METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

1

VLIV VODÍKU NA MIKROSTRUKTURNÍ VLASTNOSTI SLITIN NA BÁZI Ti

EFFECT OF HYDROGEN ON MICROSTRUCTURE FEATURES OF Ti-BASED ALLOYS

Monika Losertová, Vladimír Dostál, Ivo Szurman

VŠB-Technical University of Ostrava, faculty of Metallurgy and Material Engineering, 15 av. 17.listopad, 708 33 Ostrava, Czech Republic,mlosertova@vsb.cz

Abstract

Study of hydrogen effect on microstructure and mechanical features of TiAl- and NiTi-based alloys was performed. Specimens of polycrystalline alloys were observed in following states: as-received and after heat treatments in different atmospheres. The annealing in flowing argon or hydrogen gas was realised at 750°C for 4 hours in case of Ti-46Al-1.5Mo-0.2C and Ti-44Al-8Nb alloys and at 850°C for 1 hour in case of NiTi, NiTi-Cu and NiTi-Fe alloys. Microstructure was studied using light and scanning electron microscopies. Measurements of microhardness completed the metallographic study of Ti-Al based alloys. The NiTi wires were tensile tested in air at room temperature and strain rate of 3 x 10-4 s-1 and fracture surfaces were observed using SEM. AFM topography of martensite plates in NiTi as well as of lamellar structure in TiAl studied in order to determine the hydrogen influence. The quantity of hydrogen was determined by gas desorption at 700°C in vacuum desorption apparatus. Abstrakt

Bylo provedeno studium vlivu vodíku na mikrostrukturní a mechanické vlastnosti slitin na bázi TiAl a NiTi. Vzorky polykrystalických slitin byly studovány v různém stavu mikrostruktury: dodaném nebo žíhaném v průtoku argonu nebo vodíku. Vzorky byly tepelně zpracovány při teplotě 750 °C po dobu 4 hodin u slitin Ti-46Al-1.5Mo-0.2C a Ti-44Al-8Nb, respektive při 850 °C po dobu 1 hodiny u slitin NiTi, NiTi-Cu a NiTi-Fe. Pozorování mikrostruktury TiAl-slitin bylo provedeno pomocí optické a elektronové rastrovací mikroskopie a doplněno měřením mikrotvrdosti. Dráty NiTi-slitiny byly zkoušeny v tahu při pokojové teplotě a rychlosti deformace 3 x 10-4 s-1 na vzduchu lomové plochy byly studovány pomocí SEM. Topografií AFM martenzitické struktury NiTi, jakož i lamelární struktury TiAl slitin byl sledován vliv vodíku na mikrostrukturu. Množství vodíku bylo stanoveno měřením v desorpčním vakuovém zařízení při teplotě 700 °C.

Keywords: TiAl-based alloy, NiTi-based alloy, hydrogen effect, hydrogen embrittlement.

1. Hydrogen in metals and alloys Hydrogen atom is an element with the lowest atom diameter of all periodic system

elements. Hydrogen presence in materials, especially in metals and alloys, could significantly affect material properties. Titanium is highly reactive element and can react at the surface with water or water vapour under the formation of TiO2 and of atomic hydrogen. The following equation is used for a general explanation:

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

2

xMe + y H2O → MexOy + 2yH (1) where Me … Al, Ti, Si, V

By this reaction hydrogen atom can penetrate into the material and lead to embrittlement

by a reduction of the atomic forces in the lattice. Furthermore, brittle crack propagation and fracture along grain boundaries and/or cleavage planes occur. Also, ordered intermetallics are well-known to be susceptible to hydrogen embrittlement in hydrogen gas [1,2], especially when they contain the elements such as Ni or Fe [3,4], which can act as catalysts for hydrogen dissociation.

1.1. Hydrogen in γγγγ-TiAl based aloys

Intermetallic alloys based on γ-TiAl have many potential applications in the 600-800°C temperature range in the aerospace, automotive and turbine power generation industries thanks to its excellent properties, such as low specific weight and high temperature mechanical properties. However, in all these applications a considerable amount of water vapour (up to 20 vol.% H2O) can be encountered in the service environment. TiAl easily absorbs hydrogen under ambient conditions (eq.(1)) and thus suffers hydrogen embrittlement. The embrittling influence of hydrogen on TiAl as well as on Ti3Al was confirmed in a number of investigations at room temperature.

Furthermore, it has been reported that hydrogen embrittlement sensitivity of TiAl alloys strongly depends on microstructure that could be formed by near γ, duplex (DP), near lamellar (NL) and fully lamellar (FL). Alloys with two-phase structures consisting of the major γ (TiAl) and minor α2 (Ti3Al) phases are the most intensively studied materials. Normally, the presence of the lamellar microstructure would be expected to provide easier diffusion paths for hydrogen through the lamellar boundaries, although, in earlier work with the same materials [5], the values of the apparent diffusion coefficients calculated indicate that lattice diffusion is the preferred diffusion path for hydrogen in titanium. 1.2. Hydrogen in NiTi based aloys

NiTi-based alloys are used as important commercial shape-memory materials for applications such as fine medical wires, electrical switches, pipe couplings, electronic connectors etc. NiTi-based alloys could be deteriorated by environmental embrittlement like other intermetallic compounds. Hydrogen penetration according to eq.(1) into NiTi-based alloys attached with its high diffusivity induces NiTiHx hydrides formation (x = 1.0 and 1.4) [6].

On the one hand, the hydrogen could affect mechanical properties (hydrogen-induced embrittlement), the harmful phenomenon occurs when the amount of absorbed hydrogen exceeds over 50 mass ppm. On the other hand, hydrogen atoms influence the martensitic transitions of NiTi-based alloys as well as the twin boundary mobility, which is responsible for the high damping capacity of these materials. In this case, the hydrogen effect leads to internal friction peak PA-M about 7.5 times higher than structural damping level [7,8].

It is thus of interest to determine the hydrogen behaviour in these alloy systems, where although many investigations have been carried out, some mechanisms of hydrogen effect remain ambiguous. This paper reports the results of microstructure and mechanical properties study of fully lamellar TiAl-based alloys and NiTi-based alloys before and after annealing in flowing argon or hydrogen gas.

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

3

2. EXPERIMENT

TiAl-alloy rods of nominal composition (in at.%) Ti-46Al-1.5Mo-0.2C and Ti-44Al-8Nb (Table 1.) were prepared using plasma melted master alloys and vacuum induction melting. Disc specimens cut from the rods were annealed at 750°C for 4 hours and furnace-cooled down to 400 °C and then air-cooled to room temperature. Heat treatment was performed in flowing argon or hydrogen gas in order to compare the influence of hydrogen on microstructure.

Metallographic study was carried out on the specimens polished using SiC papers P60-P1200 and diamond paste of 3 and 1 µm. Fully lamellar microstructure was revealed by etching agent of the following composition: HF+ HNO3 +H2O (10: 5: 85).

The microstructure after different heat treatments was studied by optical microscopy (OM) and scanning electron microscopy (SEM), the composition alloys were determined by EDAX Philips XL30. The grain size was examined.

Microhardness measurements were carried out from the edge to the middle of the specimens by step of 1 mm. Measurement results were drawn in the graph. The quantity of hydrogen was determined by gas desorption at 700°C in vacuum apparatus.

Ingots of NiTi and NiTi-Me (Me= Cu or Fe) based alloys with nominal composition as summarised in Table 2 were prepared by vacuum induction melting using carbon crucible. The material was hot forged to wires with diameter of 2.6 mm. All specimens were annealed at 850°C for 1 hour in flowing argon or hydrogen gas in order to compare the influence of hydrogen on microstructure. Wires of Z composition with length of 10 mm were tensile tested at room temperature and strain rate = 3 x 10-4 s-1 in air. The hydrogen contents were determined by gas desorption at 700°C in vacuum apparatus.

Transformation temperatures Ms of specimens measured by resistometric method are only of informative nature. Tensile tests were carried out on wires with operating length of L0

= 50 mm at room temperature and engineering strain rate of 3x10-4 s-1 in air. Microstructure was observed on longitudinal sections of specimens using OM. The

polished specimens were etched in solution of HF+HNO3+CH3COOH (1:5:5). Feature of fracture surfaces was studied using SEM Philips XL30. Martensite

microstructure observation was completed with topography of martensite plates using atomic force microscopy (AFM). 3. RESULTS AND DISCUSSION 3.1 TiAl-based alloys

Microstructure of TiAl-1.5Mo-0.2C specimens in the state as received was consisting of elongated dendritic grains having diameter of 47 µm, length of over 200 µm (Table 1.) and fully lamellar structure with γ+α2 phases (Fig.1.a). Microhardness reached the average value of 626 HV0.05.

Heat treatment in flowing argon or hydrogen allowed growing of grain sizes to 79 or 67 µm, respectively (Table 1., Fig.1. b, c). The effect of hydrogen on the microhardness was not evidential because the grain growth due to annealing led to lowering of microhardness values in the both cases. Slightly higher average value (527 in comparison with 516 HV0.05) for hydrogenated state may not be related directly to hydrogen presence. Quantity of hydrogen in the structure of TiAl-1.5Mo-0.2C was not evaluated. However, according to study in [1] the penetration and diffusion of hydrogen could be influenced by possible formation of carbides in the microstructure being deep traps for hydrogen atoms.

Contrariwise, microhardness values of Ti-44Al-8Nb alloy hydrogenated increased even thought increasing grain size compared to specimens as-received and heat treated in argon

METAL 2007

22.-24

.5.200

7, H

rade

c na

d M

orav

icí

____

____

____

____

____

____

____

____

____

____

____

____

____

____

____

____

____

____

___

4

Fig. 1. O

ptical m

icro

grap

hs of Ti-46

Al-1.5M

o-0.2C

alloy

: a) w

itho

ut hea

t treatm

ent b

) after an

nealin

g at 750

°C for

4 h in

flo

win

g ar

gon

and c

) af

ter an

nealin

g at 750

°C for

4 h in

flowin

g hy

drog

en.

a b

c

Fig. 2. O

ptical m

icro

grap

hs of Ti-44

Al-8N

b allo

y: a) witho

ut hea

t tre

atm

ent b

) after an

nealin

g at 750

°C for

4 h in

flo

win

g ar

gon and

c) after ann

ealing

at 7

50 °C for

4 h in

flo

win

g hy

drog

en.

a c

b

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

5

Table 1. Composition and features of TiAl- based alloys Specimen Composition

[at. %] Heat treatment Microhardness

HV 0,05 Grain size

[µm] Without 626±51 EL: th. 47, l.>200

750 °C, 4h, in Ar(g) 516±58 EL: th. 79, l.>200 8 Ti-46Al-1.5Mo-0.2C 750 °C, 4h,in H2(g) 527±79 EL: th. 67, l.>200

Without 644±33 EQ: 53 750 °C, 4h, in Ar(g) 601±31 EL: th. 52, l.>200 9 Ti-44Al-8Nb 750 °C, 4h, in H2(g) 657±37 EL: th. 150, l.>200

EQ = equiaxed grain, EL = elongated grain: th. = thickness, l.=length (Table 1.). Change of the grain morphology from equiaxial to elongated microstructure was observed after annealing whereas the hydrogenated microstructure contained three times greater grains than this one as-received (Fig.2.). Nevertheless, the average microhardness value was higher for hydrogenated specimen (Table 1. and Fig.3.) that being in conflict with Hall-Petch’s relation. Thus, it shows indirectly hydrogen presence and its influence on the mechanical properties of the Ti-44Al-8Nb alloy. The grain growing more important during hydrogen treatment couldn’t be explained reliably. Hydrogen content in the structure of the Ti-44Al-8Nb was evaluated to be as high as 77 mass ppm.

It is possible that hydrogen penetration into the structure would differ for the both type of TiAl-based alloys due to the alloying elements.

Microstructures of etched NiTi and NiTi-Me specimens (Table 2.) were consisting of martensitic structure with homogenous or heterogenous size and distribution of martensitic plates (Fig.4.). The NiTi2 precipitates in B19‘ martensite was observed using light and atomic force microscopies (Fig.5.) and analysed by EDAX in the Z specimen [9,10], while the Ni3Ti4 in B19‘ martensite was detected by TEM study [10] in D specimen.

Deformation under tension of the Z specimens in martensitic state proceeded via martensite variant reorientation and mechanical characteristics including

Nominal composition

[at %] Martensite feature

Alloy Ni Ti Cu Fe

Ms [°C]

Type

A 49.8 50.2 - - 0 B19‘ C 50.25 49.75 - - -50 B19‘ D 50.6 49.4 - - +50 B19‘ E 49.0 50.0 1.0 - 0 B19‘ F 49.0 50.0 - 1.0 -70 R Z 48.5 51.5 - - 61 B19‘

626 644

516

601

527

657

0

100

200

300

400

500

600

700

800

Microhardness HV 0.05

without heat

treatment

heat treatment

in argon

heat treatment

in hydrogen

Ti-44Al-8NbTi-46Al-1.5Mo-0.2C

Fig.3. The average microhardness values for different heat treatment: as-received without heat treatment, annealed at 750 °C for 4 hours in argon and annealed at 750 °C for 4 hours in hydrogen.

Table 2. Nominal composition and Ms temperatures of experimental NiTi and NiTi-Me alloys.

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

6

Fig. 4. Optical micrographs of A alloy: a) as-received – heterogenous distribution of coarse and fine martensite plates, b) annealed in hydrogen - appearance of fine and long martensite plates.

a b

Fig. 5. AFM images of martensite structures of Z specimens a) non-hydrogenated and b) hydrogenated.

a b

second yield point σP were determined (Table 3.). The tensile stress-strain curves may be divided into several deformation stages as indicated in the Fig.6. The initial deformation in stage I is elastic and precedes the apparent non-linear deformation. Stage II is an inelastic deformation corresponding to either stress-induced martensitic transformation or martensite reorientation, depending on the starting structure. At the end of this stage, fully oriented martensite is produced. The plastic deformation of oriented martensite occurring over second yield point in the stage IV leads to fracture at ultimate strength.

In the Fig.6., a flat stress-plateau in stage II occurred under tension for specimens non-

hydrogenated, while no flat stress-plateau and quickly strain hardening in the stage III till the premature fracture were observed for specimens hydrogenated.

In the case of hydrogenated structure, the stages III and IV could not be reliably distinguished, so value of the second yield point σP was not evaluated.

Based on stress-strain curves, microstructure and fractographic observations, the following conclusions about hydrogen effect on the NiTi alloy were drawn: decrease of ultimate tensile strength in hydrogenated samples was related with slight change of fracture

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

7

patterns from high ductile fracture with deep dimples to ductile pattern with flatter dimples and microcracks. No doubt, hydrogen had an effect on the mobility of dislocations and the move of twin boundaries during the deformation that manifested by change of stress-strain curve. Table 3. Tensile properties of Z alloy as function of heat treatment conditions (in air, at room temperature and strain rate = 3 x 10-4 s-1).

Mechanical properties [MPa] Heat treatment

condition

Yield strength, σ0.2

Second yield point, σ.P

Ultimate tensile strength, σUTS

As-received Not measured 778 873 850 °C,1h, Ar(g) Not measured 618 784 850 °C,1h, H2(g) Not measured Not measured 622

Fig. 6. Stress-strain curves of NiTi deformed under tensile via martensite reorientation as function of heat treatment showing that the stress-strain curves are sensitive to the annealing condition: a) as-received without heat treatment, b) heat treatment at 850 °C for 1 h in Ar and c) heat treatment at 850 °C for 1 h in H2(g). Tested at 293 K.

Although, the Z specimen seemed to have more regions with residual austenite, the effect

of hydrogen on martensitic transition inhibition and hydride formation was no evident from metallographic observation of these specimens. Hydrogen content measured in A and Z specimens ranged from 2500 to 2700 mass ppm.

4. CONCLUSION

The study of microstructures and mechanical properties TiAl- and NiTi-based alloys in the state before and after heat treatment in flowing hydrogen gas was performed. The following conclusions about hydrogen effect on the Ti-based alloys could be drawn: • effect of hydrogen on the microhardness of TiAl-1.5Mo-0.2C alloy was not evidential

because the grain growing due to annealing

0

100

200

300

400

500

600

700

800

900

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

Strain [%]

Stress [M

Pa]

b) 850 °C, 1h, Ar

a) as-received

c) 850 °C, 1h, H2(g)

I IVIIIII

METAL 2007 22.-24.5.2007, Hradec nad Moravicí ___________________________________________________________________________

8

• microhardness measurements demonstrated the hydrogen influence on the Ti-44Al-8Nb alloy. The grain growing more important during hydrogen treatment couldn’t be explained reliably

• assumption that hydrogen atoms in NiTi-based alloys suppress the martensitic transition, was not verified from the microstructure study

• hydrogen influenced stress-strain curve and tensile properties in NiTi-based alloys.

More experimental testing and observation are needed for the both types of microstructure in order to confirm the hydrogen effect on hydride formation and lamellar or martensitic structure, respectively, including tensile testing, TEM observation and resistometric measurements. Acknowledgement

This study is part of the research project „Processing and properties of high purity and structure defined special materials“, MSM 6198910013, solved at Faculty of Metallurgy and Material Engineering of VŠB–Technical University of Ostrava, Czech Republic. References [1] LOSERTOVÁ, M. Interaction of hydrogen in nonferrous metals and their alloys.

(Inaugural dissertation.) VŠB-Technical University of Ostrava. Ostrava 2003. 138 p. [2] LOSERTOVÁ, M. Hydrogen effect on fracture surfaces of intermetallic alloys. In Proc.

Intern. Conf. Fractography 2006, Košice: Emilena, Slovakia, 2006, p. 414-418. [3] ZHU, J.H., LIU, C.T. Environmental effects in NiTi-based alloys. Scripta Materialia.

1999, 41, 1, p. 55. [4] ZHU, J.H., LIU, C.T., CHEN, C.H. Effect of iron additions on environmental

embrittlement of NiTi-base alloys. Intermetallics, 2004, 12, p. 859. [5] SOUBEYROUX, J.L. et al. Structural study of hydrides NiTiHx (x=1.0 and 1.4). Journal

of Alloys and Compounds. 1993, 196, p.127. [6] BISCARINI, A. et al. Martensitic transitions and mechanical spectroscopy of Ni50.8Ti49.2

alloy containing hydrogen. Acta Materialia. 1999, 47, 18, p. 4525. [7] BISCARINI, A. et al. Extraordinary high damping of hydrogen-doped NiTi and NiTiCu

shape memory alloys. Journal of Alloys and Compounds. 2003, 355, p. 52. [8] BISCARINI, A., COLUZZI, B., MAZZOLAI, G., MAZZOLAI, F.M. Diffusion of

hydrogen in the shape memory alloy Ni47Ti40Hf10Cu3. Journal of Alloys and Compounds, 2005, 404-406, 8, p. 261.

[9] LOSERTOVÁ, M., ŠTĚPÁN, P. The effect of hydrogen on the embrittlement of NiTi alloy. Acta Metallurgica Slovaca, 2006, 12, 4, p. 420-426.

[10] SZURMAN, I. Metallurgical possibilities of modification of microstructure and transformation performances of NiTi and NiTi-Me based shape memory alloys. PhD thesis. 2007. (in press).