mechanical properties of ti-based glassy and nanocomposite ... · • alloys: ti-13v-11cr-3al,...

Mechanical properties of Ti-based

glassy and nanocomposite alloys

J. Sort Departament de Física

Universitat Autònoma de Barcelona

1

Universitat Autònoma de Barcelona, Spain

A. Hynowska, J. Fornell, E. Pellicer, A. Concustell, S. González, E. Rossinyol,

S. Suriñach, M. D. Baró

OCAS N.V., Zelzate, Belgium

N. Van Steenberge

IFW, Leibniz Institute for Solid State and Materials Research Dresden,

Germany

A. Gebert, J. Das, J. Eckert

Collaborators

2

• Introduction:

General Overview: Composites vs. Metallic Glasses

• Results and discussion:

Case studies:

A) Strain hardening in nanocomposite Ti60Cu14Ni12Sn4Nb10 alloy.

B) Mechanical behavior of Ti40Zr25Ni8Cu9Be18 metallic glass.

C) Mechanical behavior of Ti60Zr10Cu38Pd12 metallic glass.

• Conclusions

Outline

3

4

Requirement of biomaterials (for orthopedic applications)

• High strength and low Young’s modulus (in bone: 4 to 30 GPa) avoid loosening of the implant. • Biocompatibility: host response and the materials degradation

- not toxic elements: Ni, Co, Al, Be, V, … • High corrosion and wear resistance. • Good Osseointegration (surface chemistry, surface roughness and topography)

Conventional

biomaterials Limitations

Stainless steel

(Co-Cr) alloys

- Ni, Co, Cr toxic effect

(dermatitis, carcinogeicity…)

- Too high Young’s modulus

Ti-6Al-4V

- Release of Al and V

(long term health problems),

V is toxic.

- Not high shear strength.

- Limited implants life (10-15 years).

Up to now…

Introduction

5

Why is titanium so much used in biomedical field?

Introduction

• Besides good anticorrosion behavior and biocompatibility:

Strong, yet light weight: Ti is 56% as dense as steel with yield stress twice that

of stainless steel. High strength-to-weight ratio. Density similar to bone.

Flexible: Ti elastic modulus and coefficient of thermal expansion not far from

human bone.

Easily workable: Ti can be machined using conventional metal processing tools.

Others: non-magnetic (allows NMR, no interactions with magnetic fields, 7th

most abundant element in Earth).

6

Introduction

Young’s modulus (GPa)

The Young’s modulus of different implant materials

7

Introduction

Ti-based crystalline materials

• Hexagonal close-packed (hcp), or -Ti, typically found at room temperature.

• Body centered cubic (bcc), or -Ti, typically found above 1156 K.

• Titanium can retain the -phase at room temperature after allotropic

transformations.

8

Introduction -Ti vs. -Ti phase alloys

• and near- alloys: Ti-2.5Cu, Ti-5Al-2.5Sn, Ti-8Al-1V-

1Mo, Ti-5Al-5Sn-2Zr-2Mo, …

• + alloys: Ti-6Al-4V, Ti-6Al-6V-4Sn, Ti-8Al-1Mo-1V,

Ti-6Al-2Sn-2Zr-2Cr-2Mo, …

• alloys: Ti-13V-11Cr-3Al, Ti-10V-2Fe-3Al,TiFe-3.85Sn …

type Ti alloys are getting attention because

of their lower Young’s modulus (E ≈ 55-100

GPa) as compared to type Ti alloys (E ≈

100-150 GPa).

Alloying elements

• stabilizers

Al, O, N

• stabilizers

Mo, V, Nb, Ta, W, Fe,

Mn, Cu, Ni, Cr

•Neutral

Zr, Si, Sn

9

H. J. Rack, J.I. Qazi, Mater. Sci. Engi. C 26 (2006) 1269-1277

• The fatigue limit of ultra-fine grained commercial purity titanium depends strongly on its

microstructure.

• Strengthening of commercial titanium occurs after equal channel angular pressing

(ECAP) in combination with other deformation processes.

Not only the composition but also the microstructure is important!

Introduction

10

• Metallic glasses (MGs) are amorphous metallic alloys

i.e. do not exhibit long-range order.

• Unique properties

• Lower Young’s modulus (elastic softening)

• Large elastic elongation

• Higher strength and fracture toughness

• Promising tribological and wear properties

• High fatigue limits and corrosion resistance

• Applications: • Biomedical

• Electronic devices

• Sporting goods

• Aerospace technologies.

Why metallic glasses? Introduction

11

Introduction

Metallic glasses vs. other materials

• Metallic glasses exhibit high yield strength compared to other materials, but limited

plasticity at room temperature.

• Ti-based metallic glasses exhibit rather large Young’s modulus. Mg-based metallic

glasses show lower Young’s modulus but they are biodegradable and dissolve at high rates

in simulated body fluids.

12

Ti-based metallic glasses

• High strength

• High elastic limit

• Low Young’s modulus

• Excellent corrosion resistance

• Good bioactivity of Ti element

Suitable biomaterials for orthopedic implants

- First Ti-based BMGs contained toxic elements (i.e., Ti-Zr-Ni-Be system)

[A. Peker, W.L. Johnson, US Patent 5, 288, 344 (1994)].

Introduction

13

Introduction

• First Ti-based metallic glasses: Ti-Zr-Ni-Cu-Be

Mei Jinna, PhD Thesis (2009)

These materials can

be fabricated in large

sizes and show

reasonable

compressive

plasticity

BUT

Beryllium is highly

toxic!

14

• First Ti-based metallic glasses: Ti-Ni-Cu base

Introduction

Mei Jinna, PhD Thesis (2009)

• These alloys exhibit similar yield stress as the Ti-Zr-Ni-Cu-Be system, but

plastic strain is much lower (i.e., they are very brittle).

• Moreover, Ni and Cu are not so good in terms of biocompatibility.

• New non-toxic Ti-based BMGs developed in recent years:

Ti-Zr-Cu-Pd-Sn [F.X. Qin et al., Mater Trans. 48 (2006) 515]

Ti-Zr-Cu-Pd [F.X. Qin et al.,Intermetallics. 15 (2007) 1337; S.L. Zhu et al., Mater. Sci. Eng. A 459 (2007) 233].

Introduction

How do metallic glasses deform?

Plastic flow in metallic glasses (MGs) is accompanied by dilatation (i.e., creation

of excess free volume).

Single atomic jumps

Spaepen, Acta Metall. 1977;25:407. Shear transformation zones Argon, Acta Metall. 1979;27:47

Falk and Langer . Phys. Rev. E 1998;57:7192.

Tk

vvkfc

B

ff2

sinh2 0000 kB Boltzmann constant

kf temperature-dependent rate constant

f volume fraction of potential flow units

is the shear stress is the shear strain rate

0 0: activation volume for a flow event and : atomic volume.

:γ

cf = exp (- v*/<vf>) flow defect concentration

Plastic flow equation:

15

16

At room temperature, the excess free volume tends to coalesce into

shear bands, leading to local viscosity drops

Consequences in the nanoindentation experiments:

Serrations (pop-in events) in the loading curves

Inhomogeneous plastic flow occurs for T< 0.8Tg. Premature fracture

occurs, unless prevented by partial nanocrystallization.

Shear bands inside and around an

indent performed on a Ti-based MG

Introduction

• How do metallic glasses deform?

17

Strategies to enhance mechanical properties

• To refine the microstructure of -Ti alloys towards the nanometer

scale (to increase hardness keeping a low Young’s modulus).

• To find new families of metallic glasses, free from toxic/allergic

elements, with good glass forming ability (to manufacture samples with

reasonable sizes) and high hardness combined with low Young’s

modulus.

• To perform suitable heat treatments of existing metallic glasses to

tailor the microstructure and avoid their premature failure (partial

nanocrystallization to form nanocomposites).

Introduction

18

CASE STUDY # 1:

“Hardening mechanisms in a

Ti60Cu14Ni12Sn6Nb10 nanocomposite alloy”

A. Concustell et al., J. Mater. Res. 24 (2009) 3146-3153.

19

Nano-composite alloy:

Micrometer size β-Ti dendrites.

Nanoestructured eutectic matrix.

Good mechanical properties: high strength, large plasticity

AIMS of the WORK:

• Study of the mechanical behaviour by nanoindentation and compression tests:

evidence for strain hardening.

• The contribution of the different constituent phases to the overall strain hardening.

• Find out the microstructural mechanisms responsible for this strain hardening.

Results & Discussion Ti60Cu14Ni12Sn6Nb10

20

Processing: Ti60Cu14Ni12Sn6Nb10 = 3 mm rods

Characterization

X-ray Diffraction (XRD) Phase identification

Scanning Electron Microscopy (SEM) Analysis of the microstructure of the as-cast and deformed specimens.

Transmission Electron Microscopy (TEM) Microstructural characterization and deformation mechanisms

Nanoindentation: A diamond pyramidal-shaped (Berkovich-type) indenter Load control mode; Forces of 1.5 and 500 mN Hardness calculated by Oliver-Pharr method

• As-cast sample • Deformed samples

Compressed to different strain levels

Results & Discussion

Arc melting + Copper mold casting

Ti60Cu14Ni12Sn6Nb10

21

XRD

SEM

Intermetallic

Dendrites

Eutectic

Ti Cu Ni Sn Nb

Dendrites 58.28 10.22 2.74 5.36 23.38

E. matrix 64.35 11.72 10.49 1.86 11.57

E. rod 72.15 10.47 3.07 1.45 12.84

Results & Discussion Ti60Cu14Ni12Sn6Nb10

22

22

Continuous work hardening

Compression tests Nanoindentation

fracture

E = 75 MPa

Yield strength (as cast) = 1400 MPa

Fracture strength = 2200 MPa

Strain rate: 1.8*10-4 s-1

0 2 4 6 8 10 12 140

500

1000

1500

2000

2500

Str

ess (

MP

a)

Strain (%)

Results & Discussion Ti60Cu14Ni12Sn6Nb10

23

23

dendrite eutectic intermetallic

Results & Discussion Ti60Cu14Ni12Sn6Nb10

24

24 Fmax: 1.5 mN

40 60 80 100

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

F (

mN

)

h (nm)

dendrites

eutectic

As cast

40 60 80 100

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

dendrites

eutectic

F (

mN

)h (nm)

12% deformed

Results & Discussion Ti60Cu14Ni12Sn6Nb10

25

In the as-cast state, dendrites

are harder (solution hardening)

than eutectic matrix.

Eutectic matrix strengthens

more than the dendrites as

deformation proceeds.

The hard CuTi2 intermetallic

phase remains unaltered.

0 2 4 6 8 10 12 145,5

6,0

6,5

7,0

7,5

8,0

8,5

Fmax

= 1.5 mN

H den

H eut

H CuTi2

H (

GP

a)

(%)

Results & Discussion

Nanoindentation results Ti60Cu14Ni12Sn6Nb10

26

• Broadening of the XRD peaks: - Grain size refinement in the different phases (grain boundary hardening) - Increase of microstrains

Results & Discussion

X-ray diffraction results Ti60Cu14Ni12Sn6Nb10

27

27

2 0 n m

HRTEM

Eutectic

-Ti

rod

200 nm

dislocations

Inverse

FFT

Dislocation-induced strain

hardening !?

Eutectic

matrix

Results & Discussion TEM results Ti60Cu14Ni12Sn6Nb10

p = 9%

28

Dark field

Evidence for a martensitic transformation

Monoclinic B19’ phase

Austenite Martensite

The B19’ phase is

located at the

eutectic matrix

NiTi B2 phase is almost

supressed at 12% deformation

This phase transformation

probably contributes to the local

hardening of the eutectic matrix.

Results & Discussion

XRD & TEM results

Cubic B2 phase

Ti60Cu14Ni12Sn6Nb10

29

CASE STUDY # 2:

“Mechanical behaviour of Ti40Zr25Ni8Cu9Be18

metallic glass: a nanoindentation study”

J. Fornell et al., Int. J. Plast. 25 (2009) 1540-1559.

30

Metallic glass rods ( = 3 mm) prepared by Cu-mold casting

Characterization

Structural and thermal properties investigated by X-ray diffraction and

differential scanning calorimetry.

Nanoindentation tests: Nanoindenter XP (MTS) and UMIS (Fischer-Cripps Lab.)

• Diamond pyramidal-shaped (Berkovich-type) tip.

• Load control mode: - Range of forces: 4 - 500 mN.

- Range of loading rates: 0.4 – 6.4 mN/s

• Displacement control mode: - Range of loading rates: 2.6 – 20 nm/s

• Hardness evaluated using the method of Oliver and Pharr at the end of load holding segments.

Results & Discussion Ti40Zr25Ni8Cu9Be18

31

dt

dh

h

1

• Pop-in events observed during loading

• Inhomogeneous plastic flow expected since:

49.0gT

RT Deformation map from:

Schuh et al., Acta Mater. 2007, 55, 4067

H (GPa) E (GPa)

Nanoindentation Compression Nanoindentation

6.9 (10 mN) 6.3 (100 mN)

98 105

PMax = 100 mN

Results & Discussion

Nanoindentation results Ti40Zr25Ni8Cu9Be18

32

nCMy 0

Application of the Mohr-Coulomb yield criterion:

194.02sin

2cos

C

CCM

Simulations confirm pressure-sensitive yielding:

• The elastic (Herzian) solution is far from the

experimental data.

• The Tresca criterion overestimates the maximum

penetration depth.

• The Mohr-Coulomb criterion allows for proper

adjustment of the experimental nanoindentation data. Finite element simulations,

Strand7 software, developed by G+D Computing Pty Ltd.

C is the fracture angle (39.5º ≠ 45º)

Results & Discussion

Mohr-Coulomb, not Tresca!

M-C is the internal friction coefficient)

Compression and Finite-element simulations

Ti40Zr25Ni8Cu9Be18

33 33

• Displacement and circumferential stress ( θθ ) contour mappings

• Application of the Mohr-Coulomb

yield criterion results in an extended

plastic zone beneath the indenter

In agreement with:

Narashiman Mech. Mater. 2004,36, 633

• Similar conclusions about yielding of

metallic glasses (obtained from

simulations) by:

Vaidyanathan et al., Acta Mater. 2001, 49, 3781

Anand and Su, J. Mech. Phys. Solids, 2005, 53, 1362

Schuh et al., Acta Mater. 2007, 55, 4067

Results & Discussion

Finite-element simulations Ti40Zr25Ni8Cu9Be18

34

CASE STUDY # 3:

“Mechanical behaviour of Ti60Zr10Cu38Pd12

glassy and nanocomposite materials”

J. Fornell et al., J. Mech. Behav. Biomed. Mater. 4 (2011) 1709-1717.

35

Sample:

Ti40Zr10Cu38Pd12 metallic glass prepared by arc-melting and subsequent

copper mould casting (rods = 2 mm)

Heat treatments:

Annealing was performed for 30 min, in vacuum, at:

Tann,1 = 713 K (Tg < Tann,1 < Tx1)

Tann,2 = 738 K (Tx1 < Tann,2 < Tx2)

Tann,3 = 923 K (Tann,3 > Tx2)

Results & Discussion Ti40Zr10Cu38Pd12

Addition of Nb:

Fabrication of = 2 mm rods with composition (Ti40Zr10Cu38Pd12)1-xNbx ( x = 0, 2, 3, 4)

by suction casting.

36

Mechanical characterization:

– Uniaxial compression tests of the Ti-based bulk metallic glass

and devitrified material (strain rate 1.8·10-4 s-1).

– Nanoindentation tests: UMIS (Fischer-Cripps Lab.)

• Diamond pyramidal-shaped (Berkovich-type) tip.

• Load control mode

• Maximum load: 250 mN

• Finite element simulations of nanoindentation curves using the Strand7

software, developed by G+D Computing Pty Ltd.

Results & Discussion Ti40Zr10Cu38Pd12

Structure and thermal stability:

Structural and thermal properties investigated by X-ray diffraction and

differential scanning calorimetry.

37

• XRD and TEM (SAED pattern)

reveal that the as-cast sample is

fully amorphous

• Glass transition temperature: Tg = 685 K

• Crystallization temperatures: Tx1 = 720 K

Tx2 = 795 K

Amorphous nature and thermal stability of the Ti60Zr10Cu38Pd12 sample

Results & Discussion

XRD and DSC results Ti40Zr10Cu38Pd12

38

105.02sin

2cos

F

FCM

º45º42F

Evidence for the Mohr-Coulomb yield criterion

As-cast alloy

• Upon compression, reasonable

plasticity, fracture at ~ 2.7% total

strain. “Tough” behaviour

expected since J.J.

Lewandowski et al., Phil. Mag.

Lett. 85 (2005) 77

• Serrated flow behaviour

Shear band activity.

• Dimple size in the fracture

surface around 15-20 m.

• Fracture angle 42º. The Mohr-

Coulomb coefficient is therefore

0.105, in agreement with other Ti-

based MGs (J. Sort et al., Int. J.

Plast. 25 (2009) 1540).

Results & Discussion

Compression test

Ti40Zr10Cu38Pd12

39

• The Tresca yield criterion (typical of

polycrystalline alloys) overestimates

the penetration depth.

• The Mohr-Coulomb criterion allows

for proper adjustment of the

experimental nanoindentation data

using:

M-C = 0.105

0 (cohesion) = 0.9 GPa

E (Young’s modulus) = 100 GPa

nCMy 0

Indentation response of as-cast alloy

Results & Discussion

Finite-element Simulations Ti40Zr10Cu38Pd12

40

As-cast TANN1 = 738 K

TANN2 = 923 K TANN2 = 923 K

• The glassy structure of the as-cast

alloy developes into a nanocomposite

material at TANN1 and a fully crystalline

alloy at TANN2

Results & Discussion

XRD and TEM results after heat-treatments Ti40Zr10Cu38Pd12

41

• Relatively high Poisson’s ratio

(some plasticity expected)

• Relatively low Young’s modulus

(ETi-6Al4V = 110 GPa)

• E and G increase after crystallization,

in agreement with other metallic

glasses (“elastic softening”, due to

free volume and the highly cooperative

shear under the action of small stress).

(T.C. Hufnagel et al. PRB 73 (2006) 064204)

E: Young’s modulus

G: Shear modulus

K: Bulk modulus

: Poisson’s ratio

Results & Discussion

Elastic properties vs. Annealing temperature Ti40Zr10Cu38Pd12

42

• The hardness of Ti60Zr10Cu38Pd12 is

larger than for Ti-6Al-4V.

• The reduced Young’s modulus of

Ti60Zr10Cu38Pd12 is lower than for Ti-6Al-

4V.

• Both H and Er tend to increase with

TANN, due to several microstructural

effects:

• Reduction of free volume during

structural relaxation.

• Crystallization of high-strength

phases, such as: CuTi2, CuZr2

(intermetallic phases).

• The wear resistance of Ti60Zr10Cu38Pd12 (H/Er ~ 0.06 - 0.07) is higher than for Ti-6Al-4V (H/Er = 0.04).

Results & Discussion

Mechanical properties vs. Annealing temperature Ti40Zr10Cu38Pd12

43

Influence of Nb addition

X = 0 X = 3 X = 4

Metallic glass Nanocomposite Polycrystalline

Results & Discussion (Ti40Zr10Cu38Pd12)1-xNbx

Nanocrystallization induces a

drastic increase of plasticity

44

• A few issues to be considered (from a mechanical viewpoint):

Search for processing routes to induce strengthening (e.g., grain size

refinement) of nanocomposite materials without compromising the Young’s

modulus.

Search for metallic glasses with non-toxic elements, reasonable sample

sizes, low Young’s modulus, large hardness and enhanced plasticity.

Search for nanocomposite materials (nanocrystals embedded inside

metallic glass matrices) with enhanced plasticity:

In-situ growth of the composite materials (particles dispersed during casting)

Thermally-induced nanocrystallization

Mechanically-driven nanocrystallization

In conclusion …

Acknowledgements

Financial support from the 2009SGR-1292, MAT-2007-61629 and BioTiNet research projects