On the development of theoretical and experimental tools for materials design of high strength steels and cemented

carbides

Annika BorgenstamMaterials Science and Engineering

KTH

Mission of competence center Hero-m 2 Innovation

To develop tools and competence for fast, intelligent, sustainable and cost efficient product development for Swedish industry. Continuous scientific breakthroughs are exploited to enable design of materials from atomistic scales to finished products.

Multi length scale engineering approach

Experimental capabilities

Hero-m 2 Innovation

Hero-m: 2007-2017Hero-m 2 Innovation:2017-2022E.g. 21 million/yearin-kind+cash

www.hero-m.mse.kth.se

The general materials design system

”Materials Genome”databases

ICME - Integrated Computational

Materials Engineering

models

Materials designmethod

Development time for new materials can be decreased from 10-20 years to 3-4 years.

Process

Structure

Properties

PerformanceCohen’s reciprocity

Engineering(Design)

Science(Understand)

Performance

Properties

Structure Translator

CompositionProcessingHeat treatment

Creator

Materials Design:The needed knowledge structure

Translator

Performer

The recipe:

Based on van der Zwaag et al.

SKF Aerospace

Research programme for Hero-m 2i

Materials Design projects in four application areas• Hard Materials (HM)• Powder Based Materials (PM) • High Strength Steels (HSS) • Advanced Stainless Steels (AdvSS).

Generic projects• Ab-Initio • Calphad• Structure Modelling• Structure Characterization • Property Modelling

Materials design projects

• Develop methods for materials design which allows an accelerated development of new materials

• Opportunity to use and test tools and models developed in Hero-m.

• To address highest priority activities for structure/property modeling.

• Educate graduate and undergraduate students in Materials design.

Examples• High strength steels• Cemented carbide

Real system

Linkage system

Creators

- …..- Cold rolling- Austenitization- Quenching- Intercritical annealing- Quenching

Formation of austenite

Start of martensiteformation

Structure

MS Calc

Design of a martensitic TRIP steel

Martensitic formation under applied stress

Fraction of martensite

Formation of carbides

Mss

MS Frac

- Austenite- Ferrite/Martensite

Design goal - Combination of high strength and elongation

Real system

Linkage system

Creators

- …..- Cold rolling- Austenitization- Quenching- Intercritical annealing- Quenching

Start of martensiteformation

Structure

MS Calc

Design of a martensitic TRIP steel

- Austenite- Ferrite/Martensite

Design goal - Combination of high strength and elongation

Thermodynamics-based modeling of the martensite start temperature

• There is a variety of models in the literature: (i) Empirical models, (ii) Neural network models and (iii) thermodynamics-based models

• Model driving force for martensitic transformation:

• −∆𝐺𝑚 = 𝐺𝑚𝐹𝐶𝐶 −𝐺𝑚

𝐵𝐶𝑇

•

−∆𝐺𝑚 = 𝐺𝑚𝐹𝐶𝐶 −𝐺𝑚

𝐵𝐶𝑇

T0

G

MS

martensite

austenite

−∆𝐺𝑚

Borgenstam et al., 1997

Advanced genome databases

Model Model

1 2 3, , ...

Entries (numbers)α α αa a a 1 2 3, , ...

Entries (numbers)β β βa a a

Calculated physical quantities

Model..

...

Entries (numbers)

Raw data from experiments, computations and general experience

New thermodynamic descriptions

Binaries and ternaries (finished and ongoing work):

Steel: Fe–Cr, Fe–Ni, Cr-Ni, Fe-Cr-Ni

Fe–Mn, Mn–C, Fe-Mn–C, Fe-C

Al–Fe, Al–C, Al–Mn

Cemented carbides: Co–Cr, Co-C, Cr–C, C-Co-Cr

Unaries: Cr, Ni, Mn, Co, C, Al, hcp-Fe •Better descriptions at low T•More physcially based models -easier to link to ab initio•Improve extrapolations for metastable states•Improve description of ordering•Improve magnetic description

Bigdeli et al., 2016 Naghari et al., 2014

Revised magnetic model

• Use separate R-K polynomials for each magnetic state for each phase

• No contribution to Gibbs energy when T is negative

• Use effective/local magnetic moment and not mean magnetic moment.

Xiong et al., 2012

Calculation of start temperature of martensitic structure

Stormvinter et al., 2012

∆𝐺𝑚(𝐿𝑎𝑡ℎ)𝛾→𝛼

= 3640 − 2.92𝑀𝑆 + 346400𝑥𝐶

2

1−𝑥𝐶− 16430𝑥𝐶 − 785.5𝑥𝐶𝑟 + 7119𝑥𝑀𝑛 −

4306𝑥𝑁𝑖 + 350600𝑥𝐶𝑥𝐶𝑟

1−𝑥𝐶, [J/mol]

C

CrCNiMnCrC

C

Cs

x

xxxxxx

x

xMG

144170051043574297011500

1750002100

2plate

m

Effect of thin-film austenite grain size on Ms

m

exsurfF 1 exp( C ΔG A )

0

S

11

mex ch ch m

T surfM

ln 1 FΔG ΔG ΔG A

C

Dispersed g:isolated, limited autocatalysis similar to particles

F: transformed number fraction of grainsAsurf: surface area

transformed

untransformed

0.18C–5.08MnIA 680°C 5min,Quenched to -60°C

Chen et al. Acta Metall. 1985

1μm

Huyan et al., 2018

Effect of austenite grain size on Ms

19

Influence of g size (area):

ex ch chT Ms

1

4crossΔG ΔG ΔG 415.1A

1

1m

ex msurf

ln 1 FΔG A

C

Yang & Bhadeshia, Scr. Mater. 2009Jimenez-Melero et al., Acta Mater. 2009van Bohemen, Morsdorf, Acta Mater. 2017

Huyan et al., 2018

Real system

Linkage system

Creators

- …..- Cold rolling- Austenitization- Quenching- Intercritical annealing- Quenching

Structure

Models needed for design of a Martensitic TRIP steel

Fraction of martensite

MS Frac

- Austenite- Ferrite/Martensite

Fraction of martensite formed below Ms

𝑓 𝑀% = 100 −100

1 + 0.05 ∗𝛥𝐺100

b𝛥𝐺 = 𝛥𝐺 𝑇 − 𝛥𝐺 Ms

b=0.006*Ms+1.205

Huyan et al., 2014

● Extend existing models to provide a semi-empirical prediction of Ms with a solid foundation in CALPHAD-thermodynamics

● Fe-based binary and some ternary systems for: Fe,Cr,Ni,Mn,C,Cu,Co,N,Mo,Al,Si,V,W,Ti,Nb

● Lath, plate and epsilon martensite● Ms, Mf and phase fraction including effect of austenite grain size

Martensite formation model for steels currently under development at Thermo-Calc Software

Lath Martensite:

Plate Martensite:

Thermo-Calc property models-Results from Ms model

Heat map of Ms Contour plot of Ms- Lath, Plate

Mass%

Cr

Mass% C

Mass%

Cr

Mass% C

Jeppsson et al., 2017

Real system

Linkage system

Creators

- …..- Cold rolling- Austenitization- Quenching- Intercritical annealing- Quenching

Formation of austenite

Structure

Models needed for design of a Martensitic TRIP steel

Formation of carbides

- Austenite- Ferrite/Martensite

Modelling of austenite growth during intercritical annealing of martensite

Retained γ

thin filmCementite

Only influence of cementite on austenite formation is considered, not aiming for exact cementite formation and dissolution

DICTRA:

650oC 60 scementite rods

Fe-0.2C-4.72Mn (wt%) Luo et al. Acta Mater. 59 (2011) 4002

Martensitelath

Growth of retained γ thin films Nucleation of γ at θ/α

Huyan et al., 2017

10-7

10-5

10-3

10-1

101

103

105

107

0.0

0.1

0.2

0.3

0.4

0.5

0.6

setup Bsetup A+M

setup A

0.2C-4.7Mn (wt%)

923 K

1γ199α

1γ198α1θ

1γ198α1θ+M

1θ0γ199α

EXP

Auste

nite

Fra

ctio

n

Time [s]

setup O - γα 0.000

0.005

0.010

0.015

0.020

0.025

0.030

10-7

10-5

10-3

10-1

101

103

105

107

0.00

0.02

0.04

0.06

0.08

0.10

Time [s]

We

igh

t F

ractio

n o

f M

nW

eig

ht

Fra

ctio

n o

f C

1γ199α

1γ198α1θ

1γ198α1θ+M

1θ0γ199α

EXP

0.2C-4.7Mn (%)

923 K

Fraction of γ as function of timeChange in C and Mn content in gwith time

Modelling of austenite growth during intercritical annealing of martensite

Huyan et al., 2017

Microstructure design tools

• Equilibrium structures and driving forces for non equilibrium (thermodynamics)

- Stresses (external and internal)

- Interfaces

• Phase field

• Prisma

• DICTRA

Stormvinter et al.

Ye et al.

Phase-field simulations of martensite compared to experiments

Yeddu et al.

Malik et al.

Kolmskog et al.

Effect of Tensile Loading- 3D

Austenite

Martensite-V1

Martensite-V2

Martensite-V3

x

y

z

Phase Field ModelingMalik et al., 2013

750 MPa500 MPa250 MPa

Cemented carbides

• Used for rock drilling and metal cutting applications

• WC + metallic binder (usually Co)

• Co is toxic, longtime exposure may cause serious health problems.

• Co may become banned in EU.

• Expensive and uncertain raw material supply

• Substitute Co

• Using materials design to tailor properties for metal cutting and rock drilling applications

Real system

Linkage system

Translators

- Binder volume fraction

- WC grain size

- Binder & overallcomposition

Hall-Petch

Work hardening

Solid solution

hardening

Precipitation hardening Martensite

formation

Hardness

Grain Growth model

MS Calc

Models needed for design of alternative binders in cemented carbides

Toughness

Designing cemented carbides with respect to hardness

10/14/2018

32

• What is the hardness of such composite material?

• Parameters:

- Hardness of the binder

- Hardness of the carbides

- Constrained effects

- Size of the non-spherical carbides

- Binder chemistry

- Diffusivities

- Sintering time

- Sintering temperature

- …

ICME framework for designing cemented carbides with respect to hardness

33

Binder chemistry - Equilibrium

Binder volume fraction

Binder Hardness 𝑯𝑩

Primary Carbides

- WC grain size

Metallic Binder

Mean-free

binder path 𝝀

Intrinsic Hardness 𝑯𝟎

Solid Solution Strength. 𝑯𝑺𝑺𝑯

Carbide Hardness 𝑯𝑾𝑪

Cemented Carbide Hardness 𝑯𝑪𝑪

TC

TC

Exp./Carbide growth Model

Model

Exp. Model

Cummulative from 𝑯𝟎+𝑯𝑺𝑺𝑯+𝑯𝑴𝒔Exp./Hall-

Petch

Model

Ms-Temp

Ms-Calc

Martensitic Hardness

Contribution 𝑯𝑴𝒔

Model

Binder chemistry - Kinetic

DICTRA

Constrained Binder &

Thermal stress

FEM/Exp.

GeneticAlgorithm

Binder Hardness

10/14/2018

34

acr(c,graph)acr(c,graph)

Tungsten and Carbon binder solubility [at%]

wt%

Ni /

(wt%

Ni+

wt%

Fe)

Binder chemistry - Equilibrium

Binder chemistry - Kinetic

DICTRA

wt%

Ni /

(wt%

Ni+

wt%

Fe)

Cummulative from 𝑯𝟎+𝑯𝑺𝑺𝑯+𝑯𝑴𝒔

acr(c,graph)

Binder Hardness 𝑯𝑩

Mean-free

binder path 𝝀

Intrinsic Hardness 𝑯𝟎

Solid Solution Strength. 𝑯𝑺𝑺𝑯

Model

Exp. Model

Ms-Temp

Ms-Calc

Martensitic Hardness

Contribution 𝑯𝑴𝒔

Model

Constrained Binder &

Thermal stress

FEM/Exp.

Binder Hardness [𝑯𝑽]

Walbrühl et al., 2017

E.g. solution hardening (use the ”compound energy formalism”). For

!parameters adjustable as taken are they Here

constants. elastic andparameter lattice in mismatch of

ncombinatio a represent parameters the and

models classical In

stress Yield

:unit Formula

A

mn

AyyyAyyy

yy

VaNCMM

ik

i

m

k

k

iijk

ij

k

n

ji

k

SSH

y

SSH

yyij

ij

jiy

b

3/2

),,...)((

'''''''''

'''

21

s

sss

Modelling of solid solution strengthening in multicomponent alloys, HSSH

Walbrühl et al., 2017

Modelling of solid solution strengthening in multicomponent alloys, HSSH

0

100

200

300

400

500

600

700

800

900

1000

0 100 200 300 400 500 600 700 800 900 1000

Measured Hardness (Vickers)

Classical Alloys

HEA

Calc

ula

ted h

ard

ness (

Vic

kers

)

Measured hardness (Vickers)

Walbrühl et al., 2017

Carbide hardness

10/14/2018

37

Primary Carbides- WC grain size

Carbide Hardness 𝑯𝑾𝑪

Exp./Mean-field Kampmann-Wagner

approach

Exp./Hall-Petch500

1000

1500

2000

2500

3000

0 20 40 60 80 100

Tungsten Carbide (WC) Hardness [HV]

Binder chemistry - Equilibrium

WC grainsize [µm]

Bonvalet et al., 2018

Composite hardness and its optimization

10/14/2018

38

acr(c,graph)

wt%

Ni /

(wt%

Ni+

wt%

Fe)

Composite Hardness [HV]

Binder

WC

Carbide Hardness

𝑯𝑾𝑪

Binder Hardness

𝑯𝑩

• All the models are integrated in a common platform (Python)

• Input:

- Kinetic and thermodynamic databases (TC)

- Compositions

- Sintering time and temperature

- Initial grain size distribution (before sintering)

• Output:

- The composite hardness

• The optimization of the design parameters is performed through a genetic algorithm scheme

Cemented Carbide Hardness 𝑯𝑪𝑪

Design parameters

Bonvalet et al., 2018

Thank you for your attention!

Researchers involved: Dr Manon Bonvalet, Prof Gustav Amberg, Ass Prof Peter Hedström, Fei Huyan, Dr Lars Höglund, Dr Pavel Korzhavyi, Dr Hemanth Kumar, Dr Amer Malik, Reza Naraghi, Prof Malin Selleby, Dr Albin Stormvinter, Dr Peter Kolmskog, Dr Ye Tian, Dr Johan Jeppsson, Dr Wei Xiong, Dr Jiayi Yan, David Linder, Dr Martin Wahlbrühl and Prof John Ågren