Microstructure From Processing Evaluation and Modelling Diffusionless Transformations

  • Slides: 39
Download presentation
Microstructure From Processing: Evaluation and Modelling Diffusionless Transformations: Lecture 6 Martin Strangwood, Phase Transformations

Microstructure From Processing: Evaluation and Modelling Diffusionless Transformations: Lecture 6 Martin Strangwood, Phase Transformations and Microstructural Modelling, School of Metallurgy and Materials

Transformation without diffusion • • Is it possible for a phase transformation to occur

Transformation without diffusion • • Is it possible for a phase transformation to occur without diffusion? As temperature decreases however the difference in free energy between austenite and ferrite increases

Transformation without diffusion • Is it possible for phase transformation to occur without diffusion?

Transformation without diffusion • Is it possible for phase transformation to occur without diffusion? • Consider c. p. planes in an fcc structure FCC A A B C HCP C A A B B C C A • Slip on every other {111} gives a full hcp phase

Thermodynamics • • For fcc – hcp transformations the interfacial energies are low and

Thermodynamics • • For fcc – hcp transformations the interfacial energies are low and the volume change minimal Hence little driving force is needed – what is the situation in iron and steels?

Bain strain • Shockley partials accomplish the fcc - hcp transformation, but the fcc

Bain strain • Shockley partials accomplish the fcc - hcp transformation, but the fcc - bcc transformation requires a different shear - the Bain strain Compression along [001] and expansion along converts bct closer to bcc Two fcc units cells (only some atoms shown) Central bct unit cell highlighted

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α Cα Wt

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α Cα Wt % C Cγ C 0

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α’ Cα Wt

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α’ Cα Wt % C Cγ C 0

Thermodynamics • For fcc to bcc (or bct) there is a need for strain

Thermodynamics • For fcc to bcc (or bct) there is a need for strain energy (in the remaining austenite) • The fcc/bcc interface is not coherent and so there will be interfacial energy needed • Thus martensite tends to form when ~ 1000 J mol-1 (depends on alloy composition) – this can be used to predict MS (martensite start temperature)

Growth rate • The interface between parent and displacive product moves at the speed

Growth rate • The interface between parent and displacive product moves at the speed of sound (3 - 7 km/s) in the material so that the transformation is effectively instantaneous once sufficient driving force is acquired • The driving force increases with increased temperature below Ms and is needed to provide interfacial energy and matrix strain energy (fcc and bcc/bct have different molar volumes) • Interfacial and strain energy increase with increased martensite volume fraction so that volume fraction increases with decreasing T; often does not go to completion due to strain concentration in increasingly finer regions along with loss of interface coherency

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature below MS, but not on time held at that temperature • A rational OR between parent and product phases (in ferrous martensites the K-S OR is often shown) – this is not predicted from the Bain strain • A coherent and glissile Habit Plane (often irrational) • The combination of a fixed OR, a coherent Habit Plane and a volume change leads to surface relief (often tent-like relief) • Atomic correspondence (key to shape memory alloys SMAs)

Macroscopic studies • Bain strain + rotation works macroscopically, but microscopically needs to work

Macroscopic studies • Bain strain + rotation works macroscopically, but microscopically needs to work for planes and does not give an IPS • IPSs are seen in shear and twinning and the intersection of two IPSs is an ILS, hence if there is a macroscopic IPS and a microscopic IPS these would give an ILS microscopically (Bain strain + rotation), but an IPS macroscopically

Proposed mechanisms • These give rise to the commonly seen ‘lath’ (dislocation / slip)

Proposed mechanisms • These give rise to the commonly seen ‘lath’ (dislocation / slip) martensites and ‘twinned’ martensites

Strength of ferrous martensites • Why is martensite so hard? • Carbon is trapped

Strength of ferrous martensites • Why is martensite so hard? • Carbon is trapped in solution – this leaves C around dislocations and so pins them effectively • Displacive transformations generally involve dislocations so that the material is effectively heavily work hardened • The structure is very fine (sub-micron) acting as very fine grains • There a fewer easier slip systems in bct than bcc so that it is more difficult for slip as tetragonality increases

Ferrous martensite tetragonality • As [C] increases then the strain in the matrix increases

Ferrous martensite tetragonality • As [C] increases then the strain in the matrix increases leading to increased austenite a and martensite c lattice parameters, but a small decrease in martensite a lattice parameter

Lath martensites (e. g. DP strip steels) • Inhomogeneous slip would be the second

Lath martensites (e. g. DP strip steels) • Inhomogeneous slip would be the second (microscopic) IPS • Often seen for low carbon contents (< 0. 5 wt % depending on overall composition) • {111}γ habit plane (composed of {335} γ and {557} γ)

Lath martensites

Lath martensites

Twin martensites (e. g. engineering steels • • Often seen for medium carbon contents

Twin martensites (e. g. engineering steels • • Often seen for medium carbon contents (0. 5 – 1. 0 wt % depending on overall composition) • {225}γ habit plane Increasing carbon level hinders slip and so the second (microscopic) IPS is now twinning

Twin martensites

Twin martensites

Plate martensites (e. g. tool steels) • Often seen for high carbon contents (>

Plate martensites (e. g. tool steels) • Often seen for high carbon contents (> 1. 0 wt) or alloy steels • {259}γ {3 10 15} γ habit planes

Ferrous martensite strength • As [C] increases the strength of martensite increases before levelling

Ferrous martensite strength • As [C] increases the strength of martensite increases before levelling out • The strength of martensite increases, but the steel strength plateaus as the amount of martensite reaches a constant level

Martensite volume fraction • MS is related to a critical value of driving force;

Martensite volume fraction • MS is related to a critical value of driving force; C is an austenite stabiliser and so reduces MS

Martensite volume fraction • • MS is related to a critical value of driving

Martensite volume fraction • • MS is related to a critical value of driving force; C is an austenite stabiliser and so reduces MS • Other alloying elements also affect MS and this can be predicted by thermodynamic models (Thermo-Calc, MT-DATA) as well as a series of empirical equations, e. g. Higher carbon contents have been associated with lower toughness / ductility and poor weldability leading to a carbon equivalent (max. usually 0. 4 wt %): • The amount of martensite depends on the undercooling below MS

Martensite volume fraction Koistinin-Marburger equation: β ~ 0. 011

Martensite volume fraction Koistinin-Marburger equation: β ~ 0. 011

Martensite volume fraction This gives three temperature regimes • • • MS > MF

Martensite volume fraction This gives three temperature regimes • • • MS > MF > room temperature – full martensite formation • MS > room temperature > MF – partial martensite formation • Room temperature > MS > MF – no martensite formation For the second and third cases sub-zero quenching (quenching to temperatures below room temperature) can lead to more martensite formation

Stress induced martensite and MD • For the condition where • Room temperature >

Stress induced martensite and MD • For the condition where • Room temperature > MS > MF – no martensite formation • Elastic deformation puts strain energy into the austenite that can add to the driving force so that the critical value is exceeded and martensite forms • The limit would be the yield stress of austenite and so the defines a maximum MD – used in rail applications and Hadfield’s manganese steel (~ 12 wt % Mn) • Plastic deformation will increase the dislocation density and can stop martensite formation (mechanical stabilisation in C-Mn steels) or provide martensite nucleation sites (often seen in austenitic stainless steels)

Long range diffusion – interfacial equilibrium ? Displacive transformation diffusionless

Long range diffusion – interfacial equilibrium ? Displacive transformation diffusionless

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature below MS, but not on time held at that temperature • A rational OR between parent and product phases (in ferrous martensites the K-S OR is often shown) – this is not predicted from the Bain strain • A coherent and glissile Habit Plane (often irrational) • The combination of a fixed OR, a coherent Habit Plane and a volume change leads to surface relief (often tent-like relief) • Atomic correspondence (key to shape memory alloys SMAs)

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature

Martensite characteristics • Athermal transformation – amount of martensite formed is dependent on temperature below MS, but not on time held at that temperature (X for bainite) • A rational OR between parent and product phases (in ferrous martensites the K-S OR is often shown) – this is not predicted from the Bain strain (✓ for bainite) • A coherent and glissile Habit Plane (often irrational) • • (✓ for bainite) The combination of a fixed OR, a coherent Habit Plane and a volume change leads to surface relief (often tent-like relief) (✓ for bainite) Atomic correspondence (key to shape memory alloys SMAs) (✓ for bainite substitutional)

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α C α

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α C α Wt % C Cγ C 0

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α’ C α

Driving Force Gibbs free energy, G (per atom) at T 2 GΥ-α’ C α Wt % C Cγ C 0

Thermodynamics • For fcc to bcc (or bct) there is a need for strain

Thermodynamics • For fcc to bcc (or bct) there is a need for strain energy (in the remaining austenite) • The fcc/bcc interface is not coherent and so there will be interfacial energy needed • Thus martensite tends to form when ~ 1000 J mol -1 (depends on alloy composition) – this can be used to predict MS (martensite start temperature)

Thermodynamics • Bainite forms at temperatures above Ms when the driving force is insufficient

Thermodynamics • Bainite forms at temperatures above Ms when the driving force is insufficient for a fully displacive product to be ‘stable’, but at too low a temperature for a fully diffusional transformation Bainite forms initially as a displacive product (i. e. as martensite), but at temperatures where carbon can diffuse (and precipitate) out of the bct product (sub-unit) before this re-transforms back to austenite • • The loss of C from solution reduces the tetragonality of the product reducing strain and interfacial energies so that the critical driving force ( ) is around 400 J mol-1 • The process can then begin again with displacive formation of another saturated sub-unit followed by C diffusion – the sequential displacive – diffusion process now has a time-dependence – even isothermally

Types of bainite UB – coarser carbides can be resolved optically occasionally and is

Types of bainite UB – coarser carbides can be resolved optically occasionally and is sometimes called ‘feathery’; carbide OR is that with austenite LB – etches darkly – in SEM / TEM parallel carbides are seen (OR between cementite and ferrite)

Upper bainite • TEM This consists of laths (lenticular units) of ferrite with carbon-rich

Upper bainite • TEM This consists of laths (lenticular units) of ferrite with carbon-rich material at the boundary

Lower bainite • This consists of laths (lenticular units) of ferrite with carbon-rich material

Lower bainite • This consists of laths (lenticular units) of ferrite with carbon-rich material at the boundary and precipitated within the ferrite lath SEM LB

Lower bainite • TEM This consists of laths (lenticular units) of ferrite with carbon-rich

Lower bainite • TEM This consists of laths (lenticular units) of ferrite with carbon-rich material at the boundary and precipitated within the ferrite lath

Incomplete reaction What happens if carbides do not form? Carbon still diffuses into the

Incomplete reaction What happens if carbides do not form? Carbon still diffuses into the remaining austenite which it will stabilise This reduces the driving force for transformation until it is < 400 J mol -1

Summary Mechanism of bainite formation compared to pearlite and martensite Structures of upper and

Summary Mechanism of bainite formation compared to pearlite and martensite Structures of upper and lower bainite Thermodynamics and kinetics driving bainite formation How could incomplete reaction (austenite retained) be used to generate useful property mixes?