High Temperature Shape Memory Alloys: From Materials Design to 

Device ReliabilityDevice Reliability

Ibrahim KaramanIbrahim KaramanDepartment of Materials Science and EngineeringTexas A&M University, College Station, TX 77843

Graduate Students: A. Evirgen, K.C. Atli, J.A. Monroe, J. Ma, C. Hayrettin, O. Karakoc

Collaborators: D.C. Lagoudas, R. Arroyave, T. Baxevanis, D. Hartl, R.D. Noebe, J. Mabe

How SMAs work?

Courtesy of Dr. Y. Liu

Courtesy of R.D. James

Zn45Au30Cu25

How SMAs work?

Courtesy of Dr. Y. Liu

www.adaptamat.com

SMA‐Enabled Applications and History

Variable Geometry Chevron(2005 Flight Test). Noise reduction and fuel economy

Future Applications, supersonic aircrafts (?)

Variable Area Fan Nozzle(2008) Farnborough

2000 2010 2020• Compact (large energy density)

• large shape change• large actuation force

Reconfigurable Rotor Blade ¼ Scale Wind Tunnel Test

• large actuation force• Lightweight• High Temperature• Relatively inexpensive¼ Scale Wind Tunnel Test

4Ma, Karaman, Noebe, Inter Mater Rev (2010)

Adaptive Trailing Edge(2012 Flight Test)

y p• Reliable

High temperature shape memory alloys: defining transformation temperature above 100 °Ctransformation temperature above 100 °C

Practically: something that works at a higher temperature than NiTiPractically: something that works at a higher temperature than NiTi

Ma, Karaman, Noebe, Inter Mater Rev (2010)

osit

ion,

ci

tyde

com

poer

elas

tic

ME

, pha

se d

no

supe

e w

ay S

M

rsib

ility

,n

issu

es,

Onl

y on

e

p, ir

reve

oxid

atio

n O

Cre

ep oMa, Karaman, Noebe, Inter Mater Rev (2010)

Actuation fatigue is a problem - Not reliable

Ni50Ti50

= 150 MPa Atli, Ph.D. Dissertation, Texas A&M University,2012

• Easy dislocations generation during reversible transformation, poor stability• Thermo-mechanical training is needed to stabilize shape memory response!

NiTiX alloys are less prone to actuation fatigue

Atli, Karaman, Noebe, Mat. Sci. Eng. A, 2013

How can we improve the actuation response of HTSMAs?response of HTSMAs?

• Use conventional metallurgy approaches

– Solid Solution Hardening (quaternary additions)• Atli, Karaman, Noebe et al., Met Trans A 2010, Acta Mat 2011, MSEA 2014. 

–Work hardening (thermo‐mechanical processing)• Atli, Kockar, Karaman, et al., Scripta Mat. 2006, J. Nuc. Mat. 2007, Acta Mat. 2010 Acta Mat 2011 Scripta Mat 20112010, Acta Mat. 2011, Scripta Mat. 2011

– Precipitation hardening• Evirgen Karaman Noebe et al Func Mat Letters 2012 Acta Mat 2013Evirgen, Karaman, Noebe, et al., Func. Mat. Letters 2012, Acta Mat. 2013, Scripta Mat. 2013, Acta Mat. 2014, Scripta Mat. 2014, Acta Mat. 2015

–Grain size refinement• Atli, Kockar, Karaman, et al. , Acta Mat. 2010, Acta Mat. 2011, Scripta Mat. 2011a.

T A&M U i it D t t f M h i l E i i

Nano-precipitates in SMAs form hierarchical microstructures

Demonstrated formation of coherent nanoprecipitates in NiTi(Hf,Zr)Texas A&M University Department of Mechanical Engineering

p p ( , )

Ni4Ti3(Ni57Ti43)

Ni-rich NiTi SMA, Ni52Ti4852 48

NiTiHf high Temperature SMAs• The initial

compositions should be Ni-i hrich

• The precipitates

100 m100 nm20 nm 5 nm

precipitates are richer in Ni than the matrix

Ni50Ti50

= 150 MPa = 150 MPa

Ni50.3TiHf20

Under 200 MPa

Significantly improved actuation fatigue response100 Cycles

Aged at 550C, 3 hrs

y

T A&M U i it D t t f M h i l E i i

Nano-precipitates in SMAs form hierarchical microstructures

Texas A&M University Department of Mechanical Engineering

Ni4Ti3(Ni57Ti43)

D i fDesign of nano-precipitates provides new opportunities in SMAs for functional

Ni-rich NiTi SMA, Ni52Ti48

SMAs for functional stability and superelasticity

100 m100 nmNiTiHf high Temperature SMAs FeMnNiAl SMAs

CoNiGa and CoNiAl-based shape memory super alloys

Fine precipitates Large variantsAging at low temperatures or for

Large precipitates Refined variantsAging at high temperatures or for longAging at low temperatures or for

short durations Aging at high temperatures or for long durations

200

150o

550oC

600oC Evirgen, Karaman, et al., Scripta Mat., 2013, Acta Mat. 2014, Scripta Mat. 2014.

100

50

Ms(

o C) 450oC

500oC

FC from 700oC to 100oC

280

240

200550oC

600oC0

100806040200A i Ti (h)

400oCNi50.3Ti29.7Zr20

200

160

120Ms (

o C)

450oC

500oCAging Time (h)

80

40

0302520151050

Ni50.1Ti24.9Hf25

• Great flexibility to modify transformation temperatures (same trend in Af), and hysteresis (dissipative mechanisms)

302520151050Aging Time (h)

600

580

560(ºC

)

160 150

140 130 120

110

110

100

80

90

Region 1

Af

Nanoprecipitates540

520

500

empe

ratu

re 0 110 100

80

70 6

0

8070

80

5090 suppress functional 

degradation480

460

Te

1 10 100 1000

80

Region 2 60

50

5

40 600

580 100 90

80

80 30

20

70Ms

degradation

Time (min)560

540

520ratu

re (º

C) 80

60 40

30

2

10

Region 1

2

70

50

New NiTiHf and  520

500

480

460

Tem

per 30

10

0

0

-20

-40 Region 2 -30

-10

-10-20

1020

NiTiZr alloys are less prone to actuation 

Evirgen, Karaman, Noebe, Acta Materialia, 2015

460

1 10 100 1000Time (min)

0 10prone to actuation fatigue

Work Output of Comparison

103

104

Elect rost rict ive ceramics

Magnet ic SMAs (MSMAs) (Field induced phase t ransformat ion) TWSME in

Ni Ti Hf

102

10

ss (M

Pa)

Magnet ost rict ive mat erials

Shape memory alloys (SMAs)

100

Ni50.3Ti29.7Hf20

101

tion

Stre

s

TWSMEin SMAs

Piezolect ric ceramics

El t t i

10 MJ/ m 3

00 MJ/ m 3 TWSME inNi24.5Ti50.5Pd25

TWSME inNi49 9Ti50 1

10-1

100

Act

uat

Piezolect ric polymers

MSMAs (Field induced variant

Elect roact ive polymers

Shape memory polymers

1 MJ/ m 3

Ni49.9Ti50.1

10-2

0.01 0.1 1 10 100 1000

p y ( reorient at ion)

10 J/ m 3

100 J/ m 31 kJ/ m 3

10 kJ/ m 3

100 kJ/ m 3

Actuation Strain (%)

Atli, Karaman, Noebe, Smart Materials and Structures, 2015

Microstructure generation• Multiple RVEs of SMA matrix and

elastic Ni4Ti3 precipitates

Micromechanics• Periodic boundary conditions

for stress, strain, and Ni-t ti

Coherency stress field• Mismatch between the lattice

parameters of the two phasesconcentration

Ni-depletion (3D Fick’s law)• Representative of precipitate

formation

Thermo-mechanical response prediction Thermal cycle microstructure and obtain response

AppliedLoad

2

3

4

5

6

Strain [%

]

150 MPa

ExperimentPrediction

Ni-content distributionTypical stress field under macroscopic load

‐1

0

1

‐50 ‐25 0 25 50 75 100 125 150

Temp [°C]

Current work• Given solutionized material response and calorimetric data after heat treatment,

Predict precipitated material response

Precipitation RelationsSolutionized Material Response

Response Prediction600

580 160

150 141

110

10

90

Af

2

3

4

5

6

rain [%

]

Experiment

Calibration

580

560

540

520

500mpe

ratu

re (

ºC) 140 130

120

0

110

00

100 80

70 6

0

8070

80

Region 1

0

90

Af

‐1

0

1

‐50 ‐25 0 25 50 75 100 125 150

Str

Temp [C]

480

460

Tem

1 10 100 1000Time (min)

80

70

Region 2 60

50

50

40

0123456

Strain [%

]

‐10

‐50 ‐25 0 25 50 75 100 125 150Temp [°C]

What is next?What is next?• Design of HTSMAs with operating temperatures between 500°C to 800°C – Discovered a new class of materials that works around 600°C

• Better understanding of fatigue and fracture g gmechanisms in HTSMAs

• Accelerated Alloy and Processing DesignAccelerated Alloy and Processing Design• Ultra‐high temperature shape memory ceramics