ByRune LagneborgTadeusz SiweckiStanislaw Zajac

Bevis HutchinsonSwedish Institute for Metals Research

Stockholm, Sweden

The Role Of Vanadium

In Microalloyed Steels

Reprinted from:The Scandanavian Journal of Metallurgy

October 1999

THE ROLE OF VANADIUM IN MICROALLOYED STEELS

Rune Lagneborg, Tadeusz Siwecki, Stanislaw Zajac and Bevis HutchinsonSwedish Institute for Metals Research

S-11428 Stockholm, Sweden

ABSTRACT

The overall objective of the present paper is to review the role of V in microalloyed steelswith a particular address as to how it affects microstrucural evolution and mechanicalproperties. Its role in thermomechanical controlled processing (TMCP) is emphasised. Thereview is largely based on work carried out at the Swedish Institute for Metals Research(SIMR) during the last 25 years but includes also reference to other relevant, published work.

A specific aim is to demonstrate the present scientific knowledge of the subject. Therefore theunderstanding and interpretation of essential phenomena related to microstructure formationand properties are thoroughly examined, ranging from the influence of microalloying onprevention of austenite grain growth and recrystallization, to precipitation in ferrite and itseffect on strength. Within the well-known thermodynamic database Thermocalc a specialmicroalloy database has been developed at SIMR, allowing reliable predictions of phaseequilibria and thermodynamic functions for phase transformations in microalloyed steels. Acomprehensive account is given of the role of V in the most important processing steps thatthe microalloyed steels are subjected to, viz. TMCP, continuous casting and welding.

Compared to the other microalloying elements, Nb and Ti, V exhibits essential differences. Inparticular the solubility of its carbonitrides is much larger and the solubility of its nitride isabout two orders of magnitude smaller than its carbide, contrary to Nb but similar to Ti. Foroptimal alloy design and thermomechanical processing proper allowance must be made forthese differences. To reach the maximum ferrite grain refinement, ~4 µm, repeatedrecrystallization in a series of rolling reductions is used in TMCP of V-microalloyed steels, socalled Recrystallization-Controlled-Rolling (RCR), as opposed to traditional controlledrolling of Nb-steels where heavy rolling at low temperatures in the non-recrystallizationregime is the means of attaining grain refinement. RCR presents some important advantages,in particular a more economical hot rolling practice by allowing low reheating and highfinishing temperatures.

As compared to Nb, V has certain further advantages as a microalloying element due to itsgreater solubility in autstenite. The tendency for hot cracking of cast slabs is much lesspronounced and dissolution of coarse V(C,N) compounds is more easily achieved prior to hotrolling than for the corresponding NbC.

It is demonstrated that the relatively large solubility of V(C,N) and the much lower solubilityof VN than VC makes V an eminent choice for strong and easily controllable precipitationstrengthening. A corollary of the difference in VN and VC solubilities is that N becomes anessential microalloying element in V-steels, because it largely determines the density of

V(C,N) precipitation and thereby the degree of precipitation strengthening. Moreover, sincepure ferrite dissolves more N than C the total N-content of the steel is normally dissolved inthe ferrite before V(C,N) precipitation whereas only a fraction of the C-content, given by theaustenite/ferrite or ferrite/cementite equilibrium, is dissolved in ferrite. Hence, by preciseadditions of N this circumstance facilitates the control of V(C,N) precipitation strengthening.

Recent studies at SIMR have shown that the precipitation strengthening increasessignificantly with total C-content of the steel, an effect previously not recognised. IncreasingC-content delays the pearlite formation and thereby maintains for a longer time the highercontent of solute C in ferrite corresponding to the austenite/ferrite equilibrium as compared tothat of ferrite/cementite, allowing new nucleation C-rich V(C,N) to occur when most N hasbeen consumed.

High N-contents, >0,010%, in V-microalloyed steels are often considered to causeunsatisfactory HAZ weldability due to solute N which has been claimed to impair toughness.Recent studies show, however, that such sweeping statements are incorrect and that fullysatisfactory welding results can be achieved also at high N-contents by selective choice ofwelding parameters. Furthermore, it has been shown that when a lowering of the impacttoughness is observed at high N-levels and high heat input it cannot be explained by solute Nbut rather by changes in the HAZ microstructure.

TABLE OF CONTENTS

1. PROLOGUE ......................................................................................................................1

1.1 The discovery of vanadium and the evolution of its use..........................................1

1.2 Statistics of the use of vanadium...............................................................................2

1.3 The development of V-microalloying for structural steels ......................................2

1.4 Scope of the present review .......................................................................................3

2. ALLOY SYSTEM AND THERMODYNAMIC BASES.................................................4

2.1 Thermodynamic considerations................................................................................5

2.2 Thermodynamic description of the Fe-V-C-N system .............................................7

2.3 The Fe-Ti-C-N system................................................................................................8

2.4 The Fe-Nb-C-N system ............................................................................................10

2.5 The Fe-Al-B-C-N system .........................................................................................11

2.6 Chemical composition of precipitated particles in multi-microalloyed steels......12

2.7 Thermodynamic Driving Force for Precipitation..................................................13

2.8 Practical consequences of different solubilities of V carbides and nitrides .........14

2.9 Summary of chapter.................................................................................................14

3. MICROALLOYING EFFECTS IN AUSTENITE........................................................15

3.1 Precipitation and dissolution of microalloy-carbonitrides in austenite................15

3.2 Grain growth inhibition and grain coarsening ......................................................19

3.3 Effects on recrystallization ......................................................................................23

3.4 Effects on the austenite-ferrite transformation......................................................24

3.5 Summary of chapter ................................................................................................25

4. PRECIPITATION IN FERRITE ...................................................................................26

4.1 Interphase precipitation ..........................................................................................27

4.2 The mechanism of interphase precipitation ...........................................................29

4.3 General precipitation...............................................................................................32

4.4 Effects of precipitation on strength ........................................................................34

4.5 Summary of chapter ................................................................................................38

5. PRINCIPLES AND PRACTICE OF THERMO-MECHANICAL CONTROLLEDPROCESSING, TMCP....................................................................................................40

5.1 General principles of recrystallization-controlled rolling, RCR, and controlled

rolling, CR, processing.............................................................................................40

5.2 Mechanical properties and microstructure of TMCP material............................41

5.3 Effect of nitrogen content ........................................................................................44

5.4 Application of TMCP practice ................................................................................46

5.5 Microstructure development during TMCP rolling ..............................................52

5.6 Summary of chapter ................................................................................................54

6. CAST STRUCTURES.....................................................................................................55

6.1 Microstructural features .........................................................................................55

6.2 Ductility during casting ...........................................................................................57

7. WELDABILITY ..............................................................................................................61

7.1 Structure and properties of coarsed grained HAZ................................................61

7.2 Effect of nitrogen and welding parameters on HAZ toughness of V-steels..........62

7.3 Toughness-Microstructure Relationships ..............................................................65

7.4 Summary of chapter ................................................................................................67

8. HOW IS V BEST USED IN MICROALLOYED STEELS - EXECUTIVESUMMARY......................................................................................................................68

8.1 Austenite conditioning by thermomechanical processing .....................................68

8.2 Precipitation strengthening .....................................................................................70

8.3 Cast structures .........................................................................................................71

8.4 HAZ weldability.......................................................................................................71

8.5 Interplay between general trends in steel technology and the technology of V-

microalloyed steels ...................................................................................................72

ACKNOWLEDGEMENTS .................................................................................................73

REFERENCES .....................................................................................................................74

1

1. PROLOGUE

1.1 The discovery of vanadium and the evolution of its useVanadium was discovered by the Swedish scientist and doctor Nils Gabriel Sefström in 1830.In his studies of ductile iron originating from iron ore of the Taberg mine he obtained aresidue which he concluded contained a compound with a previously unknown element,vanadium. At this point, Sefström’s master, the famous Swedish chemist Jöns JacobBerzelius, took up interest in the new element. He announced Sefström’s discoveryinternationally and also initiated a large research programme on vanadium salts. Sefström’sand Berzelius’ work was, however, confined to chemical studies of a large number ofvanadium compounds. It was only some 30 years later that vanadium was isolated as a metaldue to work by the English chemist, Sir Henry Roscoe.

The early use of vanadium was as a chemical compound and was based on the investigationsby Sefström, Berzelius and Roscoe. Examples of early applications of vanadium compoundare: black colouring of ink and fixing aniline black dyes for fabrics. The function of vanadiumsalts as catalyst for many chemical reactions was discovered 1900 in Germany and hasbecome the most important application of vanadium as a chemical.

At the turn of the century the technology of process metallurgy had developed such thatcommercial quantities of ferroalloys could be produced. In order to exploit the opportunitiesfor V as an alloying element in steel the first production unit was set up in South Wales, GreatBritain, and a special assignment was given to Professor Arnold of Sheffield College toinvestigate effects of V alloying in various steels.

Arnold’s research together with studies in steel works in the Sheffield neighbourhood laid thefoundation for a whole range of tool and die steels, from high speed steels to both cold workand hot work die steels. Hence, in these applications the hardness of V carbide and its stabilityat high temperatures played an essential part. The importance of V-alloying in tool steels hasnot diminished with time. The development of new processing technologies, especiallypowder metallurgy, has made it possible to increase the hard phase content in high speed andcold work steels considerably. Hence, in recent years this has rather increased the use of V intool steels.

The usefulness of V in engineering steels was early recognised. Work in the first years of thecentury both in England and France showed that carbon steels exhibit a considerable increasein strength by V-alloying, particularly when quenched and tempered. In the U.S. the use of Vfor alloying of automotive steels had, as it appears, an accidental beginning. Henry Ford Iwatched a motor race in which a French car crashed. He examined the wreck and found acrankshaft made from Swedish iron that suffered less damage than expected. Afterexamination in his laboratories it was reported that the steel contained V. As a result Fordgave instructions to use V-steels in critical components in Ford cars ‘in order to better resistshock and fatigue on American roads’. At that time, ~ 1910, an understanding of thebeneficial effect of V was missing. With today’s knowledge we can confidently say that theobserved strengthening was due to fine precipitation of V-carbonitrides during cooling ortempering. The improved impact resistance was due to a refined grain structure resulting fromgrain growth inhibition by V-nitrides at the quenching or normalizing temperature.

Other important applications of V-alloying, primarily developed, during the first two thirds ofthis century, are high temperature power plant steels, rail steels and cast iron.

2

The most exciting and today the largest use of V is in high strength low alloy (HSLA)structural steels; another name for the same group of steels and more related to their alloyingis ‘microalloyed steels’. The development of these materials started in the 50’s and a briefhistorical account of this is given below.

Beside its application in alloying of steel and cast iron V is an essential alloying addition inTi-alloys for aircraft and aerospace applications. Although V-base alloys have been used innuclear breeder reactors because of their nuclear, physical and high temperature mechanicaland corrosive properties and are considered as a candidate in future fusion reactors, they haveat present no important application as structural materials.

1.2 Statistics of the use of vanadiumThe world consumption of vanadium is today (1999) some 34000 tonnes per annum, of which28000 tonnes in the Western world.

The predominant application of vanadium is for alloying of steel and cast iron. It amounts toabout 85 % of the total consumption, a share that appears rather stable over time. Theremaining part goes to chemicals and to alloying of titanium. World-wide about 65% of thevanadium for alloying of steel is used in microalloyed structural steels, whereas the remaining35% is used in V-alloyed steels, such as high speed and tools steels, high temperature lowalloy steels, etc.

From this follows that the consumption of vanadium and the production of steel are stronglyinterrelated. The rather constant level of steel production over the last 20 years hasconsequently resulted in a similar constant consumption of vanadium, at a level of 0,04 kg pertonne steel produced. However, since 1992 this level has increased to about 0,05 kg now,most probably reflecting an increased use of V-microalloyed structural steels.

1.3 The development of V-microalloying for structural steelsThe development of the microalloyed structural steels and the related development ofcontrolled rolling started in the late 1950’s. With the broad introduction of welded structuresafter the Second World War strengthening of steel by increasing the carbon content becamepractically prohibitive in view of carbon’s detrimental effect on toughness of weldedstructures and on weldability. Therefore the new knowledge that grain refinement canimprove both these properties and at the same time enhance strength was a strong stimulus fordeveloping new hot rolling practices and new steels. With time it was also recognised thatstrengthening by microalloy precipitation could be substituted for strengthening by carbonwhile weldability is improved.

In the past normalizing heat treatment was an absolute requirement for steel plate and moststructural sections used in structures for heavy-duty service. In the 60’s and 70’s V-microalloyed steels with 0.15-0.20% and 0.10-0.15% V and somewhat enhanced N-content0.010-0.015% found wide use for this type of applications, e.g. thick walled gas pipe lines.

In the early 60’s Bethlehem Steel developed and marketed a series of V-N steels with max.content of C and Mn of 0.22% and 1.25%, respectively, with yield strengths ranging from 320to 460 MPa and to be used in the as-rolled condition. The steels were offered in practically allproduct forms, plate, strip and structural sections.

3

An early advanced V strip steel was VAN80, developed at Jones & Laughlin around 1975with a typical yield strength of 560 MPa. This was one of the first steels where in-lineaccelerated cooling was applied in a controlled manner to enhance grain refinement andstrengthening by microalloy precipitation.

Around 1965 the development of controlled rolling had reached a level such that it wasintroduced in commercial production. By microalloying with Nb it was possible to preventrecrystallization during hot rolling, thereby flattening the austenite grains gradually more withreduction and in this way the austenite was conditioned to transform to very fine-grainedferrite. In order to increase strength it was soon found advantageous to also microalloy withV. This type of controlled rolled Nb-V steels, with about 0.10%C, 0.03%Nb and 0.07%V, hasbeen widely adopted for pipelines.

A controlled rolling route later developed, ~ 1980, was recrystallization-controlled-rolling. Byallowing the worked austenite to recrystallize after each rolling reduction it was foundpossible to reach a similar grain refinement as with the former method. This process results inhigher finishing temperatures; it therefore requires lower rolling forces and can accordinglybe practised in weaker rolling mills as well as offering superior productivity. It was found thatTi-V microalloyed steels are especially suited for this process since both these elements onlyretard recrystallization weakly and therefore allow repeated recrystallization during rolling.

1.4 Scope of the present reviewThe aim of the present paper is to review the role of V in microalloyed structural steels andour understanding of how it affects microstructure evolution and mechanical properties. Itsrole related to thermo-mechanical processing is especially addressed.

The paper covers V-microalloyed structural steels with ferrite-pearlite microstructures havinglow to medium carbon contents. Hence, the many important uses of V-alloying in other steels,e.g. tool and high speed steels, low-alloy creep resistant steels etc., are omitted.

The review is mainly based on work carried out at the Swedish Institute for Metals Researchduring the last 25 years but includes also reference to other relevant, published work. Itshould be emphasised, though, that the intention is not to attempt to critically review all theavailable literature in the area.

A specific aim is to show the present scientific understanding of the subject. Therefore wehave devoted the first 3 chapters following the introduction to our view as to how essentialphenomena and properties of V-microalloyed steels should be understood and interpreted,ranging from thermodynamic predictions of relevant phase equilibria in V-steels, tomicroalloying effects in austenite and precipitation in ferrite and its effect on strength. Thelast 3 chapters are devoted to a comprehensive account for the role of V related to the mostimportant processing steps that microalloyed structural steels are subjected to, viz.thermomechanical processing, continuous casting and welding.

4

2. ALLOY SYSTEM AND THERMODYNAMIC BASES

The strengthening effect of microalloying additions may be produced by the dispersionstrengthening effect of fine carbo-nitride particles or by grain refinement, i.e. inhibition ofgrain growth by carbo-nitride particles, or by a combination of these two effects. In order tomaintain a fine austenite grain size prior to transformation, particles that remain undissolvedin the austenite, or particles that will precipitate during hot rolling are required. To producethe very fine particles that are responsible for dispersion strengthening (i.e. particles that are2-5 nm in diameter), it is necessary that these should be freshly precipitated during or aftertransformation to ferrite.

To achieve the desired metallurgical states, a detailed knowledge of the solubilities of themicroalloy carbides and nitrides is required, together with a knowledge of their precipitationbehaviour. An understanding of the role of different microalloying elements can be gainedfrom the solubility product data summarised in Fig. 2.1 from recent thermodynamicalevaluation. Despite a simplification the individual solubilities of the microalloyed carbidesand nitrides offer clear directions for the selection of microalloying additions for specificpurposes. It is seen that TiN is extremely stable and can withstand dissolution at hightemperatures during reheating prior to rolling or during welding. Niobium nitride and carbidehave relatively low solubilities and may precipitate out in the later stages of rolling.Vanadium on the other hand, has a rather high solubility in austenite even at temperatures aslow as 1050°C. It is also seen that the nitrides are substantially less soluble than thecorresponding carbides. This is especially true for titanium and vanadium where thedifferences are particularly pronounced.

Fig. 2.1 Solubilities of microalloy carbides and nitrides.

In normal circumstances the solubility relationships shown in Fig. 2.1 represent asimplification, as steel may contain more alloying elements with high affinity for carbon andnitrogen which alter the solubility of the microalloyed carbides and nitrides.

5

2.1 Thermodynamic considerations

2.1.1 Approximate expression for the solubility productAs was stated above the solubility of carbides and nitrides in austenite and ferrite is usuallyexpressed as a solubility product in terms of wt pct of microalloy element and carbon and/ornitrogen. The temperature dependence of the solubility product is expressed by the Arrheniusrelationship in the form,

log ks= log [M][X]= A-B/T, 2.1

where ks is the equilibrium constant, [M] is the dissolved microalloy (wt%), [X] is the content(wt%) of nitrogen or carbon, A and B are constants and T is absolute temperature.

Several authors have made efforts to establish the solubility of carbides, nitrides and carbo-nitrides in austenite (see compilation in 2.1). All of these investigations expressed their resultsin a form of solubility products. However, the results proposed by various authors as the"best-line" relationships sometimes differ considerably. An analysis of published dataconcerning the solubility of VC, VN, NbC, NbN, TiN and AlN, Fig. 2.2, has shown that thereexist more than ten different solubility equations for each carbide/nitride, and the spreadbetween them is significant, in most cases greater than 150°C. It should be pointed out thatthe situation is even more complicated when more than one microalloy element is present andwhen the ratio of carbon to nitrogen is changed. The influence of carbon and nitrogen oncarbo-nitride formation has been handled in different ways by different authors, (2.2). Forexample, for Nb-microalloyed steels the nitrogen was often considered to modify the effectivecarbon concentration as [C+12/14N] (e.g. 2.3).

The analysis of the solubility data in terms of solubility products implicitly assumes that theactivity coefficients of microalloying elements as well as carbon and nitrogen are equal tounity and the microadditons are treated as a dilute solute and any interactions between solutesin the system are neglected. Therefore the individual solubility equations apply only to aparticular composition investigated and are not capable of predicting the solubility at differentcompositions, where the effects of solute interactions become significant. The carbides andnitrides are also non-stoichiometric and, hence, their composition can vary when precipitatedin steels.

The similarity of crystal structure of carbides and nitrides enables them to show mutualsolubility. Except for (V,Zr)N, all other carbides and nitrides formed by Ti, Nb, Zr and Vshow continuous or extended mutual solubility. For V and Zr the large difference in atomicsize enables their compounds only to show limited mutual solubility. It is necessary toemphasise that aluminium differs from the microalloying elements V, Nb and Ti because itdoes not form a carbide in steel, but only the nitride, AlN, having a different crystal structure(hexagonal close packed). Therefore, there is no mutual solubility with the carbides/nitrides ofV, Nb and Ti, all of which have NaCl cubic crystal structure.

6

Fig. 2.2 Solubility data for (a) VC and VN, (b) NbC and NbN, (c) TiN and (d) AlN (2.1).

2.1.2 Thermo-Calc modelIn order to accurately predict the austenite/ferrite carbo-nitride equilibria, expressionsdescribing the chemical potentials of the transition metals and carbon and nitrogen inaustenite and metal carbo-nitride are required as a function of composition and temperature.To describe the thermodynamics of austenite, ferrite and the non-stoichiometric carbo-nitridephase respectively, the Wagner formalism (2.4) for dilute ternary solutions or a sublattice-subregular solution model proposed by Hillert and Staffanson (2.5) can be used. The Wagnerformalism is in fact identical to the regular solution model if the second-power terms areneglected.

In the sublattice-subregular model proposed by Hillert and Staffansson the partial Gibbsenergy of component M in each individual phase is expressed as,

G RT lna ln GM M ME

M= = +RT x (2.2)

where a is the activity and EMG is the excess Gibbs energy.

7

The real crystalline structure of nitrides and carbides is expressed in the form of twosublattices, one for the substitutional elements and one for the interstitial elements. Most ofthe interstitial sites are usually vacant and vacancies (Va) must thus be treated as an additionalelement. In principle, an interstitial solution cannot extend beyond the composition where allthe interstitial sites are occupied. To take fcc V-N as an example, the model of the solutionextends from pure V, which may be regarded as a V1Va1 to a compound V1N1. A generalThermo-Calc database for the thermodynamic properties of microalloyed steels has recentlybeen developed at the SIMR (2.6), which comprises the thermochemical parameters for multi-component systems of HSLA steels. In the Thermo-Calc approach (2.7), the properties ofelements and phases are described by the mathematical representation of their thermodynamicproperties and the phase equilibria and complete phase diagrams are calculated by minimisingthe Gibbs energy. The database can also be used to calculate metastable equilibria since thedata can be extrapolated from the regions where the phases involved are stable. Interactionsbetween different atoms in higher-order systems are described by the mixing parameters.This model is now successfully utilised to calculate the volume fraction and compositions ofparticles at different temperatures as well as the driving force for precipitation. The analysisof the alloy systems of microalloyed steels described below was made with the aid of theThermo-Calc model.

2.2 Thermodynamic description of the Fe-V-C-N system

2.2.1 The Fe-V-C Ternary SystemIn the Fe-V-C ternary system the following phases have been considered (including avanadium-rich eutectic); liquid, ferrite, austenite, VC1-x carbide, V2C1-x carbide, V3C2 carbide,sigma (σ) phase, graphite and metastable phase cementite. Both V2C and VC carbides haveappreciable ranges of homogeneity due to the carbon deficit (2.8). The carbon atoms arerandomly distributed in the interstitial sublattice at high temperatures. The solubility of VCcarbide in the Fe-V-C system has been studied rather extensively (2.9). The experimental iso-carbon activity data in austenite were described by the ternary interaction parameter, LFe V C

fcc, : ,

in the form -(7645.5 + 2.069T)[1+(yFe-yV)] (2.10).

2.2.2 The Fe-V-N Ternary System

The Fe-V-N ternary system consists of two bcc phases (ferrite (α) V(b-V)), two fcc phases(austenite (γ), vanadium mononitride (VN)), two hcp phases (iron nitride (ε), vanadiumdinitride (V2N)), liquid (L), Fe4N(γ´) and so-called sigma phase (σ), Fig. 2.3 (2.11). Inaddition to these phases, N2 gas is involved in some equilibria.The VN phase has a NaCl type of structure, which can be described as fcc V with all theinterstitial sites filled with N. Since this phase is isomorphous with austenite from acrystallographic point of view, the solubility of VN in austenite is described by a miscibilitygap which separates the Fe-rich fcc phase (i.e. austenite) from Fe-poor fcc phase (i.e. VN).Several reports were presented previously on the equilibrium nitrogen solubility and nitrideformation in austenite (2.12) and ferrite (2.13). Considerable experimental uncertainty stillremains which makes some difficulty in the accurate determination of the interactionparameters. This is mainly due to the quite narrow austenite single phase region in theexperimental temperature range. The ternary interaction parameters LFe,V:C were evaluated byOhtani and Hillert (2.11) for liquid phase and fcc. Using these parameters it was possible toestimate the solubility product of VN in austenite and ferrite. The following equations wereobtained, in austenite; log (wt%V)(wt%N)= - 7600/T – 10.34 + 1.8lnT + 7.2x10-5 T (2.3)and ferrite; log (wt%V)(wt%N)= - 12500/T + 6.63 – 0.056lnT + 4.7x10-6 T (2.4)

8

Fig. 2.3 The Fe-V-N ternary system at 800°C.

2.2.3 Solubility and composition of V(C,N) in V-microalloyed steelsThermo-Calc calculations of precipitation of V(C,N) in steels microalloyed with vanadium isshown in Fig. 2.4. For these calculations the parameters for the carbon, nitrogen andvanadium interaction with other elements present in steels have been taken from the SSOLand TCFE Thermo-Calc databases (2.7). For V-microalloyed steels the effect of manganeseon the activity of V is especially important. Mn is known to increase the activity coefficient ofV and at the same time decrease the activity coefficient of C. Also shown in this figure is themole fraction of nitrogen in carbo-nitrides at various temperatures and for various nitrogencontents from hyperstoichiometric levels to zero. This shows that vanadium starts toprecipitate in austenite as almost pure nitride, until practically all the nitrogen is consumed.When the nitrogen is about to be exhausted there is a gradual transition to form mixed carbo-nitrides, Fig. 2.4(b). Nitrogen rich vanadium carbonitrides may precipitate during orsubsequent to the γ-α transformation. Precipitation during the transformation is a result of thesolubility drop of the vanadium carbo-nitrides associated with the transformation fromaustenite to ferrite at a given temperature.A significant feature of the solubility of vanadium carbide in austenite is that it isconsiderably higher than those of the other microalloy carbides and nitrides, suggesting thatvanadium carbide will be completely dissolved even at low austenitising temperatures.

2.3 The Fe-Ti-C-N systemThe calculated precipitation of TiN and TiC in steels microalloyed with titanium, is shown inFig. 2.5. Also shown in this figure is the mole fraction of nitrogen in carbo-nitrides at varioustemperatures for various nitrogen contents from hyperstoichiometric levels to zero. Titaniumforms an extremely stable nitride TiN that is virtually insoluble in austenite and may beeffective in restricting its grain growth during processing and welding. Only a small titaniumaddition (~ 0.01%Ti) is required for this purpose. If larger contents are present the excesstitanium can precipitate at lower temperatures as carbide TiC which contributes aprecipitation hardening effect. This shows that the solubilities of titanium nitride and titaniumcarbide are significantly different and the precipitates formed in the austenite are almost purenitrides, until practically all the nitrogen is consumed. It is worth noting that in the Ti-steel

9

containing 0.01%Ti or less for controlling austenite grain size there is usually sufficientnitrogen to combine all Ti as TiN. When the nitrogen is about to be exhausted there is agradual transition to form mixed carbo-nitrides, Fig. 2.5 (b).

Fig. 2.4 Example showing the precipitation of nitrides, nitrogen rich carbonitrides andcarbides in 0.10% V steels at various nitrogen contents.

Fig. 2.5 Calculated precipitation of nitrides, nitrogen rich carbonitrides and carbides in0.01%Ti steels at various nitrogen contents.

Titanium exerts a strong affinity not only for carbon and nitrogen but also to other elementssuch as oxygen and sulphur. In Ti-microalloyed low-carbon steels the carbosulphide Ti4C2S2or sulphide TiS may form which remain undeformed during hot rolling (2.14). According toFig. 2.6, Ti4C2S2, is more stable than MnS and the formation of MnS is restrained by theaddition of Ti to the steels. It is also clear from this figure that the formation of Ti-sulphidesor carbosulphides will decrease the amount of Ti available for TiN and TiC. Therefore forproper calculation of the precipitation of microalloyed carbides and nitrides the effect of allelements present in the steel must be considered.

10

Fig. 2.6 Calculated precipitation of MnS, Ti4C2S2 and TiN in low manganese steel.

2.4 The Fe-Nb-C-N systemFor Nb-steel, Fig. 2.7 where the difference in solubility between the nitride and the carbide isrelatively small, mixed carbo-nitrides form at all nitrogen contents, even at thehyperstoichiometric levels (2.15). Hence, for realistic contents of carbon and nitrogen in steel,relatively pure nitrides cannot form at all.

Nb(C,N) is stable at low temperatures in austenite but dissolves at higher temperatures, forexample, during re-heating prior to rolling. Nb(C,N) can precipitate quite readily in austeniteunder deformation (strain induced precipitation) and the particles so formed resist graingrowth and even recrystallization of the austenite during the intermittent deformation at lowertemperatures. The deformed austenite structure transforms to fine grained ferrite on cooling,which imparts both high strength and toughness to such ‘controlled rolled’ steels. A furtherprecipitation of the remaining niobium as smaller particles during cooling gives rise toadditional strengthening.

Fig. 2.7 Example showing the precipitation of Nb carbonitrides in 0.03% Nb steels at variousnitrogen contents.

11

2.5 The Fe-Al-B-C-N systemAluminium is not usually classified as a microalloying element. The precipitation of AlN can,however, have strong effects on the properties of microalloyed steel, which are at least assignificant as those produced by Nb, V or Ti additions. Al-killed steels have long been knownto show improved strain ageing resistance and enhanced deep drawability, and for many yearsthe term ”aluminium grain refined” has been synonymous with grain size control (2.16).Aluminium is normally dissolved in austenite at high temperature prior to rolling but thenitride phase is thermodynamically stable at lower temperatures. However, AlN nucleateswith some difficulty in steel and has a widely different morphology from other nitrides (2.17).AlN precipitation in ferrite has been extensively studied in connection with the developmentof favourable crystallographic textures. Considerably less investigation has been made intothe precipitation of AlN in austenite. It is well known that the kinetics of AlN precipitation inaustenite are generally slow and dependent upon thermal history.

Boron additions enhance the hardenability of heat-treatable steels and assist in the control ofstrain ageing in sheet steel for deep drawing (2.18). Boron has long been known to have apowerful influence when added as a microalloying element to steel but its usage has oftenbeen restricted because of variability of its behaviour in practice. This problem of variabilityis mostly due to the strong affinity that boron has for oxygen and nitrogen and the very lowcontents of boron that are desired (~0.001%), such that the true level of active boron has beendifficult to control in practice (2.19).

Fig. 2.8 Precipitation of BN and AlN in B-bearing Al-killed steels (2.18).

When considering steels containing boron and aluminium the situation is more complicatedthan for other microalloying elements 2.06), Fig. 2.8. The solubility of BN is smaller thanthat of AlN for temperatures of practical importance so that BN might be expected toprecipitate preferentially on thermodynamic grounds. However, the total content ofaluminium in typical killed steels is at least an order of magnitude greater than the boronadditions which are normally used and for this reason AlN often tends to be stabilised.

12

2.6 Chemical composition of precipitated particles in multi-microalloyed steelsThe carbides and nitrides of V, Ti and Nb show extensive mutual solubilities which arisesfrom the fact that they have the same cubic crystal structure and have very similar latticeparameters.The precipitation of carbo-nitrides in multiple microalloyed steels with V, Nb, Ti and Al isshown in Fig. 2.9(a), where the soluble components are expressed as a function oftemperature. In Fig. 2.9 (b) the mole fraction of Ti, Nb and V in M(C,N) is also included.

As expected, the major part of the precipitate at the highest temperatures is TiN whilst furtherprecipitation at lower temperatures is predominantly Nb(C,N). Thus, the primary precipitatesto be formed in austenite are composite (Ti,Nb)-nitrides. The Ti and Nb contents of thosedepend on steel composition and precipitation temperature. Thermodynamic calculationsimply that at the high temperature of 1200°C, primary (Ti,Nb,V)N particles contain ~20% ofNb and ~5% of V for multiple microalloyed steels. It is clear from this figure that the volumefraction of microalloy nitrides at high temperatures will be considerably larger in the Ti-Nb-Vsteel than in the single microalloyed steels. It may also be seen that AlN (with close packedhexagonal structure) has little or no solubility for other microalloying elements and starts toprecipitate above 1200°C in the presence of 0.035% Al, Fig. 2.9 (a).

Fig. 2.9 Example showing the precipitation of nitrides and nitrogen-rich carbonitrides inmultiple microalloyed steel (a) and the mole fraction of Ti, Nb and V in M(C,N) (b).

It should be mentioned, however, that the uniform composition of second phase particlessuggested from the thermodynamic calculations is often not the true situation in practice. Theparticles are frequently cored, reflecting the high temperature stability of a titanium-rich andnitrogen-rich compound in the interior of the particle with relative enrichment of Nb and Vtowards the surface of the particle. This suggests that these primary composite nitrides can actas nucleation centres for subsequently precipitation at relatively low temperatures, forexample forming a NbN-deposit around a core of TiN. Similar results have been obtained forNb-V steels where niobium-vanadium nitrides have shown Nb and N enrichment in theprecipitates, in accord with the higher thermodynamic stability of NbN. These coring effectsmay be a result of very slow diffusion of Ti in carbo-nitrides or the existence of a miscibilitygap as explained further in Chapter 3.

13

Another important aspect of multiple microalloying is that the niobium or vanadium tied up as(Ti,Nb,V)N particles is not available for subsequent thermo-mechanical treatment. Fig. 2.9 (b)shows that a significant fraction of added niobium may remain undissolved even at highreheating temperatures and cannot contribute to retardation of recrystallization and/orprecipitation strengthening. Vanadium, on the other hand, exhibits higher solubility and istherefore normally fully available for precipitation strengthening in ferrite.

(a)

(b)

Fig. 2.10 (a) Chemical driving force, ∆Gm/RT, for precipitation of VC and VN in 0.12% Vsteel. (b) Effect of equilibrium and metastable carbon content in ferrite on driving force forprecipitation of V(C,N) at 650°C.

2.7 Thermodynamic Driving Force for PrecipitationThe precipitation process proceeds at a perceptible rate only if there is a driving force, that is,a free energy difference between the product and parent phases. This driving force enters thesteady state nucleation rate in a central way and must be known with some accuracy ifnucleation rates are to be calculated, or even estimated. The chemical driving force fornucleation of VN and VC in V-microalloyed steels, determined from the HSLA database areillustrated in Fig. 2.10. It can be seen that, as the temperature is decreased the driving force

3.5

4

4.5

5

5.5

6

0 50 100 150 200 250 300

Nitrogen content in ferrite, ppm

DR

IVIN

G_F

OR

CE

for V

(C,N

)

equilibrum carbon content in ferrite

250 ppm carbonin ferrite

14

increases monotonically and changes slope after the austenite to ferrite transformation. Thedominating effect of N on the driving force is clearly seen in this figure.

2.8 Practical consequences of different solubilities of V carbides and nitridesVanadium is the most soluble of the six elements discussed here and does not readilyprecipitate in austenite. Since V(C, N) has high solubility in austenite it hardly affects the hotdeformation process but rather precipitates during cooling in ferrite, thereby increasing thestrength level of the steel by precipitation hardening. For hard precipitates such as microalloycarbonitrides bypassing of dislocations is expected to occur by bowing between the particles(Orowan-mechanism) under all practical conditions. This means that there is only oneparameter, which determines the precipitation strengthening namely the interparticle spacing.The smaller is the interparticle spacing the larger is the strengthening effect.

In precipitation reactions, the decisive factor which minimises the interparticle spacing (andmaximise the precipitation strengthening) is the rate of nucleation since it determines theparticle density. In turn, the most important parameter controlling the nucleation frequency isthe chemical driving force for precipitation. Hence, the strong effect of N in increasing thisdriving force, Fig. 2.10(a), is the explanation of the well-known increase in strength by N inV-microalloyed steels (for a full discussion see Chapter 4). It should be noted, however, thatthe driving force can also be raised by V-addition, Fig. 2.10(b), but the effect perconcentration unit is much less. Moreover, V-additions increase alloying costs significantlywhereas the usage of N at the present levels is free.

A combination of V with other microalloy elements in the same steel can naturally lead tointeractions since they compete for the same interstitial elements and may also form mixedcompounds with one another. For example, the stable TiN phase may incorporate a significantamount of V and even AlN has sometimes been detected in complex titanium nitride particles.Since the (Ti,V)N particles are rather coarse such an effect reduces to some extent thecontribution which the V can be expected to make towards precipitation strengthening.

2.9 Summary of chapter

• The solubility of V(C,N) in austenite and ferrite is much larger than that of othermicroalloying elements.

• The solubilities of the carbides and nitrides of V differ widely; the solubility product ofVN is about two orders of magnitude lower than that of VC. This implies that N has adecisive role in V-microalloyed steels, especially for the enhancement of the drivingforce for precipitation.

• Interaction of V with other microalloy element leads to the formation of mixedcarbonitrides and decreases the amount of vanadium dissolved in austenite.

• To accurately predict the austenite/ferrite carbo-nitride equilibria, expressions describingthe chemical potentials of the transition metals and carbon and nitrogen in austenite andferrite as well as metal carbo-nitride are required as a function of composition andtemperature. Interactions between alloying elements influence significantly the solubilityof V(C,N).

15

3. MICROALLOYING EFFECTS IN AUSTENITE

In commercial V-microalloyed steels small amounts of Ti (~ 0.01%) are commonly added toprevent excessive grain coarsening at high temperatures. The technical background to this isthat Ti reacts with the nitrogen in the steel to form a fine dispersion of very stable TiN. Forthis to occur in typical HSLA plate and strip steels the normal levels of N in BOF and EAFsteel making, i.e. 0.004 – 0.015%, are sufficient. However, to attain the fine TiN sizenecessary for effective grain growth control fast cooling is needed during the solidification ofthe steel, as in continuous casting of slabs.

However, the TiN precipitation may by affected by a second microalloy addition, such as V,in various ways. This alone, but especially together with the formation of a second precipitate,may affect the grain coarsening behaviour significantly. Therefore these questions, arising inTi-bearing V-steels, are discussed in some detail below.

3.1 Precipitation and dissolution of microalloy-carbonitrides in austeniteIt is well established that TiN forms as a fine precipitate dispersion early after solidification inTi-V and Ti-Nb HSLA-steels having less than about 0.02%Ti. At lower temperatures whenthe solubility of Nb(C,N) and V(C,N) has been exceeded many experimental studies haveshown that these phases co-precipitate on already existing TiN-particles to a large extent. Ofcourse, as the driving force for carbonitride formation increases at lower temperatures in theaustenite regime the probability for nucleation of new carbonitride, rich in V or Nb andeventually pure V- or Nb-carbonitrides, increases.

Fig. 3.1 Calculated amounts of Ti, V and Nb precipitated as (Ti,M)N in a Ti (a), Ti-V (b) andTi-Nb (c) steel. GCT is the experimentally determined grain coarsening temperature ofcontinuously cast 220 mm slabs (2.6).

Fig. 3.1 shows results from calculations for Ti-, Ti-V-, and Ti-Nb steels using thethermodynamic database developed for microalloyed steels (3.1,3.2). The graphs show thefractions of Ti, V and Nb precipitated as nitrides at equilibrium in the austenite. From this wecan see that the precipitation curves, for Ti are the same for all three steels. Hence, we canconclude that the temperature for initial precipitation and complete dissolution should beequal for all three steels. Furthermore, as long as all V or Nb on lowering the temperature co-

16

precipitates on pre-existing TiN, we should expect the density of precipitates also to be thesame in the steels; of course the individual particles will grow larger in the V- and Nb-steelsas these elements co-precipitate. However, at lower temperatures V- and Nb-carbonitrides areexpected to form as new precipitates, and indeed this is proven experimentally (3.3,3.4). Inaccordance with this reasoning it is generally found experimentally that larger particles thatformed at high temperatures are high in Ti and low in V or Nb, whereas smaller ones formedat low temperatures are low in Ti and high in V or Nb (3.3,3.4). This is demonstrated in Fig.3.2 for a Ti-V steel showing a gradual decline of the Ti-content in (Ti,V) N as particle sizedecreases (3.3). Along the same lines Fig. 3.3 shows how the spread in V and Ti in a (Ti, V)N-population increases with decreasing reheating temperature; as expected the widest spreadis observed in the strand-cast condition (220 mm) (3.5). The low N-steel with nearstoichiometric Ti/N ratio exhibits a considerably smaller spread due to the fact that nearly allN is consumed to form TiN.

Fig. 3.2 Relationship between composition of(Ti,V)N-inclusions in a Ti-V-steel with high Vand N, and the particle size. The steel wasmade as a laboratory cast but was cooled at arate equivalent to that of 220 mm concast slab(3.3).

Fig 3.3 Spread in composition of (Ti,V)N-particles in two Ti-V-steels with low andhigh N in concast condition and for varyingreheating temperatures (3.5).

A general conclusion is that the stability of TiN itself is not expected to change by additionsof other microalloying additions, such as V or Nb. However, the separately formed V- andNb-rich carbonitrides are less stable and dissolve at lower temperatures. This is exactly whathas been found experimentally (3.6), but has sometimes incorrectly been stated as a decline ofthe stability of TiN.

Beside the rate of nucleation the density of precipitates is also controlled by coalescencewhich occurs at a constant volume fraction of particles after the completion of theprecipitation reaction. The rate of particle coarsening during coalescence is given by

)(4

333

XXRTXDV

rrM

MCNM

MMmo γ

γγσ−

=− (3.1)

17

where r is the mean particle radius at time t, ro the initial mean radius, σ is the interphaseenergy, Vm the particle molar volume, DM

γ the diffusivity of the rate controlling element M,X M

γ the solubility of M in austenite and X MCNM the concentration of M in the (Ti,V, Nb) (C,

N). From this we can directly learn that in our case Ti will normally be rate controlling as aresult of its much lower solubility compared to V or Nb. It is only for very high contents of Vor Nb in the carbonitrides that these elements become rate controlling. So again we expect nodifference in resistance to coarsening of TiN in Ti-steels and Ti-rich nitrides in V- and Nb-steels. Again we have not found any experimental results opposing this statement.

Fig. 3.4, showing the solubility curves for TiN in austenite for various temperatures,demonstrates that the Ti/N ratio affects the TiN stability in various ways (3.2). Changing theTi and N contents along the stoichiometric lines results in the largest change in solubility.Hence, using over-stoichiometric contents of N, as indicated in the graph, gives rise to:

(i) a higher dissolution temperature of TiN

(ii) a smaller fraction of dissolved TiN for a given temperature increase

(iii) a reduction of the Ti-content in austenite in equilibrium with TiN, resulting in slowercoarsening.

The usefulness of over-stoichiometric N-contents in Ti-V-steels for grain size control in theHAZ of welds has been demonstrated (3.2,3.7).

Fig. 3.4 Effect of excess nitrogen on dissolved Ti (3.2).

An interesting feature of (Ti,V)(C,N) precipitation found in austenite of Ti-V steels is that theparticles exhibit compositional gradients such that the external parts are richer in V and theinterior is richer in Ti (3.3,3.6). Electron microscopy and microanalysis of the (Ti,V)-carbonitrides formed in austenite show furthermore that the compositional gradients of theparticles are not always smooth. The outer V-rich part occurs commonly as dendrites (3.5). Atfirst this may appear natural in view of the sequential precipitation of Ti and V duringcooling. However, a prerequisite is then that the interdiffusion of Ti and V in the (Ti,V)-

18

carbonitride is substantially lower than in austenite. Otherwise there will be ample time forhomogenisation of the elements to occur. Unfortunately, the scatter of reported diffusioncoefficients in refractory carbides and nitrides is very large. For instance, the self diffusion ofTi in TiC is slower by 1010 than Ti in austenite, whereas the interdiffusion of Nb in (Ti,Nb)Cis only slower by 103 (3.8,3.9). This implies that in the former case any compositionalgradient formed during precipitation of (Ti,V)N will be frozen whereas in the latter case therewill be time for full equilibration in normal processing. In view of the fact that it isinterdiffusion that is significant in the present situation it seems that we should give moreconfidence to the reported results on (Ti,Nb)C, and hence that homogenisation will occur ofthe (Ti,V)-particles. This, in turn, would imply that Ti-rich and V-rich carbonitrides can co-exist in equilibrium, and consequently leads to the conclusion that there is a miscibility gap inthe (Ti,V)(C,N)- system. In fact, this was exactly the result of a recent thermodynamicanalysis of the Ti-V-N system (3.10). It should be borne in mind, though, that only smallshifts in the input data for the thermodynamic analysis might cause a change from full inter-solubility to de-mixing.To illustrate the time scale and the temperature dependence of the changes of the precipitationthat may occur in the austenite range we have found it instructive to compute

• the time for completion (90%) of the TiN precipitation from an initial state of all Ti and Nin solution

• the time for TiN dissolution (90%) from an initial state of full equilibrium between TiNand austenite at 1000°C

• the time for 50% increase of initial precipitate size by coalescence.

The computations have been carried out for different particle dispersions, expressed asinterparticle spacings, and for the cases of precipitation and dissolution allowance is made forsoft impingement of adjacent solute-enriched zones. The results are displayed in Figs. 3.5 and3.6. The rates of coalescence and dissolution depend critically on the solubility of TiN inaustenite. The solubility data available in literature exhibit a considerable scatter, as wasdemonstrated in chapter 2. In the selection of data for the present computations our preferencehas been to use data that produce reasonable agreement with observed ripening of TiN (3.5),although this has implied that we have chosen somewhat lower solubilities than the mean ofthe scatter band.

From Fig.3.5 we may note that very little dissolution of TiN takes place, even at the highesttemperatures. At 1450°C, for instance, an initial precipitate size of 5.0 nm equilibrated at1000°C only shrinks to 4.80 nm. The time for dissolution is about an order of magnitudefaster than the time for precipitation, Fig.3.5. The explanation is the very minute dissolutionof TiN as compared to the amount of TiN transferred for complete precipitation. Obviouslythis effect outweighs the lower Ti-gradient in dissolution as compared to precipitation. Thecomputations of coalescence in Fig.3.6 show furthermore that coalescence does not producesignificant particle coarsening for normal heating temperatures and times except for the finestparticle dispersion, particle radius 5 nm and spacing 175 nm, and above about 1300°C.Hence, these computations show that the grain coarsening temperatures of about 1200°C andeven lower that has been observed in thermo-mechanically processed steels and will bediscussed below cannot be accounted for TiN-dissolution or TiN-ripening.

19

Fig. 3.5 Time for 90% completed precipitation and dissolution of TiN as a function oftemperature in a 0,01%Ti-0,01%N steel at given particle spacings (175, 1750, and 7000nm).Precipitation starts with all Ti and N in solution, whereas the dissolution starts in a statecorresponding to full equilibrium between austenite and TiN at 1000°C. Labels at data pointsindicate changes in particle radius.

Fig 3.6 Time for 50% increase of the initial TiN size by coalescence as a function oftemperature. The initial state corresponds to complete precipitation of TiN for 3 givenparticle spacings, 175, 1750, and 7000 nm.

3.2 Grain growth inhibition and grain coarseningIt is a common view that abnormal grain growth is invariably associated with ripening anddissolution of the restraining particles. However, this is normally not the case for Ti-microalloyed steels since the TiN-dispersion is virtually unaltered at temperatures belowabout 1250oC (3.3). To give a sound understanding of the grain coarsening behaviour of thesesteels and all the associated important and often intriguing experimental findings it isnecessary to go back to the basics of grain growth and its dependence of impeding particles.

20

The growth rate for an individual grain of diameter D in a distribution of grain sizes with themean size D and containing a dispersion of small particles with volume fraction f and meanradius r is given by (3.11)

)8311(

rf

DDM

dtdD ±−= ασ (3.2)

where α is a constant (~ 1), σ the grain boundary energy, and M the grain boundary mobility.

Thus the impeding effect of particles enters by the term rf

83 .

From Eq.(3.2) the following conclusions, essential in abnormal grain growth, can be drawn:

(i) Normal grain growth reaches a limiting grain size

frD

98

limit = (3.3)

(ii) However, if, in an assembly of grains with mean size D limit, there is one grain largerthan 1.5 D limit it can grow and with increasing speeds as it grows. Hence, a fewindividual grains can in this fashion cause abnormal grain coarsening.

(iii) The tendency for abnormal grain growth diminishes with increasing D and increasing3f/8r. In fact, there is a limiting mean grain size beyond which abnormal grain growth istotally prevented

D abnormlimit =

fr

38 (3.4)

Thus, a perfectly stable precipitate dispersion causing a total arrest of normal grain growth isnot a guarantee for grain coarsening by abnormal growth. This is the explanation why we seegrain coarsening temperatures in Ti and Ti-V steels of 1200-1250oC, despite the fact that wedo not observe measurable dissolution or ripening of TiN at these temperatures.

It is tempting to believe that on heating Ti-microalloyed into the austenite range normal graingrowth occurs and the austenite reaches its limiting grain size set by the TiN-dispersion, Eq.(3.3). However, there is ample evidence from early investigations by Siwecki, Sandberg andRoberts (3.4) that the ferrite-austenite transformation produces an austenite grain size wellabove D limit. In support for this it has been found that the average grain size at the graincoarsening temperature is about the same in spite of large differences in amounts of TiN (3.4).A corollary of this conclusion is that the grain coarsening temperature should be a function ofthe heating rate through the ferrite-austenite transition. With decreasing heating rate the grainsize of the transformed austenite increases; hence the driving force for abnormal grain growthdecreases, and as a result the grain coarsening temperature is raised. This explanation isconfirmed by experiments the results of which are summarised in Fig. 3.7. If the grain sizeinstead was controlled by the TiN dispersion, then the grain coarsening temperature should beindependent of heating rate.

21

Fig. 3.7 Effect of heating rate on the graincoarsening temperature (GCT) for a 0,036V-0,013Ti-0,011N steel. Encircled numbersrefer to the size of small austenite grains atthe GCT (3.4).

Fig. 3.8 The relationship between austenitegrain coarsening temperature and N-contentfor continuously cast slabs (220 mm) andsmall laboratory ingots of 0,01%Ti HSLAsteels(3.2)

Under the very high heating rates in welding the ferrite-austenite transformation should leadto particularly fine austenite grains and therefore a strong driving force for abnormal graingrowth. Still we know that Ti-microalloyed steels containing a fine dispersion of TiN exhibitvery good HAZ toughness as a result of fine austenite grain structure. This apparent anomalyis due to the fact that it takes a considerable time for the few growing grains consuming all thefine grains and thereby causing an abnormal grain size. The short dwell periods at hightemperatures during welding are not long enough for this to occur. On the other hand theoverwhelming majority of the grains will be totally inhibited from coarsening by normal graingrowth since they, although finer than for slower heating rates, are still likely to be largerthanDlimit for normal grain growth. Hence, the overall result will be a fine austenite structure.The theoretical explanations given above are therefore also consistent with the observedrefined austenite in the HAZ of weldments in Ti-micoalloyed steels.

Fig. 3.8 summarises measurements of grain coarsening temperatures for steels with varyingTi/N ratios and cast at different cooling rates (3.2). The larger cooling rates in laboratoryingots produce a finer precipitation and a particle density increasing with the N content. Theslower cooling in the continuously cast slab produces coarser TiN, and increasingly so withincreasing N content due to the fact that precipitation commences at higher temperatures. Theincreasing precipitate density in the former case progressively prevents abnormal graingrowth, and thus causes a rise in GCT. The decreasing TiN density in the slabs, on the otherhand, gives the opposite behaviour.

Again, assuming that normal grain growth up to the limiting size controlled by the TiNdispersion would determine the average grain size before abnormal growth sets in, we noticethat this would cause problems in explaining the observations in Fig. 3.8. As a matter of fact,if 1/ D were to be more decisive than 3f/8r in Eq. (3.2) for controlling abnormal growth, thenwe should even expect GCT to be lowered with N content for the laboratory ingots as D limit isreduced with increasing particle density. In a similar way, this line of reasoning would giveincreasing GCT for the slab materials.

22

The GCTs in Fig. 3.8 are assigned to both Ti and Ti-V microalloyed steels (3.5,3.12),indicating that all observations can be understood in terms of the Ti- and N-contents and thecooling conditions during casting. So for GCT in as cast material, V additions to a Ti-steelseem to have no or only marginal effect.

It has long been known that GCT of hot rolled Ti-microalloyed steel is lower than in the castcondition (3.4). It is also found that as the number of reheatings increases as processingproceeds the grain coarsening resistance deteriorates. This behaviour can readily beunderstood by the refinement of the average grain size engendered by multiple reheating, cf.Eq. (3.2).

However, Roberts et al (3.4) have convincingly shown that the grain refinement during hotrolling cannot alone explain the observed reductions in GCT for Ti-V steels in hot rolledcondition as compared to cast. They designed experiments carefully to produce TiNdispersions and austenite grain sizes with only marginal differences, but widely different V-carbonitride precipitation by adopting varying cooling through the austenite-ferritetransformation. Fig 3.9 shows the measured GCT for steels processed in this way, coveringcooling rates from 220 mm concast slab to water quenching of 10 mm plate. Thus theobserved reduction in GCT of about 130oC cannot in this case be accounted for by a reductionin grain size. Roberts et al propose that the occurrence of bimodal precipitate distributions ofV-carbonitrides and (Ti, V)-nitrides and the rapid dissolution of the former on heating causessufficient change in the grain size distribution so as to be conducive to abnormal graingrowth. The results of Fig. 3.9 demonstrate also that GCT is further reduced as the differencein density of the two precipitate dispersions increases.

Fig. 3.9 The effect of cooling rate during previous processing on the grain coarseningtemperature (GCT) of a 0,036V-0,013Ti-0,011N-steel. Encircled numbers are the averagediameters of the ´matrix´ austenite grains at the GCT (3.4).

23

3.3 Effects on recrystallizationOne of the most essential features of microalloying additions is their influence onrecrystallization during controlled rolling, either to prevent recrystallization and therebyproducing ‘pancaked’ austenite grains during hot deformation and conditioning the austeniteto form fine ferrite grains during transformation, or to minimise their effect onrecrystallization thereby allowing repeated recrystallization to occur during multipledeformation and a gradual refinement of austenite and the subsequent ferrite. The first case isthe controlled rolling route for the Nb-bearing structural steels, and the second is the route,recrystallization controlled rolling, adopted for the Ti-V structural steels (cf chapter 5). Theeffect of the microalloying elements on recrystallization differs widely, especially so for Vand Nb, as demonstrated in the classical graph due to Cuddy (3.13), Fig. 3.10; it shows therecrystallization stop temperature as a function of the microalloy content (atom %), asdissolved when the hot deformation starts.

The mechanism by which the microalloying elements act to raise the recrystallization stoptemperature has been the subject of much debate throughout the history of microalloyed steelsand numerous investigations have been performed. Mechanisms based either on solute drag(3.14,3.15,3.16) or particle pinning have been proposed (3.17,3.18,3.19). The view mostcommonly upheld today seems to be that strain-induced microalloy carbonitrides formedduring the hot deformation suppress recrystallization during the interpass time. For Nb-steelsconvincing evidence has been presented that the pinning force created by the observed Nb-carbonitride dispersion at the recrystallization stop temperature exceeds that of the drivingforce for recrystallization (3.20,3.21).

Still this leaves us without a full explanation as to why the microalloy elements act sodifferently, Fig. 3.10. A partial explanation could be gleaned from solubility diagrams, Fig.2.4 and 2.7, from which we can see that most Nb can be precipitated in the temperature range1000-800oC (cf Fig 3.10) for a 0.03%Nb steel whereas only a minor part of V can beprecipitated between 900-800oC (cf Fig. 3.10) for a 0.10%V steel. However, it appears thatthis alone cannot account for the large difference between the Nb- and V- curves in Fig. 3.10.

Fig. 3.10 The increase in recrystallization-stop temperature with increase in the level ofmicroalloy solutes in 0,07C-0,25Si-1,40Mnsteel (3.13).

Fig. 3.11 Dependence of ferrite grain sizeon the total area of austenite grainboundary per unit volume. Data points arefor Ti-V and V microalloyed steels. Curvesrefer to Nb-microalloyed steels (3.23,3.24).

24

A detailed interpretation of the figure is more difficult than one might first think, because ofthe complex experiments behind the results. In particular, the recrystallization stoptemperature has been determined by carrying out a multipass deformation in a temperaturerange of about 50oC above the recrystallization stop temperature. This leaves us, referringagain to the solubility diagrams in Fig. 2.4 and 2.7, with a situation where, considering thefull temperature range for deformation, very little strain–induced V-carbonitride precipitationcan take place. The Nb-steels, on the other hand, are open for Nb-carbonitride precipitationduring the full deformation cycle.

The conclusion of this discussion is that the large differences between the microalloyelements in suppressing recrystallization could possibly be explained by the particle pinningmodel and the varying precipitation characteristics of the different elements. However, a fullquantitative explanation of the differences in impact on recrystallization is still lacking.

3.4 Effects on the austenite-ferrite transformationThe primary parameters controlling the ferrite grain size produced in the austenite-ferritetransformation are the effective austenite grain boundary area, i.e. austenite grain boundaryarea/unit volume (Sv), and the cooling rate.

However, for Nb-microalloyed steels it has been shown that for a given Sv the ferrite grainsize is smaller when transformed from unrecrystallized, deformed and flattened austenitegrains than from recrystallized, equiaxed grains (3.22). Traditionally this difference has beenaccounted for by the additional ferrite nucleation that takes place in the deformation bands ofheavily deformed austenite. Interestingly though when the same type of experiments areconducted for V- and Ti-V-microalloyed steels no difference is found (3.23,3.24). Deformed,unrecrystallized austenite grains transform to ferrite with virtually the same grain size as thatproduced from equiaxed, recrystallized austenite, Fig. 3.11. Thus, for the V-steels the ferritegrain size is found to be independent of austenite grain shape and processing method.

Fig. 3.12 The effect of N on the refinementof ferrite during the austenite-ferritetransformation in Ti-V-(Nb)-N steels (3.25).

Fig. 3.13 The effect of V and high N inrefining the polygonal ferrite grain sizeproduced during the austenite-ferritetransformation (3.5,3.26).

25

As can be seen from the figure the V-steels occupy an intermediate position between the linesfor recrystallized and unrecrystallized Nb-steels. An essential result is that the same degree offerrite grain refinement can be achieved for the V-steels as for the Nb-steels, viz. about 4 µm,provided the effective austenite grain boundary area is large enough.

An important and interesting feature of Ti-V and V-steels is that additional N above a level of~ 0.003% further refines the ferrite during the austenite-ferrite transformation (3.5,3.25,3.26),as shown in Fig. 3.12 for recrystallization controlled processing and in Fig. 3.13 fornormalizing. This behaviour has been explained partly as a result of interphase precipitationof V-carbonitrides which slows down the austenite-ferrite transformation and thereby givesmore time for ferrite nucleation, and partly as a result of more profuse ferrite nucleation along‘scalloped’ austenite grain boundaries (3.5,3.26).

3.5 Summary of chapter

• V-microalloyed steels are frequently co-alloyed with Ti to inhibit excessive austenitegrain growth both during hot working and welding, thereby ensuring a refinedmicrostructure in the finished product. This is attained by small controlled contents of Tiand N resulting in a fine effective dispersion of TiN after casting, especially for processeswhere cooling is fast such as continuous slab casting and thin slab casting. It isdemonstrated that the TiN-particles are extremely resistant to coarsening due to their highstability and therefore low solubility. Practically no coarsening occurs below 1300°C,even for the finest dispersions. V and Nb co-precipitate on TiN formed at highertemperatures. This does not, however, reduce the stability of the original TiN dispersionon subsequent reheating.

• It is shown that in heavily hot worked Ti-microalloyed steels abnormal grain growth ofaustenite can still take place on prolonged heating at temperatures below 1300°C despitethe stability of the TiN-dispersion. The driving force for this growth is primarily a fineaustenite microstructure, e.g. formed by heavy hot working. Due to the mechanism ofabnormal grain growth the short heating periods during welding are not sufficient toproduce grain coarsening and under these conditions the TiN-dispersion is capable ofeffectively preventing excessive grain growth.

• It appears that the widely different abilities of V, Nb, and Ti to retard recrystallizationduring hot working is associated with the varying solubilities of the carbonitrides of theelements and the resulting temperature ranges where effective microalloy precipitationcan occur to affect recrystallization.

• Experiments have provided conclusive evidence that the Recrystallization-Controlled-Rolling route (RCR), involving repeated recrystallization during working, can produceferrite grain sizes in Ti-V-steels with virtually the same size as those produced fromheavily deformed, unrecrystallized Nb-microalloyed austenite.

• Vanadium especially in combination with N refines the ferrite during the austenite/ferritetransformation.

26

4. PRECIPITATION IN FERRITE

As regards strengthening, the effective V-carbonitrides are those formed in ferrite during thelatest passage through the austenite-ferrite transformation. At equilibrium the circumstancesare such that a certain, small portion of the vanadium in microalloyed steel should precipitatein austenite, especially if the contents of vanadium and nitrogen are high. However, thekinetics of V(C,N) formation in austenite are sluggish and for processing at finishingtemperatures higher than 1000oC and for normal steel compositions virtually all vanadiumwill be available for precipitation in ferrite. Some solute vanadium can be forfeited duringcontrolled rolling as a result of strain-induced V(C,N)-precipitation, or by the formation ofalloyed (Ti,V)N particles during casting and reheating of Ti-V-microalloyed steels, asdiscussed in chapter 3.

However, compared to the other microalloying elements vanadium has a higher solubility andtherefore remains in solution to a much larger extent during processing in the austenite range.Accordingly, in steels where especially efficient precipitation strengthening is wantedvanadium is the preferred choice.

Precipitation of V-carbonitrides can occur randomly in ferrite in the wake of the migratingaustenite-ferrite (γ-α) boundary – general precipitation – or by interphase precipitationcharacterised by the development of sheets of particles parallel to the γ/α-interface formedrepeatedly with regular spacing. Many investigations have shown that for compositionstypical of structural steels the general precipitation takes place at lower temperatures,typically below 700oC, and the interphase precipitation at higher temperatures.

It is now a well established fact, demonstrated in several electron microscopy studies, thatV(C,N)-particles are also formed in the pearlitic ferrite (4.1). Because of the lowertransformation temperatures of pearlite this type of precipitation is usually finer. Bothinterphase and general precipitation have been observed.

V-carbonitrides are sometimes observed in a fibrous morphology when cooling is slow or onholding at high temperature in the γ/α range. The typical feature of this mode of precipitationis that the fibres are perpendicular to the γ/α interface. Furthermore it occurs sparsely and isnever a dominant microstructure. The view of the present authors is that this mode ofprecipitation is a distorted form of eutectoid γ → α + V(C,N) transformation. By realising thatthis type of decomposition is driven by a V-gradient ahead of the α/V(C,N) interface andparallel to it given by the γ/V(C,N)- and γ/α-equilibria and producing the necessary lateralredistribution of V from γ into α+V(C,N), it may be shown by analysing isothermal sectionsof the Fe-V-C system that this kind of eutectoid can only form from certain austenitecompositions corresponding to relatively low supersaturation. It places furthermore,restrictions on the appearance of the phase diagram. In particular the (γ+α)/γ equilibriumsurface must slope with increasing V-content to give room for this type of eutectoidtransformation (4.2).

In view of the sparse occurrence of the fibrous mode of V(C,N) precipitation in the steelsconsidered here we will not deal with it further in the present paper.

27

4.1 Interphase precipitationFigure 4.1 shows the typical morphology of interphase precipitation of V(C,N) inmicroalloyed 0.10%C-0.13%V steels. Already from its appearance one can conclude that sucha microstructure is formed in sheets parallel to the γ/α-interface by repeated nucleation ofparticles as the transformation front moves through the austenite. For this type of steelcomposition it is normally observed in the temperature range 800-700oC. Despite its clearrelation to the migrating γ/α-interface there has been discussions in the literature whether theinterphase precipitation nucleates in the boundary, ahead of it in the austenite or behind it inthe ferrite. However, firm experimental evidence now exists showing that it actually nucleatesin the interface (4.1). By electron microscopy of steels with high alloy contents that stabilisethe austenite to room temperature it has been possible to directly observe V(C,N)-particles inthe γ/α-boundary. Also from a very general viewpoint it is to be expected that, at hightemperatures where the chemical driving force for precipitation is low, nature chooses thesites where nucleation is energetically most favoured, viz. the interface. At lowertemperatures where the driving force is large we should expect general nucleation in theferrite matrix to occur.

Fig. 4.1 Electron micrographs illustrating the effect of N in 0,10C-0,13V steels on the spacingof the interphase rows and the density of V(C,N) precipitates after isothermal transformationat 750°C for 500 s, (a) 0,0051%N (b) 0,0082%N (c) 0,0257%N (d) 0,0095%N, 0,04%C, (4.4).

Sometimes it has been argued that the carbon concentration ahead of the moving interfaceduring γ/α-transformation should favour formation of microalloy-carbides in this location.However, under normal circumstances the carbon diffusion driven γ/α-transformation takesplace under local equilibrium (with respect both to C and V) in the interface, pseudo-para-equilibrium according to Hillert (4.3). The consequence of this is that the chemical drivingforce for V (C,N)-precipitation is the same both in the ferrite and in the austenite close to theinterface. Hence, this situation and the narrow region of the high carbon austenite togetherwith the rapidly moving γ/α-boundary will not make austenite the preferred site of nucleation.Instead, heterogeneous nucleation at the interface will be the energetically favoured site.

28

At high transformation temperatures, ~800oC, for typical compositions of V-microalloyedstructural steels the interphase precipitation consists of irregularly spaced, and often curvedsheets of V(C,N)-particles. With decreasing temperatures the incidence of curved rows ofprecipitates diminishes and the dominant mode is regularly spaced, planar sheets of particles,Fig. 4.1. From about 700oC the interphase precipitation is commonly found to be incomplete,and random precipitation from supersaturated ferrite after the γ/α-transformation takes overprogressively with decreasing temperature.

A characteristic feature of interphase precipitation is that it becomes more refined at lowertemperatures, as confirmed by many investigations (3.2,4.4). This is demonstrated for theintersheet spacing in 0.10%C-0.12%V steels in Fig.4.2. The interparticle spacing within thesheets of precipitation and the precipitate size decreases also with temperature, as expected.We may also notice that the particle spacing is notably shorter within a sheet than betweensheets. Figs. 4.1 and 4.2 show that the intersheet spacing is affected considerably by thenitrogen content of the steel. As is evident it is diminished to almost one third at 750oC onincreasing the nitrogen content from 0.005 to 0.026%.

Fig. 4.2 The effect of transformation temperature (a) and N-content (b) on intersheet spacingof V(C,N) interphase precipitation, B5 0,10C-0,12V-0,0056N, A5 0,10C-0,12V-0,0051N, A140,10C-0,12V-0,014N, B25 0,10C-0,06V-0,025N, C9 0,04C-0,12V-0,0095N, A25 0,10C-0,12V-0,026N, (4.4).

V(C,N), which has a f.c.c. structure, forms in ferrite as semi-coherent discs with theorientation relationship, (001) α // (001)V(C,N) , (110) α//(100) V(C,N) , first established by Bakerand Nutting (4.5). The discs are parallel to (110) α. Many electron microscopy studies haveshown that V(C,N) in interphase precipitation exhibits a single variant out of the threepossibles in the B-N orientation relationship(4.6). This crystallographic selection is given twoalternative explanations (4.7,4.8,4.9). In cases where the ferrite and the austenite into which itgrows are related by the Kurdjumov-Sachs relationship, (111)γ// (110) α , the variant will be

29

chosen that makes the close packed planes of all three phases parallel. This choice willminimise the free energy for nucleus formation. In cases without a specific crystallographicγ/α-relationship, as for incoherent interfaces, it is suggested that the variant is chosen that putsthe disc plane as close as possible to the γ/α-boundary, thereby again minimising the freeenergy. From this we realise that the selection of one variant of the B-N orientationrelationship is another confirmation of the nucleation of the interphase particles in the γ/α-boundary. The observed selectivity is not compatible with nucleation in the austenite or theferrite.

An essential characteristic of the V(C,N) precipitation in these types of steels is theconsiderable variation of the different modes of precipitation within the same sample, andindeed within the same grain. This has been observed in many investigations, but has beenemphasised in particular by Smith and Dunne (4.6). Not only is there a variation of thedifferent modes of interphase precipitation but, as these authors have found, generalprecipitation occurs both at low and high temperatures and is commonly formed jointly withinterphase precipitation in the same grain. They even found general precipitation, confirmedby the occurrence of all three Baker-Nutting crystallographic variants of V(C,N), at atransformation temperature as high as 820°C. The explanation they offered for this apparentanomaly is that the first-formed ferrite may grow too rapidly for interphase precipitation tooccur, leaving the ferrite supersaturated for general precipitation subsequently (4.6).

4.2 The mechanism of interphase precipitationThe mechanism of interphase precipitation has been the subject of considerable discussions.The models that have been proposed to explain the phenomenon fall broadly into twocategories: ledge mechanisms and models based on solute diffusion control. Honeycombe andcoworkers were among the first to study interphase precipitation more profoundly (4.10,4.11).They suggested that interphase particles form heterogeneously on γ/α-boundaries therebypinning their migration normal to the boundary. Local breakaway leads to formation ofmobile ledges. The ledges move sideways while the remaining part of the released boundaryis stationary and enables repeated particle nucleation to occur, forming a new sheet. Hence, inthis mechanism the intersheet spacing will be determined by the ledge height.

From this follows one of the main drawbacks with the ledge mechanism, viz. to produce acredible explanation of the observed variation of the intersheet spacing with temperature, andsteel composition, especially N, V and C. It is hard to see how these parameters shouldgenerate a corresponding variation of the ledge height.

Among the models based on diffusion control the solute-depletion model due to Roberts(4.12) is the most prominent and promising one, as it appears to the present authors. Two ofus have recently developed the model further, and especially have been able to put it in ananalytical form giving it predictive capacity (4.13). We will therefore dwell briefly on thisand discuss some of its predictions in relation to experimental observations.

The quantitative description of the solute-depletion model due to Lagneborg and Zajac (4.13)considers a ferrite grain growing into austenite where the growth is controlled by carbondiffusion in austenite while maintaining local equilibrium at the interface. At a given point ofthis growth the interplay is analysed between the nucleation of V(C,N)-particles in the γ/α-interface, the accompanying growth of V-depleted zones around the precipitates, and thecontinued migration of the γ/α-boundary away from the precipitate sheet. The growth rate of aV-depleted zone is infinitely large directly after nucleation but declines gradually with time

30

according to a parabolic growth law of type, particle radius ∝ (time)½ . The growth of theferrite obeys a similar parabolic time-law, but its rate can be considered constant for the shortdistance corresponding to the intersheet spacing relative to the size of the ferrite grain. Theimplication of this is that the moving γ/α-interface will after nucleation initially be in a V-depleted zone, but will eventually catch up the growing depleted zone. The boundary has nowreturned into material with the original V-content, and nucleation of a sheet of particles willrepeat itself. Hence, this gives the condition for calculating the intersheet spacing.

Carrying out this quantitative analysis gives the following expression for the intersheetspacing

� is = 2So . KK

1

2 (4.1)

where So is the distance grown by the ferrite at the point analysed, K2 and K1 are theproportionality factors in the parabolic growth laws for the V-depleted zone around theprecipitate, and for ferrite, respectively, Eqs. (4.2) and ( 4,3) below

tCC

CCDr

VCNV

VCNV

VCNVv

v ⋅−

−⋅⋅=∞

/

/42 102 α

ααα (4.2)

where r is the radius of the depleted sphere at a point with 99% of the original V-content( CV

∞α ), Dvα the diffusion coefficient of V in α, CV

∞α the original V-content in α, C α/VCN the V-

content in α at the α/VCN-interface, and CVCNV the V-content of the V(C,N)-particle.

tCCCC

CCDS

cccc

ccc ⋅

−−=

∞

∞

))(((

/

2_/2 )

αγααγ

γαγγ (4.3)

where S is the distance from the point of α-nucleation to the location of the γ/α-interphase attime t, Cc

αγ / the C-content of γ at the γ/α-interphase, Cc∞γ the original C-content of γ, and Cc

α

the C-content of α in equilibrium with γ.

Fig.4.3 shows the first predictions made with the model. The agreement with observations oftemperature dependence of the intersheet spacing is almost perfect. Also, agreement of theabsolute values is very satisfactory. The noticeably smaller spacings observed for the high-Nsteel as compared to the low-N steel cannot explicitly be accounted for by the expressionsabove. However, we do know, also quantitatively, that the chemical driving force for VN-precipitation is much larger than for VC. This means that profuse nucleation occurs for lowerV-concentrations, and thus we should expect that going from C-rich V(C,N) to N-rich willlead to smaller intersheet spacings.

The computations shown in Fig.4.3 also predict that the intersheet spacing around 700oC willfall below the approximate size of observed precipitates. Physically, this implies that thegrowth rate of the γ/α-interphase exceeds that of the V-depleted zone. Hence, the γ/α-boundary escapes, leaving the ferrite in its wake supersaturated with respect to V(C,N). Inexcellent agreement with experiments it is predicted that general V(C,N)-precipitation occursbelow about 700oC for steel compositions referred to in (4.4).

31

Fig. 4.3 Experimentally measured and computed intersheet spacings in interphaseprecipitation of a 0,10%C-0,13%V steel as a function of the transformation temperature.

Above 800oC the (V(C,N)-precipitation becomes increasingly irregular and sparse forcompositions typical of V-microalloyed structural steels, 0.10%C-0.10%V (4.4,4.6). Again,this can be understood from the model. It predicts an accelerated increase of the intersheetspacing in this temperature range; at 850oC the predicted value for the steel examined inFig.4.3 exceeds 500 nm. At the same time the spacing within the sheet will also increase.Under those circumstances it will be very difficult to perceive the observed microstructure asintersheet precipitation. To this should be added that around 850oC the solubility limit ofV(C,N) is approached for the compositions considered.

The computations of Figs.4.3 are made with the assumption that the ferrite grain has grown toa size of 5µm (So=5µm). For a final grain size of 10µm this means that the γ/α-transformationhas proceeded to 50%. Hence, the model predicts that the spacing is directly proportional tothe ferrite growth, or to the degree of transformation. In fact, the model predicts that in theearly stages of transformation the ferrite growth will be too rapid for V(C,N)-nucleation totake place and it is only later, when the growth rate has declined sufficiently, that thecondition of interphase nucleation will be fulfilled. In the first part of the γ/α-transformationthe ferrite will therefore be supersaturated with respect to V(C,N) directly behind the movingγ/α-interface and hence general precipitation will occur. Smith and Dunne (4.6) made exactlythis statement based on their microscopical observations and without the assistance of amodel. Beside this, however, no attention seems to have been paid in the literature to thisessential feature of interphase precipitation. The general statement often made, that a largevariation of precipitation modes are observed, is probably largely a reflection of this specificcharacteristic of the phenomenon.

Any mechanism for interphase precipitation based on nucleation of particles in the interfacemust be able to account for the release of the boundary from the dense particle population inthe intersheet. For the chemical driving force of the γ/α-reaction at 750oC, estimates indicatethat particle spacings below about 50nm will permanently pin the boundary (4.14). As we cansee from the electron micrographs in Fig.4.1 the spacings are in that range. However, theyalso show that the spacings vary a great deal and that there are gaps allowing the boundary tobulge locally and then sweep out the rest of it from the particle row. At any rate this processwill temporarily hold the boundary and consequently slow down the average migration rate.We suggest that this accounts for the observed reduction of the γ/α-transformation rate in V-steels.

32

4.3 General precipitationFor compositions typical of V-microalloyed steels, 0.10%C-0.10%V, general precipitation ofV(C,N) occurs in a temperature range from about 700oC and below. As already discussed inthe previous section this transition from interphase to general precipitation can be predictedexcellently by the solute-depletion model (4.13). We have also already seen that generalprecipitation may also occur locally above that temperature and have discussed how it shouldbe understood.

Experimentally it is well established that VN has a considerably lower solubility than VC,both in ferrite and austenite. The thermodynamic background to this is the much largerchemical driving force for formation of VN. This larger chemical driving force makes N-richV(C,N) the preferred precipitation as long as there is sufficient nitrogen in the matrix, as hasbeen demonstrated both by thermodynamic analysis (4.15) and by experiment (4.15,4.16) inferrite as well as in austenite. It is only when the nitrogen content falls below about 0.005%that the initial V(C,N) starts to increase its C-content, Fig.4.4.

Fig. 4.4 Thermodynamic calculationsof the variation in the composition ofVCxNy with content of N remaining insolution during precipitation in ferrite(4.15).

Fig. 4.5 Growth of V(C,N)- precipitates aftertransformation at 650°C (a) as a function of the N-content of the steel, and (b) as a function ofholding time (4.4).

A technically very important finding is that the precipitate size diminishes considerably withincreasing nitrogen content in the steel, as is shown in Fig.4.5 (4.4). This is accompanied by aconcurrent increase of precipitate density, although that is more difficult to measurequantitatively, Fig.4.6. These effects must be accounted for by an increased nucleationfrequency as a result of the larger chemical driving force for precipitation of nitrogen-richV(C,N). For the temperature and holding times in the experiments of Fig. 4.5 supersaturationwith respect to V(C,N) still remains. Therefore the observed difference in precipitate growthbetween high and low-N steels cannot be explained by differences in coalescence. Instead, itmust be interpreted as resulting from the denser particle nucleation in the high-N material,thereby producing an earlier soft impingement of V-denuded zones and so slowing down the

33

precipitate growth. This reasoning is substantiated by the computed curve for precipitategrowth inserted in Fig. 4.7, where no allowance is made for impingement of solute depletedzones. Evidently, the computation can perfectly account for particle growth in the low-N steelwith less dense precipitation, whereas in the high-N steel where we do expect effects ofimpingement, the growth is less than half of this. The size of the depleted zone correspondingto half the original solute content is double the size of a particle growing withoutimpingement, cf. Eq. (4.2). According to the calculations in Fig.4.7 this gives a depleted zonewith a diameter of 23 nm after an ageing time of 500 s at 650°C. Comparing this with theprecipitate densities in the micrographs of Fig.4.6 shows clearly that impingement of denudedzones has taken place long before 500 s ageing in the high N-steel, whereas it has hardlystarted in the low N steel.

Fig. 4.6 Electron micrographs illustrating the effect of N on the density of V(C,N)-precipitatesduring isothermal transformation at 650°C for 500s.

Fig. 4.7 Computation of V-diffusion controlled growth of V(C,N) at 650°C; no allowance ismade for impingement between adjacent particles. Comparison with experimental values attwo levels of N.

In numerous investigations of V-microalloyed steels it has been shown that the observedprecipitation strengthening originates seemingly only from N-rich V(C,N), despite the factthat there is abundant V to combine with the carbon dissolved in ferrite. The normalexplanation of this behaviour has been that when the first formed N-rich precipitates haveconsumed practically all nitrogen the chemical driving force for formation of C-rich V(C,N)is too small for profuse precipitation to occur and hence no added strength is observed.However, this issue is complex, and recently it has been shown that under certain conditions

34

the C-content in the steel can add significantly to precipitation strengthening. This questionwill be treated in the following section.

4.4 Effects of precipitation on strengthSince normal contents of V in structural steels dissolve in austenite at relatively lowtemperatures, unlike the other microalloying elements, and therefore can contribute fully toprecipitation strengthening when the steel is cooled into the α-range, V is usually thepreferred element when precipitation strengthening is wanted. A modest addition of 0.10%Vcan bring about a strength increase beyond 250 MPa, and in special cases even up to 300MPa. This was demonstrated in an early work by Roberts et al. in 1980 (4.17), Figs.4.8 and4.9. These graphs show another essential feature of V-steels, viz. that nitrogen in smallcontents adds significantly to precipitation strengthening. Similarly, enhanced coolingthrough the γ/α-transformation and afterwards augments the particle hardening, Fig.4.9.Theresults of Fig.4.8 show also that normalizing at 950oC does not allow all V to go into solutioncompletely and hence produces only modest strengthening.

Fig. 4.8 Effect of processing method(laboratory simulation) on the precipitationstrengthening derived from V(C,N). Basecomposition of steel is: 0,12%C-0,35%Si-1,35%Mn-0,095%V-0,02%Al (4.17).

Fig. 4.9 Influence of cooling rate on thestrength contribution from V(C,N)precipitates. Base composition as for Fig.4.8 (4.17).

In a number of investigations at the Swedish Institute for Metals Research(3.2,4.4,4.18,4.19,4.20) the strengthening effects of V have been carefully examined inisothermally heat treated steels of varying V, N and C-contents. Fig.4.10 shows the influenceof V, N and ageing temperature on precipitation strengthening.The technically very important effect of N on the strengthening of V-steels has sometimesbeen interpreted such that only VN forms as precipitates, and that the remaining V does notcombine with C, possibly due to the larger solubility and lower chemical driving force forVC. However, this is a rather loose statement and is not an entirely correct description of thephenomenon. A physically more correct account would be as follows. In the present materialwith hard non-penetrable V(C,N)-precipitates the particle strengthening occurs by the Orowan

35

mechanism – bowing of dislocations between particles – and in that case the decisiveparameter is the interparticle spacing in the slip plane. In turn, that is determined by thedensity of the precipitates. This density is controlled by the nucleation frequency and thenumber of nuclei it creates until the supersaturation has diminished so that nucleation dies.The essential parameter governing the variation in nucleation is the chemical driving force forV(C,N)-precipitation, and there we know that it varies relatively strongly with the N-content,as shown in Fig.2.10 (4.21). To put it simply this is a consequence of the larger stability, orlower free energy, of VN than of VC. Hence, the larger driving force will cause a denserprecipitation in the high-N steels than in the low-N steels. Electron microscopy has clearlyconfirmed this, as shown in Fig.4.6 (4.20). When nearly all N has been consumed inprecipitation of N-rich V(C,N) the process can follow one of the following lines:

(i) in the case of continuous cooling, the temperature may have dropped sufficiently for V-diffusion to essentially cease, and therefore continued precipitation stops (in line withthe traditional explanation stated above).

(ii) the decomposition continues by co-precipitation of C-rich V(C,N) on previously formedN-rich V(C,N).

(iii) if the carbon content in solution is sufficiently large (cf. below) the driving force maybe large enough for continued nucleation, but now of C-rich V(C,N).

Obviously it is only case (iii) that will contribute significantly to further precipitationstrengthening, whereas cases (i) and (ii) will not.

Fig. 4.10 Effect of N, V and transformationtemperature on the precipitationstrengthening in 0,1%C-V-N steels afterisothermal ageing at different temperaturesfor 500 s (3.2).

Fig. 4.12 Solvus lines for C in ferritecalculated using Thermocalc for equilibriumwith cementite and austenite, respectively(4.19).

Recent investigations (4.18,4.19,4.20,4.21) have demonstrated very clearly that theprecipitation strengthening of V-steels increases significantly with the C-content of the steels.Plots of the observed strengthening ∆Rp against the C-content, and for comparison also N-

36

content, are given in Figs.4.11(a) and (b) (4.20). These results show that carbon raises ∆Rp by~ 5.5MPa for every 0.01%C in the steel, whereas nitrogen raises it by ~ 6 MPa for every0.001%N. This important effect of carbon has not previously been noticed. The reason ispresumably that there has always been an uncertainty in the procedure of evaluating theprecipitation strengthening as to how the contribution of pearlite should be deduced anddeducted from the measured yield strength. This procedure was improved in the present case,and furthermore nano-hardness measurements were performed in order to unequivocallydemonstrate the precipitation hardening in the ferrite (4.18,4.19,4.21).

Fig. 4.11 Deduced values of precipitation strengthening for isothermally transformed V-steels(650°C/500s). (a) as a function of N-content (b) as a function of C-content (4.20).

That carbon dissolved in ferrite should contribute to precipitation strengthening is, of course,no surprise. As a matter of fact all the traditional graphs showing precipitation strengtheningvs. N-content exhibit a considerable residual strength when the curves are extrapolated to a N-content of 0, cf. Figs.4.8 and 4.10.The strong effect of the total carbon content, however, is atfirst sight puzzling, as the C-content of ferrite is controlled by either of the phase equilibria,γ/α or α/cementite. The key to the understanding of the influence of the total C-content is asfollows.

(i) The considerable difference in solubility of C in ferrite between the γ/α and α/cementiteequilibria below A1, shown in Fig.4.12. The solubility at 600oC is 5 times larger in γ/αthan α/cementite. The austenite phase also acts as a reservoir of carbon, which can flowinto the ferrite and maintain its high saturation even if precipitation of V(C,N) isoccurring.

(ii) The fact that the total C-content affects the kinetics of the transformation of γ to α andeventually to α + cementite. The essential point here is that the C-content displaces thepearlite formation to longer times, Fig.4.13 (4.21). The implication of this is that alarger chemical driving force for V(C,N)-precipitation will remain longer and hencepromote profuse nucleation. It should also be recognised that the shift to lower C-content in α and therefore lower driving force as a result of the cementite formationoccurs very quickly, less than a second for typical microstructures, because of the fastdiffusion of carbon in α.

37

Hence, this chain of interrelations leads with increasing total carbon content to graduallylonger dwell times during which nucleation occurs under the high chemical driving force.Consequently, we do expect denser V(C,N)-precipitation as the C-content is raised. This isvery clearly demonstrated by the electron micrographs of Figs.4.14a, b and c (4.20).

Fig. 4.14 Electron micrographs showing V(C,N) precipitation after isothermaltransformation at 650°C in 0,12%V-0,013%N steels with different C-contents (4.20). (a)0.04% C, (b) 0.10% C and (c) 0.22% C.

As a matter of fact this line of reasoning can be carried steps further and be shown to accountfor the observed precipitation strengthening in quantitative terms. According to classicalnucleation theory the decisive factor controlling the nucleation frequency is

)3

16(exp2

23.

G

V

m

m

kTN

∆⋅−∝ σ (4.4)

where N� is the rate of nucleation, σ the α/V(C,N) surface energy, Vm the molar volume ofV(C,N), ∆Gm the molar free energy for formation of V(C,N) from α of a given composition,and k and T have their usual meanings.

Integration of this expression over the precipitation period with the variable driving force asthe supersaturation declines gives the density of precipitates (numbers of particles per unitvolume). We will simplify and assume that the initial nucleation rate according to Eq (4.4)will be proportional to the final particle density. Let us consider the observed precipitationstrengthening of the 0.10C-0.12V steel in Fig.4.10 produced by isothermal ageing at 650°Cfor 500 s and compare with computed precipitate densities by using the chemical drivingforces of Fig.2.10. In view of the observed profuse nucleation in the high-N steels, Fig.4.6,and the consequent early completion of precipitation – soft impingement has occurred before100s at 650oC, cf. Fig.4.7, i.e. before pearlite formation sets in − it seems likely that theparticle density in these steels will be controlled by the driving force for high-C-content inferrite corresponding to γ/α-equilibrium. Similarly, the relatively sparse particle density in thelow N steels suggests that most nucleation takes place after pearlite has formed, and thereforewe should in this case use the driving force for the low C-content in ferrite according to theα/cementite equilibrium. By adopting a value for the V(C,N)-α interphase energy of0.325J/m2 , close to the values suggested values for Nb(C,N) (4.22), we can compute relativevalues of the planar interparticle spacing for the two extreme N-contents for the 0.10C-0.12V

38

steel and from that a strength increase of 70%. This computed value is virtually a perfectmatch with the observed strengthening of 71% (Fig.4.10, 0,10C-0,12V, ageing at 650°C).

Fig. 4.13 Pearlite start time duringisothermal transformation at 650°C for V-Nsteels with C-contents varying between 0,10and 0,22% (4.21).

Fig. 4.15 Deduced values of precipitationstrengthening in steels isothermallytransformed at 650°C (4.19).

A quantitative comparison between the precipitation strengthening of 0.10%C and 0.22%Csteels in Fig.4.11 is considerably more complicated since the large difference in duration ofhigh C-content in ferrite before pearlite formation must then be accounted for, cf. Fig.4.13.But clearly in qualitative terms, the longer time for profuse nucleation in the 0.22%C steelwill give rise to a much larger strengthening.

It is known that the precipitation strengthening of Ti-V steels is notably smaller than for V-steels with the same V-content. This is evident from Fig.4.15 (4.19). It is readily shown -along the same lines as above - that the reduction of effective N and V due to formation of(Ti,V)N in the austenite cannot alone account for the substantial difference in strengtheningbetween the 0.22C-0.12V-Ti and 0.22C-0.12V steels in Fig.4.15. It is possible that theremaining part can be explained by the faster γ/α-transformation and hence earlier pearliteformation in the Ti-V steel as compared to the V steel (4.19).

4.5 Summary of chapter

• Precipitation strengthening in microalloyed steels benefits from the γ/α-transformationbecause the chemical driving force for precipitation of microalloy carbonitrides increasessuddenly and strongly as γ is transformed to α - also reflected in the widely differentsolubilities in the two phases – and thereby gives rise to profuse precipitation in theferrite. V-microalloyed steels are particularly suited for generating large precipitationstrengthening because of their ability to dissolve large quantities of V-carbonitrides atrelatively modest temperatures in the γ-range due to the larger solubility as compared toother microalloy carbonitrides. This again creates conditions for a high driving force andtherefore dense precipitation.

• At low supersaturations for V-carbonitride precipitation, corresponding to temperaturesbetween A3 and A1 and slow cooling, nature takes advantage of easier nucleation in

39

interfaces and carbonitrides are formed repeatedly in the moving γ/α-boundary , so calledinterphase precipitation. On lowering the temperature the interface will gain speedrelative to the rate of precipitation, implying that the boundary escapes the precipitatesleaving supersaturated ferrite in its wake for subsequent general precipitation to occurbehind the migrating γ/α-boundary. For compositions typical of microalloyed structuralsteels general precipitation takes place below about 700°C and interphase precipitation athigher temperatures. However, there is a considerable variation of the different modes ofprecipitation; so is general precipitation commonly also found at high temperatures jointlywith interphase precipitation.

• A model for interphase precipitation with predictive capacity is presented. It considers thegrowth of the V-depleted zone around a V(C,N)- particle and the simultaneous motion ofthe γ/α-boundary in which it was nucleated. The growth rate of the depleted zone isinitially high but declines, the γ/α-interface catches up and enters material with theoriginal V-content, and nucleation of V(C,N) will repeat itself. The model exhibitsexcellent agreement with observed values of intersheet spacing, its temperaturedependence and the transition from interphase to general precipitation.

• Experiments show clearly that the V(C,N)-precipitation becomes denser and the particlesfiner with increasing N-content. This originates from the fact that the chemical drivingforce for this reaction increases as more N is dissolved in ferrite and therefore increasesthe nucleation rate. It is unambiguously shown that the lower growth rate of V(C,N) inhigh N-steels, resulting in a smaller particle size, is caused by an earlier depletion of V inthe surrounding ferrite due to the larger density of precipitates.

• V(C,N)-precipitation generates considerable strengthening at moderate additions of V; aV-content of 0,10% engenders in steels with 0,010%N typical of EAF steels precipitationstrengthening up to 250MPa. Enhanced cooling through the γ/α-transformation and in theferrite range causes significant further increases in strength, particularly at high N-contents.

• In steels with hard and for dislocations non-penetrable particles such as V(C,N) the onemicrostructural parameter determining the precipitation strengthening is the interparticlespacing. Therefore to obtain the largest possible strengthening the nucleation rate ofV(C,N), which determines the density of the precipitate population, should be maximised.The strong effect of N in precipitation strengthening of V-microalloyed steels is due to itsability to enhance V(C,N)-nucleation.

• It has been shown that the first V(C,N) to be formed is N-rich. It is only until almost all Nhas been consumed that the C-content of the carbonitride rises. To benefit from continuedprecipitation strengthening of C-rich V(C,N) beyond this point, this implies that theremust be sufficient C in solution and chemical driving force for fresh V(C,N)-nucleation tooccur. It is demonstrated that such a situation is at hand when the ferrite C-content iscontrolled by the γ/α- rather than the α/cementite-equilibrium; in the former case theferrite C-content is 5 times larger than in the latter at 600°C.

• Recent studies have demonstrated that the precipitation strengthening in V-microalloyedsteels increases significantly with the total C-content of the steel, an effect previously notrecognised, ∆Rp ~5,5MPa per 0,01% C as compared to ~6MPa per 0,001% N. This isexplained by the fact that the C-content of the steel delays the pearlite formation andthereby maintains the higher content of C dissolved in ferrite corresponding to the γ/α-equilibrium for a longer time, cf. previous paragraph.

40

5. PRINCIPLES AND PRACTICE OF THERMO-MECHANICAL CONTROLLEDPROCESSING, TMCP

5.1 General principles of recrystallization-controlled rolling, RCR, and controlledrolling, CR, processing

Good combination of strength, fracture toughness and weldability of V-microalloyed HSLAsteels in the as hot rolled condition can be obtained by careful choice of thermo-mechanicalcontrolled processes (TMCP) (5.1-5.13). The improvement in properties of plates, longproducts, strips and forgings is associated with different strengthening mechanisms of whichthe most important is grain refinement whereby both strength and toughness are improved atthe same time. In order to obtain optimum ferrite refinement, it is necessary to maximise thearea of austenite grain boundary per unit volume at the on-set of phase transformation (5.8,5.9). This may be achieved either by low temperature controlled rolling (CR) or byrecrystallization controlled rolling (RCR), when the finish rolling temperature is relativelyhigh (see Fig. 5.1).

Fig. 5.1 Thermo-Mechanical Controlled Processes. 1) Recrystallization controlled rollingand accelerated cooling 2), controlled rolling and accelerated cooling and 3)recrystallization controlled rolling or controlled rolling with air-cooling.

The aim of recrystallization controlled rolling (RCR) is to make use of grain refinement byrepeated recrystallization of austenite. The fine austenite structure is retained during inter-passdelays and during cooling down to Ar3 by a suitable dispersion of second phase particles thattogether with accelerated cooling (ACC) lead subsequently to a minimisation of the ferritegrain size after transformation. The concept of RCR is attractive in that it is a relativelyuncomplicated and high productivity process and can be applied on conventional mills. Thisprocedure is intrinsically more economical than controlled rolling at low finish rollingtemperatures (lower than the recrystallization stop temperature, RST) and also lends itselfwell for use in mills which are not sufficiently strong for low temperature controlled rollingpractice and for rolling of long products with high inherent finishing temperature.

41

In very many commercial HSLA steels processing is controlled in such a way as to ensurethat precipitation occurs during or after thermo-mechanical treatment. The precipitatedparticles play an important role in control of structure and final properties of microalloyedsteels. V, Ti and/or Nb are essential for successful TMCP processing, both for enhancingstrength through precipitation hardening and for providing stable particles to inhibit graingrowth in austenite.

Accelerated controlled cooling after hot rolling alters the microstructure of plate from ferrite-pearlite to fine-grained ferrite-bainite and consequently increases the strength without a lossin low temperature toughness. While precipitation strengthening of Nb(C,N) or V(C,N) is alsoenhanced by accelerated cooling, the deleterious effect of this on impact toughness iscounteracted by improved grain refinement.

Proper processing of microalloyed HSLA steel should aim to combine the maximum grainrefinement with an effective utilisation of the strengthening potential of the microadditions.V(C,N) precipitation in austenite is usually rather slow unless the finish rolling temperature isbelow 850oC (controlled rolling). Thus, on transformation from hot worked austenite, theamount of vanadium available for precipitation in ferrite, and thereby the potential hardeningcontribution is considerable.

The key to success with TMCP-processing is to define rolling schedules combining amaximum degree of microstructural refinement with acceptably low rolling loads, good shapecontrol and high productivity. In this context, computer models for calculation ofmicrostructure development and precipitate evolution during hot rolling (5.14-5.17) areinvaluable for design of rolling schedules.

5.2 Mechanical properties and microstructure of TMCP materialThe mechanical properties and the final microstructure of V-(Ti-Nb)-microalloyed steels aredependent on the following process parameters:

- Reheating temperature (Treh),- Rolling schedules: reduction (Red) and finish rolling temperature (FRT),- Cooling parameters: accelerated cooling rate (ACC) and finish accelerated cooling

temperature (FACT),- Steel chemistry.

The slab reheating temperature has a strong influence on the strength, toughness andmicrostructure of microalloyed steel in the as thermo-mechanical controlled processedcondition. A low slab reheating temperature gives finer austenite grains, refines the finalmicrostructure of the materials and as a consequence improves the low temperature toughnessfor steels processed in a similar manner. This is mainly attributed to the more profuse fineprecipitation remaining after low temperature reheating which more effectively resists graingrowth of austenite.

However, the yield stress and tensile strength decrease because lower reheating temperaturereduces the amount of dissolved vanadium (or niobium) in the austenite and accordingly thepotential for precipitation hardening after cooling. For Ti-V-N microalloyed steels it wasfound that a reduction in reheating temperature from 1250o to 1100oC reduced the yield stressby about 40MPa while at the same time decreasing the ductile-brittle transition temperatureby about 15oC (5.5).

42

Fig. 5.2 Effect of finish rolling temperature (FRT) on the strength, impact toughness andferrite grain size of Ti-V-(Nb)-N steel (5.9).

Rolling schedules, the finish-rolling temperature and deformation in the final pass areimportant TMCP-parameters, affecting strength and toughness. Effects of the finish rollingtemperature on the microstructure and mechanical properties for various microalloyed steelsare summarised in Fig. 5.2 using data from various sources (5.1, 5.4, 5.8, 5.9).

Fig. 5.2 shows that the very best combinations of low temperature toughness and tensilestrength are obtained in Ti-V-(Nb)-N steels for finish rolling temperatures close to Ar3 (i.e.conventional CR practice). However, an almost equally good combination of strength andtoughness is obtained for a FRT of 950oC by recrystallization controlled rolling. Theseobservations are in good agreement with the results reported by Chilton and Roberts (5.18)for V- N steels that FRT has little marked effect on the mechanical properties.

On-line accelerated cooling (ACC) processes after hot rolling, from recrystallized or non-recrystallized austenite are very important because the ACC rate and the finish ACCtemperature have a strong influence on the final microstructure and mechanical properties.The effect of cooling rate and finish cooling temperature on the final microstructure, yieldstrength and toughness of Ti-V-(Nb)-N-steels after RCR are illustrated in Fig. 5.3. It isapparent from the figure that the final microstructure and mechanical properties are greatlyinfluenced by cooling rate.

The yield stress of Ti-V-(Nb)-N-steels rises as the cooling rate increases, although thevariation of strength with cooling rate is smaller at higher rates than lower (<7oC/s). At veryhigh cooling rates (15oC/s) the austenite transforms to a ferrite / bainite microstructure. Thefinal ferrite grain size depends only weakly on the finish accelerated cooling temperature(FACT) in the range of 400-600oC although the second phase consists of bainite when FACTis below 500oC. Yield strength, on the other hand, depends on the finish accelerated coolingtemperature and increases with decreasing FACT down to 500oC or even lower. The highestyield stress was observed in the 0.09V microalloyed steel, (without Ti), RCR processedfollowed by ACC to room temperature as shown in Fig. 5.3. The appearance of bainite in thestructure of Ti-V-N and Ti-V-Nb- steels changes the dependence of yield strength on FACT.

43

Fig. 5.3 Effect of cooling rate from finish-rolling temperature of 1030oC to finish-coolingtemperature (FCT) on the ferrite grain size, yield stress and toughness of Ti-V-(Nb)-N

steels. Final reduction was 25%.

Impact transition temperatures of RCR+ACC material increase, as the cooling rate becomeshigher despite the fact that the ferrite grain size decreases. This is a result of the precipitationstrengthening of V(C,N) or (V,Nb) (C,N) and at low FACTs the increasing volume fraction ofbainite in the microstructure. Cooling rate is a principal factor influencing the size distributionof precipitates. A greater proportion of smaller particles and reduced particle spacing havebeen observed when the cooling rate was increased (5.5), so producing more efficientprecipitation strengthening.

The cooling rate subsequent to RCR (during γ�α transformation) should be, in principle,restricted to a maximum 10-12oC/s and the finish accelerated cooling temperature should notbe lower than 500oC if the formation of martensite or bainite is to be avoided. However, directquenching (DQ) and tempering (T) processes can also be utilised in the production of heavyplates of high strength steels.

44

It has been shown that a sufficiently high cooling rate is the most important factor of the DQprocess, which controls the transformation during quenching (5.12).Finish rolling temperature and cooling start temperature or delay time prior to quenching maybe quite important parameters with respect to control of the final austenite grain size andtoughness. However, the influence of these parameters on yield stress and hardness is weakerthan that of cooling rate (see Fig. 5.4).

The optimum combination of strength and toughness properties of steels microalloyed withV-Mo-B are obtained for the direct quench and tempering process with a finish rollingtemperature (FRT) at about the recrystallization stop temperature (~940oC) followed byquenching with the highest practical rate (~30oC/s) and then tempering at 680oC. Moreover,the direct quench and tempered steel has almost equally good mechanical properties (strengthand toughness) as those achievable by conventional off-line quenching and tempering(RQ+T).

Fig. 5.4 Effect of cooling rate from FRT or reheating temperature (920°C) to RT as wellas tempering at 680°C on the yield strength of V-Mo-B steel (FRT=920°C and 970°C,

respectively) with total reduction ~50% (5.12). RQ denotes off-line quenching.

5.3 Effect of nitrogen contentThe positive effect of higher nitrogen content in V-microalloyed steel on the yield strength,notch toughness and weldability was reported during the 1960s by Melloy (5.19). Siwecki etal. (5.1) have also shown that nitrogen in V-microalloyed steel is conducive to improved grainrefinement and raised yield stress (see Fig. 5.5). As shown in this figure, the cooling rate is animportant parameter affecting yield stress and microstructure. However, the processconditions of final rolling temperature do not strongly affect the yield stress. Heavy plate(40mm thick), thin plate (10mm) or 10mm bar of V-steel can be processed by controlledrolling or RCR with attractive strength properties only weakly dependent on FRT. The reasonfor this, as can be deduced from Fig.5.5(b), is that higher temperature RCR rolling provides a

45

higher precipitation strengthening effect, but that low temperature CR processing gives a finergrained structure although with less precipitation strengthening and these two effects almostcancel one another.

Fig. 5.5 Dependence of yield stress (a) and precipitation strengthening (b) on nitrogenlevel in 0.09V steel with 0.12C and 1.35Mn content for the various processing methods and

cooling rate (5.1). (Cooling rates during simulations correspond to natural air cooling of40mm thick plate (0.24oC/s), 15 mm plate (0.88oC/s) or 10 mm bar (5oC/s))

Nitrogen's effect in connection with recrystallization controlled rolling and acceleratedcooling is demonstrated in Fig. 5.6. Material with higher N content is significantly strongerthan that with low N processed in the same manner, although with some sacrifice oftoughness.

It may be seen that the strengthening potential of vanadium can be utilised effectively only athigher levels of nitrogen and that increased cooling rate has a profound effect especially atthese higher levels. For one thing, ferrite grain sizes of low nitrogen steel are significantlylarger than those of the higher N steel, in particular for slow cooling rates. In particular, withTi-microalloyed steels the effectiveness of nitrogen in the low temperature controlled rolledcondition is much smaller than in the RCR and ACC condition, (5.5, 5.8). For low N steel(0.01Ti-0.08V-0.003N) the yield stress after low temperature CR was found to be almostidentical with that following RCR+ACC processing since almost no precipitation of VN inaustenite could take place. However, for higher N steel (0.013%N) the yield stress was~50MPa less after the CR treatment, which is attributed to the loss of precipitation hardeningeffect due to premature formation of coarse VN particles in austenite.

Some results illustrating the effect of processing method by recrystallization hot rolling,controlled rolling and normalising (the processes were followed by air cooling at 0.9oC/s) onthe precipitation strengthening contribution to yield strength, ∆Rp, for V-N steels arepresented in Fig. 4.8 (5.1).

46

Reheated to 1200°C

LOW

ER Y

IELD

STR

ENG

TH (N

/mm

2 )

V–Nb–NV–NbV–N–Nb–Ti

V–Nb–TiV–N

660

640

620

600

580

560

540

520

50020 30 40 50 60 70 80 90

SLAB THICKNESS (mm)

Fig. 5.6. The effect of nitrogen, vanadium (niobium) and cooling rate on the yield stressand impact toughness of Ti-V-(Nb)-N-steels after recrystallization controlled rolling (5.8).

For RCR steels, ∆Rp increases linearly with nitrogen content. For normalised material,however, the precipitation hardening contribution to yield stress saturates at high nitrogenlevels because of the increasing amount of V(C, N) remaining undissolved at the normalisingtemperature. Fig. 5.5 also demonstrates that ∆Rp is lower after processing by CR withFRT=800oC than after RCR with FRT=950oC. This difference in ∆Rp can be reconciled withthe loss of precipitation hardening potential when deformation-induced precipitation of V(C,N) accompanies rolling at low finishing temperatures.

5.4 Application of TMCP practiceResults of full scale industrial plate, long product and strip processing using TMCP practicesobtained at high (RCR) and low (CR) finish rolling temperature practice with and withoutapplication of accelerated cooling are discussed below.

5.4.1 Heavy plate rollingChanges in the mechanical properties and the microstructure of Ti-V, and Ti-V-Nb-microalloyed steels were studied as a function of the finish rolling temperature, platethickness and cooling parameters. The influence of these parameters on the yield strength,toughness and ferrite grain size for various microalloyed steels are shown in Fig. 5.7 and 5.8.The 0.01Ti-0.08V steels described in the previous section have also been rolled to 15-25 mmplates on a commercial mill at SSAB Oxelösund. The steels were rolled using RCR practicewith and without application of ACC. These were also CR processed with a low finish rollingtemperature (800oC) such that the austenite was in a substantially deformed condition prior totransformation. Table 1 summarises the important parameters relevant to their processing.

Table 1. Process parameters for full-scale production.

ProcessTreh(oC)

FRT(oC)

No ofpasses

Time(min)

Thickness(mm)

Cooling rate(oC/s)

FCT(oC)

RCR 1250 1050 10 3 min. 25 0.4 -RCR+ACC 1250 1050 10 3 min. 25 7 600CR 1250 800 13 8 min. 20 0.5 -

47

Mechanical properties of the as hot rolled plates of 0.01Ti-0.08V-steel are summarised in Fig.5.7. Toughness is in all cases excellent (ITT40J <-80oC) but best for the CR treatment becauseof the very fine ferrite grain size and smaller precipitation hardening contribution. The finestgrain structure is obtained from the CR route and this should be capable of some furtherrefinement with application of accelerated cooling. However, despite the finer grain size ofthe CR plates these have slightly lower strength than the RCR+ACC ones as a result ofpremature precipitation of VN in austenite during the final low temperature CR-rollingpasses.

Fig. 5.7 The mechanical properties of the commercially processed 0.01Ti-0.08V-0.013Nsteel plates produced by different full scale processing routes (5.5).

Very similar results to these reported by Zajac et al (5.5) have been obtained by Lee et al.(5.7) on a 0.015Ti-0.08V-steel with a base composition: 0.09C - 0.3Si - 1.5Mn - 0.005N. Themechanical properties of 12 mm thick plate (manufactured at POHANG (5.7)), processed byRCR followed by ACC with the following parameters: Treh=1250°C, FRT=950°C,ACC=5°/s and FACT=550°C, were Re=~360 MPa, Rm=490 MPa and E-40°C=300 J.

RCR + ACC was applied by Rautaruukki Oy for heavy plate production. Tamminen (5.10)reported that good mechanical properties of Ti microalloyed steels were obtained. Moreover,flatness as well as residual stress levels of these plates was better than those produced by CRpractice.

Changes in the mechanical properties of heavy plates of Ti-V and Ti-V-Nb steels processedby low temperature controlled rolling have also been studied. The results in Fig. 5.8demonstrate that the yield strength of CR-processed steels increases with decreasing the FRTfrom 840oC to ~730oC. Strength and toughness depend not only on FRT but also on the platethickness and cooling rate. As shown in Fig. 5.8, the strength increases with decreasing platethickness, whereas the toughness decreases at the same time. Bodnar et al (5.20) have alsoreported an increase of strength with decreasing thickness of RCR rolled beams (44 mm -> 15mm) of V-steel. The best toughness was obtained for 30mm thick beam, whereas for bothlower and higher beam thicknesses deterioration in toughness were observed.

48

Fig. 5.8 Mechanical properties and ferrite grain size of the commercially CR-processed0.01Ti-0.04V (HS-350) and 0.01Ti-0.085V-0.04Nb (W-500) steels as a function of: (a)-

FRT and (b)-plate thickness (5.11).

Important advantages of the RCR process are its high productivity and lower mill loading ascompared to CR. The measured rolling loads for the two types of processing when rollingidentical plates are compared in Fig. 5.9.

Fig. 5.9 Comparison of observed rollingloads during RCR and CR rolling of Ti-V-N-steel (5.8).

Fig. 5.10 Effect of carbon content in the Ti-Vand V microalloyed steels on the strength vs.toughness of specimens processed in a similar

manner (5.9).

It is evident that the maximum loads are ~ 25% higher for the CR process due to the lowertemperatures and the accumulated deformation below the recrystallization stop temperature.The longer time necessary for CR is mainly due to the delay for cooling from 1100°C to920°C at the 40mm gauge but partly also due to the greater number of finishing passes. Forthese reasons, RCR processing offers intrinsically higher productivity than does CR for agiven plate product while at the same time making more efficient use of the microalloyingadditions for precipitation strengthening.

49

5.4.2 Long productsCarbon content and alloying elements such as Mn and Ni stabilise the austenite and lower theAr3-temperature. This suppression of the γ → α transformation temperature leads to arefinement of the final structure by decreasing the growth rate of the ferrite grains. Thedifference between chemistry of the steels which are intended for long products and platesteels lies in the higher carbon contents, ~>0.2% C, of the former steels. The amount ofpearlite is greater and they show a stronger tendency to develop bainitic or other acicularmicrostructures. Since processing of sections and other long products necessitates relativelyhigh finishing temperatures, the RCR processing route together with appropriate micro-alloying should lend itself well for obtaining improved mechanical properties (5.9, 5.13).

Figure 5.10 shows a plot of impact transition temperature against yield strength for the longproduct steels and comparably treated plate steels (FRT=1000-1050oC, ACC=8oC/s andFCT=500-550oC). The impact transition temperature is seen to be strongly dependent on thecarbon content of the steel. Improvement in toughness of long products while maintaining orenhancing the strength level remains an important area for further research and development.

5.4.3 Strip rollingThe principal difference between the steels which are intended for strip and plate steels lies inthe usually lower carbon contents (≤0.08%C) of the former steels. Since the processing ofstrip necessitates relatively low finishing temperatures (usually ≤900oC), the CR processingroute together with appropriate microalloying should lend itself well for obtaining improvedmechanical properties. Water cooling for obtaining high strength steel following hot rollingwas first introduced on strip mills about 35 years ago. A maximum refinement of ferritemicrostructure was achieved by increase of cooling rate after hot rolling. High cooling ratesuppresses the γ→α transformation start temperature and increases the rate of ferritenucleation. Kovac et al. (5.21) claimed to have achieved RCR conditions in hot rolled strip ofTi-V microalloyed steel using with FRT of 930oC followed by laminar cooling. A yieldstrength of 550 MPa was achieved, for 8-10mm thick strips, along with a 35J transitiontemperature at –40oC .The results of microstructure and strength properties of V-Nb-N strip steels processed viadirect charging of thin slab (50mm) showed that a yield strength of ~600MPa and elongationof 24% were obtained for 3.5 – 6.0 mm strips with hot rolling finishing at 900oC and coilingat a temperature of ~600oC (5.22). The rolling parameters and the thickness of strips areimportant for refinement of austenite contributing to ferrite grain size strengthening, whereasnitrogen content shows a very strong effect on yield stress due to precipitation strengthening(see Fig. 5.11) and some further ferrite grain refinement (Fig.3.12).

50

Fig. 5.11 Strengthening effect of (a) ferrite grain size and (b) nitrogen content to the yieldstress of the commercially processed HSLA strip steels microalloyed with V and Nb.

Reheated to 1200°C

LOW

ER Y

IELD

STR

ENG

TH (N

/mm

2 )

V–Nb–NV–NbV–N–Nb–Ti

V–Nb–TiV–N

660

640

620

600

580

560

540

520

50020 30 40 50 60 70 80 90

SLAB THICKNESS (mm)

Fig. 5.12 Effect of steel chemistry and slab thickness on lower yield strength (5.23).

Crowther (5.23) has reported the results of a study on high strength V-microalloyed stripsteels produced by thin slab casting. It was found that a higher reheating temperature of1200oC produced the highest strengths with FRT ~850oC followed by cooling at ~18oC/s andcoiling at ~600oC. The effect of slab thickness on the mechanical properties of strips has alsobeen studied. Results of the yield strength v.s. slab thickness for a number of microalloyedsteel combinations (Ti-V-Nb-N) are shown in Fig. 5.12. As can be seen, the Ti-free steelsexceed comfortably a yield strength of 550MPa for the studied slab thickness although the Ticontaining steels failed to meet this strength level in some cases. With increasing slabthickness some reductions in grain size and precipitation strengthening were observed whichtended to decrease the yield stress but contributed to improved Charpy toughness.

The results also show that Nb addition to a V-N steel and similarly, N addition to a V-Nbsteel has little influence on the precipitation strengthening. This can be explained in thefollowing way. In nitrogen-alloyed steel, Nb will tend to remove N by Nb(C, N) formationduring rolling in the lower austenite temperature range. This not only removes nitrogen butalso tends to co-precipitate some of the V present, and as a result reduces the potential forV(C, N) precipitation strengthening in ferrite. Little additional effect of Nb on strength istherefore seen. The reduced strength in Ti containing steels is similarly in part due to the

51

presence of the large Ti-rich particles, such as (Ti, V)(C, N) or (Ti, Nb)(C, N) which alsoreduce the potential for precipitation strengthening in ferrite.

5.4.4 ForgingIn forging the reheating temperature, range of forging temperature, and reductions are largelygoverned by the material flow necessary to reach the final shape of components, often withcomplex geometries. Consequently, when attempting to attain adequate mechanical propertiesof the component directly after forging we are restricted to much fewer processing variablesas compared to rolling. In fact, the steel chemistry and the cooling following forging are theonly parameters available in practice.

High strength ferrite-pearlite V-microalloyed steels have aroused considerable interest overthe last decades. Strength properties equivalent to those of quenched and tempered (Q-T)steels can be reached by adopting controlled air or forced air cooling, hence completelyescaping the elaborate heat treatments of Q-T steels.

Microalloying with V is an efficient method of strengthening medium-carbon (0.40-0.50% C)ferrite-pearlite steels (5.24-5.26). From electron microscopy it has long been known that thisoccurs by fine N-rich V-carbonitrides in the polygonal ferrite as well as in the ferrite-lamellaeof pearlite. The temperature dependence of the solubility of V-carbonitrides makes dissolutionpossible at convenient temperatures and thereby gives potential for a considerableprecipitation strengthening. V-microalloying is therefore commonly preferred to othermicroalloying additions in these types of steels. The effect of V- and N-additions forprecipitation strengthening of these steels is clearly demonstrated in Fig. 5.13 (5.26).

According to these results 5 parts per weight of N are equivalent to one part of V. Enhancedcooling is an efficient method to increase precipitation strengthening; raising the cooling ratefrom 1°C/s to 4° C/s increases the tensile strength of a 0.4C-1.2Mn-0.10V steel by 120 MPa(5.24, 5.25).

Fig. 5.13 Effect of precipitation strengthening by V and N on Rm. All data have beennormalized to 0.3Si, fp=0.8 and a section size of 20mm Φ. All steels contain 0.7%Mn and

were austenized at 1200-1250°C (5.26).

52

Although other properties are equivalent to those of Q-T steels the toughness of themicroalloyed ferrite-pearlite steels is clearly not. Hence, to expand the use of these steelsimprovement of their toughness is vital. By microalloying with Ti or alternatively raising theAl-content somewhat above normal levels and with close control of N it is possible to reducethe austenite grain size to below 50 µm. This refines the final ferrite-pearlite microstructureand thereby increases toughness. By combining this method with a reduction of the C-contentand an increase of the Mn-content to dilute the pearlite with respect to cementite, it has beenpossible to double the Charpy impact toughness of the standard V-microalloyed steel whilemaintaining the level of strength (5.25, 5.26).

5.5 Microstructure development during TMCP rollingKnowledge of the microstructure development in association with hot rolling is veryimportant for optimising rolling schedules with the aim of improving properties of TMCPsteels. The optimisation of rolling schedules for TMCP-processing by recrystallization hotrolling requires detailed knowledge pertaining to the coupling between rolling schedules(reduction, temperature and pause time) and development of microstructure.

With intelligent design of the rolling schedules it is possible to obtain a very fine and uniformaustenite microstructure. A computer model for calculation of microstructure evolution duringhot rolling is therefore extremely useful for optimising rolling schedules.

The computer model MICDEL developed at the Swedish Institute for Metals Research hasbeen extended to many different steel compositions and permits comparison to be made of theeffect of various material parameters (5.14-5.16). Fig. 5.14 shows examples of predictedmicrostructure evolution during realistic full scale hot rolling of 0.01Ti-V-N and 0.01Ti-V-Nbsteels having different levels of vanadium, niobium and nitrogen, when subjected to the samerolling schedule (5.15). The data refer to rolling of 25 mm plate in 11 passes, starting from1100°C. The initial grain sizes after reheating were taken to be 20 µm for both 0.01Ti-V-Nsteels and 55µm for the Ti-V-Nb steel. For the lower alloyed steel, 0.04V-0.01N-0.01Ti, theeffect of abnormal grain coarsening is demonstrated by using an initial grain size of 500 µm.As evident from Fig. 5.14 the steel with high vanadium and nitrogen is characterised by themost effective microstructural refinement.

The model is based on the concept that the resulting structure is dependent on the staticrecrystallization, static recovery and grain growth of austenite. The static recrystallizationkinetics, and recrystallized austenite grain size, are influenced strongly by: strain, strain rate,temperature and pre-existing grain size as well as steel chemistry. Grain growth of austenitefollowing static recrystallization is described in accordance with Hillert´s theory, where thenormal growth rate of spherical` grains with radius R is dependent on average grain size, theinhibition caused by second phase particles (Zener parameter), the specific grain boundaryenergy and mobility of the grain boundary.

The microstructure evolution routine has also been used to compute austenite microstructuredevelopment during strip rolling of 0.08V-0.018N or similar HSLA steels (5.17), which havebeen processed by recrystallization controlled rolling in one of the mini-mills of NUCORSteel. A general observation is that the grain size existing after thin slab casting or slabreheating has almost no effect on the final austenite grain size resulting from recrystallizationrolling, as long as the rolling schedule involves more than 3 - 4 passes. After large reductions,recrystallization takes place via copious nucleation of new grains and refinement of thestructure can be achieved.

53

The nucleation of recrystallization decreases strongly as the level of strain is reduced, andhence small rolling reductions at high temperatures can cause abnormally large recrystallizedgrains. This is most serious at the end of the hot rolling sequence when a small `sizing` passapplied for control of thickness or shape may destroy the fine-grained austenite structurecreated by RCR processing. A study of sizing passes on the final microstructure of Ti-V-N-steel (5.27) shows that undesirable coarse grain structures can be expected (Fig. 5.15). Nograin coarsening was observed following a strain of 3% for any of the combinations ofholding time and temperature.

Fig. 5.14 Development of austenitemicrostructure for Ti-V-(Nb)-N-steels

during simulation of industrial hotrolling of plates (5.15).

Fig. 5.15 Processing window for avoidingcoarse grained austenite structures in Ti-V-Nsteel after the final sizing pass. The surface in ε-T-t space separates the original fine structurefrom coarse-grained regions created byrecrystallisation or grain growth (5.27).

Figure 5.15 demonstrates the result in ε-T-t space where a surface is defined which separatesthe safe working region for the final processing from that region where undesirable coarsegrain structures can be expected (arbitrarily defined at an austenite grain size of 30 µm).Coarsening of the austenite can occur as a result of grain growth at high temperatures or, atlower temperatures, from the ‘critical strain anneal’ phenomenon. With small reductions infinal sizing passes, nucleation of new grains takes place only infrequently at a few favouredsites followed by extensive growth. Instead of refinement, this process leads to a mixture ofrecovered pre-existing grains of approximately 10 µm diameter together with some very muchcoarser grains. Holding time at temperature is also important for microstructure coarseningand a practical conclusion is that accelerated cooling applied rapidly after the final pass can beof great benefit in reducing the danger of structure coarsening.

54

5.6 Summary of chapter

• Recrystallization controlled rolling (RCR) or low temperature controlled rolling (CR)followed by accelerated cooling (ACC) lend themselves well to plate, long products aswell as strip production of V-(Nb)-Ti microalloyed steels.

• RCR+ACC is a superior alternative to CR as a TMCP-process for manufacturing V-microalloying steel plate and long product with high strength and toughness. Advantagesof the RCR process are its high productivity and lower mill loading as compared to CR, aswell as more efficient utilisation of V in precipitate strengthening.

• Accelerated cooling increases the strength of the V-microalloyed HSLA steel productboth by refinement of microstructure (ferrite, bainite) and by generating smaller anddenser microalloy precipitates

• A computer routine for prediction of microstructure evolution during TMCP processing isavailable and good agreement between calculated and observed final microstructure of Ti-V-(Nb)-N microalloyed HSLA steel plate has been obtained.

• In order to prevent the occurrence of grain coarsening following the final "sizing pass" inthe Ti-V-microalloyed steel it is recommended that the finishing temperature should below and that reductions smaller than 10% should be avoided in this pass, preferablyfollowed by accelerated cooling.

55

6. CAST STRUCTURES

6.1 Microstructural featuresMuch less attention has been devoted to the microstructures of as-cast slabs than those ofrolled products for obvious reasons. On the one hand, sampling of these thick gauges(typically 220mm) is troublesome and on the other hand their influence on final properties isoften only indirect. Nevertheless, some knowledge of these structures is necessary, especiallyin connection with down-stream processing such as hot rolling. Broadly, the as-cast structurescontain two types of features; those which are persistent and may directly affect the finalproperties and others which are modified beyond recognition during subsequent stages of themanufacturing process. Among the persistent features can be listed inclusion morphology,TiN dispersion and segregations which can lead to banded structures and inhomogeneousdispersions of microalloy carbonitrides. Transient microstructures include grain structuresand microalloy precipitates in the form of eutectics or other coarse carbonitrides, which maybe refined by recrystallization or dissolved by diffusional processes.

In fact, the distinction between such persistent and transient features is not always clear cut.For example, when comparing conventional slab casting and reheating prior to rolling withthin slab casting and direct hot rolling it is advisable to consider more carefully the influenceof as-cast structures. The absence of the γ−α−γ phase transformation prior to rolling and thesmaller total reduction may mean that the initial grain structure still has some influence on thefinal grain size. Similarly, the lower temperature and shorter holding time before rolling maynot be sufficient to dissolve eutectic alloy carbides, with implications for precipitationstrengthening in the final product. Few systematic studies on thin slab structures have so farbeen reported so it is not possible to draw clear conclusions but it is probably wise to assumeat least some differences from the behaviour in traditional processing.

Continuously cast slabs of structural steels usually show evidence of columnar solidificationnear to the surfaces with equi-axed structure towards the centre. However, this is often notapparent in the final ferrite+pearlite or bainite microstructure due to the sequence oftransformations δ−γ−α on cooling. Figure 6.1 is taken from the work of Bruce (6.1) on220mm thick cast slabs of a V-Ti-microalloyed steel (0.13%C,1.46%Mn,0.042%V,0.01%Ti).

200x 25x 25x

Fig. 6.1 Optical micrographs of as-cast HSLA steel slab. (a) grain structure, (b) prioraustenite grain boundaries, and (c) dendritic structure of the original δ-ferrite.

56

At higher magnification the microstructure is a relatively uniform mixture of ferrite andpearlite (Fig. 6.1(a)) but low magnification (Fig.6.1(b)) reveals the prior austenite grainboundaries decorated by continuous networks of ferrite. In this sub-surface region theaustenite grains are seen to be columnar with a thickness of about 1mm and are severalmillimetres in length. The original dendritic structure of this steel was revealed by a heattreatment involving austenitizing at 950oC followed be slow cooling to 770oC, waterquenching and tempering at 400oC. After etching in nital the dendrites appear as lightcoloured ferrite in a background of darker tempered martensite (Fig 6.1(c)). It is apparenthere that the as-cast δ-ferrite grains were considerably larger than the subsequent austenite inthis case. There are indications that titanium additions can assist this austenite refinement oralternatively hinder grain growth in the newly transformed austenite (6.1). On the other hand,giant austenite grains centimetres in diameter may also occur in such microallyed steel,especially near to corners. These give the appearance of secondary recrystallization or of acritical strain recrystallization phenomenon arising from bending, thermal effects or, possiblyfrom strains caused by the peritectic reaction.

The as-cast structures of thin (50mm) slabs of V-Nb microalloyed steel(0.06%C,1.49%Mn,0.11%V,0.017%Nb) have been examined by Siwecki (6.2). The columnaras-cast structure was almost continuous except for a layer of finer equi-axed grains at thesurface (Fig. 6.2(a)). Also here, the austenite grain boundaries were visible at lowmagnification decorated by ferrite films (Fig.6.2(b)). The austenite grains tended to becolumnar except close to the surface and were smaller by a factor of about two than those inthe CC slab described by Bruce (6.1).

25x 25xFig.6.2 Optical micrographs of as-cast thin (50mm) slab showing (a) original dendritestructure, and (b) prior austenite grains.

Vanadium in cast slabs of microalloyed steel is found in carbo-nitride particles that may adoptat least three types of morphology. Fine scale precipitates around 100nm in diameter arethought to arise during cooling through the low temperature austenite range (Fig.6.3(a)). Insteels containing also Nb these may occur as mixed (V,Nb)(C,N) phases. Steels alloyed withboth V and Ti (Zajac et al. (6.3)) contain large ‘dendritic’ particles with arms extending inthree mutually perpendicular directions (Fig. 6.3(b)). At the centres of these crosses can be

57

found cuboidal TiN particles on to which the arms of almost pure VN have grown. Their sizeindicates that diffusion must have occurred over relatively long distances and accordingly thatthese structures formed at higher temperatures in the austenitic range, when the solubilitylimit was exceeded during slab cooling, nucleating on the pre-existing TiN precipitates. Thethird type of V(C,N) phase which has been found in cast slabs is much coarser still andappears to be a eutectic product similar to the NbC eutectic phases that have been reported byPriestner and coworkers (e.g. 6.4,6.5) in solidified niobium microalloyed steels. Occurrenceof such products in these steels is attributed to microsegregation which leads to enrichment ofthe alloy elements within the liquid phase so that solidification terminates with a eutecticreaction in inter-dendritic regions An example of such a phase in a cast slab structure(0.09%C, 1.41%Mn, 0.013%N, 0.08%V, 0.01%Ti) is shown in Fig.6.3(c) from unpublishedwork at SIMR.

(a) (b) (c)

Fig. 6.3 V(C,N) phases precipitated in as-cast HSLA steel slab. (a) and (b) replicas in TEM,and (c) optical micrograph.

The vanadium-rich carbonitride phases normally dissolve during the prolonged hightemperature reheating of slabs before hot rolling in conventional processing. Even the coarseeutectic phases are seldom apparent in final rolled products. It can be supposed that themicroalloy addition is therefore made available for precipitation strengthening as intended.For direct rolling of thin slabs, both the holding time and temperature prior to hot rolling areless and it is possible that some of the coarse particles remain with detriment to the finalproperties. This seems to be the case for some Nb-microalloyed steels (6.5); as yet, thesituation with regard to V-steels has not been clarified to our knowledge.

6.2 Ductility during castingProbably the most important practical consequence of the structure and chemistry in as-caststeels is how these affect ductility of the slabs when the strand is bent and unbent at hightemperature in the production process. It is well known that a trough in ductility exists over acertain range of temperature where surface cracking of the strand is likely to occur and thatthe problem is exacerbated in microalloyed as compared to plain C-Mn steels. The lowductility region usually lies in the range 700oC – 1000oC but the most critical factor is theupper temperature at which ductility deteriorates since bending of the slab, which is coolingcontinuously, must be complete before any part of it (face, edge or corner) falls below thistemperature.

58

Magn. 200x

Fig. 6.4 Example of cracking along an austenite grain boundary in a cast slab ofTi-V-microalloyed steel.

The normal appearance of these cracks is that they follow austenite grain boundaries as shownby the example of a corner crack in a slab of Ti-V-microalloyed steel in Fig.6.4 from Bruce(6.1). In this case the ferrite films delineating the original austenite grain boundary haveformed after the cracking took place. The cracking along austenite grain boundaries at hightemperature and rather low rates of deformation is an example of typical creep rupturebehaviour which is known (6.6) to become more serious at large grain sizes as in the case ofthese cast products. As the grain boundaries act as softer material than the grain interiors,deformation and sliding concentrate at these, leading to void initiation at triple junctions orsecond phase particles lying in the boundaries. Microalloyed steels have been postulated tobe sensitive to such cracking for several reasons (6.7,6.8), viz:

• Void initiation at coarse carbonitride particles located on austenite grain boundaries• Hardening of the grain interiors by fine scale carbonitride precipitation• Inhibition of dynamic recrystallization by precipitation on deformation substructures

thus preventing relaxation of stresses built up at grain boundaries.

An additional effect in steels comes into play when the temperature falls below Ar3 and ferritestarts to form, commencing along the austenite grain boundaries. At a given temperature thestrength of ferrite is only about 60% that of austenite so the tendency for deformation toconcentrate to the boundary regions becomes magnified with rapid localised rupture as aconsequence. Only when the structure becomes mainly ferritic at lower temperatures can thedeformation become more uniformly spread and ductility be restored.

A number of investigations have been reported comparing the high temperature ductilitybehaviour of different types of steels. These have mostly made use of tensile testing or insome cases bending to simulate more closely the steelworks situation. The most reliable datacome from tests where the steel is melted and solidified and then cooled directly to the testingtemperature but in some cases a very high prior austenitization temperature was used toachieve coarse grain structures having the microalloying elements in solid solution.

Results of four sets of hot tensile tests are compared in Fig.6.5, all using different ranges ofsteel chemistry. Reduction in cross-sectional area at failure was the measure of ductility.

59

Figure 6.5(a) is from Mintz and Abushosha (6.9) for samples pre-heated at 1330oC while Fig.6.5(b) refers to solidified samples (6.10) as does Fig.6.5(c) from the work of Karjalainen et al.(6.11). An alternative measure of hot ductility in Fig.6.5(d) is the length of the largest crackfound in 50mm thick slabs bent at various temperatures after casting, from Crowther etal.(6.12). These various results are not easy to compare in absolute terms of deformation ortemperature that can have been affected by the particular conditions of cooling or strain ratesthat were applied. However, they show a fairly consistent trend when different steelcompositions are compared in each case. Plain C-Mn steels show the least sensitivity tocracking, both in terms of the depth of the ductility trough and the range of temperature overwhich it exists. The most susceptible steels are those microalloyed with Nb and this appearsto be particularly true when the upper temperature of the embrittlement range is considered.As mentioned above, this is the most relevant parameter for steelworks practice. Vanadiummicroalloyed steels show an intermediate behaviour between the plain and the Nb steels. It isseen that higher nitrogen contents in the V-steels extend the low ductility trough, especiallytowards the higher temperatures. According to Mintz and Abushosha the hot ductility of V-steels are superior to Nb-microalloyed ones provided that the V and N levels in the steelexpressed in weight % are such that [V][N]<1.2x10-3. A curious feature in some of thesereports is that the presence of V in Nb-microalloyed steels seems actually to improve theresistance to hot cracking even though the level of Nb remained the same.

Fig. 6.5 Comparative hot ductility data for steels from various literature sources.(a) 6.9 (b)6.10 (c) 6.11 and (d) 6.12.

Interpretation of the ductility trough results in Fig. 6.5 is not entirely straightforward. Thecracking data of Crowther et al.(6.12) were considered to be partly a result of ferrite formationalong austenite grain boundaries. This seems reasonable in view of the temperatures thatwere all below 900oC. All the tensile tests show that loss of ductility commences at much

60

higher temperatures, however. The effect of the microalloying elements is, of course, not dueto a change in the Ar3 temperature but must be associated with precipitation of carbonitrideparticles. This is fully compatible with the lower solubility of Nb than V in austenite suchthat precipitation of NbC occurs at higher temperatures. It can also be reconciled with theeffect of nitrogen on the driving force for VN as discussed in chapter 2. However, fullagreement concerning the mechanism of ductility loss still seems to be lacking. The idea thatdynamic recrystallization is inhibited by Nb is in qualitative agreement with manyobservations on hot working and could indeed be true for the hot tensile test conditions.However, the strains involved in these tests are much larger than those in the real steelworkssituation. Dynamic recrystallization would not normally be expected for the strains of lessthan 1% which occur in practice. Strengthening of the austenite grain interiors by precipitatedparticles which in turn raises loading on the grain boundaries is suggested by some test resultsand the same effect may possibly become magnified when softer grain boundary ferritedevelops. Fine precipitates are expected to have a greater strengthening effect than coarseones so would be more deleterious. In one study at least (6.12), the beneficial effect of V-additions to the hot ductility of Nb-microalloyed steels was seen to be associated with anincrease in size of the precipitated particles.

The higher casting speeds in thin slab casting appear to make further demands upon the hotductility of microalloyed steels. As a consequence of this, carbon contents in the peritecticrange, ~0.07 – 0.17%, are avoided. Furthermore, it has been demonstrated that Nb-microalloying deteriorates the surface quality of hot rolled strips because of surfacemicrocracking during thin slab casting (6.13). V-microalloying even when combined with anenhanced N-content is capable of remedying these problems. First the loss in strength byreducing the C-content below 0.07% can be satisfactory compensated by proper V/Nmicroalloying. Furthermore the hot cracking during casting in Nb-steels leading to impairedsurfaces can be avoided in V-steels.

6.3 Summary of chapter

• Grain structures in as-cast steels are not strongly dependent on the presence of V as amicroalloying element. Vanadium exists in carbonitride particles having a wide range ofsizes as a result of precipitation in different temperature ranges during solidification.

• The reheating treatment of slabs prior to conventional hot rolling is normally adequate toensure that V is re-dissolved and is available to provide precipitation strengthening in thefinal rolled products. In the case of direct hot rolled thin slabs there remains someuncertainty as to whether complete dissolution of V(C,N) takes place.

• The tendency for hot cracking of cast slabs is more severe in Nb-microalloy steels than inV-steels. High N-levels increase the susceptibility for hot cracking. Addition of V to Nb-steel appears to have a beneficial effect associated with coarsening of the (Nb,V)(C,N)particles in the cast material.

61

7. WELDABILITY

7.1 Structure and properties of coarse grained HAZIn recent years the need for structural steels having satisfactory low-temperature heat affectedzone (HAZ) toughness has resulted in a general reduction in the carbon, sulphur and nitrogencontent (below 40 ppm N) and microadditions of Ti and/or Ca. However, such low nitrogencontent is very costly and also troublesome to achieve in practice. Many features of variousmicroalloyed steels have been studied at the Swedish Institute for Metals Research andparticular attention has been paid to the important role of the microalloying elements V, Ti,Nb as well as N in the HSLA steels and their effects on properties of weldments (7.1-7.5).High nitrogen (wt%N>0.009) in HSLA steel containing V has sometimes been considered ashaving an unsatisfactory weldability due to "free" nitrogen which has been claimed to impairtoughness in the HAZ. However, some of the results obtained lately show that the V steelwith high nitrogen can have a good HAZ strength and toughness by careful selection ofmicroadditions and the welding parameters. The affinity of microalloying elements tonitrogen decreases in the order Ti, Al, Nb and V, but it is Nb(C,N) and VN which engenderthe greatest precipitation hardening.

Fig. 7.1 Austenite grain size vs. heating rate for various HSLA-steels subjected to weldsimulation (7.1).

Austenite grain coarsening during welding which depends critically on the heat input andsteel chemistry has a significant effect on the toughness of steel by controlling the ductile tobrittle fracture transition temperature. The results of grain coarsening studies (7.1, 7.2) forvarious microalloyed steels are shown in Fig. 7.1 as the average austenite grain size afterwelding simulation at varying heating rates.

The tendency for austenite grain growth at different temperatures demonstrating the effect ofdifferent microalloying elements is shown in Fig. 7.2 (7.2). It is shown there that the averageaustenite grain size of the Ti-bearing steels is ~6-15 times smaller than that of C-Mn steelafter 30 min. holding at 1350°C. It can also be seen that the Ti-microalloyed steel hasexcellent resistance to grain growth for temperatures up to 1300°C. The presence of V and/orNb in Ti-steel deteriorates grain coarsening resistance. However, these steels are still betterthan the C-Mn steel.

62

Fig. 7.2 Effect of peak temperature and steel chemistry on the austenite grain size (7.2).

The results of grain coarsening studies (7.4, 7.5) for commercial 0.01Ti-0.08V plates with low(0.003 % N) and high (0.013 % N) nitrogen contents show that extensive grain growth occursjust above 1250°C in low N steel whereas the high nitrogen steel preserves small grains up to1350°C.

7.2 Effect of nitrogen and welding parameters on HAZ toughness of V-steelsSeveral investigations showed that V-microalloyed steels gave increased impact transitiontemperature (ITT) with increasing nitrogen content for high heat input, single pass welds, butrather constant ITT values for low and medium heat input using multi-run welding, as shownin Fig. 7.3 (7.3). During multi-run welding the VN, which is dissolved in the earlier passes,re-precipitates in the subsequent passes and these particles inhibit grain coarsening and allowtransformation to fine grained ferrite which results in high toughness.

Fig. 7.3. Effect of N on toughness of heat affected zone of welded 0.1 % V steels (7.3).

63

Another method of mitigating the effect of nitrogen in HSLA steel is by microaddition oftitanium. TiN is stable up to very high temperatures, where niobium, vanadium and/oraluminium nitrides are dissolved. For this reason, titanium additions to a weldable steel can bea useful instrument to restrict grain growth (see Fig. 7.1) and thereby increase the toughnessin the HAZ adjacent to the fusion line.

Commercial 0.01Ti-0.08V-N steel plates with high (0.013 % N) and low (0.003 % N)nitrogen contents and produced by RCR, RCR+ACC and CR routes have been subjected totemperature cycles simulating the heat affected-zone of 25 mm plates during high, mediumand low heat-input welding in order to investigate the influence of plate productionprocessing route, nitrogen content and heat input on HAZ toughness and microstructure (7.4,7.5). Changes in the impact toughness of the HAZ of 0.01Ti-0.08V-N steels obtained fromthermal simulation of HAZ are shown in Fig. 7.4(a), whereas the results of full scale weldingtrials are shown in Fig. 7.4b for comparison. As seen here, there is virtually no effect of theprevious TMCP parameters on the toughness of the coarse grained HAZ after welding.

a) b)

Fig. 7.4 Impact transition temperature at 40J of: (a) simulated HAZ and (b) HAZ after fullscale welding of low (0.003N) and high (0.013N) nitrogen plates with different heat input(7.4, 7.5).

Combined effect of heat input and N-content for V-Ti-N steels is shown in Fig. 7.4. Theseresults do not allow us to directly associate the low toughness after high energy welding with"free" nitrogen. On the contrary, as shown in Fig. 7.5, higher N steel (0.013 %N) exhibits thehighest HAZ-toughness for the cooling time of 10s (low heat input). For this heat input(cooling condition ∆t8/5≤30s) one does not expect precipitation of VN in austenite duringcooling to the transformation temperature and the amount of "free" nitrogen in ferrite shouldbe higher than in the slower cooled specimens. With increasing cooling time precipitation ofV(C,N) should be more abundant reducing the amount of nitrogen in solid solution. However,it was these specimens which exhibited low toughness.

Quite low impact toughness in the HAZ of 0.08V-0.05Nb-0.01Ti steel was reported byMitchell (7.6). This low toughness (40J at –5oC) is most likely not due to free nitrogenbecause the nitrogen level had almost the stoichiometric ratio to Ti, and should therefore allhave been combined as TiN. However, some improvement in HAZ toughness (40J at -25oC

64

for synthetic HAZ at ∆t 8/5=50secs) was obtained by addition of 0.25%Mo to the V-Nb-Tisteel.

Fig. 7.5 Effect of heat input and nitrogen content on the HAZ impact transition temperatureof RCR + ACC, 0.01 Ti - 0.08 V plates.

The effect of vanadium content on the HAZ toughness of welds has been reported in anumber of investigations (7.6-7.8). No significant effect of vanadium up to 0.1% on the HAZtoughness of synthetic/thermally simulated welds was reported by Hannerz et al. (7.7) forcooling times ∆t8/5< 100secs, whereas some reduction of toughness was observed with higherV level.

Mitchell et al. (7.8, 7.9) reported on improvement in the HAZ impact toughness in multipasswelds with increasing V level up to 0.16%, although the CTOD transition temperatureincreased slightly at the same time. It was also reported that higher heat input (>2kJ/mm) hada positive effect on the CTOD toughness and a slight negative effect on 40J ITT. It wassuggested that the improvement in as welded toughness is connected with the formation of afavourable microstructure, intragranular ferrite. This is in general agreement with our ownobservation, that the HAZ microstructure at high heat input is usually a mixed structure offine ferrite with second phases (FS), grain boundary (PF(G)) and intragranular ferrite (PF(I))(see Fig. 7.6). The poorer toughness of the high heat input HAZ in high N steel is consideredto be caused by the presence of a significant volume fraction of coarse grain boundary ferritein association with V(C, N) (7.4). Note also that hot rolling conditions do not affect HAZstructures in these steels to any significant extent.

According to Swedish standard welding practice 25mm plates are normally welded in two ormore passes. The investigation (7.4) of multi-pass welding of 0.01Ti-0.08V-N steels shows apositive effect of the second pass, which allows good toughness even at high heat input. Amarked improvement in HAZ toughness after the second pass was observed as compared tothe single pass welding with high heat input, cooling time ∆t8/5 = 100s. The ITT40J of 0.013Nsteel plate decreased by 50 degree to -48°C and for 0.003N steel plate reached -60°C.

65

Fig. 7.6. Relation between volume fraction of microstructure constituents and the heat input(cooling time ∆t8/5 from 1350°C) for Ti-V steel with low and high nitrogen contentprocessed by RCR+ACC and by normal rolling (NR) (7.4) .

7.3 Toughness-Microstructure RelationshipsConcerning the results in Figs. 7.4 and 7.5, the drastic changes in 40J ITT with varying ofheat input / cooling time, it can be concluded that the notch toughness properties of coarsegrained HAZ of the 0.01Ti-0.08V-N steels can be explained on of the basis of theirmicrostructures. For high nitrogen steel the results show that the presence of coarse grainboundary ferrite in the HAZ makes the most significant contribution to the overall low notchtoughness properties rather than "free nitrogen".

The excellent notch toughness properties of the HAZ of high nitrogen plates subjected tothermal simulations of heat input with ∆t8/5<40s and particularly at ∆t8/5=10s can beattributed to the characteristics of the fine ferrite with second phase (FS) (see Fig. 7.7). Thusthe cooling time which gives 50% of fine FS-phase is sufficient to confer good impact notchtoughness (see Fig. 7.6 and 7.8) and this seems to be independent of nitrogen.

66

Fig. 7.7 Microstructure of HAZ of the Ti-V-N steels with low – 0.003% and high –0.013% nitrogen content after thermal simulation at 1350°C and cooling down with∆t8/5=10s (a), 40s (b) and 100s (c) (7.4).

The coarse, intergranular ferrite structure developed in the coarse grained HAZ of platesubjected to the simulation with ∆t8/5>40s was responsible for the poor low temperature notchtoughness properties. The transformation temperature of plates of 0.003%N and 0.013%Nafter heating to 1350°C and cooling down with the rate equivalent to welding with low,medium and high heat input was studied by Zajac et al (7.4). The results showed that thetransformation start temperature of the high N steel is more than 75°C higher than that of lowN steel for low heat input and 45°C higher for high heat input welding conditions. Althoughthe reason for this is not entirely clear, it provides an explanation for the occurrence of thenetworks of coarse ferrite grains seen in the high N steel.

Altogether, the impact toughness values found can mainly be understood by considering thegeneral microstructure; the detrimental effect of high heat input is caused by coarse grainedmicrostructure with assistance of precipitates. Some effort has been expended to study theparticle size distribution and its chemistry in the HAZ of HSLA steels (7.3, 7.4). These resultshave shown that the TiN particles in the HAZ of high N 0.08V-steel plate had an average sizeof ~24 nm at a top temperature 1350°C, whereas in the low nitrogen steel plate only a fewlarge cuboids, ~100nm, were observed. However, if the specimen was cooled slowly down to800°C and then quenched, a number of very small particles, less than 5 nm in diameter wereobserved. This suggests that these small TiN particles re-precipitate during the cooling stageof the thermal cycle. In a specimen of high N steel cooled slowly (∆t8/5=100s) to roomtemperature, additional round particles of V(C,N) were observed.

67

Fig. 7.8 Relationship between ITT40J of HAZ and volume fraction of ferrite with secondphase - FS.

In high heat input welding exceeding 10 kJ/mm, toughness becomes sometimes critical at thecoarse grained HAZ due to formation of the upper bainite and associated martensite-austeniteislands. The problem can be solved by deliberate addition of Ti (ca 0.015%) or variouscombinations of Ti, B, REM and Ca. Steels modified by these microadditions showsatisfactory HAZ toughness at -20oC for the high input of 15 kJ/mm. (7.3-7.9). As discussedpreviously, fine particles of TiN or other compounds restrict the austenite grain coarseningand act as nucleation sites of ferrite, which results in a decreased fraction of upper bainite.

The results presented above show that proper microstructure, and therefore good toughness,could be obtained for high nitrogen steel by balancing heat input to provide a finetransformation product.

7.4 Summary of chapter

• High nitrogen in V microalloyed steel can be compatible with good weldability, asregards HAZ hardness and toughness, by careful selection of microadditions andwelding parameters.

• There is a general trend of decreasing HAZ toughness with increasing heat input(cooling time). This effect is very small in the low nitrogen steels that showed only aslight deterioration in toughness. The transition temperature of the high nitrogensteels (0.013 % N) rises drastically for cooling times above 30-40s.

• The austenite in the HAZ of V- and Nb-steels coarsens significantly during welding.By Ti-microadditions to V- and Nb-steels this coarsening can be minimised. Anincrease in nitrogen content in Ti-V-N plates, above the stoichiometric level for TiNsuppresses grain coarsening of the HAZ.

• Plate production processing route seems to have insignificant effect on HAZtoughness.

68

8. HOW IS V BEST USED IN MICROALLOYED STEELS - EXECUTIVESUMMARY

The rapid development and use of welded structures after World War II soon made it clearthat strengthening of structural steels by increased C was unacceptable because of low tough-ness and poor weldability. In this technical environment the development of microalloyedsteels and the thermomechanical processing of such steels took off with a great thrust. Inretrospect one can say that the basic idea behind this development was, and in fact still is, tocapitalise on the effects of

(i) grain refinement and(ii) microalloy precipitation

Grain refinement is unique in that it adds both strength and toughness to the steel whereasmicroalloy precipitation increases strength at the expense of some loss in toughness.Weldabilty is normally maintained in both cases but, when grain refinement can be retainedduring welding, it is substantially improved. In short this is the background to the arrival ofnew structural steels with steadily improved strength, toughness and weldability over the last40 years resulting from the development of microalloyed steels and routes forthermomechanical processing of these.

In such a general description of the development one can see a common ground for all themicroalloyed steels, alloyed with Nb, V or Ti or combinations of them. However, when itcomes to the real development of steels with optimised composition and processing routeproper account must be taken to the differences between the various elements. The mostsignificant properties of V as compared to Nb and Ti are:

(i) The solubility of V(C,N) is much larger; this is particularly evident at highertemperatures, implying that the solution temperature is lower or alternatively that largeramounts of V can be dissolved at a given temperature.

(ii) As opposed to Nb but similar to Ti, the solubilities of the carbides and nitrides of Vdiffer widely; the solubility product of VN is about two orders of magnitude lower thanthat of VC. This implies that N as a microalloying element has a decisive role in V-microalloyed steels, especially for the enhancement of their precipitation strengthening.

8.1 Austenite conditioning by thermomechanical processingThe primary objective of thermomechanical processing of microalloyed steels is to conditionthe austenite to produce as refined ferrite as possible. Today it is a well documented fact thatone can reach the same level of grain refinement in V-steels as in Nb-steels, a ferrite grainsize of ~4µm. However, the processing routes to achieve this are totally different because ofthe widely different solubilities of Nb(C,N) and V(C,N) at the temperatures for hot rolling.Nb(C,N) precipitates already in austenite during hot rolling and thereby retards recrystalliza-tion effectively. Eventually recrystallization stops and on continued rolling theunrecrystallized austenite grains are gradually more flattened. Mainly because of theincreasing austenite grain boundary area this produces after transformation a much refinedferrite microstructure.

69

V offers no effective resistance for austenite recrystallization during hot rolling since its largersolubility engenders no impeding precipitates. The absence of hindering particles offers onthe other hand a new opportunity for grain refinement: Repeated austenite recrystallizationafter each rolling reduction generates, after a sufficient number of passes, a very efficientaustenite grain refinement which after transformation can produce a ferrite grain size equallyfine as with the former method, ~ 4µm, see Fig. 3.11.

Controlled rolling (CR) of Nb-steels by extending to low temperatures in the range of non-recrystallization was already developed and applied in the mid 60’s. Beside the fact that it wasfirst on stage its primary advantage appears to be - given the provision that the mill is strongenough for the large rolling forces at low temperatures – that it is relatively easy to control.Recrystallization-Controlled-Rolling (RCR), which is the name given to the latter process,was developed later and was first practised ~ 1980. Although traditional CR of Nb-steels isprobably dominant RCR of V-microalloyed steels offers important advantages whichprobably are still not fully utilised:

(i) economical hot rolling practice (high productivity) by low reheating and highfinishing temperatures.

(ii) ferrite microstructure can be further refined by forced cooling and by somewhatenhanced N-content (typical of EAF steel making).

(iii) V is retained in solution before the austenite/ferrite transformation due to its largersolubility and the higher finishing temperature in RCR, and therefore generates a con-siderable precipitation strengthening in the austenite/ferrite and ferrite ranges. Forcedcooling and enhanced N-content add significantly to the strengthening. Precipitationstrengthening of 250-300MPa may be reached in V-N-microalloyed steels.

(iv) The repeated recrystallization in RCR breaks down prior austenite microstructureefficiently. Traditional CR characterised by a high recrystallization stop temperaturemay retain large austenite grains, typical of the beginning of abnormal grain growthduring reheating. During continued rolling such grains will gradually be flattened butcould still produce a local, deviating microstructure causing detrimental properties.

V-microalloyed plate steels are commonly co-alloyed with Ti, ~ 0.01%. Provided coolingduring casting is sufficiently fast, as in concast slabs, this creates a dispersion of fine, verystable TiN-particles that effectively prevents austenite grain coarsening at high temperatures.Although this effect is primarily used to improve the HAZ weldability of the steel, it is also anefficient means of ensuring the finest possible grain size during RCR processing since itprevents austenite grain growth between the rolling passes and after finishing. Ti-addition istherefore often a preferred choice in V-microalloyed plate steels. Ti-V-N- plate steels with ~0.01%Ti and ~ 0.01%N offer very competitive combinations of strength, toughness andweldability, cf. chapter 5.

The very stable TiN dispersion of the Ti-microalloyed steels, showing no measurablecoarsening below about 1300oC, prevents coarsening of the bulk of the austenite micro-structure well above that temperature. During short heating periods, such as in HAZ of welds,this gives a secure control of the grain growth inhibition. But it is important to recognise thatduring prolonged heating, such as during reheating, there is risk for initiation of abnormalgrain growth despite the stability of the TiN dispersion, However, this is a matter of a few

70

exceptional grains, large enough to grow through the TiN-dispersion, that will slowly con-sume the fine grained microstructure, Therefore, unless reheating is not excessively long, thisphenomenon creates only streaks of coarse microstructure.

8.2 Precipitation strengtheningOf the different microalloying elements used, V is clearly the one best suited for strong andreliable precipitation strengthening, the principal reason being the larger solubility product ofits carbonitrides resulting in a lower solution temperature and a larger capacity to dissolvethem at elevated temperatures. A special feature of V as compared to Nb is that its nitride ismuch less soluble than its carbide and this gives N a very important role in V-steels,especially in their precipitation strengthening.

For hard particles such as microalloy carbonitrides the sole parameter determining pre-cipitation strengthening is the particle spacing. Therefore to obtain the largest possiblestrengthening the nucleation rate of V(C,N), which determines the density of the precipitatedispersion, should be maximised. Experimentally it is well proven that the V(C,N)-pre-cipitation becomes denser and finer with increasing N-content. Fundamentally this is due toan increasing chemical driving force for this reaction and this causes in turn a larger nuclea-tion rate. The strengthening by enhanced N is quite considerable; in a 0.12%V-steel it istypically 6 MPa/0.001%N and is rather independent of the C-content of the steel. Instead ofadding strength this effect can be used to save alloying costs. As an example the presentresults, Fig. 4.10, indicate that by adding N from a level of 0.005% to 0.015% makes itpossible to reduce V from 0.12% to 0.06%.

As a result of the lower solubility of VN than of VC the first formed V(C,N) is N-rich for allnormal N-contents, and it is only when almost all N has been consumed that the C-content ofthe carbonitride starts to increase. To benefit from continued precipitation strengtheningbeyond this point there must be sufficient C in solution and chemical driving force for freshV(C,N)-nucleation to occur. It has been shown that such a situation exists when the ferrite C-content is controlled by the austenite/ferrite rather than the ferrite/cementite equilibrium; inthe former case the C-content is 5 times larger than in the latter at 600oC.

A long upheld view is that the total C-content of the steel has little, if any, influence on theprecipitation strengthening by microalloy carbonitrides in view of fact that the content of dis-solved C in ferrite is normally controlled by the austenite/ferrite or the ferrite/cementiteequilibria. However, recent studies have shown that the precipitation strengthening of V-microalloyed steels increases significantly with the total C-content, ~ 5.5 MPa/0.01% C, cf.Fig. 4.11(b). The explanation is that the C-content of the steel delays the pearlite(ferrite+cementite) formation and thereby maintains the higher C-content in ferrite given bythe austenite/ferrite equilibrium for a longer time.

The present results show clearly essential differences between N and C when it comes to thepractical use of them for V(C,N) precipitation strengthening. Beside the fact that the N-con-tent largely determines the V(C,N) density and thereby the size of precipitation strengthening,a very important advantage with N is that the content of N dissolved in ferrite at the onset ofprecipitation equals, or very closely so, the total N-content of the steel. The reason for this isthat pure ferrite dissolves much larger contents of N than C. The solubility of N at theeutectoid temperature is about 0.10% whereas it is 0.02% for C. Differently expressed, thisimplies that a very large N-content for a steel, of say 0,02%, will be dissolved in ferrite over awide temperature range, ~ 800-350oC. This circumstance together with the much lower

71

solubility of VN as compared to VC makes the combination of V and N especiallyadvantageous in precipitation strengthening. The C-content on the other hand is difficult touse and control for the continued precipitation of C-rich V(C,N) after the consumption ofmost N. This is due to the fact that the solute content of C in ferrite, which is the relevantparameter for the precipitation, depends in a complex manner on the transformations ofaustenite to ferrite, pearlite and eventually to cementite. Hence, not only does N addsignificant precipitation strengthening to the steel but it also facilitates the control of thestrengthening.

8.3 Cast structuresA practical problem with processing of microalloyed steels is their greater susceptibility forcracking as the continuously cast slabs are bent and unbent in the hot as-cast condition.Although the mechanism causing this loss of ductility is not fully clarified, it is certainlyexacerbated by precipitation of microalloy carbonitrides in some form. The higher solubilityof V as compared to Nb in austenite implies that the temperature range where precipitationcan occur is removed to lower temperatures. Accordingly, the hot cracking problem is lesssevere with V-steel. This view is supported by several investigations where solidified samplesof steels were bent or tensile tested at elevated temperature. Higher N contents in V-steelsincrease the susceptibility for hot cracking to occur since the temperature range over whichprecipitation of VN occurs is extended to higher temperatures. Nevertheless, V is still lessharmful than Nb in the respect and it has furthermore been found that additions of V to Nb-microalloyed steels can actually reduce the tendency for hot cracking in these. This isapparently due to a change in morphology of the (Nb,V)(C,N) precipitation.

8.4 HAZ weldabilityHigh N-contents, >0.010%, in V-microalloyed steels are often considered to causeunsatisfactory HAZ-weldability due to solute nitrogen which has been claimed to impairtoughness. Recent investigations show, however, that such generalising statements areincorrect and that satisfactory welding results can be achieved also at high N-contents byselective choice of welding parameters.

It is true that the HAZ impact transition temperature (ITT) increases with increasing N con-tent in high heat input single pass welds (8kJ/mm), Figs.7.3 and 7.4. On the other hand it hasbeen shown that quite adequate ITT, -45oC, can be obtained in plate welds of a 0.01Ti-0.08V-0.013N steel up to heat inputs of 4kJ/mm. Moreover, multipass welding of V-N steelsproduces very low ITT, ~ -70oC, independent of the N-content up to 0.016%.

Microadditions of Ti, ~ 0.01%, provide an efficient means of suppressing grain coarsening inthe HAZ, a technology applied in the Ti-V-N plate steels. A method to further enhance thegrain coarsening resistance in these steels is to increase N above the stoichiometric level ofTiN.

Detailed analysis of results on the relationship between welding parameters, microstructureand ITT of HAZ has clearly shown that the observed increase of ITT at high heat inputs inhigh N-steels cannot be explained by solute N. Instead, the explanation lies in variations ofmicrostructure, and in particular the presence of coarse grain boundary ferrite makes the mostsignificant contribution to the lower toughness at high N levels and high heat inputs.

72

8.5 Interplay between general trends in steel technology and the technology of V-microalloyed steels

As a final remark we think that it is worth noting that the technology of V-microalloyed asdescribed above is in good accord with many of the prevalent trends in steel technology. Thefollowing general trends in steel technology we find especially worthy of mentioning in thiscontext:

• growth of EAF steelmaking, typified by high N-content, 0.007-0.011%.• limitations of traditional controlled rolling into the non-recrystallization region• cost advantages of recrystallization-controlled-rolling in a high temperature regime• recrystallization-controlled-rolling of long products where a high finishing is an

inevitable consequence of the process• application of the TiN technology for controlling austenite grain coarsening• growth of thin slab casting; in strip and plate steels processed by thin slab casting V is

often the preferred microalloying because of the greater demand of resistance to hotcracking in this process.

73

ACKNOWLEDGEMENTS

This review rests primarily on research carried out during the last 25 years at the SwedishInstitute for Metals Research (SIMR) with economical support by the U.S. VanadiumCorporation/ Stratcor (and its predecessor Union Carbide) and technical cooperation withMichael Korchynsky of that organisation. The personal interest and engagement of MichaelKorchynsky through all these years have been a prerequisite for the fulfilment of this work.The authors are indebted to him for countless technical discussions over the years, sorewarding and indeed essential for the outcome of the work.

We would also like to acknowledge the financing of some of the underlying work by theGeneral Research Programme of SIMR and by SSAB.

Finally it should be noted that many researchers at SIMR, beside the present authors, havecontributed to this review as investigators and authors of individual projects and papers. Inthis context the outstanding work of William Roberts should be given special recognition. Helaid the foundation for the research on microalloyed steels and hot working at SIMR, and hisresearch increased our understanding of the behaviour of these steels significantly and madeimproved alloy design and processing possible.

74

REFERENCES

Chapter 2

2.1 T. Gladman, The physical metallurgy of microalloyed steels. The Institute of Metals,(1997).

2.2 K.J. Irvine, F.B. Pickering and T. Gladman, Grain-refined C-Mn steels. JISI, 205,(1967) pp. 161-182.

2.3 K. Narita, Physical chemistry of the group IVa (Ti,Zr), Va (V,Nb,Ta) and rare earthelements in steel. Trans. ISIJ, 15 (1975), pp. 145-152.

2.4 C. Wagner, Thermodynamics of alloys., Addison-Wesley Co., Reading, Mass.,(1952).

2.5 M. Hillert and J. Staffanson, The regular solution model for stoichiometric phases andionic melts, Acta Chemica Sc. 24, (1970) pp. 3618-3626.

2.6 S. Zajac, Thermodynamic model for the precipitation of carbonitrides in microalloyedsteels. Swedish Institute for Metals Research, Report IM-3566, (1998).

2.7 B. Sundman, B. Jansson and J.O. Andersson, Thermocalc databank system.CALPHAD, 9 (1985) pp. 153-159.

2.8 K. Yvon and E. Parthé, Crystal chemistry of the close-packed transition-metalcarbides. Pt1. Crystal structure of the zeta-vanadium niobium and tatanium carbides.Acta Cryst., 26 (1970) pp. 149-153.

2.9 R.R. Zupp and D.A. Stevennon, The influence of vanadium on the activity of carbonin the Fe-C-V system at 1000°C., correlation of the influence of substitutional soluteson the activity coefficient of carbon in iron-base system. Trans. AIME, 236 (1966)pp. 1316-1323.

2.10 H.A. Wriedt and H. Hu, Chemical Metallurgy, A Tribute to C. Wagner., Proc. Symp.Chicago, Feb. (1981) TSM-AIME, Warrendale, PA (1981) pp. 171-194.

2.11 H. Ohtani and M. Hillert, A thermodynamic assessment of the Fe-N-V System.CALPHAD, 15 (1991), pp. 25-39.

2.12 R.W. Fountain and J. Chipman, Solubility and precipitation of boron intride in iron-boron alloy, Trans. AIME, 224 (1962) pp. 599-606.

2.13 S. Koyama, T. Ishii and K. Narita, Physical chemistry, J. Japan Inst. Metals, 37(1973) pp. 191-195.

2.14 N. Yoshinaga, K. Ushioda, S. Akamatsu and O. Akisue, Precipitation Behaviour ofSulfides in Ti-added Ultra Low-carbon Steels in Austenite. ISIJ Int., 34 (1994) pp. 24-32.

2.15 S. Zajac. and B. Jansson, Thermodynamics of the Fe-Nb-C-N System and theSolubility of Niobium Carbonitrides in Austenite, Metall. Trans. B, Vol. 29B, 1998,pp. 163-176

2.16 F.G. Wilson and T. Gladman, Aluminium nitride in steel, Int. Materials Rev., 33(1988) pp. 221-287.

75

2.17 E.L. Brown and A.J. DeArdo, Aluminium nitride precipitation in C-Mn-Si andmicroalloyed steels, Proc. Conf. Thermomechanical Processing of MicroalloyedAustenite, Eds. A.J. DeArdo, G.A. Ratz and P.J. Wray, AIME, (1982) pp. 319-341.

2.18 N. Takahashi, M. Shubata, Y. Furuno, H. Hayakawa, K. Kakuta and K., Yamamoto,Boron-bearing steels for continuous annealing to produce deep drawing and highstrength steel sheets, Ed. Bramfitt and Mangono, AIME, (1982) pp. 133-153.

2.19 H-R. Lin and G-H. Cheng, Analysis of hardenability effect of boron, Mat. Sci. andTech. 6 (1990) pp. 724-729.

Chapter 3

3.1 S. Zajac, Thermodynamic model for the precipitation of carbonitrides in microalloyedsteels, Swedish Institute for Metals Research, Internal Report IM-3566, Jan. 1998.

3.2 S. Zajac, R. Lagneborg, T. Siwecki, The role of nitrogen in microalloyed steels.Contribution to the Int. Conf. Microalloying “95”, Iron and Steel Society Inc., June11-14, 1995, Pittsburgh, PA, June 1995, pp. 321-340.

3.3 W. Roberts, Recent innovations in alloy design and processing of microalloyed steels.Contribution to the 1983 Int. Conf. on Technology and Applications of High StrengthLow Alloy (HSLA) Steels, Philadelphia, PA, Oct. 3-6, 1983, pp. 67-84.

3.4 T. Siwecki, A. Sandberg, W. Roberts, Processing characteristics and properties of Ti-V-N steels. Contribution to the 1983 Int. Conf. on Technology and Applications ofHigh Strength Low Alloy (HSLA) Steels, Philadelphia, PA, Oct. 3-6, 1983, pp. 619-634.

3.5 S. Zajac, T. Siwecki, W. B. Hutchinson, M. Attlegård, Recrystallization controlledrolling and accelerated cooling for high strength and toughness in V-Ti-N steels,Metall. Trans. A 22A (1991) pp. 2681-2694.

3.6 B. Lehtinen, P. Hansson, Characterisation of microalloy precipitates in HSLA steelssubjected to different weld thermal cycles, Swedish Institute for Metals Research,Internal Report IM-2532, June 1989.

3.7 S. Zajac, T. Siwecki and L.-E. Svensson, The Influence of Plate Production ProcessingRoute, Heat Input and Nitrogen on the HAZ Toughness in Ti-V Microalloyed Steel, inProc. Conf. on Processing, Microstructure and Properties of Microalloyed and OtherModern Low Alloy Steels, ed. A. J. Ardo, TMS, Warrendale, PA, 1991, pp.511-523.

3.8 S. Sarian, Diffusion of Ti in TiC, J. Appl. Physics 40 (1969) pp. 3515-3520.

3.9 M. L. Baskin, V. I. Tretyakoo, and I. N. Chaporova, W diffusion in monocarbides ofW, Ta and Ti in the solid solution TiC WC and TiC WCTaC, Physics of Metals andMetallography 14 (1962) pp. 86-90.

3.10 M. Prikryl, A. Kroupa. G. C. Weatherly, and S.V. Subramanian, PrecipitationBehaviour in a Medium Carbon, Ti-V-N- Microalloyed Steel, Met. Mat. Trans. 27A(1996) pp.1149-1165.

3.11 P. Hellman and M. Hillert, On the Effect of Second-Phase Particles on Grain Growth,Scand. J. Metallurgy, 4 (1975) pp. 211-219.

3.12 T. Siwecki, A. Sandberg, W. Roberts, The influence of processing route and nitrogencontent on microstructure development and precipitation hardening in vanadium-

76

microalloyed HSLA steels, Swedish Institute for Metals Research, Internal ReportIM-1582, Aug.1981.

3.13 L. J. Cuddy, The Effect of Microalloy Concentration on the Recrystallization ofAustenite during Hot Deformation, Proc. Conf. Thermomechanical Processing ofMicroalloyed Austenite, Pittsburgh PA, Aug. 1981, TMS-AIME, Warrendale, PA,1981, pp.129-140.

3.14 J. J Jonas and I. Weiss, Effect of precipitation on recrystallization in microalloyedsteels, Metals Sci., 13 (1979) pp. 238 – 245.

3.15 M. G. Akben, T. Chandra, P. Plassiard and J.J. Jonas, Dynamic precipitation andsolute hardening in a V microalloyed steel and two Nb steels containing high levels ofMn, Acta Met. 29 (1981) pp.111- 121.

3.16 M. J. Luton, Interaction Between Deformation, Recrystallization and Precipitation inNiobium Steels, Metal. Trans. 11A (1980) pp. 411-420.

3.17 J. D. Jones and A. B. Rothwell, Controlled Rolling of Low-Carbon Niobium-TreatedMild Steels, Proc. Deformation under Hot Working Conditions, ISI Publication 108,London 1968, pp.78-82, 100-102.

3.18 A. T. Davenport, L.C. Brossard and R.E.Miner, Precipitation in Microalloyed High-Strength Low-Alloy Steels, J. Metals 27 (1975) pp. 21 –27.

3.19 A. T. Davenport, Proc. Conf, Hot Deformation of Austenite, Cincinnati, TMS-AIME1977, pp.517-536.

3.20 A. J. De Ardo, Modern Thermomechanical Processing of Microalloyed Steel: APhysical Metallurgy Perspective, Proc, Conf. Microalloying ’95, Pittsburgh June 1995,Iron & Steel Soc. pp.15 – 33.

3.21 E. J. Palmiere, Precipitation Phenomena in Microalloyed Steels, Proc. Conf,Microalloying ’95, Pittsburgh June 1995, Iron & Steel Soc. pp 307 – 820.

3.22 I. Kozasu, C. Onchi, T. Sampei and T. Okita, Hot Rolling as a High-TemperatureThermo-Mechanical Process, Proc. Conf. Microalloying ’75, Ed. Korchynsky, NewYork, 177, pp.120 –135.

3.23 T. Siwecki, B. Hutchinson, S. Zajac, Recrystallisation controlled rolling of HSLAsteels. Contribution to the Int. Conf. “MICROALLOYING ‘95”, Iron and SteelSociety Inc., June 11-14, 1995, Pittsburgh, USA, pp.197-212.

3.24 R. Lagneborg, W. Roberts, A. Sandberg and T. Siwecki, Influence of processing routeand nitrogen content on microstructure development and precipitation hardening in V-microalloyed HSLA-steels, Proc. Conf. Thermomechanical Processing ofMicroalloyed Austenite, Pittsburgh, Aug. 1981 Met. Soc. AIME, pp.163 – 194.

3.25 T. Siwecki and S. Zajac, Recrystallization Controlled Rolling and Accelerated Coolingof Ti-V-(Nb)-N Microalloyed Steels, The 32nd Mechanical Working and SteelProcessing Conf, Cincinnati, USA 1990, pp.441– 451.

3.26 S. Gong, The influence of vanadium on the austenite–ferrite transformation inmicroalloyed steels, Swedish Institute for Metals Research, Internal Report IM-1488Oct. 1980.

77

Chapter 4

4.1 F.A. Khalid and D. V. Edmonds, Interphase Precipitation in Microalloyed EngineeringSteels and Model Alloy, Mat. Sci. Techn. 9 (1993) pp. 384-396.

4.2 R. Lagneborg, Private Communication.

4.3 M. Hillert, Phase Equilibria Phase Diagrams and Phase Transformations, CambridgeUniv. Press, 1998.

4.4. S. Zajac, T. Siwecki, M. Korchynsky, Importance of nitrogen for precipitationphenomena in V-microalloyed steels, Int. Symp. on Low Carbon Steels for the 90’s1993 ASM/TSM Materials Week, Pittsburgh, USA, Oct. 17-21, 1993, pp. 139-150.

4.5 R. G. Baker and J. Nutting, The Tempering of a Cr-Mo-V-W and a Mo-V Steel,Precipitation Processes in Steels, ISI Report No 64, London 1969, pp. 1-22.

4.6 R. M. Smith and D. P. Dunne, Structural Aspects of Alloy Carbonitride Precipitationin Microalloyed Steels, Materials Forum 11 (1988) pp.166-181.

4.7 R. W. K. Honeycombe, Fundamental Aspects of Precipitation in Microalloyed Steels,Proc. Int. Conf. on Technology and Applications of HSLA Steels, Philadelphia PA,Oct 1983, ASM, pp.243-250.

4.8 W. C. Johnson, C. L. White, P. E. Marth, P. R. Ruf, S. M. Tuominen, K. D. Wade, K.C. Russell, H. I Aaronson, Influence of crystallography on aspects of solid-solidnucleation theory, Metall.Trans.A, 6A (1975), pp. 911-919.

4.9 J. L. Lee and H. I Aaronson, Influence of faceting upon the equilibrium shape ofnuclei at grain boundaries, Acta Met. 23 (1975) pp. 799 - 808.

4.10 A. T. Davenport and R. W. K. Honeycombe, Precipitation of carbides ataustenite/ferrite boundaries in alloy steels, Proc. Roy, Soc. London 322 (1971) pp.191-205.

4.11 R. W. K. Honeycombe, Transformation from austenite in alloy steels, Metall.Trans.A,7A (1976) pp. 915-936.

4.12 W. Roberts, Hot deformation studies on a V-microalloyed steel, Swedish Institute forMetals Research, Internal Report IM-1333, 1978.

4.13 R. Lagneborg and S. Zajac, to be published.

4.14 P. Li and J. A Todd, Application of a new model to the interphase precipitationreaction in V Steels, Metal Trans. 19A (1988) pp. 2139-2151.

4.15 W. Roberts, A. Sandberg, The composition of V(C, N) as precipitated in HSLA steelsmicroalloyed with vanadium, Swedish Institute for Metals Research, Internal ReportIM-1489 Oct. 1980.

4.16 M. J. Schmutz, Research Report to Vanitec from Massachusetts Institute ofTechnology, 1981.

4.17 W. Roberts, A. Sandberg, and T. Siwecki, Precipitation of V(C,N) in HSLA steelsmicroalloyed with V, Proc. Conf. Vanadium Steels, Krakow, Oct. 1980, Vanitec,pp.D1-D12.

4.18 S. Zajac, T. Siwecki and W. B. Hutchinson, Precipitation phenomena in V-microalloyed 0.15-0.22%C structural steels, Swedish Institute for Metals Research,Internal Report IM-3453 Dec. 1996.

78

4.19 S. Zajac, T.Siwecki, W. B. Hutchinson, R. Lagneborg., Strengthening mechanisms invanadium microalloyed steels intended for long products, ISIJ Int., 38 (1998)pp.1130-1139.

4.20 S. Zajac, T. Siwecki, W. B. Hutchinson, R. Lagneborg, The role of carbon inenhancing precipitation strengthening of V-microalloyed steels. Contribution to theInt. Symp. “Microalloying in Steels: New Trends for the 21st Century, Sept. 7-9, 1998,San Sebastian, Spain, pp. 295-302.

4.21 S. Zajac, The role of nitrogen and carbon in precipitation strengthening of V-microalloyed steels, Swedish Institute for Metals Research, Internal Report 1997.

4.22 S. Zajac, Kinetics of precipitation during hot rolling and cooling in microalloyedsteels, Swedish Institute for Metals Research, Internal Report IM-3568, Jan. 1998.

Chapter 5

5.1. T. Siwecki, A. Sandberg, W. Roberts, and R. Lagneborg, The influence of ProcessingRoute and Nitrogen Content on Microstructure Development and PrecipitationHardening in V-microalloyed HSLA Steels. Thermomechanical Processing ofMicroalloyed Austenite, (Ed. by A.J. DeArdo, G.A. Ratz, and P.J. Wray), TMS-AIME, Warrendale, USA, 1982, pp.163-192.

5.2. T. Siwecki, A. Sandberg, and W. Roberts, Processing Characteristics and Properties ofTi-V-N Steels, Int. Conf. on HSLA Steels – Technology and Application (Ed. M.Korchynsky), ASM, Ohio, USA, 1984, pp.619-634.

5.3. M. Korchynsky, New trends in Science and Technology of Microalloyed Steels, Int.Conf. on HSLA steels -85, Beijing, 1985, ed. by J. M. Gray, Metals Park, OH, ASM,(1986), pp.251-257.

5.4. T. Siwecki and S. Zajac, Recrystallization Controlled Rolling and AcceleratedCooling of Ti-V-(Nb)-N microalloyed steels, 32nd Mechanical Working and SteelProcessing Conference, Cincinnati, OH, 1990, Warrendale, PA, ISS-AIME, (1991),pp. 441-451.

5.5. S. Zajac, T.Siwecki, B.Hutchinson and M.Attlegård, Recrystallization ControlledRolling and Accelerated Cooling as the Optimum Processing Route for High Strengthand Toughness in V-Ti-N Steels. Met. Trans., 22A (1991) pp.2681-2694.

5.6. R.M. Fix, A.J. DeArdo And Y.Z. Zheng, Mechanical Properties of V-Ti MicroalloyedSteels Subject to Plate Rolling Simulations Utilizing Recrystallization ControlledRolling, Int. Conf. on HSLA steels -85, Beijing, 1985, Ed. by J. M. Gray, MetalsPark, OH, ASM, (1986), pp.219-227.

5.7. S.W. Lee, W.Y Choo and C.S. Lee, Effects of Recrystallization Controlled Rolling onthe Properties of TMCP Steel for Ship Building application, Int. Symp. on Low-Carbon Steels for 90`s, Ed. by R. Asfahani et al. , ASM Intern. and M, M&M Soc,Pittsburgh (1993), pp.227-234.

5.8. T. Siwecki, B. Hutchinson and S. Zajac, Recrystallization Controlled Rolling ofHSLA Steels, MICROALLOYING-95, Pittsburgh, PA, 1995, Warrendale, PA, ISS-AIME, (1995), pp.197-211.

5.9. T. Siwecki and G. Engberg, Recrystallization Controlled Rolling of Steels, Thermo-Mechanical Processing in Theory, Modelling & Practice, Stockholm, 1996, ASMIntern. (1997), pp.121-144.

79

5.10. A. Tamminen, Development of Recrystallization Controlled Rolling and AcceleratedCooled Structural Steels, Thermo-Mechanical Processing in Theory, Modelling &Practice, Stockholm, 1996, ASM Internat. (1997), pp.357-368.

5.11. T. Siwecki, G. Engberg and A. Cuibe, Thermomechanical Controlled Processes forHigh Strength and Toughness in Heavy Plates and Strips of HSLA Steels, THERMEC97, Wollongong, Australia, 1997, TMS, Minerals, Metals, Materials, (1997), pp.757-763.

5.12. T. Siwecki, S. Zajac and B. Ahlblom, The Influence of Thermo-Mechanical ProcessParameters on the Strength and Toughness in Direct Quenched and Tempered Boron-Steels (Re>700Mpa), Physical Metallurgy of Direct Quenched Steels, ASM/TMSMaterials Week 92 Conference, Chicago, November (1992), pp.213-230.

5.13. A. Blomqvist and T. Siwecki, Optimization of Hot Rolling and Cooling Parametersfor Long Products, Swedish Institute for Metals Research, Internal Report IM-3101,1994.

5.14. W. Roberts, A. Sandberg, T. Siwecki and T. Werlefors, Prediction of MicrostructureDevelopment during Recrystallization Hot Rolling of Ti-V-Steels, HSLA steels,Technol. and Application, Philadelphia, PA, 1983, ed. by M Korchynsky, Metals Park,OH, (ASM, 1984), pp.67-84.

5.15. T. Siwecki, Modelling of Microstructure Evolution during RecrystallizationControlled Rolling, ISIJ International, 32 (1992) pp.368-376.

5.16. T. Siwecki and B. Hutchinson, Modelling of Microstructure Evolution duringRecrystallization Controlled Rolling of HSLA Steels, 33rd Mechanical Working andSteel Processing Conference, St. Louis, MO, 1991, Warrendale, PA, ISS-AIME,(1992), pp.397-406.

5.17. S. Pettersson and T. Siwecki, Microstructure Evolution during RecrystallizationControlled Rolling of V-Microalloyed Steels. Swedish Institute for Metals Research,Contract Report No 39.190 (1998).

5.18. J.M. Chilton and M.J. Roberts, Microalloying effects in hot rolled low-carbon steelsfinished at high temperatures, Metall. Trans. 11A (1980) p.1711-1721.

5.19. G.F. Melloy, How changes in composition and processing affect HSLA steels, MetalProgr. 89 (1966) pp.129-133.

5.20. R.L. Bodnar and E.J. Watkins, Recent developments in V-steel structural shapes.34th MWSP Conference, ISS-AIME (1992) pp. 279-290.

5.21. F. Kovac, V. Jurko and B. Stefan, Report, UEM SAV Kosice, 1990.

5.22. P.J., Lubensky, S. L. Wigman and D. J. Johnson, High Strength Steel Processing viaDirect Charging using Thin Slab Technology, MICRPOALLOYING 95, Pittsburgh,PA, 1995, Warrendale, PA, ISS-AIME, (1995), pp. 225-233.

5.23. D.N. Crowther, V containing High Strength Strip using Thin Slab Casting, BritishSteel Report (1996).

5.24. M.Korchynsky and J.R.Paules, Microalloyed Forging Steels-State of the Art Review,Intern. Congress and Expo. Detroit, USA, SAE_Paper 890801, (1989), pp.1-11.

5.25. R. Lagneborg, O. Sandberg and W. Roberts, Microalloyed Ferrite-Pearlite ForgingSteels with Improved Toughness, Proc. HSLA Steels` 85, Beijing, (1985), pp.863-873.

80

5.26. R. Lagneborg, O. Sandberg and W. Roberts, Optimisation of Microalloyed Ferrite -Pearlite Forging Steels, in conference proceeding – Fundamentals of MicroalloyingForging Steels, Golden, Colorado, (1986), pp.39-54.

5.27. F. Kovac, T. Siwecki, B. Hutchinson and S. Zajac. Finishing Conditions Appropriatefor Recrystallization Controlled Rolling of Ti-V-N-Steel, Metall. Trans., 23A (1992)pp. 373-375.

Chapter 6

6.1 H. Bruce, Transverse corner cracks in continuously cast microalled steels, SwedishInstitute for Metals Research, Internal Report IM-2568, May, (1990) (In Swedish).

6.2 T. Siwecki, Surface and structure of thin slab and precipitates in hot strip rolled steelmicroalloyed with V and Nb, Swedish Institute for Metals Research ConfidentialReport (1996).

6.3 S. Zajac, T. Siwecki, W.B. Hutchinson and M. Attlegård, Recrystallised controlledrolling and accelerated cooling for high strength and toughness in V-Ti-N-steels,Metall. Trans. A, 22A, (1991) pp. 2681-2694.

6.4 R. Priestner and C. Zhou, Simulation of microstructural evolution in Nb-Timicroalloyed steel during direct hot rolling, Ironmaking and Steelmaking, 22, (1995),pp. 326-332.

6.5 C. Zhou and R. Priestner, The evolution of precipitates in Nb-Ti microalloyed steelsduring solidification and post solidification cooling, ISIJ International, 36, (1996) pp.326-332.

6.6 R. Lagneborg, The effect of grain size and precipitation of carbides on the creepproperties of Fe-20%Cr-35%Ni alloys, J. Iron and Steel Inst., 207, (1969) pp. 1503-1506.

6.7 H.G. Suzuki, S. Nishimura and S. Yamaguchi, Characteristics of hot ductility in steelssubjected to melting and solidification, Trans. ISIJ, 22, (1982) pp.48-56.

6.8 B. Mintz and J.M. Arrowsmith, Hot ductility behaviour of C-Mn-Nb-Al steels and itsrelationship to crack propagation during the straightening of continuously cast strand,Met. Technol., 6, (1979) pp. 24-32.

6.9 B. Mintz and R. Abushosha, Influence of vanadium on hot ductility of steel.Ironmaking and Steelmaking, 20, (1993) pp.445-452.

6.10 B. Mintz and R. Abushosha, The hot ductility of V, Nb/V and Nb containing steels.Materials Science Forum, 2284-286, (1998) pp. 461-468.

6.11 L.P. Karjalainen, H. Kinnunen and D. Porter, Hot ductility of certain microalloyedsteels under simulated continuous casting conditions. Materials Science Forum, ,2284-286, (1998) pp. 477-484.

6.12 D.N. Crowther, M.J.W. Green and P.S.Mitchell, The influence of composition on thehot cracking susceptability during casting of microalloyed steels processed to simulatethin slab casting conditions,. Materials Science Forum, , 2284-286, (1998) pp. 469-476.

6.13 P.L. Lubensky, S.L. Wigman and D.J. Johnson, High strength steel processing viadirect charging using thin slab technology, Proc. Microalloying ’95, Iron and SteelSociety Inc., Pittsburgh (1995) pp.225-233.

81

Chapter 7

7.1. T. Siwecki, Unpublished work - Swedish Institute for Metals Research, (1986).

7.2. P. Hansson and X. Z. Ze, The influence of Steel Chemistry and Weld Heat Input onthe Mechanical Properties in Ti-Microalloyed Steels, Swedish Institute for MetalsResearch, Internal Report IM-2300 (1988).

7.3. P. Hansson, Influence of nitrogen content and weld heat input on Charpy and CODtoughness of the grain coarsened HAZ in vanadium microalloyed steels, SwedishInstitute for Metals Research, Internal Report IM-2205 (1987).

7.4. S. Zajac, T. Siwecki, B. Hutchinson, L-E, Svensson, and M.Attlegård, Weldability ofhigh nitrogen Ti-V microalloyed steel plates processed via thermomechanicalcontrolled rolling, Swedish Institute for Metals Research, Internal Report IM-2764(1991).

7.5. S. Zajac, T. Siwecki, and L-E. Svensson, The influence Plate Production ProcessingRoute, Heat Input and Nitrogen on the Toughness in Ti-V Microalloyed Steel, Intern.Symp. on Low Carbon Steel at Material Week 93, Pittsburgh, USA, Ed. by A. J.DeArdo, TSM, (1993), pp.511-523.

7.6. J.T.Bowker, R.F.Orr, G.E.Ruddle and P.S. Mitchell, The Effect of Vanadium on theParent Plate and Weldment Properties of Accelerated Cooled API 5LX100 Steels,35th.Mech.Working and Steel Processing Conf.,Pittsburgh,(1993)pp.403-412.

7.7. N.E. Hannerz, Weld metal and HAZ toughness and hydrogen cracking susceptibilityof HSLA steels influenced by Nb, Al,V,Ti and N, Proc. from Conf. Welding of HSLA(Microalloyed) Structural Steels, Rome (1976), pp.365-385.

7.8. P.S. Mitchell, W.B. Morrison and D.N.Crowther, The Effect of Vanadium on theMechanical Properties and Weldability of High Strength Structural Steels, LowCarbon Steels for the 90`s, ASM/TSM, Pittsburgh, USA, (1990), pp.337-334.

7.9. P.S. Mitchell, Hart, P.H.M. and W.B. Morrison, The Effect of Microalloying on HAZToughness, MICROALLOING 95, Eds. M. Korchynsky et. al., I&SS, Pittsburgh,USA, (1995), pp.149-162.