austenite. The results of
Product Method* Ref.
these studies are given in Table 2 in the form of
solubility products. The diVerences among these prod- Nb–C system
ucts are considerable and may be attributed to a log[Nb][C]=2·9­ 7500/T D 38
number of factors, foremost among which would be log[Nb][C]=3·04­ 7290/T B 38
log[Nb][C]=3·7­ 9100/T C 56
the methods used in obtaining the given solubility
log[Nb][C]=3·42­ 7900/T B 18
product as each technique has its own assumptions log[Nb][C]=4·37­ 9290/T C 57
and limitations. The techniques used in obtaining the log[Nb][C]0·87 =3·18­ 7700/T B 58
solubility products of Table 2 are classiŽ ed here as log[Nb][C]0·87 =3·11­ 7520/T E 38
A–E and are, brie y: log[Nb][C]=2·96­ 7510/T E 38
A: thermodynamic calculations log[Nb][C]0·87 =3·4­ 7200/T A 38
B: chemical separation and isolation of precipitate log[Nb][C]=3·31­ 7970/T+Q† B 59
C: equilibrating a series of steels with diVerent log[Nb][C]0·87 =2·81­ 7019·5/T A 60
niobium contents with a H –CH atmosphere log[Nb][C]=1·18­ 4880/T … 62
at various temperatures, after2 which
4 the carbon log[Nb][C]=3·89­ 8030/T … 62
log[Nb][C]=4·04­ 10230/T C 63
contents are analysed
log[Nb][C]=3·79­ 10150/T B 58
D: hardness measurements
E: statistical treatment of previous solubility Nb–N system
products. log[Nb][N]=2·8­ 8500/T B 18
log[Nb][N]=3·7­ 10800/T B 65
As was stated in the previous section, thermodynamic
log[Nb][N]0·87 =2·86 ­ 7927/T A 60
calculations of solubility products often neglect any log[Nb][N]=4·2­ 10000/T … 62
interaction between elements and particle curvature
eVects. As a result, the activity coeYcients are assumed Nb–C–N system
log[Nb][C]0·24 [N]0·65 =4·09­ 10500/T B 38
to be unity and the activities are represented by their
log[Nb][C+12/14N]=3·97 ­ 8800/T C 29
weight per cents.34 Later studies, however, by Koyama log[Nb][C+N]=1·54­ 5860/T B 38
et al.68 and Sharma et al.60 have included Wagner log[Nb][C]0·83 [N]0·14 =4·46­ 9800/T B 38
interaction parameters to account for the eVect of log[Nb][C+12/14N]=2·26 ­ 6770/T C 64
alloying elements on the solubility of Nb (CN) in
* Briefly, A: thermodynamic calculations; B: chemical separation
austenite. This leads to a more realistic solubility
and isolation of precipitate; C: equilibrating a series of steels
product, since it incorporates non-unity activity with different niobium contents with a H2 –CH4 atmosphere at
coeYcients. various temperatures, after which the carbon contents are
Although techniques such as chemical separation analysed; D: hardness measurements; E: statistical treatment of
and methane equilibration indirectly account for previous solubility products.
chemical interactions, both have their own limitations. † Q=[Mn](1371/T­ 0·9)­ [Mn]2 (75/T­ 0·0504).

International Materials Reviews 2003 Vol. 48 No. 6


11 Solubility products for various Nb precipitates
in austenite (after Nordberg and Aaronson)7,38
hardness is proportional to the amount of niobium
dissolved into austenite, and subsequently precipi-
tated in ferrite as NbC.38 Although this does occur,
all of the carbon and nitrogen in the alloy is not
necessarily associated with the precipitate. Also,
diYculties arise in separating this hardness increment
from those due to other mechanisms such as grain
size strengthening, solid solution strengthening, and
dislocation strengthening.
A comparison between some of the solubility
relations listed in Table 2 is shown in Fig. 11.7 This
Ž gure depicts the solubility product in austenite K
versus temperature. The results shown by Fig. 11
indicate two important points. First, the solubility
product is substantially lowered as the compound
becomes enriched in nitrogen. Second, the solubility
of a precipitating compound in austenite is decreased
as the vacancy content becomes smaller. Solubility
products play a vital role in understanding the physi-
cal metallurgy of microalloyed steel, especially those
aspects which are concerned with precipitation related
phenomena. Solubility products can be plotted in
either of two ways. Consider the case of the solubility
product of NbC in austenite at some given temper-
ature. If the abscissa (C) and ordinate (Nb) axes have
a linear scale, then any given solubility isotherm will
have the shape of a hyperbola. If, however, the axes
have a logarithmic scale, then the solubility isotherm
will be a straight line. Both approaches have been
widely used in the literature.
A hypothetical solubility isotherm for NbC in
equilibrium with austenite at 1000°C is shown in
Fig. 12a.7 The solubility isotherm gives the locus of
niobium and carbon products which represents the
limit of solid solubility of NbC in austenite at 1000°C,
i.e. any combination of products located above the
line will be in the c+NbC phase Ž eld at 1000°C. The
straight line with a positive slope on the diagram
represents the stoichiometric ratio. The precipitation
of NbC can be followed by considering the slightly
curved line which is nearly equidistant from the
stoichiometric line and passes through the point with
the niobium and carbon coordinates which describe
the composition of the steel. A schematic illustration
is also presented in Fig. 12b. Consider a steel of
composition given by point B which is reheated to
DeArdo Niobium in modern steels 379
a relationship at 1000°C; b interrelationships among
isotherms, steel composition, processing, and potential
amount of precipitation
12 Hypothetical solubility diagrams describing
equilibrium between NbC and austenite (after
DeArdo et al.7 )
1300°C and is very slowly cooled to 900°C. Since
1300°C is the solution temperature for the NbC under
consideration here, the NbC can be assumed to be
completely dissolved after reheating. If it is further
assumed that precipitation occurs during the cooling,
then the composition of the austenite in equilibrium
with NbC would move along the curved line which
passes through point B and which is equidistant from
the stoichiometric line. The distance moved along the
curved line through point B is proportional to the
volume fraction of precipitate formed as a result of
the cooling, assuming equilibrium is established at all
temperatures. In other words, the distance moved
along the curved line through B during cooling is
proportional to the supersaturation and would also
be proportional to the volume fraction of precipitate
formed if equilibrium prevailed. Therefore, if steel B
is reheated to 1300°C and then cooled to and rolled
at 900°C, the distance BC will be proportional to the
volume fraction of precipitation of NbC formed in
austenite. This precipitation would be of two types:
that formed during cooling and that formed during
or after rolling. The Ž rst type is not likely to be
signiŽ cant, since precipitation in recrystallised austen-
ite is normally very sluggish, as will be discussed
below. The latter type is the strain induced precipi-
tation. A second interesting case would occur if steel
International Materials Reviews 2003 Vol. 48 No. 6
380 DeArdo Niobium in modern steels

As stated above, solubility products can be in u-

enced by the presence of elements that do not directly
participate in the precipitation reaction. This eVect
occurs through the in uence of third and higher order
elements in solution on the interaction coeYcients
(and hence on the activities) for niobium and carbon
in austenite. Third elemental solutes that raise the
activity of niobium or carbon through a positive
interaction coeYcient decrease their solubility, while
those that decrease the activity through negative
coeYcients increase it. Interaction coeYcients for
carbon, nitrogen, and niobium in liquid steel at
1600°C are given in Table 3.75 Unfortunately, similar
data for these species in austenite at 1000°C, for
example, do not exist at this time. If it were assumed
that the data of Table 3 apply to austenite at 1000°C
also, then certain solutes would be found to increase
the solubility of carbon and niobium in austenite,
leading to less than expected NbC, while others would
decrease it, leading to more.
Examples of this eVect have been studied by
Koyama et al.68 who have examined the in uence of
several elements on the solubility of NbC in austenite.
By way of an example, the solubility product for NbC
in otherwise unalloyed austenite is about 5×10Õ 3 at
13 Amount of niobium carbide available for 1150°C. An addition of manganese increased log K
precipitation at 923 K (after solution treatment by about 5% per %Mn added, while an addition of
at 1373 K) as function of degree of deviation silicon reduced log K by about 45% per %Si added.
from stoichiometry r: positive values of r
The solubility product at 1150°C for a steel that
indicate C rich compositions, negative values
Nb rich (after Wadsworth et al.73 ) contained 1·5%Mn and 0·4%Si was ~4×10Õ 3; the
addition of manganese and silicon acted to reduce
the solubility product by about 20%.
of composition A in the schematic Ž gure is considered. The latest and perhaps the most accurate solubility
If this steel is reheated at 1300°C and hot rolled at product for nominally stoichiometric NbC in austen-
900°C, then two arrays of NbC particles would be ite in a steel containing 0·08C–1·5Mn–0·008N–0·02Nb
expected. The Ž rst would be the precipitates that was published by Palmiere et al.76 in 1994. The
survived the reheating treatment (volume fraction following expression was derived from solubility data
proportional to AB) and the second, strain induced obtained from reheated and quenched samples as
precipitates (volume fraction proportional to BC). measured using the atom probe facility of an atom
Several investigators have used solubility diagrams probe Ž eld ion microscope (APFIM)
to help explain the physical metallurgy of microal-
loyed steels and, in particular, the precipitation phen- log[Nb][C ]=2·06­ 6700/T . . . . . (20)
omena.70–72 Understanding of precipitation in these
steels has been substantially increased by the work of It is important to note that a comparison of this
Wadsworth et al.,73 Roberts et al.,45 and Keown and product with several earlier ones indicates that the
Wilson.74 The study by Wadsworth et al.73 concluded earlier products substantially overestimated the
that the supersaturation that could be developed
between any two temperatures is a strong function of
the position of the steel relative to precipitate stoichi- Table 3 Influence of several selected elements on
ometry on the solubility diagram. This eVect, illus- interaction coefficients for C, N, and Nb
trated in Fig. 13, shows that the largest possible in liquid iron at 1600°C75
supersaturation occurs when the microalloying Solute C N Nb
element and the interstitial are present in the steel in
the stoichiometric ratio, and deviations from this Al 5·3 5·2 …
ratio will lead to decreasing supersaturations. C 6·9 5·86 ­ 23·7
Cr ­ 5·1 ­ 10 …
The work of Roberts et al.45 presents an interesting
H 3·8 … ­ 1·5
way of considering the precipitation of carbonitride Mn ­ 2·7 ­ 8·1 …
precipitates in microalloyed steels. They suggest a Mo ­ 4 ­ 4·9 …
thermodynamic analysis which attempts to predict N 5·86 0·8 …
the change of x and y in VC N with changes of Nb ­ 23·7 ­ 26 ­ 26
temperature and/or extent of x precipitation.
y Their Ni 2·9 1·5 …
model predicts that nitrogen rich precipitates are the O ­ 22 4·0 ­ 54
Ž rst to form and that nitrogen plays a central role in P 7 6·2 …
controlling the precipitation until it is completely S 6·5 1·4 ­ 5·8
consumed. Si 9·7 5·9 …

International Materials Reviews 2003 Vol. 48 No. 6


14 Solubility for each pure compound, NbC, TiC,
and TiN in austenite and ferrite80,82
amount of niobium in solution in austenite and
underestimated the dissolution temperature.76 This
product has been used successfully in subsequent
research.77
Solubility of NbCN in ferrite
It is well known that the solubility of various elements
is higher in austenite than in ferrite.78,79 An example,
published in 1994, for the solubility of various
microalloyed carbides and nitrides in austenite and
ferrite in an ultralow carbon steel is shown in
Fig. 14.80,82 At about the same time, solubility prod-
ucts for various microalloyed precipitates in ferrite
were published by Taylor.83
It is of interest to compare the expected behaviour
for MC in austenite and ferrite for the various
microalloying elements. One way that this can be
accomplished is by comparing the solubility limits for
the various carbides in austenite and ferrite at a given
temperature, for example, near a typical Ar . The
result of using this approach is given in Table3 4, for
which the solubility limits in austenite and ferrite
were calculated at 800°C using the products suggested
by Taylor.83 This table shows that the limiting product
for NbC is reduced by a factor of nearly 20 when
going from austenite to ferrite at 800°C. A more
encompassing discussion of solubility products, cover-
ing a wide variety of precipitates, can be found in the
recent work of Gladman.81
Table 4 Comparative solubility limits for MC in
austenite and ferrite at 800°C83
Solubility product [M][C]×104
NbC in c 8·9×10Õ 5 the three octahedra shown in Fig. 15 all have a cube
NbC in a 4·5×10Õ 6
4
TiC in c 1·7×10Õ
5
TiC in a 3·0×10Õ
3
VC in c 7·9×10Õ
3
VC in a 1·1×10Õ
DeArdo Niobium in modern steels 381
Although the use of solubility products is helpful
in understanding the behaviour of microalloying
elements in austenite and ferrite, there are some
important restrictions that must be recognised when
applying them. The question is one more of validity
than accuracy. It is indeed true that precipitation
reactions are controlled by solubility relations; how-
ever, the major question is what should be considered
the proper matrix composition. In most cases, the
bulk composition is used to calculate, for example,
the dissolution temperature of a precipitate in austen-
ite. However, many precipitate reactions are governed
by a local composition rather than a bulk compos-
ition. Hence, the in uence of segregation on precipi-
tation must be considered when applying any
solubility equation. This point has been discussed in
some detail by Palmiere et al.77
Crystallography of precipitation: orientation
relationships and lattice matching
The crystallography of precipitation in steel has been
reviewed by Jack and Jack84 and Davenport and
Honeycombe.85,86 There are two aspects of the crys-
tallography of precipitation that are of interest in this
discussion. These are (i) the orientation relationship
that exists between the crystal structure of the precipi-
tate and that of the matrix and (ii) the degree of
lattice registry between the precipitate and the matrix.
Carbonitrides of niobium, vanadium, and titanium
can precipitate in both austenite and ferrite. Several
studies (e.g. Refs. 87–90) have shown that when these
carbonitrides precipitate in austenitic stainless steel,
they do so such that the lattice of the precipitate,
which has a NaCl crystal structure, is parallel to the
fcc lattice of the parent austenite, i.e.
[100] d [100]
M(CN) c
[010] d [010]
M(CN) c
Davenport and co-workers have provided direct evi-
dence that this same relationship holds for the strain
induced precipitation of NbC in austenite in a mic-
roalloyed steel.85,223 When NbC N precipitates in
ferrite85,86,223 or martensite91 it x does
y so with the
Baker–Nutting orientation relationship92
[100] d [100]
NbC a
[011] d [010]
NbC a
The ‘parallel’ and Baker–Nutting orientation relation-
ships can be simply illustrated through the use of the
appropriate metal-atom octahedra.84 The metal-atom
octahedra for austenite, NbC N , and ferrite are
shown in Fig. 15.84 The structuresx y of the austenite
and NbC N are positioned to represent the parallel
x yrelationship and the structures of the
orientation
ferrite and NbC N are positioned to represent the
x y
Baker–Nutting orientation relationship. Note that
plane at the base.
The octahedra shown in Fig. 15 also permit the
calculation of lattice misŽ t strain or the lattice strain
required for coherency. This is done by calculating
the linear strain (e) in the matrix lattice parameter
International Materials Reviews 2003 Vol. 48 No. 6
382 DeArdo Niobium in modern steels

precipitate and the matrix. The strains illustrated in

Table 5 are very large and most likely cannot be
easily accommodated by a few interfacial dislocations.
This probably explains why NbC particles are always
found on boundaries, sub-boundaries, or dislocation
arrays in either austenite or ferrite.
The orientation relationship that is observed
between the NbC and ferrite can be used to dis-
tinguish the NbC which had nucleated in austenite
from the NbC which had nucleated in ferrite. As was
noted above, NbC forms in austenite with a parallel
orientation relationship and in ferrite with the Baker–
Nutting orientation relationship. Therefore, all NbC
precipitates that show the Baker–Nutting relationship
must have formed in the ferrite. The NbC that forms
in austenite will not have the Baker–Nutting relation-
ship with the ferrite.
When austenite transforms to ferrite or martensite,
it does so with the Kurdjumov–Sachs orientation
relationship93,223
(111) d (110)
c a
[110] d [111]
c a
The consequence of this relationship is that when the
matrix transforms from austenite to ferrite, the orien-
tation of the original austenite and the precipitates
that formed in that austenite would be related to
the ferrite by the Kurdjumov–Sachs relationship.
Therefore, the precipitates that formed in the austen-
ite can be identiŽ ed because they will have
15 Metal atom octahedra for a austenite, b NbCN, the Kurdjumov–Sachs orientation relationship with
and c ferrite (adapted from Jack and Jack84 ) the ferrite when observed at room temperature. If the
orientation of the austenite grains were to change
after precipitation had occurred, e.g. by grain rota-
that would be required to bring the two lattices into tions accompanying deformation or recrystallisation,
coincidence at the matrix/precipitate interface, i.e. then there would be no rational crystallographic
relationship between the NbC that had formed in the
L ­ L
%e= p m 100 . . . . . . . . . (21) austenite and the Ž nal ferrite matrix. In principle, this
L change in austenite grain orientation after precipi-
m
where L is the appropriate length of octahedron of tation would not appear to be a very frequent occur-
p
the precipitate and L is the appropriate length of rence, since most of the precipitation that forms
m in austenite is strain induced, and these precipitates
octahedron of the matrix.
Examples of the required strain of the matrix for act to suppress subsequent recrystallisation of the
several types of precipitates are given in Table 5. The deformed austenite in which they formed. (Here-
magnitudes of the elastic matrix strains (e =0·255 for inafter, NbCN will be used to denote NbC N .)
x y
NbC in austenite and e=0·105 and 0·563 for NbC in
ferrite) required for lattice registry would appear to
rule out any large degree of coherency between the Precipitation of NbCN: typical morphologies
and distributions
The precipitation of NbCN in austenite and ferrite is
Table 5 Lattice mismatch for NbCx Ny precipitates heterogeneous in nature, i.e. it always occurs in con-
in austenite and ferrite junction with crystalline defects such as grain bound-
Required linear strain in matrix, %* aries, incoherent twin boundaries, stacking fault
Orientation boundaries, subgrain boundaries, or dislocations. One
relationship NbC NbC0·8 NbN0·8 reason why the precipitation forms in this fashion is
c matrix the rather large mismatch between the lattice of
[100]ppt d [100] 25·5 26·6 23·0 NbCN and the matrix, austenite18,23,31,33–35 or fer-
c
[010]ppt d [101] 25·5 26·6 23·0 rite,15,23,29,34,36–42 as can be seen in Table 5. These
c
[001]ppt d [001] 25·5 26·6 23·0 crystalline defects are sources of dislocations which
c
a matrix can act to cancel some of the elastic strain which may
[100]ppt d [100] 56·3 57·7 53·1 develop during the formation of the precipitates.
a
[011]ppt d [010] 10·5 11·5 8·4 In one of the few studies that have been conducted
a
[011]ppt d [001] 10·5 11·5 8·4 on the precipitation of NbCN in recrystallised austen-
a
*All matrix strains are tensile (+). ite, Santella has shown that the NbCN forms almost
International Materials Reviews 2003 Vol. 48 No. 6

16 Strain induced precipitation of NbCN in
austenite in steel containing 0·09C–0·07Nb:
specimen reheated to 1250°C, rolled 25%, and
held at 950°C, and air cooled to room
temperature (RT) – centred dark field electron
micrograph using (111) NbC reflection (after
Santella 29 )
exclusively on grain boundaries29 as is the case for
precipitation in casting.42 Much more research has
been carried out on NbCN precipitation which has
formed in deformed austenite and in ferrite. Good
illustrations of strain induced precipitates of NbCN
in deformed austenite are available in the litera-
ture94–98,223 and an example is presented in Fig. 16.29
In every case, the precipitates appear to decorate
what was once a grain or subgrain boundary in the
prior austenite.98
When NbCN precipitates in ferrite, the nature of
the precipitate distribution is related to the nature of
the austenite to ferrite transformation. If the ferrite
that forms has a polygonal morphology, the NbCN
precipitation will have the ‘interphase’ distri-
bution.85,86 During the interphase precipitation, the
NbCN forms along the advancing austenite/ferrite
phase boundary. When the boundary moves to a new
location, the precipitates are left behind in a sheet-
like array. The Ž nal microstructure consists of numer-
ous sheets of precipitates, where each sheet denotes
the location of the interphase boundary during the
course of the transformation. This form of precipi-
tation in niobium steels has been observed by Gray
and Yeo99 and others,85,100,101,181–184 and an example
is given in Fig. 17.29 It should be noted that this form
of precipitation only occurs at very high temperatures
in ferrite. When precipitation occurs substantially
after transformation is complete, the precipitate has
a more uniform or general distribution.99 These pre-
cipitates are responsible for the precipitation harden-
ing eVect of NbCN.
Similar general distributions occur in ferrite which
has an acicular or bainitic character.85,181–184
Furthermore, there is great similarity in distributions
and morphologies between the NbCN that forms in
acicular ferrite and the NbCN that forms during
secondary hardening in the tempering of a quenched
steel.91 An example of this general type of precipi-
tation of NbCN in acicular ferrite is given in Fig. 18.29
DeArdo Niobium in modern steels 383
17 Interphase precipitation of NbCN in ferrite in
steel containing 0·09C–0·07Nb: specimen
reheated to 1250°C, hot rolled to 1000°C, and
air cooled to RT – bright field electron
micrograph (after Santella 29 )
18 General precipitation of NbCN in austenite in
steel containing 0·09C–0·07Nb: specimen
reheated to 1250°C, hot rolled to 1000°C, and
air cooled to RT – centred dark field electron
micrograph using (111) NbC reflection (after
Santella 29 )
Niobium thermomechanical
processing and austenite
conditioning
Background
Although niobium can play many important roles in
steel, none is more important than austenite con-
ditioning. Here, austenite conditioning means having
the terminal hot rolled austenite with the proper
microstructure and composition to allow the desired
Ž nal ferrite microstructure to be achieved after suit-
able cooling. For simple ferrite–pearlite steels, this
means controlling (increasing) the crystalline defect
content of the austenite since these defects can act as
nucleation sites for ferrite upon transformation. This
was the original view of austenite conditioning and
was widely discussed in the early 1980s. These defects,
as shall be seen, include grain boundary area,
deformation band area, and incoherent twin bound-
ary area per unit volume. This catalytic eVect has
been discussed previously.102–104 This view of austen-
International Materials Reviews 2003 Vol. 48 No. 6
384 DeArdo Niobium in modern steels
19 Isothermal precipitation of NbCN at
900–1200°C in undeformed austenite in steel
containing 0·07C–0·04Nb–0·010N (after
Simoneau et al.121 )
ite conditioning and its beneŽ ts has since been
expanded to include acicular ferrites, bainites, and
martensites.105 Furthermore, the beneŽ ts of austenite
conditioning have been extended to high strength
strip and sheet with improved stretch forming, e.g.
the dual phase and TRIP (transformation induced
plasticity) steels. Here, the distribution of low temper-
ature transformation products and their contribution
to work hardening is controlled by the defect structure
of the austenite.106–113 This work shows that the Ž ner
the austenite, the Ž ner the distribution of low temper-
ature transformation products, the higher the work
hardening rate and the resulting stretch formability.
Role of processing
Before the role of niobium in austenite conditioning
can be discussed, it is important brie y to review the
various commercial rolling practices being used. Since
niobium can act as either a solute or precipitate in
austenite, each with diVerent eVects, it is important
to recognise where the niobium would be during
rolling, and this will vary with the rolling practice.
For precipitation to occur in austenite, there must
exist adequate levels of both supersaturation and
interpass time. Hence, slow processing in austenite,
typical of reversing plate and heavy section rolling,
20 Influence of strain level on kinetics of NbCN
precipitation at 900°C in steel containing 0·17C–
0·04Nb–0·011N (after Lebon et al.115 )
International Materials Reviews 2003 Vol. 48 No. 6
21 Time–temperature–precipitation (TTP) diag-
ram showing effect of deformation of Nb(C,N)
in austenite (after Watanabe et al.25 )
where the rolling involves many light passes with
long interpass times and long total rolling times, is
rather ideal for extensive static precipitation of NbCN
in austenite. However, fast processing of austenite,
such as is found with strip or bar rolling, where both
the interpass times and the total rolling times are
short, is not conducive for extensive precipitation. It
is, therefore, not unreasonable to assume that niobium
will in uence the behaviour of austenite Ž rst as a
solute and then as a precipitate in plate and structural
rolling, whereas it is expected that niobium would
largely remain in solid solution during strip and bar
rolling.104,114
Behaviour of niobium in austenite
When a niobium bearing low alloy steel is in the
austenite phase Ž eld, the niobium will be in the solid
solution matrix and in the precipitated NbCN, the
partitioning depending on temperature. At equi-
librium, this partitioning of niobium between the
matrix and precipitate will be controlled by the
solubility relations discussed above. Since one of the
primary requisites of a successfully conditioned
22 Influence of per cent deformation on Nb
precipitation in steel containing 0·06C–0·041Nb
and 0·06%N (after Hoogendorn and Spanraft 65 )

23 Precipitation of Nb(CN) and recrystallisation
kinetics during isothermal holding in Nb
bearing steel (after Ouchi et al.123 )
austenite is the presence of a large number of crystal-
line defects that can act as sites for ferrite nucleation
during cooling (i.e. a total near-planar crystalline
defect surface or boundary area per unit volume S ),
v
the motion of subgrain and grain boundaries associ-
ated with static recrystallisation and grain growth
after hot deformation must be retarded. Evidence for
this retardation by niobium, both in solid solu-
tion115–117 and in precipitate,118–120 can be found in
the literature.
The diVerent eVects of NbCN precipitation during
hot rolling have stimulated a large number of studies
concerned with the kinetics of precipitation in austen-
ite as in uenced by composition, strain, strain rate,
temperature, and overall heat treatment. This group
of studies of precipitation kinetics has utilised a wide
variety of techniques including: chemical analy-
sis,20,26,28,43,63,121 electrical resistivity,121,122 X-ray
diVraction,96 quantitative electron microscopy,98  ow
curves,117 and hardness testing.121,123,124 The precipi-
tation that has been studied in these experiments is,
with one exception, the strain induced precipitation
which occurs during and/or after deformation.
The exception is the dynamic precipitation which
accompanies deformation and which has been studied
by Jones and Rothwell119 and Akben et al.125
DeArdo Niobium in modern steels 385
24 Recrystallisation – precipitation – temperature –
time (RPTT) diagram for 0·10C–1·24Mn–0·09Nb
steel after solutionising at 1250°C and hot
rolling 50% at 950 C (after Hansen et al.98 )
Studies of the precipitation in recrystallised austen-
ite have shown that the kinetics is very slug-
gish.25,115,121,122 The results of the study by Simoneau
et al.121 on precipitation rates at temperatures above
900°C are shown in Fig. 19. The kinetics of NbCN
precipitation at 900°C is shown in Fig. 20; these data
are from the work of LeBon et al.115 Finally, the
precipitation kinetics at low temperatures, below
950°C, has also been determined by Watanabe et al.;25
these results are shown in Fig. 21. These studies
illustrate the slow rate of precipitation in recrystallised
austenite; it takes several thousands of seconds at
900°C for 50% of the potential precipitates to form.
Also, by combining the results of Simoneau et al.121
and Watanabe et al.,25 the overall precipitation behav-
iour does appear to conform to C curve kinetics.
The precipitation rate is remarkably sensitive to
the level of strain imparted prior to aging, as is
illustrated by the results of LeBon et al.115 (Fig. 20)
and Hoogendorn and Spanraft65 (Fig. 22). The strain
induced precipitation of NbCN in austenite appears
to follow C curve kinetics. Three C curves, each based
on a diVerent technique, are shown in Figs. 21, 23,
and 24. These results are respectively from the work
of Watanabe et al.,25 Ouchi et al.,123 and Hansen
et al.98 In all three cases, the nose of the C curve
appears to be in the temperature range 900–950°C.
The overall composition of the steel appears to
have a strong eVect on the kinetics of precipi-
tation.25,117,124 For example, the presence of molyb-
denum appears to shift the C curve to lower
temperatures and shorter times,25 whereas an increase
in manganese level acts to shift the C curve to longer
times.125 These eVects are probably caused by the
in uence of these elements on the activities of solute
niobium and carbon, as discussed above.
Studies of the precipitation of NbCN in deformed
austenite are subject to certain problems when the
temperatures of deformation (and subsequent hold-
ing) are raised. The problem arises because the density
of crystalline defects in the austenite does not vary
International Materials Reviews 2003 Vol. 48 No. 6
386 DeArdo Niobium in modern steels
25 Relative integrated intensity versus aging time
for samples deformed 60% at indicated
temperatures (after Davenport et al.96 )
continuously with temperature over a large temper-
ature range, say, 800å T å 1200°C. For a given set
of conditions of composition, temperature, strain, and
strain rate, the austenite deformed at high temper-
atures may experience static recrystallisation between
passes if the interpass time is suYciently long. This
would be expected at large strains, high temperatures,
and high strain rates. The austenite grain structure
deformed at low temperatures will be highly elongated
and the austenite deformed at intermediate temper-
atures will have a mixed equiaxed plus elongated
grain structure. Behaviour of this kind would be
expected to give a discontinuous relationship between
the density of crystalline defects that can act as
nucleating sites for NbCN precipitation and the
deformation temperature. This type of austenite
deformation behaviour has been observed in micro-
alloyed steels46,96 and has been shown to in uence
the results of precipitation kinetic studies.98 An
example of this eVect is shown in Fig. 25, from the
work of Davenport et al.96 Figure 25, where the
relative integrated intensity from NbC diVraction
peaks is assumed to be roughly proportional to the
volume fraction of precipitate, indicates that there are
high rates of precipitation in the temperature range
950–1100°C and lower rates at temperatures which
are either above or below this range. Davenport
et al.96 have noted that the lowering of precipitation
kinetics which results from higher temperature
deformation and holding may be due to two eVects.
The Ž rst is the reduction in solute supersaturation
with increasing temperature; this alone would lead to
a decrease in precipitation rate. The second is the
change in deformation structure of the austenite near
1100°C. Austenite rolled below this temperature exits
the rolls with an elongated, deformed microstructure
that contains numerous sites for strain induced pre-
cipitation. Austenite rolled above this temperature
quickly achieves the statically recrystallised state; this
structure contains few sites for strain induced precipi-
tation; hence, low precipitation kinetics can be
expected.
The studies of precipitation kinetics described so
far have involved ‘static’ precipitation, i.e. the austen-
International Materials Reviews 2003 Vol. 48 No. 6
26 TTP curves for dynamic precipitation of NbCN
in austenite (after Akben et al.125 )
ite has been deformed in one operation and then aged
in a second operation. A diVerent approach to study-
ing strain induced precipitation has been developed
by Jonas and co-workers.117,125 This approach enables
the kinetics of strain induced precipitation to be
determined in a ‘dynamic’ context, i.e. where the
austenite is being strained and aged at the same time.
This approach yields C curves describing the dynamic
precipitation and is based on the analysis of hot  ow
curves. C curves describing this dynamic precipi-
tation125 are shown in Fig. 26.
Principles of austenite conditioning
There are two diVerent approaches available for
austenite conditioning during hot rolling. They are
known as conventional controlled rolling (CCR) and
recrystallisation controlled rolling (RCR). The pur-
pose in this paper is to describe these practices, to
show how they diVer from conventional hot rolling
(CHR), and to indicate the microalloying additions
required for their respective implementation.104,114,126
In addition, the requirements of the mill equipment
suitable for these practices are discussed.
Conditioning of austenite
Obtaining Ž ne ferrite grain sizes from the transform-
ation of austenite requires high rates of ferrite
nucleation, and low rates of growth and subsequent
coarsening. Conditioning of austenite means that the
microstructure of the austenite has achieved, through
controlled hot deformation, the proper predetermined
metallurgical state or condition prior to transform-
ation that will embody these requirements. It was
shown earlier that high ferrite nucleation rates result
from having a large number of potential nucleation
sites and a high nucleation rate per site.103,127,128 The
sites for ferrite nucleation are austenite grains and
incoherent twin boundaries and deformation
bands.103,129–134 The density of these sites per unit
volume is expressed as the total interfacial area of the
near-planar boundaries per unit volume and has the

27 Ferrite grain sizes produced from recrystallised
and unrecrystallised austenite at various Sv
values 136
units mm2/mm3 or mmÕ 1. This stereological concept
was Ž rst discussed by Underwood,135 and was later
adopted by Kozasu et al.129 to describe austenite.
Kozasu et al. called this measure of the total interfacial
area per unit volume, the parameter S . The para-
v 95% the reheat temperature
meter S can be considered as an approximate austen-
v
ite ‘grain size’, and its magnitude is an indication of
the degree of austenite conditioning. The principal
goal of thermomechanical processing and austenite
conditioning is to maximise S . The in uence of S
on ferrite grain size is shown inv Fig. 27.136 v
There are two entirely diVerent approaches to
increasing S . In the Ž rst, the initially recrystallised
v
reheated austenite grains undergo repeated recrystal-
lisation during subsequent hot deformation leading
to grain reŽ nement. This deformation would take
place at temperatures above T
95%
in Fig. 28. Here,
one set of equiaxed grains is replaced by a new set of
Ž ner equiaxed grains. Since these Ž ne grains would
have a strong tendency to coarsen during the interpass
time, this Ž ne grain size can be retained only if a
suitable mechanism is available to suppress grain
28 Schematic illustration of austenite micro-
structures resulting from various deformation
conditions:49 T95% and T5% are temperatures for
95 and 5% recrystallisation respectively
DeArdo Niobium in modern steels 387
29 Schematic representation of austenite micro-
structure when deformed above or below
recrystallisation-stop temperature of aus-
tenite, with corresponding description of Sv :
superscripts GB, DB, TB, and NPD denote
contribution to total Sv from grain boundaries,
deformation bands, twin boundaries, and near-
planar defects139
coarsening. The process that involves repeated recrys-
tallisation and inhibition of grain coarsening is called
recrystallisation controlled rolling (RCR), and was
originally proposed by Sekine and co-workers.137,138
Clearly, the lower the T , the larger will be the
processing window between
and the minimum Ž nishing temperature of T .
Hence, steels designed for RCR conditioning must 95%
have a low recrystallisation-stop temperature and a
pre-existing grain coarsening inhibition system. This
practice is well suited for production conditions
that require high Ž nishing temperatures, e.g. under-
powered rolling mills, heavy section and heavy plate
rolling, and forging.
The second approach involves substantial deforma-
tion below T , where, for example, the thickness at
T is three 5%
or four times the Ž nal gauge. In this
5%
method, the grain at T is deformed and stays
5%
unrecrystallised through the interpass time for all
subsequent passes. Hence, there is a change in grain
shape and the occurrence of transgranular twins and
deformation bands. The process that involves
repeated  attening or ‘cold working’ of the grains by
repeated deformation below T is called conven-
5%
tional controlled rolling (CCR). Clearly, the higher
the T , the greater the number of passes that can be
used 5%and the more eVective is the practice. Hence,
steels designed for CCR conditioning must have a
high recrystallisation-stop temperature.
Although there are two diVerent approaches to
austenite conditioning, i.e. RCR and CCR, they both
have the same objective of resulting in structural
reŽ nement in the Ž nal plate, coil, beam, or forging.
The diVerences between RCR and CCR can be easily
understood with the aid of Fig. 29 which shows the
diVerent paths used by each to achieve high values
of S .139
v increase in S for the RCR practice comes
The
v
exclusively from the increase in grain boundary area
per unit volume which results from a decrease in
average grain volume, while the increase in S for the
v grain
CCR practice results from the increase in
boundary area per unit volume resulting from a
International Materials Reviews 2003 Vol. 48 No. 6
388 DeArdo Niobium in modern steels
30 Grain size evolution through succeeding
passes depending on thermomechanical
processing161,224
change in grain shape and through the addition of
the transgranular twins and deformation bands.
Physical metallurgy of thermomechanical
processing: driving forces and retarding
forces
It is evident from the above that temperatures such
as T and T are signiŽ cant in the consideration
95% for thermomechanical
of steels 5% processing. Steels that
are to be used for RCR processing must achieve,
through microalloying, two essential features. First,
they must have a relatively low T to enable the
15–21 passes needed, for example, 95%for wide plate
rolling to take place considering the usual drop in
temperature of about 300–400°C from the reheat to
the Ž nishing temperature. Second, they must have a
pre-existing grain boundary motion inhibition system
that is weak enough to permit static recrystallisation,
but suYcient to suppress post-recrystallisation grain
coarsening. The diVerence in high temperature rolling
behaviour of austenite during conventional hot rolling
(CHR) and RCR is shown schematically in
Fig. 30.161,224 The principal diVerence between the
two is that the RCR steel has a grain coarsening
inhibitor, whereas the CHR steel does not.
Grain coarsening and its driving force
Fine grain sizes result from high rates of ‘nucleation’
and low rates of ‘growth’ during recrystallisation and
from the inhibition of subsequent grain coarsening.
Grain coarsening occurs in polycrystalline metals and
alloys because of the need to approach an equilibrium
condition. This condition requires that the grain
boundaries move to reduce curvature and to meet at
certain angles.140 The non-equilibrium forces on grain
boundaries cause them to move toward their respect-
ive centres of curvature. This boundary motion causes
the smaller grains to disappear and the larger grains
to grow with a net increase in grain size with, for
example, holding time at a suitable temperature. The
pulling force per unit area of grain boundary is
F=DG/V . . . . . . . . . . . . (22)
M in grain sizes that can be pinned by these two very
where DG is the lowering of the free energy caused
by the boundary migration and V is the unit volume
M to any boundary
of material. Equation (22) applies
whose motion leads to a lowering of the free energy.
International Materials Reviews 2003 Vol. 48 No. 6
For grain coarsening, the DG arises from the boundary
curvature and, in the case of Ž ne grained austenite,
leads to a driving force of approximately 100 kPa.141
For recrystallisation, the DG is a result of the dislo-
cation density diVerence between the unrecrystallised
and recrystallised grains. A typical driving force for
the recrystallisation of austenite as a result of a single
low temperature rolling pass is approximately
20 MPa,49,77,142,143 or about 200 times larger than
that for coarsening.
The coarsening of austenite grains has been treated
by several investigators.144–146 The approach adopted
here is that of Gladman.147–149 Gladman suggested
that grain coarsening is accompanied by two types of
boundary motion: the inward motion of boundaries
of shrinking grains and the outward motion of bound-
aries of growing grains.147–149 An energy balance
showed that only grains with a size larger than 4/3
that of the average could grow. This clearly shows
the importance of homogeneity in grain size distri-
bution. According to this view, very narrow grain size
distributions would be expected to show little, if any,
grain coarsening, independent of average grain size.
Coarsening inhibition using particles
While the tendency for grain coarsening will always
be present in real steels, it is important to realise that
this coarsening tendency can be slowed or eliminated
by the presence of suitable forces that can inhibit
boundary motion. Two of the more important mech-
anisms are particle pinning and solute drag. When a
boundary intersects a particle, a portion of the bound-
ary is eliminated. To pull the boundary away from
the particle requires the re-creation of boundary area,
and, hence, the application of a force. This force is
the force for grain coarsening. Gladman’s treatment
of grain coarsening also included the eVects of
particles.147–149 Gladman’s analysis is based on the
earlier work of Zener.150 His energy balance consider-
ing the shrinking of small grains, the expanding of
large grains, and the pulling of boundaries away from
particles, led to what is considered the Gladman
equation
r =6R f [p(3/2 ­ 2/Z)]Õ 1 . . . . . . (23)
c 0
where r is the critical or maximum particle size
c
required to suppress coarsening, R is the radius of
0
the grain of average size, f is the volume fraction of
particles, and Z is the ratio of radii R/R of growing
0
to average grains. Hence, for any average grain size,
there exists a combination of particle size and particle
volume fraction that can suppress coarsening. The
Gladman equation is shown in Fig. 31, assuming Z=
1·5.149 The negatively sloping lines represent combi-
nations of r and f that can pin all grain sizes which
c
fall above the line. The Ž nest grain sizes that can be
stabilised by particles in real steels are shown in the
lower hatched area. The upper hatched area shows
the Ž nest grains that can be pinned by large, widely
spaced inclusions. The nearly two orders of diVerence
diVerent particle distributions is striking.
The particles best suited for inhibition of grain
coarsening are those that have the smallest size and
the largest volume fraction. This requires that the

31 Effect of particles on inhibition of grain
growth149
particles have a low solubility in austenite and them-
selves coarsen slowly. The volume fraction will be
governed by the bulk composition, the appropriate
solubility product, and the temperature, while the
particle size will be controlled by the rate of coarsen-
ing of the particles. The particle coarsening rate due
to Lifshitz–Slyozov–Wagner151,152 is given as
r3­ r3 =8V 2cC*Dt[9kT ]Õ 1 . . . . . . (24)
o
where r is the average particle size at time t, r is the
o recrystallised matrix. This can be estimated by using
initial average particle size, V is the atomic volume
of the precipitate, c is the particle/matrix surface
energy, C* is normally taken to be the concentration
of the metal atom in solution at the particle interface,
D is the diVusion coeYcient of the metal atom in the
matrix, t is the coarsening time, k is Boltzmann’s
constant, and T is the absolute temperature. Equation
(24) shows that particle coarsening rates will be
reduced when c, C*, and D are as low as possible.
The choice of particle will determine these three
parameters. The structure and local composition of
the particle/matrix interface determine c, the con-
trolling factor being the lattice matching across the
interface. Once the particle is chosen, the matching is
essentially set. The more coherent the interface, the
lower the c. The magnitude of C* is determined by
the composition of the particle, the nature of the
particle, and the solubility product. Systems that have
low concentrations of metal atoms in the matrix in
equilibrium with the particles obviously have a low
C*. Two requisites for low C* are a low solubility
product for the particle in austenite and a rather high
interstitial level. This means that systems with low
solubility products will have not only large volume
fractions of particles, but also low rates of coarsen-
ing.149,153 Alternatively, low particle coarsening rates
would be favoured by steels that are hypostoichi-
ometric in the interstitial element, i.e. on the right-
hand side of Fig. 13.
Coarsening inhibition using solute drag
When solute atoms are located on and attached to a
grain boundary, the migration rate of the boundary
DeArdo Niobium in modern steels 389
resulting from a given driving force will be in uenced
by the solute atoms. The theory of solute drag
has been the subject of numerous studies and
reviews.154–156 Most of this work was conducted in
recrystallisation experiments where, as discussed
above, the driving forces are rather high. Grain coars-
ening, on the other hand, takes place at very low
driving forces.
According to Cahn’s theory,155 the rate of grain
growth or coarsening G can be given as
G2=(2sV n/t)/(l +aCê ) . . . . . . . (25)
M
where s is the speciŽ c grain boundary energy per
unit volume, V is the molar volume of austenite, n
M
is the isothermal grain coarsening law exponent, t is
the coarsening time, l is the reciprocal boundary
mobility of ‘pure’ austenite, a is the reciprocal bound-
ary mobility at unit solute concentration, and Cê is
the bulk solute concentration. The linear relationship
predicted by this theory between G2 and Cê has, in
fact, been observed.156 The coarsening rate will fall
with increasing solute levels. The choice of solute will
in uence the values of s, V , n, and a; hence, each
M
solute would be expected to exert its own particular
in uence on G.
Recrystallisation: driving force for
recrystallisation
The driving force for static recrystallisation can be
taken to be the diVerence in dislocation density
between the deformed or recovered matrix and the
the approach of Keh and Weissmann157 who related
the increase in  ow stress caused by work hardening
to the increase in dislocation density. Their
expression is
Ds=s +amb(Dr)1/2 . . . . . . . . (26)
0
where Ds is the observed increase in  ow stress
caused by work hardening, s is the initial yield
0
strength, a is a constant, m is the shear modulus, b is
the Burgers vector, and Dr is the change in dislocation
density. The driving force for recrystallisation is then
given by
F =mb2Dr/2 . . . . . . . . . . (27)
RXN
As mentioned above, driving forces resulting from
individual hot roll pass simulations have been
found to be on the order of 20 MPa using this
technique.49,77,142
Recrystallisation: pinning forces resisting
recrystallisation
When reheated austenite is deformed at a lower
temperature where the austenite is supersaturated
with respect to the precipitating species, a temperature
will be reached where strain induced precipitates will
form in the deformed or recovered austenite thereby
preventing the nucleation of recrystallisation that
would otherwise occur during the interpass time in
the absence of precipitation.49,77,142,159 The temper-
ature where this initially occurs is shown as T in
Fig. 28. At a somewhat lower temperature, the95% sup-
pression of nucleation is complete resulting in T in
5%
International Materials Reviews 2003 Vol. 48 No. 6
390 DeArdo Niobium in modern steels

Fig. 28. The most homogeneous austenite micro-

structures result from rolling either above T or
below T . 95%
5%
To retard recrystallisation, the pinning forces devel-
oped by the strain induced precipitate must be larger
than the 20 MPa driving force available to cause
recrystallisation. The pinning force can be given as
F =4rsN . . . . . . . . . . . (28)
PIN S
where r is the particle radius, s is the interfacial
energy of the austenite grain boundary, and N is the
S have
number of particles per unit area. Three models
been proposed for N in calculating the pinning forces
S
exerted by arrays of microalloy precipitates.98,158
These are the rigid boundary, the  exible boundary,
and the subgrain boundary models. Recent work has 32 Effect of deformation temperature on
shown that the subgrain model of Hansen et al.98 is softening behaviour in 0·002%C and 0·002C–
probably the most realistic.49,77,142 This model pre- 0·097Nb steels (from Yamamoto et al.162 )
dicts the pinning force to be
FS =3sf l(2pr2)Õ 1 . . . . . . . . . (29) (ii) that they can generate retarding forces that are
PIN V small enough to allow static recrystallisation to occur
where s is the austenite grain boundary energy, f is
V but large enough to suppress grain coarsening
the volume fraction of particles, l is the subgrain size, between passes. Hence, both the dispersion character-
and r is the particle radius. From equation (29) it
istics and particle stability at the start of hot rolling
appears to be rather simple to measure an average r
are important. The behaviour of austenite during
and to measure or calculate the f . An example of
the precipitates causing this pinning V eVect in a
RCR processing is shown schematically in Fig. 30.
niobium steel is shown in Fig. 16.29
It is immediately apparent from Fig. 16 that the Role of niobium in thermomechanical
particles are not randomly distributed, but appear to processing
have formed on the subgrain boundaries of deformed The value of an alloying or microalloying addition
and recovered austenite. Clearly, an average value for to a steel from an austenite conditioning standpoint
N would be inappropriate, since the distribution of for controlled rolling derives from its ability to gener-
S
particles is so localised. It has been found that true ate suYciently large pinning forces to retard the
localised pinning forces at the sites for preferential motion of crystalline defects such as dislocations,
precipitation are far higher than those expected from subgrain boundaries and grain boundaries during
a uniform distribution of particles.49,77,142 Equation recovery and recrystallisation, that would otherwise
(29) shows that high pinning forces will be promoted occur to lower the free energy of the system. Hence,
by high values of f and small particle sizes. Those of interest here is a quantitative assessment of how a
locations with veryV high localised values of f , i.e. given addition can retard recovery and recrystallis-
V
grain and subgrain boundaries, as shown in Fig. 16, ation. This is accomplished through the use of the
will have especially high pinning force magnitudes. well known ‘double hit test’,77,142,159,160 where a series
In order to have the largest number of passes of samples is reheated to a given temperature and
possible below T in Fig. 28, it is clearly necessary then cooled to the temperature of interest. One speci-
to have T in the 5% correct temperature range. This men is deformed continuously to a large strain and
5%
means that the strain induced precipitation should the resulting  ow curve acts as a reference for the
begin at a temperature which is well up in the range interrupted tests to follow. Other specimens are
of temperatures where the rolling passes actually deformed to a constant pre-strain, often approximat-
occur. If the precipitate forms at too high a temper- ing a rolling pass strain, after which they are unloaded,
ature, it will not be able to generate a suYciently held at temperature for various delay times, then
high pinning force to suppress static recrystallisation reloaded past yielding. Some measure of softening is
during rolling at lower temperatures. If it forms at recorded for each holding time and plots of fractional
too low a temperature, T will be too close to the softening versus holding time at a given temperature
5% enough passes to occur
Ž nishing temperature to allow are constructed.77,142,159,160 Earlier work on austenite
to achieve a suYciently high S . has shown, approximately, that the initial 25% soften-
v
The idiosyncrasies of RCR processing will not be ing is caused by static recovery while the remaining
discussed here in great detail, as they have been 75% is due to recrystallisation.77,142
presented elsewhere.104,114,225 SuYce it to say that This technique has been used in numerous previous
two systems seem particularly well suited for engen- studies,77,142,159,160 and niobium has repeatedly been
dering RCR processing. These are either undissolved shown to strongly retard the progression of both
TiN particles or a combination of undissolved NbCN recovery and recrystallisation of austenite.49,77,142,161
particles plus solute molybdenum.104,114,225 The pre- An example of the retarding eVect of solute niobium
requisites for a system capable of allowing RCR is shown in Fig. 32 from the work of Yamamoto
processing are: (i) that they are able to survive the et al.162 In the experiment leading to Fig. 32, the
reheating process with an adequate dispersion, and niobium is assumed to be in solution in the austenite.
International Materials Reviews 2003 Vol. 48 No. 6

33 Comparison of effects of Nb, V, and Ti on
softening behaviour in 0·002%C steels (from
Yamamoto et al.162 )
Hence, the solute niobium was responsible for delay-
ing the onset of recrystallisation at 850°C from 20 s
in plain carbon steel to 400 s in the niobium steel for
a true pre-strain of 0·69 applied at a strain rate of
10 sÕ 1. A comparison of the solute retarding eVects
of the three microalloying elements niobium, titanium,
and vanadium is shown in Fig. 33.162 It is obvious
that vanadium has the weakest retarding eVect, with
titanium intermediate and niobium the strongest. It
is also clear that the behaviour exhibited in Fig. 33
represents the same trend as can be observed in the
Cuddy diagram in Fig. 34.158
A comparison of the relative retarding eVects of
niobium as a solute and as a precipitate is shown in
Fig. 35.162 In Fig. 35, the niobium in the 0·002%C
steel is in solution, whereas the niobium in the
0·019%C steel is present as precipitates. The precipi-
tation has caused an order of magnitude longer
delay for the onset of recrystallisation and an even
much larger retardation in the progress or rate of
recrystallisation.
A somewhat diVerent example of this softening
behaviour is shown in Fig. 36142 where the behaviour
of three steels is compared. The diVerence in the
nature and magnitude of the in uence of solute
niobium and precipitated niobium is clear in Fig. 36.
The arrows shown in Fig. 36 denote the earliest
appearance of strain induced precipitation in austen-
ite. Solute niobium is capable of suppressing recrystal-
34 Increase in recrystallisation temperature with
increase in level of microalloy solutes in 0·07C–
1·40Mn–0·25Si steel158
DeArdo Niobium in modern steels 391
35 Effect of C content on softening behaviour in
0·10%Nb steels at 900°C (from Yamamoto
et al.162 )
lisation for short times while precipitated niobium
can do so for much longer times. Hence, either solute
or precipitated niobium would be expected to sup-
press recrystallisation in strip mill rolling which has
short interpass times, while solute niobium alone
could not in a reversing plate mill with much longer
interpass times. In this case, precipitated niobium
would be needed.
When a deformed, supersaturated solid solution of
niobium in austenite is held at diVerent temperatures,
there is a competition concerning kinetics between
precipitation and recrystallisation and their inter-
action. Several studies have attempted to treat this
rather complex situation (e.g. Ref. 117). One example
of such work is shown in Fig. 37,142 where the behav-
iour of plain carbon steel is compared to that of a
niobium steel.
In an attempt to explain the behaviour shown in
Fig. 34, Cuddy tried to calculate the pinning force
required by NbC particles to suppress recrystallis-
ation.158 This was done by equating the driving force
for recrystallisation with the pinning force generated
by the particles. In Cuddy’s work, the volume frac-
tions required for the pinning force estimate were
calculated using published solubility equations. He
found that the assumed homogeneously distributed
array of particles could suppress recrystallisation. In
later work, Palmiere et al.77 conducted a similar
experiment, but where the characteristics of the pre-
cipitate array were taken from experimental obser-
36 Static restoration behaviour of three steels at
900 and 1000°C: arrows denote first observable
precipitation (from Kwon and DeArdo142 )
International Materials Reviews 2003 Vol. 48 No. 6
392 DeArdo Niobium in modern steels
R¾s and R¾f refer to start and finish of recrystallisation
respectively, in HSLA steels; Rs and Rf refer to start and
finish of recrystallisation respectively, in plain carbon
steels; P¾s and P s refer to hypothetical precipitation start
times in deformed and undeformed austenite respectively;
P s is actual precipitation start time
37 RPTT diagram showing interaction between
precipitation and recrystallisation 142
vations. This work diVered from Cuddy’s in one
important way. In the study by Palmiere et al., real
and local volume fractions and particle sizes were
measured on austenite grain and subgrain boundaries
and in the grain interiors. The pinning forces that
resulted from these observations showed that the
forces calculated at the grain boundaries were well in
excess of the driving force and could easily be respon-
sible for the observed suppression of recrystallisation.
These results are summarised in Fig. 38.77
While it is now possible to explain why the niobium
in the Cuddy diagram (Fig. 34158) behaves the way it
does, i.e. that the recrystallisation-stop temperature
increases strongly with increasing bulk niobium con-
tent, it remains to be explained why the microalloying
elements diVer in eVectiveness from one another. One
possible explanation is given in Fig. 39,161,224 where
the temperature dependence of the driving force for
precipitation, i.e. the supersaturation, is plotted versus
temperature for various microalloying precipitating
systems. Of all the possible precipitating systems, only
NbC can have high supersaturations over a large
portion of the typical hot rolling temperature range.77
E3: 0·08C–1·45Mn–0·02Nb; E4: 0·08C–1·45Mn–0·09Nb
38 Comparison between FPIN and FRXN versus
deformation temperature: data to right of FRXN
will result in complete suppression of
austenite recrystallisation and data to left of
FRXN will result in partial or fully recrystallised
austenite microstructure77
International Materials Reviews 2003 Vol. 48 No. 6
MAE=microalloying elements
39 Precipitation potential of various microalloying
systems161,224
Niobium and transformation of
austenite
Presented here are the manifestations of niobium
beyond, and in opposition to, the ferrite grain reŽ ning
eVects attributed to the austenite conditioning
described above. The high S that results from con-
trolled rolling by itself wouldvlower the hardenability
of the austenite. In order to obtain lower transform-
ation temperatures and low temperature transform-
ation products, this eVect must be overcome by a
combination of higher alloying and/or accelerated
cooling. When relatively large amounts of niobium
are in solution in austenite at the transformation
temperature, i.e. high niobium contents, low carbon
contents, and high reheating temperatures, it often
has an important eVect on the CCT (continuous
cooling transformation) diagram and subsequent
transformation during continuous cooling. This is
manifested as both lower transformation start temper-
atures and a higher probability of achieving non-
polygonal ferrite microstructures, especially at higher
cooling rates. The Ž rst eVect has been reviewed earl-
ier,7,20,65,163,164 and a good example is shown in
Fig. 40.165
More recent work has shown that more than just
the temperature is altered by solute niobium. The
increase in hardenability also means that for the same
40 Corrected Ar3 temperatures of microalloyed
steels with standard austenite grain size of
100 mm165

41 Effects of Nb, V, and Ti on volume fraction of
bainite and ferrite grain size in accelerated
cooled steels165
general conditions, the niobium steels will exhibit
larger amounts of low temperature transformation
products such as acicular ferrite, Widmanstätten fer-
rite, and bainitic ferrite, particularly at higher cooling
rates.165,166 An example of this eVect is shown in
Fig. 41.165
Particularly striking about the eVects shown in
Figs. 40 and 41 is the in uence on properties (see
Fig. 42).165 While it is true that the niobium addition
had a strong eVect on ferrite grain size (Fig. 41), the
principal eVect was on the nature of the ferrite. At
0·04%Nb, about 80% of the ferrite was bainitic
(Fig. 41). It seems obvious from Fig. 42 that the
strength increased directly with the amount of bainite.
Several similar examples have also been shown else-
where.166–171 The question then arises as to how the
solute niobium is responsible for the improvement in
fracture appearance transition temperature (FATT).
Is it the grain reŽ nement of 20% of the structure that
is polygonal ferrite, or is it the presence of low carbon
bainite? In simple terms, this eVect on toughness is
probably due to the niobium acting as both a harden-
ability agent and a grain reŽ ner.
42 Effects of addition of Nb, V, and Ti on tensile
strength and Charpy V 50% FATT of a air
cooled, b accelerated cooled, and c direct
quenched steels after controlled rolling 165
DeArdo Niobium in modern steels 393
The transformation products resulting from the
decomposition of austenite containing niobium as a
solute were found to have a surprisingly high dislo-
cation density. This was to be expected for acicular
ferrite and bainite, which, by their nature, exhibit very
high dislocation densities.168,169 Not expected were
the rather high dislocation densities found in what
is usually called polygonal ferrite in niobium
steels.167,168,172–174 This often overlooked character-
istic of niobium steels has very important ramiŽ -
cations regarding mechanical properties as will be
shown below.
Niobium and strengthening
There is no doubt that the addition of niobium to
most low alloy steels results in higher yield strengths.
This is true for both plate and strip products.
Remaining elusive is the precise way that niobium
causes this eVect. The analysis of strength usually
starts with the expanded Hall–Petch equation, where
a linear additivity of strength components is assumed
YS =[YS +DYS +DYS +DYS ]
obs P–N SS texture disl
+DYS +k DÕ 1/2 . . . . . (30)
pptn y a
where YS is the observed yield strength; YS is
the latticeobsfriction stress; DYS , DYS , DYSP–N ,
and DYS SS
are the stress increments texture by solid
caused disl
pptn
solution, texture, dislocation, and precipitation; and
k DÕ 1/2 represents the contribution by the ferrite
y a size.
grain
The universal use of this equation has been ques-
tioned lately from several perspectives. First, the
general accuracy of this approach has been contested
and new summation approaches have been sug-
gested.175–179 It appears that the linear additivity
approach sometimes overestimates the observed yield
strength, and other means of summing, such as a root
mean square summation, might be more appro-
priate.175–178 However, recent study of several 350
grade strip steels containing niobium showed that the
linear approach was reasonably accurate.172,173,180
Second, if the equation is inaccurate or invalid, then
its use may lead to erroneous interpretations as to
how niobium actually strengthens ferrite. For
example, the precipitation hardening increment has
often been determined in the literature by subtracting
all other components from the measured yield
strength. Clearly, this would lead to incorrect values
of precipitation hardening if the linear summation
law were incorrect or inappropriate.
Some of the potential problems encountered in
ascribing attributes to niobium in terms of contri-
butions to strength are as follows. Consider the
contribution caused by grain reŽ nement when what
is taken to be polygonal ferrite is formed at decreasing
transformation temperatures, e.g. as a result of accel-
erated cooling in a controlled rolled niobium steel
that still has a portion of the initial niobium remaining
in solution. While there is little doubt that there is a
reduction in ferrite grain size with falling transform-
ation temperature, there is also a subtle change in
both the nature of the ferrite and in the grain bound-
aries.174 As the transformation temperature decreases,
there is an increase in grain shape irregularity and
International Materials Reviews 2003 Vol. 48 No. 6
394 DeArdo Niobium in modern steels
boundary rugosity, and a higher proportion of the
grain boundaries assume a lower angle character.
Hence, it seems most unlikely that the same Hall–
Petch slope would be valid for both the high angle
grain boundaries found in the larger polygonal ferrite
formed at higher transformation temperatures and
the low angle grain boundaries found in the smaller
non-polygonal ferrite formed at the lower temper-
atures. This would lead to an overestimation of the
contribution of grain size to strength at lower trans-
formation temperatures.
The precipitation of NbC in ferrite is often taken
to be a major contributor to the strength of ferrite.
While this is widely assumed to be the case, there is
very little direct evidence that any observed precipi-
tation actually results in signiŽ cant amounts of
precipitation hardening of ferrite by NbC that occurs
in commercial niobium steels, particularly strip steels.
There are good reasons for this lack of evidence.
First, the NbC lattice does not Ž t particularly well in
either austenite or ferrite, as was shown by the lattice
misŽ t strain in Table 5. This means that all precipi-
tation will be either semi-coherent or incoherent with
the matrix, and will require crystalline defects such
as dislocations, deformation bands, or subgrain and
grain boundaries to aid in the nucleation.
As mentioned above, there have been two types of
precipitation formed in ferrite in transformed microal-
loyed steel: interphase and general. Interphase precipi-
tation forms in the ferrite at the advancing austenite/
ferrite interface as the transformation occurs. This
leads to a sheet-like distribution where the carbides
exhibit only one of the three possible variants of the
Baker–Nutting orientation relationship.85,86,181,182,223
Importantly, interphase precipitation of NbC is usu-
ally found in ferrite formed at temperatures above
700°C. General precipitation forms from a supersatu-
rated low temperature acicular or bainitic ferrite after
the completion of the transformation, i.e. a form of
direct aging. General precipitation exhibits all three
of the variants of the Baker–Nutting orientation
relationship.181,182 Examples of interphase and gen-
eral precipitation are shown in Figs. 17 and 18.29 The
formation of interphase precipitation has been studied
in niobium steels,182–184 and its occurrence is shown
in the TTT (time–temperature–transformation) dia-
gram of Fig. 43182–184 for a steel containing 0·07C–
1·07Mn–0·036Nb. These results were later veriŽ ed by
Thillou et al.16,172,173 in a steel containing 0·08C–
1·3Mn–0·025Nb.
It is immediately apparent from Fig. 43, that only
very speciŽ c thermal paths will intersect the precipi-
tation region, even after it is converted to a CCT
diagram. For example, the chances of encountering
interphase NbC in a commercial strip product seem
highly remote, given that the strip would be water
spray cooled at 30–80 K sÕ 1 to around 650–600°C
and then very slowly cooled to room temperature.
This cooling path would completely miss the region
of the TTT or CCT diagram for interphase precipi-
tation, i.e. temperatures in excess of 700°C. Similarly,
the accelerated cooling of plate at 10 K sÕ 1 to around
600°C or below, followed by air cooling, would also
be expected to largely miss this region on the TTT
diagram. However, the air cooling of plate at below
International Materials Reviews 2003 Vol. 48 No. 6
43 Schematic TTT curves for Fe–0·036Nb–0·09C
and Fe–0·036Nb–0·09C–1·07Mn alloys: inter-
phase precipitation (IP) occurs in certain
shaded areas – combined data from Sakuma
and Honeycombe 182–184 and Thillou (0·025Nb–
0·07C–1·1Mn, dotted line)16
1 K sÕ 1 from the Ž nish rolling temperature might be
expected to penetrate the interphase region. Sakuma
and Honeycombe have suggested that the prerequisite
for the formation of interphase precipitate is a balance
between the ferrite growth rate and the supersat-
uration of NbC in ferrite.182–184 Interphase precipi-
tation is favoured by slow growth rates and high
supersaturations of NbC. It is quite important to note
that little general precipitation was found below
700°C in the Sakuma and Honeycombe study. Again,
this was also found by Thillou et al.16,172,173 In a
similar vein, little precipitation hardening was found
in studies of both laboratory processed (i.e. coiled)
and commercial 350 grade niobium high strength
strip.16,172,173,185,186
These observations regarding precipitation in strip
steels are important, but it should be kept in mind
that the niobium levels used in these steels were
0·035% or less. It is interesting to consider what
might be expected in steels of higher niobium content
and where the Nb/C ratio is closer to stoichiometry,
which would maximise the volume fraction of NbC
available for precipitation. In this regard, TEM
studies of commercial strip steels containing
0·09C–1·2Mn–0·1Nb223 and 0·08C–1Mn–0·065Nb187
revealed that little, if any, Ž ne scale precipitate was
present in either steel. On the other hand, a recent
TEM study of a high niobium plate steel designed
for high temperature processing, designated HTP
steel, containing 0·03C–1·5Mn–0·08Nb and air cooled
to room temperature after rolling, revealed that pre-
cipitation in ferrite representing approximately
80 MPa of Orowan–Ashby strengthening was
found.188,211 Also, precipitation hardening approach-
ing 100 MPa has been found in air cooled plate
containing 0·04%Nb and a ferrite–pearlite micro-
structure that exhibited a yield strength of
450 MPa.189 These values of 80–100 MPa probably
represent the maximum amount of real precipitation
hardening achievable in niobium steels with normal
compositions and processing.
It is interesting to note that the magnitude of
precipitation hardening found in the plate steels, i.e.
80–100 MPa, is fairly closely predicted by two dia-
grams representing precipitation hardening according
 DeArdo Niobium in modern steels 395

to the Orowan–Ashby theory. If one assumes: (i) part-

icle diameters of 2–3 nm and (ii) that perhaps 50–70%
of the bulk niobium might be available after rolling
for precipitation in the ferrite, then the diagrams
published by Gladman190 and Gray191 predict values
of precipitation hardening fairly close to the values
quoted for the air cooled plate steels. However, it is
extremely important to mention that the values pre-
dicted by these diagrams are maximum values that
require certain special processing conditions for a
given level of niobium to achieve its full precipitation
hardening potential. The simple presence of niobium
in a steel clearly does not guarantee any particular
level of precipitation hardening.
By the early 1980s, it was well known that the 44 Variation in work hardening parameter with
structure and properties of polygonal ferrite steels increasing retained austenite volume fraction
had been fully optimised. For example, precipitation (after Rigsbee and VanderArend 197)
hardened, controlled rolled and air cooled low carbon
ferrite–pearlite steels containing 0·04%Nb, and less
than 1·5% Mn, and with a ferrite grain size of 5 mm
could be expected to exhibit a yield strength of around
400–420 MPa in strip and light plate. However, with
the passage of time, technological applications were
demanding even higher strength levels. It was clear
that other approaches to microstructural design were
needed to reach yield strengths over 450 MPa, while
keeping other important properties at acceptable
levels.
One of the greatest changes in the steel metallurgi-
cal landscape in the past 20 years has been the
application of multiphase microstructures in both
strip and plate products. Multiphase steels include
the so called dual phase (ferrite–martensite–retained 45 Correlation between work hardening
austenite)192–194 and transformation induced plasti- parameter and mean austenite interparticle
city (TRIP) steels (ferrite–bainite–retained austen- spacing (after Rigsbee and VanderArend 197 )
ite).195,196,226 While ferrite–pearlite microstructures
can reach yield and tensile strengths of 420 and transformation products, i.e. bainite or martensite,
550 MPa respectively, the multiphase steels can reach and (ii) it has altered the nature of the Ž nal micro-
tensile strengths well in excess of 600 MPa. Neither structure, often in the presence of interrupted accel-
dual phase steels nor TRIP steels are new. The dual erated cooling, through its eVect as a solute on
phase steels originated in 1975192 as a strip or sheet hardenability. Hence, the combination of the niobium
product and were the subject of intense research over and the interrupted accelerated cooling has en-
the following decade.193,194 Carbon TRIP steels were abled tensile strengths of 700 MPa to be achieved in
Ž rst introduced as plate product in the late 1960s.195 20 mm plate while maintaining very good toughness
In these high carbon, highly alloyed steels, the TRIP (Fig. 42).165
reaction was used to attain very high strength prop- Niobium has also proven to be instrumental in
erties, with yield strength levels exceeding 1400 MPa achieving excellent properties in both hot rolled strip
and total elongations of 30%.195 Modern TRIP steels and cold rolled and annealed sheet versions of multi-
are now being developed for strip, sheet, and coated phase steels where strength, formability, and sheared
sheet applications.226 edge ductility are important. At the high strength
Multiphase microstructures are currently used rou- levels involved, with tensile strengths around
tinely in plate steels to reach beyond what is achiev- 700 MPa, the multiphase steels exhibit high levels of
able with ferrite–pearlite steels. This is accomplished all of these properties.112,226
by water spray cooling the as controlled rolled austen- In some early work on multiphase steels, the very
ite at about 10 K sÕ 1 to the water end temperature, important but perhaps underappreciated role of
and then further air cooling to room temperature. microstructural reŽ nement on formability was pre-
Clearly, for a given CCT diagram, the choice of sented (Figs. 44 and 45).197 Figure 44 shows the
water end temperature gives some  exibility to the expected relationship between the amount of retained
Ž nal microstructural microconstituents that can be austenite and work hardening rate in a so called dual
obtained. An example of this phase balance is shown phase steel containing a ferrite–martensite–retained
in Fig. 41.165 In this work, niobium has played two austenite microstructure.197 Somewhat overlooked is
important roles that are often found in this kind of perhaps the importance of the scale of the dispersion
microstructural development: (i) it has caused strong of the retained austenite (Fig. 45). Clearly, a Ž ne
reŽ nement of the Ž nal microstructure, which includes dispersion of retained austenite is advantageous in
both the polygonal ferrite and the low temperature attaining higher rates of work hardening even at
International Materials Reviews 2003 Vol. 48 No. 6
396 DeArdo Niobium in modern steels
constant amounts of retained austenite. Since the
interparticle spacing is directly related to the S of
the hot rolled austenite, the importance of niobium v
in austenite conditioning during rolling is obvious.
Previous work on TRIP steels has shown a clear
beneŽ t of niobium to the microstructures, properties,
and manufacturability of these steels.112,197 Again, the
beneŽ ts of the niobium additions originate in its dual
abilities to alter the transformation behaviour of the
hot rolled austenite and to cause microstructural
reŽ nement.
The in uence of microstructural reŽ nement on
work hardening rate and total elongation can be
traced back to early work of Ashby,198,199 in which
he studied the work hardening of plastically hetero-
geneous materials. Ashby’s work led to the equation
ds
=aG[( f b)/(De)]1/2 . . . . . . . . (31)
de
where a is a constant near 1, G is the shear modulus,
b is the Burgers vector, f is the volume fraction of
hard phase, D is the diameter of hard phase, and e is
the strain. When this equation is expressed in terms
of f and l, the inter-island distance between hard
volumes, the work hardening rate varies as lÕ 1/2.
Hence, the work hardening rate is inversely related
to the dispersion of the martensite–retained austenite
islands in the dual phase steels and to the bainite–
retained austenite islands in the TRIP steels: the Ž ner
the dispersion, the higher the work hardening rate.
The experimental veriŽ cation of this theory has been
clearly established for the martensite islands in dual
phase steels.108,200 Equation (31) explains, to some
extent, the importance of niobium in these multiphase
steels. The use of niobium allows very high values of
S to be achieved for the controlled rolled austenite.
v
Higher values of S mean smaller values of l and,
hence, higher values v of work hardening rate, total
elongation, and stretch formability.
In summary, niobium has long been recognised as
being able to increase the strength of ferrite in ferrite–
pearlite steels through various means. This has led to
steels with yield strengths near 400 MPa. Higher
strength steels have required multiphase micro-
structures involving combinations of microcon-
stituents such as ferrite and low temperature trans-
formation temperature products such as bainite and
martensite. Research over the past 20 years has shown
that niobium is also important in achieving the proper
microstructure and Ž nal properties in these newer
steels. Besides its well recognised beneŽ t to austenite
conditioning, and, hence, to microstructural reŽ ne-
ment, niobium has also been shown to play a critical
role in helping to control the evolution of Ž nal
microstructure in these steels during plate, strip, and
sheet processing.
Niobium and stabilisation
Within the past 10 years, niobium has become a
popular addition to two important classes of steel,
the ultralow carbon (ULC) or interstitial free (IF)
steels and in ferritic stainless steels. The primary
reason for making the addition of niobium was to
complement titanium in the stabilisation of carbon
International Materials Reviews 2003 Vol. 48
and nitrogen, i.e. the removal of carbon and nitrogen
from solid solution, in these steels. However, later
work has shown that the addition of niobium can
have beneŽ ts far beyond simple stabilisation. In fer-
ritic stainless steels, stabilisation is used to remove
the yield point and prevent strain aging, but it also
plays the important role of reducing or eliminating
sensitisation. Sensitisation is the intergranular cor-
rosion caused by the precipitation of chromium car-
bides at the ferrite grain boundaries, thereby depleting
the grain boundary region of suYcient chromium to
prevent grain boundary attack.201,202 In ULC steels,
the stabilisation is important mainly because it
removes the yield point after either batch or contin-
uous annealing, prevents strain aging, and aids in the
development of suitable crystallographic textures to
enhance formability.203,204
Niobium in interstitial free steels
The early IF steels were stabilised by titanium alone,
where enough titanium was added to theoretically tie
up all the carbon, nitrogen, and sulphur in the steel.203
The amount of titanium required for complete stabil-
isation was often given as
Ti=4C+3·42N+1·5S . . . . . . . . (32)
The Ž rst particle that forms with falling temperature
is TiN. The subsequent behaviour of titanium depends
upon the initial sulphide that forms in the interden-
dritic pools.205 For example, if the manganese is below
about 2000 ppm, TiS is formed. TiS is a complex
compound with a rhombohedral crystal structure that
can exhibit numerous crystal structures or polytypes,
based on the defect structure contained within the
structure. It has been found that when both the bulk
titanium content is high (over 0·06 wt-%) and the
Ti/S ratio is high (over 7), the 9R rhombohedral form
of TiS is formed. When the bulk titanium is lower
(0·04 wt-%) and the Ti/S ratio is lower (near 5), the
18R and 6R polytypes of TiS are formed. During
subsequent cooling, the diVerent polytypes of TiS
behave diVerently. The 9R polytype undergoes inter-
calation during cooling such that layers of titanium
and carbon are added to the pre-existing layers of
titanium and sulphur to form the hexagonal or H
phase Ti C S .205 This results in the following in situ
4 2 2 of TiS205
transformation
TiS(9R)+Ti+C =(1/2)Ti C S . . . . (33)
4 2 2
Hence, the formation of H phase is one way of
stabilising carbon. This might be followed by the
formation of TiC if there were more titanium and
carbon in the steel than could be incorporated into
the H phase. When the 9R TiS is predominant, the
precipitation behaviour as shown in Fig. 46205 was
suggested, depending upon the composition of the
steel.
The 18R and 6R polytypes often do not fully
transform during cooling. Thus, all the carbon, in this
case, would be tied up as TiC, and the resulting hot
band would contain a mixture of TiN, TiS, and TiC.
Also, when the manganese is high, say over 3000 ppm,
MnS is favoured over TiS, and the resulting precipi-
tates found at room temperature in the hot band
would be TiN, MnS, and TiC.
No. 6

46 Proposed precipitation behaviour in ULC steels
with TiS (9R) and Ti4 C2 S2 (H) as dominant 205
From the above, it is clear that there can be
multiple roles for niobium in these steels.205–207 In
steels exhibiting the 9R TiS polytype, niobium can
act to: (i) replace some or all of the titanium in the
intercalation of TiS to Ti C S , (ii) replace some or
4 2 2
all of the titanium used to form TiC to remove carbon
in excess of that level that could be consumed by
intercalation, or (iii) remain in solid solution.206 In
low titanium steels, where there is insuYcient titanium
to tie up all the carbon, niobium has been used to
form NbC. A similar behaviour is found where high
manganese is used.
One problem with titanium stabilised IF steels is
their inferior behaviour after galvannealing. Problems
such as the outburst phenomenon and powdering
have created concern in this important product.207
Early empirical work had shown that the presence of
niobium in a titanium treated IF steel, i.e. dual
stabilised steel, often led to marked improvement in
the overall behaviour and performance of galvan-
nealed steels during forming operations.205–207 Recent
work has shown that superior galvannealing perform-
ance, e.g. adherence and powdering resistance, may
be the result of the presence of solute niobium. When
more niobium is added in excess to that which is
necessary, along with titanium, to fully stabilise the
steel, this solute niobium has been observed to
strongly segregate to the ferrite grain and subgrain
boundaries (Fig. 47).205 This boundary segregation,
along with free surface segregation of solute niobium,
is thought to be responsible for the improved stability
of the galvannealed coating. While a complete under-
standing of the powdering phenomenon is lacking,
recent work has shown that reducing the c phase
layer thickness at the zinc/steel interface appears to
be critical in improving powdering resistance.208 It
has been found that the thickness of this c layer
appears to diminish with increasing niobium level.209
If solute niobium on the ferrite grain boundaries is
the key to better resistance to the outburst phenom-
enon, and if solute niobium at the free surface is the
key to better coating adherence, two questions arise:
(i) how much solute niobium is needed for signiŽ cant
improvement and (ii) how does one achieve this
solute level? Recent work has shown that the galvan-
nealing performance of a dual stabilised IF steel was
much superior to that of a titanium stabilised IF
steel.210 In this work, about 150 ppm niobium was
DeArdo Niobium in modern steels 397
47 Concentrations of Nb on grain boundary, on
subgrain boundary, and within matrix,
measured by atom probe field ion microscopy
(APFIM) from finish rolled and/or coiled IF
steel205
estimated to be in solution at room temperature.
Figure 48 illustrates the combinations of additions of
titanium and niobium for given levels of carbon and
sulphur that would be needed in a typical IF steel in
order to guarantee that 150 ppm niobium would
remain in solution at room temperature.211 This Ž gure
shows, for example, that for IF steels containing
30 ppm carbon, 30 ppm nitrogen, 80 ppm sulphur,
and 600 ppm titanium, a niobium addition of 200 ppm
would be required for a niobium solute level of
150 ppm. A similar steel containing 40 ppm carbon
would require a bulk niobium content of about
240 ppm to keep the 150 ppm niobium in solution. A
research programme is currently in progress to better
deŽ ne the compositional requirements of the steel
substrate for superior galvannealing.227 The actual
segregation of solute Nb to the free surface of IF
steels during normal processing has recently been
veriŽ ed.227
Finally, there are two other important eVects that
have been attributed to solute niobium that has
segregated to the ferrite grain boundaries. First, it
has been shown to be important in helping to over-
come the deleterious eVects of phosphorus in cold or
secondary work embrittlement.207,212,213 Second, it
has been shown to be important in improving the
crystallographic texture in cold rolled and annealed
IF steels.207,214–216
48 Alloy design for maintaining 150 ppm Nb in
solution211
International Materials Reviews 2003 Vol. 48 No. 6
398 DeArdo Niobium in modern steels
49 Resistance to intergranular corrosion in heat
affected zone as function of stabilisation:
results of Type 409 welded production heats
(from Franson and Fritz219)
Niobium in ferritic stainless steels
In 1997, the North American stainless steel industry
produced approximately 2·4 million tons of steel of
which about 62% was austenitic stainless and 33%
was ferritic stainless. Of the austenitic stainless steels
produced, Type 304 and derivatives were the largest
group representing about 62%, with Type 316 rep-
resenting 13%, and Type 301 representing about 10%.
On the ferritic stainless side, Type 409 represented
about 67% of the total ferritics, with Type 43X
representing about 12%, and Type 439 representing
about 4·2%. During the past 10 years, the growth in
use of niobium in ferritic stainless steels in North
America has been extremely strong. This has been
particularly true in Type 409 used mainly for auto-
motive exhaust systems.
In general, niobium is added to ferritic stainless
steels to achieve several beneŽ ts that include:
F intergranular corrosion resistance (stabilisation)
F creep resistance
F roping and ridging resistance
F oxidation resistance
F improved surface quality
F improved die wear during forming.
These advantages have been discussed and explained
elsewhere.217,218
The Ž rst stabilised ferritic stainless steel intended
for the automotive exhaust system was developed
by Allegheny Ludlum in 1961. This steel contained
12%Cr, was fully stabilised with titanium, and now
has the designation UNS S40900. Over the past
decade, dual stabilisation of Type 409 steel with
Ti+Nb has become a popular alternative to the
earlier titanium-only approach. The dual stabilised
Type 409 steels fall under the classiŽ cation of UNS
S40930. There are several reasons for this change in
stabilisation systems; these have been reviewed by
Franson and Fritz.219
The eVectiveness of Ti+Nb as a stabilisation
system is shown in Fig. 49.219 This Ž gure indicates
that while the choice of titanium alone or in combi-
nation with niobium would appear to be equivalent,
this choice has ramiŽ cations far beyond simple sensi-
tisation and stabilisation. For example, Type 409
(T409) stabilised with the combination of Ti+Nb,
designated as Type 466 (T466), shows an improve-
International Materials Reviews 2003 Vol. 48 No. 6
50 Elevated temperature tensile properties of
Type 409 (Ti stabilised) and Type 466 (Ti+Nb
stabilised) steels217,218
ment in creep or sag resistance, especially between
600 and 800°C (see Fig. 50).217,218 Also, the oxidation
resistance of dual stabilised T466 is much superior to
that of the titanium-only T409 (see Fig. 51).217,218 The
beneŽ ts of the niobium additions shown in Figs. 50
and 51 are probably caused by a combination of
solute and precipitated niobium. As noted earlier,
other beneŽ ts of niobium additions in dual stabilised
T409 include improvements in texture formation and
formability,220,221 surface quality, and die wear.217–221
To explain some of these diVerences in behaviour, a
study is currently being conducted of the precipitation
in single stabilised (Ti) and dual stabilised (Ti+Nb)
T409 ferritic stainless steel. Preliminary results indi-
cate that there are clear diVerences in the precipitation
behaviour.222 Both steels showed precipitation of TiN
that had occurred at high temperature, possibly in
the liquid or in the interdendritic pools, that was
stable during subsequent processing. In the titanium-
only steel, precipitation of TiC upon cooling occurred
between 830 and 780°C. The TiC was observed on
the pre-existing TiN, on the ferrite grain boundaries,
and in the matrix. The precipitation sequence in the
dual stabilised steel was similar. There were two
major diVerences in precipitation between the two
steels. First, the carbide in the dual stabilised steel
was a mixed (TiNb)C. Second, the precipitation in
the dual stabilised steel occurs at much higher temper-
atures upon cooling than in the titanium-only steel,
i.e. near 1200°C as compared to the 830–780°C range
for the titanium-only T409. An example of the epitax-
ial precipitation of NbC on (Ti,Nb)(C,N) in the dual
stabilised T409 is shown in Fig. 52.222 This diVerence
51 Still air cyclic oxidation resistance of Type 409,
Type 409 HP (low interstitial), and Type 466
(low interstitial, dual stabilised) steels217,218

52 Secondary electron micrograph (top) and
back scattering image (bottom) of large
cube-shaped (Ti,Nb)(C,N) particle with small
epitaxial NbC particles attached: 222 cube side
length is ~3 mm
in precipitation temperature can have important
ramiŽ cations regarding the behaviour during hot
deformation. The fact that the carbon in the dual
stabilised T409 is stabilised at temperatures near
1200°C means that most, if not all, of the hot rolling
of ferrite in this steel in a typical strip mill will occur
in the absence of solute carbon. This will not be the
case for the titanium-only steel, since the TiC does
not form until 830°C, or while the strip is on the
runout table or in the coil. The absence of solute
carbon during hot rolling of the ferrite in the
dual stabilised T409 is expected to aid the forma-
tion of beneŽ cial texture components. The eVect of
solute carbon on texture development has been
observed repeatedly in the ULC steels discussed
above.203,204,207,215,216,228
DeArdo Niobium in modern steels 399
Summary
The past 20 years have witnessed growth in both the
applications of niobium in steels and in the under-
standing of how niobium eVects its beneŽ ts.
Furthermore, niobium additions are now the basis
for new high strength multiphase steels for both strip
and plate applications. Other new applications
include stabilisation of both interstitial free steels and
ferritic stainless steels. Perhaps the most important
new Ž nding concerning the behaviour of niobium
is the role played by solute niobium in achieving a
wide range of structure, property, and performance
improvements. Since the Ž rst commercial niobium
steel was produced in 1958, the impressive beneŽ ts
attainable through the addition of this microalloying
element has spawned nearly Ž ve decades of intense
research. Some of this work was aimed at simply
documenting the beneŽ ts of improved microstructure,
properties, or processing. Others chose to delve into
the mysteries of how niobium acted to create its
powerful eVect. The story of the use of niobium in
steel is by no means complete nor concluded, to
which researchers 20 years hence will surely attest.
Acknowledgements
This paper could not have been written without the
guidance and assistance of many colleagues. These
include past and current members of the Basic
Metals Processing Research Institute, Department
of Materials Science and Engineering, University
of Pittsburgh. Special recognition is extended to
Professor C. I. Garcia and Professor M. J. Hua as
well as to Dr G. Tither, Dr M. L. Santella and
Dr H. Stuart, for their constant help, encouragement
and support. Others who were very helpful in for-
mulating this paper are W. Gao, R. Nooning,
C. Parish, M. Renavikar, R. Seal and J. Wu. The
author thanks Mr C. Parish for help with reformulat-
ing equation (31) and Mr R. Seal for help in plotting
Fig. 48. Finally, the assistance of Ms K. Yobbi in
the preparation of this manuscript is gratefully
recognised.
References
1. c. hatchett: as cited in Palache et al., Ref. 2.
2. c. palache, h. berman and c. frondel: in ‘The system of
mineralogy’, (ed. J. D. Dana and E. S. Dana), 7th edn, Vol. 1,
1082; 1944, New York, Wiley.
3. o. de souza paraiso filho and r. de fuccio, jr: in ‘Niobium’,
(ed. H. Stuart), 113–132; 1984, Warrendale, PA, TMS.
4. f. m. becket and r. franks: US Patent 2158651, 16 May 1939.
5. c. a. beiser: ‘The eVect of small columbium additives to semi-
killed medium carbon steels’, ASM preprint no. 138; 1959,
Metals Park, OH, American Society for Metals.
6. h. stuart (ed.): ‘Niobium’; 1984, Warrendale, PA, TMS.
7. a. j. deardo, j. m. gray and l. meyer: Proc. Int. Symp.
‘Niobium ’81’, (ed. H. Stuart), 685–759; 1984, Warrendale,
PA, TMS.
8. a. j. deardo et al. (ed.): Proc. Int. Conf. on ‘Thermomechanical
processing of microalloyed austenite’, 1984, Warrendale, PA,
TMS.
9. Proc. Int. Conf. on ‘HSLA steels: technology and applications’;
1984, Materials Park, OH, ASM International.
10. a. j. deardo (ed.): Proc. Int. Conf. on ‘Processing, micro-
structure and properties of HSLA steels’; 1988, Warrendale,
PA, TMS.
International Materials Reviews 2003 Vol. 48 No. 6
400 DeArdo Niobium in modern steels
11. Proc. Int. Conf. ‘Microalloying ’95’; 1995, Warrendale, PA,
ISS.
12. Proc. Int. Conf. on ‘Thermomechanical processing of steel’;
London, May 2000, The Institute of Materials.
13. w. f. smith: ‘Structure and properties of engineering alloys’;
1981, New York, McGraw-Hill.
14. r. p. elliot: ‘Constitution of binary alloys’, 1st suppl., 155;
1965, New York, McGraw-Hill.
15. s. kurokawa, j. e. ruzzante, a. m. hey and f. dyment: Proc.
36th Annual Congress of ABM, ‘Niobium diVusion in iron
and some iron base alloys’, Vol. 1, 47–63; 1981, São Paulo,
Associação Brasileira de Metals.
16. v. thillou: unpublished research, University of Pittsburgh,
PA, 1992.
17. j. geise and c. herzig: Z. Metallkd., 1985, 76, 622–626.
18. k. narita: T rans. Iron Steel Inst. Jpn, 1975, 15, (3), 145.
19. a. mclean and d. a. r. kay: Proc. Conf. ‘Microalloying ’75’,
(ed. J. Crane), 215–231; 1977, New York, Union Carbide.
20. l. meyer, f. heisterkamp and w. mueschenborn: Proc. Conf.
‘Microalloying ’75’, (ed. J. Crane), 153–167; 1977, New York,
Union Carbide.
21. g. bauer, h. renner and j. wernet: Z. Anorg. Allg. Chem.,
1954, 277, 249–257.
22. e. k. storms and n. h. krikorian: J. Phys. Chem., 1960,
64, 1471.
23. p. schwaab and g. langenscheid: Prakt. Metallogr., 1972, 9,
(2), 67–75.
24. c. j. smithells: ‘Metals reference book’, 4th edn, 155; 1967,
London, Butterworths.
25. h. watanabe, y. e. smith and r. d. pehlke: in ‘The hot
deformation of austenite’, (ed. J. B. Balance), 140–168; 1977,
New York, TMS.
26. t. mori, k. fujita, m. tokizane and k. yamaguchi: T etsu-to-
Hagane (J. Iron Steel Inst. Jpn), 1964, 50, 911–917.
27. w. a. bandi: personal communication, United States Steel
Corporation Research Laboratories, 1979.
28. l. meyer, h. e. bühler and f. heisterkamp: T hyssenforschng,
1971, 3, 8–43.
29. m. l. santella: ‘Grain growth and high-temperature hot
rolling behavior of low-alloy steel austenite’, PhD thesis,
University of Pittsburgh, PA, 1981, 49.
30. w. b. pearson: ‘Handbook of lattice spacings and structures
of metals and alloys’, Vol. 2, 124; 1958, London, Pergamon.
31. p. duwez and f. odell: J. Electrochem. Soc., 1950, 97, 299–304.
32. g. brauer and r. lesser: Z. Metallkd., 1959, 50, 8–10.
33. j. m gray: in ‘Heat treatment ’73’, 19–28; 1973, London, The
Metals Society.
34. g. brauer and j. jander: Z. Anorg. Allg. Chem., 1952,
270, 160–178.
35. r. p. elliott: T rans. ASM, 1961, 53, 13–28.
36. s. kanazawa, a. nakashima, k. okamoto, k. tanabe and
s. nakazawa: Nippon Kinzoku Gakkaishi, 1967, 31, (2),
171–176.
37. g. krapf, e. g. buyok, e. g. bandiand and l. m. melnick:
J. Iron Steel Inst., 1973, 211, 353–356.
38. h. nordberg and b. aaronson: J. Iron Steel Inst., 1968, 206,
1263–1266.
39. p. mandry and w. dornelas: CR Acad. Bulg. Sci., 1966, 263,
1118–1121.
40. c. ouchi and h. izumi: in 4th Int. Conf. on ‘Recrystallization
and related phenomena, (ed. T. Sakai and H. G. Suzuki),
79–88; 1999, Tokyo, The Japan Institute of Metals.
41. r. grimaldi, s. maneschi and n. vantini: Arch.
Eisenhüttenwes., 1967, 38, (5 ), 401–406.
42. t. mori, m. tokizane and k. okamoto: T etsu-to-Hagane
(J. Iron Steel Inst. Jpn), 1965, 51, 2034–2036.
43. s. maneschi and a. m. beccaria: Metall. Ital., 1966, 58,
(1), 19–24.
44. hoesch-estel: unpublished results, Dortmund, Germany.
45. w. a. roberts, a. sandberg and t. siwecki: ‘Vanadium steels
(Krakow)’, 19; 1981, London, Vanitec.
46. e. l. brown, a. j. deardo and j. h. bucher: in ‘The hot
deformation of austenite’, (ed. J. B. Balance), 250–285; 1977,
New York, TMS.
47. r. e. reed-hill: ‘Physical metallurgy principles’, 2nd edn, 661;
1973, New York, Van Nostrand.
48. r. e. reed-hill: ‘Physical metallurgy principles’, 2nd edn, 480;
1973, New York, Van Nostrand.
International Materials Reviews 2003 Vol. 48 No. 6
49. e. j. palmiere: ‘Suppression of recrystallization during the hot
deformation of microalloyed austenite’, PhD thesis, University
of Pittsburgh, PA, 1991.
50. c. h. p. lupis: ‘Chemical thermodynamics of materials’, 345;
1983, New York, Elsevier Science.
51. d. r gaskell: ‘Introduction to metallurgical thermodynamics’,
2nd edn, 419; 1981, New York, McGraw-Hill.
52. c. wagner: ‘Thermodynamics of alloys’, 47; 1952, Reading,
MA, Addison-Wesley.
53. j. chipman: J. Iron Steel Inst., 1955, 180, (6), 97.
54. d. a. porter and k. e. easterling: ‘Phase transformations in
metals and alloys’, 44; 1981, London, Van Nostrand Reinhold.
55. c. h. p. lupis: ‘Chemical thermodynamics of materials’, 363;
1983, New York, Elsevier Science.
56. r. p. smith: T rans. Metall. Soc. AIME, 1966, 236, (2), 220.
57. t. h. johansen, n. christensen and b. augland: T rans.
Metall. Soc. AIME, 1967, 239, (10), 1651.
58. t. mori, m. tokizane, k. yamaguchi, e. sunami and y. nakaz-
ima: T etsu-to-Hagane (J. Iron Steel Inst. Jpn), 1968, 54,
763–776.
59. h. l. andrade, m. g. akben and j. j. jonas: Metall. T rans. A,
1983, 14A, 1967–1977.
60. r. c. sharma, v. k. lakshmanan and j. s. kirkaldy: Metall.
T rans. A, 1984, 15A, 545.
61. v. k. lakshmanan and j. s. kirkaldy: Metall. T rans. A, 1984,
15A, (3), 541–544.
62. k. balasubramanian and j. s. kirkaldy: ‘Advances in phase
transitions’, (ed. J. D. Embury et al.), 37; 1988, Oxford,
Pergamon.
63. r. p. smith: T rans. Metall. Soc. AIME, Feb. 1962, 224, 190.
64. k. j. irvine, f. b. pickering and t. gladman: J. Iron Steel
Inst., 1967, 205, 161–182.
65. t. m. hoogendorn and m. j. spanraft: Proc. Conf.
‘Microalloying ’75’, (ed. J. Crane), 75–85; 1977, New York,
Union Carbide.
66. r. c. hudd, a. jones and m. n. kale: J. Iron Steel Inst., Feb.
1971, 209, 121.
67. y. d. lee: ‘Dissolution of niobium carbonitride particles during
reheating’, TMS Technical Paper F88–3, Warrendale, PA,
TMS, 1988, 1.
68. s. koyama, t. ishii and k. narita: J. Jpn Inst. Mater., 1979,
35, (11), 1089–1094.
69. r. a. swalin: ‘Thermodynamics of solids’, 2nd edn, 165; 1972,
New York, Wiley.
70. r. simoneau, g. begin and a. h. marquis: Met. Sci., 1978, 12,
(8), 381.
71. t. gladman, d. dulieu and i. d. mcivor: Proc. Conf.
‘Microalloying ’75’, (ed. J. Crane), 35–58; 1977, New York,
Union Carbide.
72. e. listhuber: Proc. of ‘Symposium on low alloy high strength
steels’, 29–34; 1970, Düsseldorf, Metallurg Companies.
73. j. wadsworth, s. r. keown and j. h. woodhead: Met. Sci.,
1976, 10, (3), 105–112.
74. s. r. keown and w. g. wilson: ‘Thermomechanical processing
of microalloyed austenite’, (ed. A. J. DeArdo et al.), 343–359;
1982, Warrendale, PA, TMS.
75. c. h. p. lupis: ‘Chemical thermodynamics of materials’, 523;
1983, New York, Elsevier Science.
76. e. j. palmiere, c. i. garcia and a. j. deardo: Metall. Mater.
T rans. A, 1994, 25A, 277.
77. e. j. palmiere, c. i. garcia and a. j. deardo: Metall. Mater.
T rans. A, 1996, 27A, (4 ), 951.
78. j. f. elliott: ‘Electric furnace steelmaking: theory and funda-
mentals’, Vol. II, 95–178; 1963, New York, Interscience.
79. w. c. leslie: ‘The physical metallurgy of steels’, 68; 1981, New
York, Hemisphere.
80. h. takechi: ISIJ Int., 1994, 34, (1), 1–8.
81. t. gladman: ‘The physical metallurgy of microalloyed steels’;
1997, London, The Institute of Materials.
82. o. hashimoto, s. satoh, t. irie and n. ohashi: Proc. Int. Conf.
on ‘Advances in physical metallurgy and applications of steel’,
95–104; 1984, London, The Metals Society.
83. k. a. taylor: Scr. Metall., 1995, 32, (1), 7–12.
84. d. h. jack and k. h. jack: Mater. Sci. Eng., 1973, 11, (1 ), 1–27.
85. a. t. davenport and r. w. k. honeycombe: Met. Sci., 1975, 9,
(5), 201–208.
86. r. w. k. honeycombe: Metall. Mater. T rans. A, 1976, 7A,
915–936.
87. j. m. silcock: J. Iron Steel Inst., 1963, 201, 409–421.

88. j. s. t. van aswegan, r. w. k. honeycombe and d. h.
warrington: Acta Metall., 1964, 12, (1), 1–13.
89. j. m. silcock: Acta Metall., 1966, 14, (5), 687–692.
90. f. h. froes, r. w. honeycombe and d. h. warrington: Acta
Metall., 1967, 15, 157–159.
91. m. tanino and k. aoki: T rans. Iron Steel Inst. Jpn, 1968, 8,
(5), 337–345.
92. r. g. baker and j. nutting: ‘Precipitation processes in steels’,
ISI Special Report 64, 1; 1959, London, The Iron and
Steel Institute.
93. g. kurdjumov and g. sachs: Z. Phys., 1930, 62, 592.
94. p. l. mangonon and w. e. heitmann: Proc. Conf.
‘Microalloying ’75, (ed. J. Crane), 59–70; 1977, New York,
Union Carbide.
95. g. gauthier and a. b. le bon: Proc. Conf. ‘Microalloying ’75’,
(ed. J. Crane), 71–74; 1977, New York, Union Carbide.
96. a. t. davenport, r. e. miner and r. a. kot: ‘The hot
deformation of austenite’, (ed. J. B. Balance), 186–203; 1977,
New York, TMS.
97. j. irvine and t. n. baker: Met. Sci., 1979, 13, 228–237.
98. s. s. hansen, j. b. vander sande and m. cohen: Metall. Mater.
T rans. A, 1980, 11A, 387–402.
99. j. m. gray and r. b. g. yeo: ASM T rans. Q., 1968, 61,
(2), 255–269.
100. a. mascanzoni and g. bussichelli: Philos. Mag., 1974, 30,
(1), 207–212.
101. s. freeman: ‘EVect of second phase particles on the mechanical
properties of steel’, ISI Special Publication no. 145, 152; 1971,
London, The Iron and Steel Institute.
102. g. r. speich, l. j. cuddy, c. r. gordon and a. j. deardo: in
‘Phase transformations in ferrous alloys’, (ed. A. R. Marder
et al.), 341–390; 1984, Warrendale, PA, TMS.
103. a. j. deardo: ‘HSLA steels’, 70–79; 1985, Port Kembla, NSW,
South Coast Printers.
104. a. j. deardo: Proc. Conf. ‘Thermec ’97’, (ed. T. Chandra and
T. Sakai), Vol. 1, 13–30; 1997, Warrendale, PA, TMS.
105. a. d. wilson, e. g. hamburgh, d. j. colvin, s. w. thompson
and g. krauss: ‘Microalloyed HSLA steels’, 259–276; 1988,
Materials Park, OH, ASM International.
106. j. m. rigsbee and p. j. vanderarend: in ‘Formable HSLA and
dual-phase steels’, Vol. 244, (ed. A. T. Davenport), 56–86;
1979, New York, TMS.
107. f. b. pickering: in ‘HSLA steels: technology and applications’,
1–31; 1984, Materials Park, OH, ASM International.
108. c. a. n. lanzillotto and f. b. pickering: Met. Sci., 1982,
16, 371–382.
109. n. k. balliger: ‘Advances in the physical metallurgy and
applications of steels’, 73; 1983, London, The Metals Society.
110. j. t. barnby: Acta Mater., 1967, (5), 903–909.
111. e. smith: Int. J. Fract., 1968, 4, 131.
112. k. hulka, w. bleck and k. papamantellos: ISS Mech. Work.
Steel Process. Conf. Proc., 1999, 37, 67–78.
113. t. heller, b. engl, g. thiemann and g. stich: Proc. Conf.
‘Thermomechanical processing of steel’, Vol. 2, 438–445; 2000,
London, IOM Communications.
114. a. j. deardo: Proc. Conf. ‘Microalloying in steels: new
trends for the 21st century’, San Sebastian, Sept. 1998, (ed.
J. M. Rodriguez-Ibabe et al.), 15–26; Zurich, Trans Tech
Publications.
115. a. lebon, j. rofes-vernis and c. rossard: Met. Sci., 1975,
9, 36–40.
116. r. coladas, j. masounave and j. p. bailon: in ‘The hot
deformation of austenite’, (ed. J. B. Balance), 341–378; 1977,
New York, TMS.
117. j. j. jonas and i. weiss: Met. Sci., 1979, 13, 238–245.
118. d. webster and j. h. woodhead: J. Iron Steel Inst., 1964,
202, 987–994.
119. j. d. jones and a. b. rothwell: ‘Deformation under hot
working conditions’, ISI Publication no. 108, 78–82; 1968,
London, The Iron and Steel Institute.
120. j. m. gray: ‘Processing and properties of low carbon steel’,
(ed. J. M. Gray), 225–247; 1973, New York, TMS.
121. r. simoneau, g. begin and a. h. marquis: Met. Sci., 1978, 12,
(8), 381–386.
122. g. begin and r. simoneau: Proc. Conf. ‘Microalloying ’75’,
(ed. J. Crane), 85–87; 1977, New York, Union Carbide.
123. c. ouchi, t. sanpei, t. okita and i. kozasu: in ‘The hot
deformation of austenite’, (ed. J. B. Balance), 316–340; 1977,
New York, TMS.
DeArdo Niobium in modern steels 401
124. r. k. amin, g. butterworth and f. b. pickering: in ‘Hot
working and forming processes’, 27–31; 1979, London, The
Metals Society.
125. m. g. akben, i. weiss and j. j. jonas: Acta Metall. Mater.,
1981, 29, 111–121.
126. a. j. deardo: Ironmaking Steelmaking, 2001, 28, (2), 138–144.
127. m. e. fine: ‘Phase transformations in condensed systems’; 1964,
New York, Macmillan.
128. d. j. swinden and j. h. woodhead: J. Iron Steel Inst., 1971,
209, 883–899.
129. i. kozasu, c. ouchi, t. sampei and t. okita: Proc. Conf.
‘Microalloying ’75’, (ed. J. Crane), Vol. 1, 120–135; 1977, New
York, Union Carbide.
130. j. d. grozier: Proc. Conf. ‘Microalloying ’75’, (ed. J. Crane),
241–250; 1977, New York, Union Carbide.
131. k. amano, t hatomura, m. koda, c. shiga and t. tanaka:
Proc. Conf. ‘Accelerated cooling of steel’, (ed. P. D. Southwick),
349–365; 1986, Warrendale, PA, TMS.
132. a. j. deardo: Proc. Conf. ‘Accelerated cooling of steel’, (ed.
P. D. Southwick), 97; 1986, Warrendale, PA, TMS.
133. t. tanaka: ‘Overview of accelerated cooled steel plate’,
187–208; 1988, New York, Pergamon Press.
134. a. j. deardo: Can. Metall. Q., 1988, 27, 141.
135. e. e. underwood: in ‘Quantitative microscopy’, (ed. R. T.
DeHoV and F. N. Rhines), 77–127; 1968, New York,
McGraw-Hill.
136. l. j. cuddy: Metall. T rans., 1984, 15A, 87–98.
137. h. sekine and t. maruyama: Seitetsu Kenkyu, 1976, 289, 43–61.
138. h. sekine, t. maruyama, h. kageyama and y. kawashima:
Proc. Int. Conf. ‘Thermomechanical processing of micro-
alloyed austenite’, (ed. A. J. DeArdo et al.), 141; 1982,
Warrendale, PA, TMS.
139. e. j. palmiere: unpublished research, University of Pittsburgh,
PA, 1990.
140. d. a. porter and k. e. easterling: ‘Phase transformations in
metals and alloys’, 263; 1981, London, Van Nostrand
Reinhold.
141. p. pouly: unpublished research, University of Pittsburgh,
PA, 1993.
142. o. kwon and a. j. deardo: Acta Metall. Mater., 1991, 39, 529.
143. e. j. palmiere, c. i. garcia and a. j. deardo: Proc. Int. Symp.
on ‘Low-carbon steels for the 90s’, (ed. R. Asfahani and
G. Tither), 121; 1993, Warrendale, PA, TMS.
144. o. o. miller: ASM T rans. Q.,1951, 43, 260–289.
145. t. gladman and f. b. pickering: J. Iron Steel Inst., 1967,
205, 653–664.
146. n. e. hannerz and f. dekazinczy: J. Iron Steel Inst., 1970,
208, 475–481.
147. t. gladman: Proc. R. Soc. (L ondon) Series A, 1966, 266,
298– 309.
148. t. gladman: ‘Recrystallization of multiphase and particle
containing materials’, (ed. N. Hansen et al.), 183–192; 1980,
Roskilde, Risø National Laboratory.
149. t. gladman: ‘HSLA steels: processing, properties and appli-
cations’, 3–14; 1992, Warrendale, PA, TMS.
150. c. zener: T rans. AIME, 1949, 175, 15.
151. j. m. lifshitz and v. v. slyozov: J. Phys. Chem. Solids, 1961,
19, 35–50.
152. c. wagner: Z. Elektrochem., 1961, 65, 581–591.
153. f. b. pickering: ‘Microalloyed vanadium steels’, 79–104; 1990,
Krakow, Poland, Association of Polish Metallurgical
Engineers.
154. k. lucke and k. detert: Acta Metall., 1957, 5, 628–637.
155. j. w. cahn: Acta Metall., 1962, 10, 789.
156. p. gordon and r. a. vandermeer: ‘Recrystallization, grain
growth, and textures’, 205; 1966, Metals Park, OH, ASM.
157. a. s. keh and s. weissmann: ‘The impact of transmission
electron microscopy on theories of the strength of crystals’,
231; 1963, New York, Wiley.
158. l. j. cuddy: in ‘Thermomechanical processing of microalloyed
austenite’, (ed. A. J. DeArdo et al.), 129–140; 1982, Warrendale,
PA, TMS.
159. g. a. wilber, j. r. bell, j. h. bucher and w. j. childs: T rans.
Metall. Soc. AIME, 1968, 242, (11), 2305–2308.
160. j. n. cordea: Proc. of ‘Symposium on low alloy high strength
steels’, 61–80; 1970, Düsseldorf, Metallurg Companies.
161. a. j. deardo: unpublished research, University of Pittsburgh,
PA, 1985.
International Materials Reviews 2003 Vol. 48 No. 6
402 DeArdo Niobium in modern steels
162. s. yamamoto, c. ouchi and t. osuka: ‘Thermomechanical
processing of microalloyed austenite’, (ed. A. J. DeArdo, et al.),
613–638; 1982, Warrendale, PA, TMS.
163. l. meyer: personal communication, Thyssen Aktien-
gesellschaft, Duisburg-liamborn, 1981.
164. j. m. gray: ‘Weldability of niobium-containing high-strength
low alloy steel’, Bulletin no. 213, Welding Research Council,
Feb. 1976.
165. s. okaguchi, t. hashimoto and h. ohtani: Proc. Conf.
‘Thermec ’88’, Vol. 1, 330–336; 1988, Tokyo, Iron and Steel
Institute of Japan.
166. h. tamehiro, r. habu, n. yamada, h. matsuda and m. nagumo:
in ‘Accelerated cooling of steel’, (ed. P. D. Southwick),
401–413; 1986 Warrendale, PA, TMS.
167. k. abe, m. shimizu, s. takashima and h. kaji: Proc. Conf.
‘Thermec ’88’, Vol. 1, 322–329; 1988, Tokyo, The Iron and
Steel Institute of Japan.
168. w. m. hof, m. k. gräf, h.-g. hillenbrand, b. hoh and p. a.
peters: in ‘HSLA steels: metallurgy and applications’, (ed.
J. M. Gray et al.) 467–474; 1986, Metals Park, OH, ASM
International.
169. p. bufalini, n. pontremoli, a devito and a. aprile: in ‘HSLA
steels: technology and applications’, 743–754; 1984, Metals
Park, OH, ASM International.
170. m. k. graf, p. a. peters, f. k. lorenz and p. schwaab: in
‘HSLA steels: technology and applications’, 801–807; 1984,
Metals Park, OH, ASM International.
171. r. l. bodnar and s. s. hansen: ISS Mech. Work. Steel Process.
Conf. Proc., 1996, 34, 513–525.
172. v. thillou, m. hua, c. i. garcia, c. perdrix and a. j. deardo:
’Microalloying in steels’ (ed. J. M. Rodriguez-Ibabe et al.),
311–318; 1998, Zurich, Trans Tech Publications.
173. v. thillou, m. hua, c. i. garcia, c. perdrix and a. j. deardo:
ISS Mech. Work. Steel Process. Conf. Proc., 1999, 37, 471–482.
174. a. j. deardo: ‘Reducing the variability of HSLA sheet steels’,
American Iron and Steel Institute/Department of Energy
Technical Roadmap Program, Project no. 9807, Sixth
Quarterly Report, Dec. 2000.
175. f. y. sun, w. xu and x ma: in ‘HSLA steels: metallurgy and
applications’, (ed. J. M. Gray et al.), 277–286; 1986, Metals
Park, OH, ASM International.
176. t. n. baker: ‘Yield,  ow and fracture of polycrystals’, (ed.
T. N. Baker), 235–273; 1983, London, Applied Science.
177. e. hornbogen and g. staniek: J. Mater. Sci., 1974, 9, 879.
178. j. irvine and t. n. baker: Mater. Sci. Eng., 1984, 64, 123–134.
179. h. kejian and t. n. baker: Mater. Sci. Eng., 1993, A169, 53–65.
180. m. hua, c. i. garcia and a. j. deardo: unpublished research,
University of Pittsburgh, PA, 1996.
181. r. w. k honeycombe: ‘Phase transformations in ferrous alloys’,
(ed. A. R. Marder et al. ), 259–280; 1988, Warrendale, PA,
TMS.
182. r. w. k. honeycombe: in ‘Processing microstructure and
properties of HSLA steels’, (ed. A. J. DeArdo), 1–14; 1988,
Warrendale, PA, TMS.
183. t. sakuma and r. w. k honeycombe: Met. Sci., 1984, 18,
449–454.
184. t. sakuma and r. w. k. honeycombe: Mater. Sci. T echnol.,
1985, 1, 351–356.
185. a. itman, k. r. cardoso and h.-j. kestenbach: Mater. Sci.
T echnol., 1997, 13, 49–55.
186. m. hua: unpublished research, University of Pittsburgh, PA,
1994.
187. t. parayil and g. ludkovsky: ‘Accelerated cooling of rolled
steel’, (ed. G. E. Ruddle et al.), 131–146; 1988, New York,
Pergamon.
188. m. hua: unpublished research, University of Pittsburgh, PA,
2000.
189. h. gondoh and t. osuka: ‘Niobium’, (ed. H. Stuart), 833–882;
1984, Warrendale, PA, TMS.
190. t. gladman: Proc. Conf. ‘Microalloying ’75’, (ed. J. Crane),
32–54; 1977, New York, Union Carbide.
191. j. m. gray: in ‘Heat treatment ’73’, 19–28; 1973, London, The
Metals Society.
192. s. hayami and t. furukawa: Proc. Conf. ‘Microalloying ’75’,
(ed. J. Crane), 311–321; 1977, New York, Union Carbide.
193. r. a. kot and b. l. bramfitt: Proc. Conf. ‘Fundamentals of
dual-phase steels: proceedings of a symposium’; 1981,
Warrendale, PA, TMS.
International Materials Reviews 2003 Vol. 48 No. 6
194. a. t. davenport: Proc. Conf. ‘Formable HSLA and dual-
phase steels: proceedings of a symposium’, 244; 1979, New
York, TMS.
195. v. f. zackay, e. r. parker, d. fahr and r. busch: T rans.
ASM, 1967, 60, 252–259.
196. z. hanzaki, p. hodgson and s. yue: Metall. T rans. A, 1997,
28A, 2405–2414.
197. j. m. rigsbee and p. j. vanderarend: ‘Formable HSLA and
dual-phase HSLA steels’, (ed. A. T. Davenport), 56–86; 1979,
New York, AIME.
198. m. f. ashby: Philos. Mag., 1970, 8, (21), 399–424.
199. k. c. russell and m. f. ashby: Acta Metall., 1970, 18, 891–901.
200. n. k. ballinger and t. gladman: Met. Sci., 1981, 15, 95–108.
201. r. p. frankenthal and h. w. pickering: J. Electrochem. Soc.,
1973, 120, (1 ), 23–26.
202. w. c. leslie: ‘The physical metallurgy of steels’, 343; 1981,
New York, Hemisphere.
203. i. gupta and d. bhattacharya: in ‘Metallurgy of vacuum-
degassed steel products’, (ed. R. Pradhan), 43–72; 1990,
Warrendale, PA, TMS.
204. a. j. deardo: Proc. Conf. ‘IF steels 2000’, 125–136; 2000,
Warrendale, PA, ISS.
205. m. hua, c. i. garcia and a. j. deardo: Metall. T rans. A, 1997,
28A, 1769–1780.
206. m. j. hua, c. i. garcia and a. j. deardo: Proc. Int. Symp. on
‘Modern LC and ULC sheet steels for cold forming: processing
and properties’, (ed. W. Bleck and Verlag Mainz), 145–156;
1998, Aachen, Verlag Mainz, Wissenschaftsverlag.
207. g. tither, c. i. garcia, m. hua and a. j. deardo: in ‘Physical
metallurgy of IF steels’, 293–322; 1994, Tokyo, Iron and Steel
Institute of Japan.
208. m. f. shi, g. m. smith, m. moore and d. j. meuleman: Proc.
Conf. on ‘Zinc based coatings systems: metallurgy and per-
formance’, 387–398; 1990, Warrendale, PA, TMS.
209. g. jenkins: unpublished research, Welsh Technology Centre,
Corus Group, 1999.
210. c. c. cheng: ISS Mech. Work. Steel Process. Proc. Conf., 2000,
38, 255–263.
211. a. j. deardo: unpublished research, University of Pittsburgh,
PA, 1999.
212. n. kino, m. yamada, y. tokunaga and h. tsuchiya: Proc.
Conf. on ‘Metallurgy of vacuum degassed steel products’, (ed.
R. Pradhan), 197–213; 1990, Warrendale, PA, TMS.
213. s. rege, c. i. garcia and a. j. deardo: ISS Mech. Work. Steel
Process. Conf. Proc., 1997, 35, 149–158.
214. y. tokunaga et al.: T etsu-to-Hagané (J. Iron Steel Inst. Jpn),
1987, 73, (2), 109.
215. r. e. hook: Proc. Conf. on ‘Metallurgy of vacuum-degassed
steel products’, (ed. R. Pradhan), 263–287; 1990, Warrendale,
PA, TMS.
216. l. j. ruiz-aparicio, c. i. garcia and a. j. deardo: Metall.
T rans. A, 2001, 32A, (2), 417–423.
217. a. j. deardo: ‘Tantalum and niobium’, 435–464; 1998, Brussels,
Tantalum–Niobium International Study Center.
218. a. j. deardo: Proc. Int. Symp. on ‘Tantalum and niobium’,
Goslar, Germany, Sept. 1995, 113–127; Brussels,
Tantalum–Niobium International Study Center.
219. i. a. franson and j. d. fritz: ‘Stabilization requirements for
Type 409 (UNS S40900) ferritic stainless steel’, SAE Technical
Paper 971005, Society of Automotive Engineers, Warrendale,
PA, 1997, 155–161.
220. y. xu, c. i. garcia, i. franson and a. j. deardo: in ‘Low
carbon steels for the nineties’, 397–404; 1993, Warrendale,
PA, TMS.
221. a. j. deardo, c. i. garcia and j. s. rege: ISS Mech. Work.
Steel Process. Conf. Proc., 1995, 33, 871.
222. r. g. nooning, jr, c. isaac garcia, m. hua and a. j. deardo:
Stainl. Steel World, 2001, paper SSW–PO165.
223. a. t. davenport, l. c. brossard and r. e. miner: JOM, 1975,
27, 21–27
224. a. j. deardo: in ‘Microalloying ’95’, 15–34; 1995, Warrendale,
PA, ISS.
225. a. j. deardo: in ‘Thermec 2000’; 2002, Warrendale, PA, TMS.
226. b. c. de cooman (ed.): Proc. Conf. on ‘TRIP-aided high strength
ferrous alloys’; 2002, Aachen, Wissenschaftsverlag Mainz.
227. a. j. deardo: Proc. Conf. ‘IF steels 2003: properties and
applications of IF steels’, (ed. H. Takechi), 240–248; 2003,
Tokyo, Iron and Steel Institute of Japan.
228. y. hosaya, t. suzuki and a. nishimoto: in ‘Metallurgy of
vacuum degassed steel products’, (ed. R. Pradhan), 291–308;
1990, Warrendale, PA, TMS.