Transformation of austenite after intercritical annealing
Although the transformation of austenite in DP steel after intercritical annealing is similar to the transformation of austenite after normal austenitizing, two features make this transformation process unique:
First, because the carbon content of the austenite is fixed by the intercritical temperature, the hardenability of the austenite phase varies with intercritical temperature. Thus, at low temperatures where the carbon content of the austenite is high, the hardenability of austenite is high. Similarly, at high temperatures where the carbon content of the austenite is low, the hardenability of the austenite is low.
Second, because the ferrite already pre-exists, transformation γ→α can proceed by epitaxial growth of this old ferrite into austenite with no nucleation step required [SPEI81A].
A whole range of morphologies and a whole range of transformation products can be formed from the austenite phase after intercritical annealing, depending on the annealing temperature, time and cooling rate. The hardenability of intercritically formed austenite is also affected by alloying elements present in DP steels [SPEI81A].
The transformation of the austenite phase into martensite in DP steels occurs at low temperatures so that the ferrite phase must plastically deform to accommodate the volume expansion (2 to 4 percent) arising from the austenite to martensite transformation. As a result, both a high dislocation density and residual stresses are generated in the ferrite phase immediately surrounding the martensite particle. The residual stress patterns are too small a scale to be directly measured, but a theoretical analysis indicates that their maximum value would be of the order of the yield strength of the ferrite (at the Ms temperature) and decay exponentially away from the martensite-ferrite interface [SPEI81A].
The martensite carbon content of the most common DP steels is in the 0.4-0.7 wt. % range, thus either a lath or mixed type martensite can be expected. These changes in morphology reflect the effect of intercritical annealing temperature on the carbon content of the austenite phase and in turn its effect on the Ms temperature [SPEI81A].
Changes in ferrite phase during intercritical annealing and cooling
In cold-rolled steels, recrystallization of the ferrite will occur rapidly and is generally complete before the steel reaches the intercritical annealling temperature, even during the rapid heating encountered on most continuous-annealing lines. Grain growth of the ferrite phase after recrystallization is generally restricted because of the pinning action of the second phase austenite particles.
Changes in the carbon content of the ferrite phase may occur during intercritical annealing:
• The solubility of carbon in the ferrite may be lower at the intercritical temperature than that originally present in the ferrite phase of the as-received material. The solubility of carbon in the ferrite decreases with increasing intercritical temperature, but may also be markedly decreased by increasing the total alloy content of the steel.
• Variations in the cooling rate from the intercritical temperature can also affect the carbon content of the ferrite phase. As the cooling rate is lowered, cementite may precipitate in the ferrite resulting in lower ferrite carbon content [SPEI81A].
Two different types of ferrite may be identified in most intercritically annealed DP steels: the ferrite which is present at the intercritical annealing temperature, “retained ferrite”, and the ferrite which forms from austenite during cooling, “epitaxial ferrite” (Figure I.4). It was shown that there is no structural interface between the two types of ferrite, and that the epitaxial ferrite is an extension of the retained ferrite grains [KORZ82].
The effect of the alloying elements
One of the reasons of adding alloying elements to steels is to increase their hardenability, this is, to delay the time required for the austenite decomposition into ferrite and pearlite (transformations that occur by diffusion). This allows slower cooling rates to produce fully martensitic structures.
In general, the alloying elements that are added to control hardenability do not markedly improve the mechanical properties obtainable in tempered martensite. It has been found that there is a strong dependence of the mechanical properties on the carbon content, whereas variations in the substitutional alloying elements (Cr, Ni, Mo) have apparently a much lower effect on the mechanical properties of steels. For high tempering temperatures, however, they may serve to retard the rates of softening [MEYR01].
Basically there are two ways in which alloying elements can reduce the rate of austenite decomposition. They can reduce either the growth rate or the nucleation rate of ferrite, pearlite or bainite [PORT92].
The main factor limiting hardenability is the rate of formation of pearlite at the nose of the C curve in the TTT diagram. To discuss the effects of alloy elements on pearlite growth it is necessary to distinguish between austenite stabilizers (Mn, Ni, Cu) and ferrite stabilizers (Cr, Mo, Si). Austenite stabilizers depress the A1 temperature (Figure I.1), while ferrite stabilizers have the opposite effect. All of these elements are substitutionally dissolved in the austenite and ferrite.
At equilibrium an alloy element X will have different concentrations in cementite and ferrite, i.e. it will partition between the two phases. Carbide-forming elements, such as Cr, Mo, Mn will concentrate in the carbide, while elements like Si will concentrate in the ferrite. When pearlite forms close to the A1 temperature the driving force for growth will only be positive if the equilibrium partitioning occurs. Since X will be homogeneously distributed within the austenite, the pearlite will only be able to grow as fast as substitutional diffusion of X allows partitioning to occur. The most likely diffusion route for substitutional elements is through the γ/α and γ/cementite interfaces. However, it will be much slower than the interstitial diffusion of carbon and will therefore reduce the pearlite growth rate.
• When X is a ferrite stabilizer there are thermodynamic considerations that suggest that X will partition even at large undercooling close to nose of the C curve. Thus, Si, for example, will increase the hardenability by diffusing along the austenite/ferrite interface into the ferrite.
• When X is an austenite stabilizer such as Ni, it is possible, at sufficiently high undercooling, for pearlite to grow without partitioning. The ferrite and cementite simply inherit the Ni content of the austenite and there is no need for substitutional diffusion. Pearlite can then grow as fast as diffusion of carbon allows. However, the growth rate will still be lower than in binary Fe-C alloys since the non-equilibrium concentration of X in the ferrite and the cementite will raise their free energies, thereby lowering the eutectoid temperature, and reducing the total driving force.
• When X is a strong carbide-forming element such as Mo or Cr, it has been suggested that it can reduce the rate of growth of pearlite, as well as proeutectoid ferrite, by a solute-drag effect on the mowing γ/α interface. These elements also partition to cementite [PORT92].
Influence of alloying elements on Continuous Cooling Transformation (CCT) diagram
Figures I.11 and I.12 give a selection of CCT diagrams [BERA96] showing the effect of carbon and manganese on the transformation kinetics during continuous cooling and on the volume fractions of the various microstructural constituents. In general, increased amounts of alloying elements lower the temperature for the start of the different transformations and reduce the reaction rates. These effects are related to the influence of the solutes on both the nucleation and growth processes. The numbers on the cooling curves represent the volume fractions transformed in the corresponding range, while the circled numbers at the bottom of these curves are the final Vickers hardness values measured on the specimens concerned.
It can be seen that ferrite and pearlite start temperatures are shifted down and to the right for both carbon an Mn alloying elements. Carbon and Mn alloying elements improve steel hardenability.
The effects of alloying elements on ferrite formation
The kinetics of the transformation of austenite to proeutectoide ferrite is retarded in general by the presence of both austenite stabilizers and ferrite stabilizers probably due to effects such as solute drag on migrating interphase boundaries and the rate of diffusion of carbon.
For instance, Mo retards the transformation and it has been suggested that Mo interacts with carbon and reduces the rate at which it can diffuse away into the austenite. Another possible contribution may be done by Mo segregations to ferrite/austenite interphase [RENG85].
The beneficial effect of B occurs, when it is in solid solution in the austenite. B delays the start of ferrite and pearlite transformations. It does not seem to affect the time to complete them, so apparently B has an interaction with austenite grain boundaries. It has been suggested that B segregates at grain boundaries, because it is able to reduce the grain boundary energy, so grain boundary became less effective as nucleation’s sites. B increases hardenability without lowering the Ms temperature. This is valuable, because lowering the Ms causes an increase of the amount of retained austenite [RENG85].
The effects of alloying elements on martensite formation
The quantity of martensite formed depends on the temperature, chemical composition and the nature, degree and rate of deformation. The elements that increase hardenability are Cr, Mo, Ni, Mn and B.
The alloying elements change the martensite transformation start temperature, MS. Andrew’s formula gives the effect of alloying elements on MS temperature in steels [LAFR99]: M S = 539 − 423C − 30.4Mn − 12.1Cr − 17.7Ni − 7.5Mo − 11Si .
MS temperature is in °C, concentrations are in wt%. If the steel is quenched rapidly enough from the annealing temperature, there is no time for the diffusion-controlled decompositions processes to occur, and the austenite transforms to martensite [PORT92]. The cooling rate required to suppress diffusional transformations is called the critical cooling rate. Alloying elements change the critical cooling rate. The critical cooling rate, necessary for the DP steel formation, decreases in a linear manner with increasing equivalent Mn content (Table I.2).
Table of contents :
I.1 Dual Phase steel microstructure formation
I.1.1 Austenite formation during intercritical annealing
I.1.2 Transformation of austenite after intercritical annealing
I.1.3 Changes in ferrite phase during intercritical annealing and cooling
I.1.4 Dual Phase steel microstructure
I.2 Martensite structure
I.2.1 Martensitic transformation
I.2.2 Martensite morphology
I.3 The effect of the alloying elements
I.3.1 Influence of alloying elements on Continuous Cooling Transformation (CCT) diagram
I.3.2 The role of different alloying elements
I.3.3 The effects of alloying elements on austenitising
I.3.4 The effects of alloying elements on ferrite formation
I.3.5 The effects of alloying elements on martensite formation
I.3.6 Segregations in Ingots and Castings
I.4.1 Tempering of ferrous martensites
I.4.2 Stages of tempering
I.4.3 Tempering reactions in DP steels
I.5 The DP steel deformation behaviour
I.5.1 Mechanical behaviour
I.5.2 Continuous yielding behaviour
I.5.3 Tensile strength
I.6 The damage mechanisms in DP steel during the ductile fracture process
I.6.1 Void nucleation
I.6.2 Void growth
I.6.3 Void coalescence
I.7 Microscopic fracture appearance in DP steel
I.8 Damage resistance of DP steel through Hole Expansion (HE)
Microstructures and mechanical properties
II.A Microstructure formation
II.A.1 Chemical composition and initial microstructures
II.A.2 Continuous Cooling Transformation (CCT) diagram for studied DP steel
II.A.3 Determination of intercritical region temperatures
II.A.4 Heat treatments
II.A.4.1 Thermal treatment cycles
II.A.4.2 Direct quenching
II.A.4.3 Rapid cooling and quenching heat treatment
II.B Mechanical properties
II.B.1 As-quenched material
II.B.1.1 Stress-strain curves
II.B.1.2 Mechanical properties evolution
II.B.2 Tempered material
II.B.2.1 Stress-strain curves
I.B.2.2 Mechanical properties evolution with tempering
Fine characterisation of the microstructure
III.1 Autotempering study
III.2 As-quenched microstructure study
III.3 Evolution of microstructure with tempering
III.4 Macrosegregation analysis
Carbon distribution analysis by NanoSIMS
IV.2 Experimental results and discussion
IV.2.1 Investigation of the as-quenched state
IV.2.2 Investigation of carbon distribution after tempering
IV.3 Understanding the carbon distribution
Damage resistance through hole expansion
V.1 Damage resistance of the as-quenched material
V.2 Hole Expansion evolution with tempering temperature
V.3 HE-ferrite fraction correlations evolution with tempering temperature
V.4 Mechanical properties: correlation between HE and UTS
VI.1 Fractography analysis of tensile test specimens
VI.2 Formation of microstructural damage during tensile testing
VI.2.1 Study of the as-quenched samples
VI.2.2 Study of the tempered samples
VI.3 Damage behaviour evolution with tempering
Modeling of DP steel damage behaviour
VII.1 Application of the existing model
VII.2 Extension to include internal martensite damage
Appendix 1: Experimental procedure
A1.2 Heat treatments
A1.3 Microstructure characterization
A1.3.1 Light microscopy
A1.3.2 Quantitative analysis
A1.3.3 Scanning electron microscopy
A1.3.4 Electron probe microanalysis
A1.3.5 NanoSIMS analysis
A1.3.6 Transmission electron microscopy (TEM)
A1.4 Mechanical characterization
A1.4.1 Tensile properties
A1.4.2 Charpy pendulum impact test
A1.4.3 Limiting Hole Expansion ratio, HE
A1. 6 Void analysis
Appendix 2: Charpy impact test
Appendix 3 : Résumé élargi de la thèse en français
A3.I Etude bibliographique
A3.I.1 La microstructure des aciers Dual-Phase (DP)
A3.I.2 La martensite
A3.I.3 Revenu de la martensite
A3.I.4 Revenu dans les aciers DP
A3.I.5 Comportement mécanique des aciers Dual Phase
A3.I.6 Absence de palier élastique dans les aciers Dual-Phases
A3.I.7 L’endommagement lors de la rupture ductile