和訳レポート 氏名（学籍番号） 中島 勝悟（137

和訳レポート
氏名（学籍番号）
中島
勝悟（137-T2736）
論文題目
Temperature rise at shear bands in metallic glasses
著者
J.J.Lewandowski,A.L.Greer
論文出典
Nature materials,VOL 5,January 2006,15-18
金属ガラス中のせん断帯における温度上昇
室温において金属ガラスのせん断帯中のみで塑性流動がみられる。せん断面でみられるこの局在
性と液体のような特徴は、せん断帯におけるせん断軟化と一致している。この軟化が局所加熱の
結果であるかについては議論中であるとともに、0.1K 以下から数千 K までの範囲で局所的に温度
が上昇すると推測されている。ここで我々は数ナノ秒以上で数 K 程度まで上昇する溶融性コーテ
ィングを基にした新しい実験方法を提案する。しかしながら、せん断帯の厚さは温度上昇にはよ
らないようである。高い降伏応力の魅力的な機械的特性を持つバルク金属ガラスを構造材料への
適用を制限する要因となっている、せん断帯形成のメカニズムと軟化の関連性を理解することが
重要である。
また、せん断帯は、例えば重合体や多結晶合金といったほかの材料でも発見されており、加工
軟化は張力において可塑的な不安定性を引き起こすことに関連するため金属ガラス中のせん断帯
が特に重要であり、構造材料としての金属ガラスのバルク材（BMGs）の可能性を制限している。
透過型電子顕微鏡はせん断帯が 10～20nm と非常に薄いことを指し示す。せん断の極端な局在性を
この構造的変化またはせん断帯での温度上昇によって説明するために二種類のモデルを比較す
る。いずれにせよ、局所的な粘性は下がり、割れ目の表面に不規則な筋模様と液体のような水滴
を引き起こす
せん断誘起による乱れはより明確に拡張し温度上昇せずとも動的平衡に到達できる。既存のせ
ん断帯での優先的なエッチングと変形が構造変化の証拠を示す。赤外線撮像による直接温度測定
はせん断の局在性とごく短時間の発生のために阻害される。測定された 0.4 および 0.25K の上昇は
最初に熱がより限定的に生成することに注目することによってせん断帯内の 650－1200K の上昇を
推定するために使用されていた。単一検出器が破断後の Zr 系 BMG で 500K までの温度上昇を発見
した。同様の金属ガラスを窒素下（酸化を避けるため）で衝撃破壊で 900K までに加熱された粒子
が放出が放出される。最終破断の間温度上昇するとはいえ、先に動作した小さなずれを有するせ
ん断帯よりもはるかに高くてもよい。
せん断帯での加熱は断熱的であると考えらえていた。せん断帯面で行われる仕事を計算から温
度が 40K 上昇するが、熱拡散の説明が不正確であり、局所的なせん断率が試料全体のせん断率よ
りもさらに高かったことに気が付かなかった。これらの影響を見直すと約 103K の温度上昇が予測
された。詳細な応力‐ひずみ測定に基づいた同様の計算は鋸歯状の流れの間で数 K、最終破断中
に 280K 上昇すると示した。試験資料の標準寸法区間における弾性エネルギーが熱としてせん断帯
に完全に現れると仮定すると、900K の温度上昇が予測された。（厚さ数マイクロメートルの非現
実的なせん断帯とする）せん断が線源の熱を生成する亀裂先端での加熱を計算して適応させ、予
測された温度上昇は数十 K ある。降伏状態下のせん断帯では、温度上昇が 0.05K と計算された
が、非常に小さ合計せん断と非現実的な値の歪率を伴う。
これら先行研究の観察結果と計算はせん断帯の温度上昇において一貫した推定を与えない。本
研究では融性のコーティングを用いてせん断帯での熱生成の定量化を中心にしている。Zr 系の
BMG Vitreloy1 の切り欠きはりの曲げ試験における先行研究で破断前に欠き底近くに大規模なせ
ん断帯が存在することを示しています。このタイプの試料はスズで被覆しており（Fig. 1a の斜線
部）、島状の模様を形成する。局所融解の検出をレーザースポットにコーティングをさらし、球状
に溶融して形成された島状模様がはっきりと見えたことで確認した。充填したところ、大規模な
せん断帯があっても破断前に切り欠き底で発見された。（Fig.2）より顕著なせん断帯に沿ってコン
トラストが局所融解が Fig.1c と同様に示します。コーティングのような溶解は変形が局所的では
ないアルミニウム合金基板では見つからなかった。また、激しい局所的な変形があるスズの破断
線のガラススライド上で見つけることができなかったが、そのガラスで全く熱が生成されていな
かった。したがって、Fig.2 中の局所融解がせん断帯付近で発生した熱に起因する。
（Fig.2b）
局所融解（詳細は Fig.3 に示す）はおおよそ半球状のガラス玉を提供する。溶融個所の半分の幅
w（Fig.2b）は島状模様（Fig.1b）が除去される領域から直接推定することができる。この毛管力
によってスズが描画された領域もまたそれによって形成された半球の体積から推定することがで
きる。ガラス球に沿ってはっきり見えるせん断帯のために、w の範囲は両方の方法で 300－
1500nm と推定される。
約 45 度のせん断角度で表面と交差するせん断帯をとると、せん断帯の周りの昇温部の真の半値幅
x は w/√2、おおよそ 200‐1000nm でとられている。
島状模様の規模はバルク溶解温度を適用するために十分に荒い。コーティングを介して熱拡散
のための特性時間 t = x2/4αSn は 14ps で、この x はコーティング厚さ（50nm）αSn はスズに熱拡散率
（4.4×10-5m2s-1）である。融解潜熱はスズの有効比熱を周囲温度から融点までの範囲でおよそ 2 倍
にする。これは温度上昇を検出するための時間分解能をとる約～30ps のコーティングを溶かすこ
とを検出するための応答時間が効果的に 2 倍になった。
溶融性被膜法の空間分解能（w に誤差）は島状模様（～100nm）の規模である。この時間的および
空間的解像度の組み合わせは赤外線撮像の最高のもの（約 11μm で 1.4 ミリ秒）および単一の赤外
線検出（約 10 マイクロ秒、100μm）よりも有意に良好である。このやり方は融解の限界を記録す
ることだけ制限されているが、単純な分析がより詳しい情報を産む。
Ref.26 に記載されているように、せん断帯は熱の面源として扱うことができる。単一帯にわたっ
てせん断を取ることで効果てきめんとなり、せん断帯は無限媒質中の厚さゼロの熱源として扱わ
れる。せん断後、t = 0 のとき、せん断帯は H（Jm-1）の熱含量を有する。周囲温度上昇を ΔT と置
き換えると、温度特性はせん断帯より関数 t と距離 x とした熱拡散方程式の薄膜溶液によって与え
られ、ρ は材料の密度、比熱 C および熱拡散 α である。距離 x での最大温度上昇 ΔTmax は、時間
tmax で a=(2πe)-1/2 で発生。
Vitreloy 1のスズの融点近傍において、ρ = 6,125 kg m−3(ref. 27), C =420 J kg−1 K−1 and α =3×10−6 m2 s−1
(ref. 28)。周囲温度25℃におけるスズの融点232℃では、ΔTmax = 207K。
最大推定半幅1μmに対する式（2）、（3）においてこれらの値を代入すると、H = 2.2kJm-2とtmax =
167nsが与えられる。xの最小値200nmでは、H = 0.4Jm-2とtmax = 7nsとなる（対応する温度特性は
fig.4に示す）。tmaxのこれらの値は溶解性コーティング法の時間分解能よりも長く安定である。
Vitreloy 1の軸に対して45度のせん断を想定した引っ張り強さ（1900MPa、ref.27）は、せん断降
伏強度が850MPaであることを意味する。せん断の仕事が完全に熱に変換されたと仮定すると、H
の推定範囲は、観察範囲内で0.5-2.6μmのせん断埋め合わせを意味する（Fig.2a）。以前の行われる
仕事の計算は私たちの範囲内においてH=2kJm-2を与える。
せん断が止まった時（t = 0のとき）せん断帯の中心における温度上昇ΔTcentreは早期分析によって
与えられ、ここのHの点で適合する。δtはせん断ずれy=Vδtを与える相対速度Vでのせん断の持続時
間である。最大Vはct～2.47×103ms-1の横方向の音速の速度とした0.9ctである。このVの最大値を
y=0.5-2.6μmでとると、δt=0.2-1.2nsでの下限推定を与える（tmaxの7-167nsの推定よりもはるかに短
く、それにより式（1）とfig.4中の特性の形状を使用する）
。式（4）より、ΔTcentreが対応する上限
推定値は3100-8300Kである（つまり、実際の最大温度は3400-8600K）。実際にはより遅いせん断
（より大きいδt）で、温度上昇は非常に低いかもしれない。熱拡散長2√αtでのΔtの下限値でさえ
も、10－20nmの許容せん断帯厚さをはるかに超える、全幅の倍である100‐240nmを与える。した
がって、前述したようにせん断帯動作は十分に断熱することができず、温度上昇はせん断帯の厚
さを制御する因子とすることはできない。溶融性コーティング法は時間的および空間的分解能が
従来の赤外線測定値をはるかに超えると、せん断帯動作中の局所加熱の測定を可能にする。熱拡
散時間が大幅にせん断時間を超えるように、簡単な計算はせん断帯内での熱含量と得られる温度
特性を評価するために使用できる。我々の測定は動作中のせん断内に著しい温度変位（数ナノ秒
の間に数千回の温度上昇）が存在し得ることを示唆し、したがって、そのようなナノ結晶とナノ
ボイドを形成するように、構造的変化が存在することは驚くべきことではない。この研究で実証
された局所加熱がせん断局在化によるものでなく、温度変化は高歪速度での延性の見かけの開始
などの現象を含めて、せん断帯動作の分析において考慮されなければならない。
Method
公称寸法3.0 mm×9.0 mm×30 mmの長方形断面の棒はBMGの公称組成Zr41.2Ti13.8Cu12.5Ni10Be22.5
（at%）のVitreloy 1より機械加工し放電したものを3.0mmの厚さの板として供給された。（ATI、ラ
グナニゲル、カリフォルニア州）ルート半径110μmのノッチはダイヤモンドワイヤーソーにより、
棒に切り入れた。材料特性（組成、無定形性、機械的特性）および曲げ試験のため試料調整のさ
らなる説明をほかの場所で見つけることができる。垂直面（Fig.3a）は2400のSiCグリット紙を使
用して鏡面仕上げに研磨した後、1μmから0.3μmのダイヤモンドペーストで研磨を行い、三回アセ
トン洗浄を行った。これらの表面上にガラス顕微鏡スライドと多結晶アルミニウム基板上にスズ
の堆積（一面の厚さは50nm）は、UHVマグネトロンスパッタ堆積であった（Snターゲット純度
99.999％、基底圧10-7Pa）
。スズ被膜された基材をLambda Physik (G¨ottingen, ドイツ)COMPexシリー
ズKrFエクシマレーザからの持続時間10-50nsで200-400mJの単パルスにさらした。局所融解のため
の効率的なエネルギーにて、溶融に従う形態の変化を調べることができた。インストロンサーボ
油圧試験機を用いて、0.0016mms-1の変位制御下での四点曲げ荷重を受けたスズ被膜された二重ノ
ッチをしたVitreloyビームは障害が起こるまで充填した。このような試験では、致命的な障害はノ
ッチのいずれか一方でのみ発生する。残りのノッチは壊滅的な破壊の直前に除荷され、充填中に
進展するせん断帯をとらえる。高解像度のSEM（JEOL 6340 FEGSEM）は特に広範囲のせん断帯付
近に存在するノッチルート領域を検査するために使用された。順次充填されるサンプルのせん断
帯の進展と確認するためとノッチ領域中のスズ被膜の形態の変化を関連づける補完的な実験を行
った。エネルギー分散分析は金属ガラスではなくスズで構成されたガラス球はせん断帯に沿って
観察したことを確認した。また、せん断帯に隣接した領域でスズが枯渇していたことも確認され
た。
LETTERS
Temperature rise at shear bands in metallic
glasses
J. J. LEWANDOWSKI*† AND A. L. GREER
Department of Materials Science and Metallurgy, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, UK
*Permanent address: Department of Materials Science and Engineering, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, Ohio 44106, USA
† e-mail: [email protected]
Published online: 18 December 2005; doi:10.1038/nmat1536
A
t ambient temperature the plastic flow shown by metallic
glasses is localized into shear bands1,2 . This localization
and the liquid-like features seen on fracture surfaces are
consistent with shear softening in the bands. The extent to
which this softening is a result of local heating has remained
controversial, with estimates of the local temperature rise
ranging from less than 0.1 kelvin to a few thousand kelvin3–11 .
Here we present a new experimental method based on a
fusible coating, which shows that the temperature rise, over
a few nanoseconds, can be as high as a few thousand kelvin;
nevertheless, the temperature rise does not seem to control
the shear-band thickness. It is important to understand the
mechanisms of shear banding and associated softening because
these are the principal factors limiting structural applications
of bulk metallic glasses, which have some attractive mechanical
properties such as high yield strength12,13 .
Although shear bands are found in other materials, for example
polymers14 and polycrystalline alloys15 , they are particularly
important in metallic glasses because the associated work-softening
leads to plastic instability in tension, limiting the potential of
bulk metallic glasses (BMGs) as structural materials. Transmission
electron microscopy16,17 indicates that the shear bands are very
thin, 10–20 nm. Two families of model compete to explain this
extreme localization of shear in terms of structural change or
temperature rise in the bands. In either case, the local viscosity is
lowered, leading to vein patterns and even liquid-like droplets on
fracture surfaces1,2 .
Shear-induced disordering18 , more specifically dilatation16,19 ,
can reach dynamic equilibrium without invoking any temperature
rise20,21 . Preferential etching22 and deformation23 at pre-existing
shear bands provide evidence for structural change. Direct
temperature measurements are hindered by the localization and
short duration of the shear; in infrared imaging, measured
rises of 0.4 and 0.25 K have been used to estimate rises
of 650–1,200 K within the bands by noting that the heat
generated was initially more confined10,11 . A single detector
found a ∼500 K rise in a Zr-based BMG after fracture4 .
Impact fracture7 of a similar glass under nitrogen (to avoid
oxidation) led to ejection of particles heated by ∼900 K. The
temperature rise during final fracture, however, may be much
a
b
c
100 nm
100 nm
Figure 1 Bend-test specimen and the fusible tin coating used to study shear
banding near notches. a, Schematic of a double-notched bend-test specimen. A tin
coating is applied on the vertical face and on loading shear banding is seen around
the notch roots. b,c, Scanning electron micrographs of the tin coating as-deposited
(b) and after melting under a laser spot (c).
higher than during the previous operation of shear bands with
small offsets.
The heating in shear bands has been regarded as adiabatic24 .
Calculations of the work done1 in the shear-band plane gave a
temperature rise of about 40 K, but incorrectly accounted for
thermal diffusion and failed to note that the local shear rate is
much higher than the macroscopic rate imposed on the sample.
Correcting for these effects3 , a rise of about 103 K was predicted. A
similar calculation, based on detailed stress–strain measurements8 ,
predicted rises of a few kelvin during serrated flow and about 280 K
15
nature materials VOL 5 JANUARY 2006 www.nature.com/naturematerials
©2006 Nature Publishing Group
LETTERS
a
1 μm
b
2w
x
1 μm
1 μm
Figure 2 Shear bands near a notch in a bend-test specimen coated with tin.
a, Secondary electron micrograph showing shear bands, some decorated with tin
beads, near the notch root. b, Schematic showing how a hot zone of half-width x
around a shear band can melt the tin coating to a half-width of w. The melted
coating forms approximately hemispherical beads of tin.
during final fracture. Assuming that the elastic energy in the gauge
section of a test sample manifests entirely as heat in the bands5 , a
rise of 900 K was predicted (for an unrealistic band thickness of a
few micrometres). Adapting calculations for the heating at a crack
tip, the shearing generates a line source of heat6 , and the predicted
rise is a few tens of kelvin. For shear bands under indentation, a
temperature rise of 0.05 K was calculated9 , but with a very small
total shear and an unrealistically low value of strain rate.
These earlier observations and calculations give no consistent
estimate of the temperature rise in a shear band. The present
work centres on the use of a fusible coating to quantify the
heat production at shear bands. Earlier work on bend testing of
notched beams of the Zr-based BMG Vitreloy 1 shows that before
catastrophic failure there is extensive shear banding near the notch
root25 . Specimens of this type were coated with tin (shaded area on
Fig. 1a), which formed an island pattern (Fig. 1b). The detectability
of local melting was verified by exposing the coating to a laser spot
(Fig. 1c); the islands that have melted and formed into spheres are
clearly visible. On loading, extensive shear banding is found at the
notch root even before failure (Fig. 2a). The contrast along the more
prominent bands suggests local melting similar to that in Fig. 1c.
Such melting of the coating was not found on aluminium alloy
substrates where the deformation is not localized. It was also not
found on glass slides at the line of fracture where there is intense
local deformation in the tin, but no heat produced in the glass. Thus
the local melting in Fig. 2a is attributed to the heat produced near
shear bands (Fig. 2b).
The local melting (more detail is given in Fig. 3) gives roughly
hemispherical beads. The melting half-width w (Fig. 2b) can be
estimated directly from the area over which the islanding pattern
(Fig. 1b) is removed. This area, from which the tin has been drawn
by capillary forces, can also be estimated from the volume of the
hemispheres thereby formed. For the bands along which beads
are clearly visible, the range of w estimated by both methods is
300–1,500 nm. Taking the bands to intersect the surface at a shear
angle of ∼45◦ , the
√ true half-width x of the hot zone around the
shear band is w/ 2, taken to be roughly 200–1,000 nm.
Figure 3 Tin coating formed into spherical beads at shear bands provides
evidence of local heating and melting of the coating. Scanning electron
micrographs showing details of shear bands after local melting of the tin coating.
Near the shear band, the as-deposited island pattern disappears and is replaced by
beads of tin. The inset shows the shear-band pattern and tin beads located in
regions removed from the large shear offset shown at top of figure.
The scale of the island pattern (Fig. 1b) is sufficiently coarse
for the bulk melting temperature to apply. The characteristic time
for thermal diffusion through the coating, t = x 2 /4αSn , where x is
the coating thickness (50 nm) and αSn the thermal diffusivity of tin
(4.4 × 10−5 m2 s−1 ), is 14 ps. The latent heat of melting roughly
doubles the effective specific heat of the tin in the temperature
range from ambient to melting; this effectively doubles the response
time for melting of the coating to ∼30 ps, which we take to be
the time resolution for the detection of a temperature rise. The
spatial resolution of the fusible-coating method (error in w ) is
the scale of the islanding pattern (∼100 nm). This combination
of temporal and spatial resolutions is significantly better than the
best reported for infrared imaging10,11 (1.4 ms; ∼11 μm) or for a
single infrared detector4 (∼10 μs; 100 μm). The method is limited
by only registering the limit of melting, but a simple analysis yields
further information.
As noted in ref. 26, shear bands can be treated as planar sources
of heat. Taking the shear across a single band to be effectively
instantaneous, the band is treated as a source of zero thickness in
an infinite medium. After shear, at time t = 0, the band is taken
to have a heat content H (J m−2 ). Using T to represent the rise
above ambient, the temperature profile, as a function of t and of
distance x from the band, is given by the thin-film solution of the
heat-diffusion equation:
T =
H
√
2ρC πα
2
1
−x
,
√ exp
4αt
t
(1)
where ρ is the density of the material, C the specific heat and α the
thermal diffusivity. At distance x the maximum temperature rise
Tmax is
H 1
(2)
,
Tmax = a
ρC x
16
nature materials VOL 5 JANUARY 2006 www.nature.com/naturematerials
©2006 Nature Publishing Group
LETTERS
1,500
0.2
H = 2.2 kJ m–2
Temperature rise, ΔT (K)
H = 0.4 kJ m–2
1,000
10
500
50
1
167
ΔT = 207 K
7
1,000
50
0
–2
–1
0
Distance, x (μm)
1
2
Figure 4 Local heating at a shear band, calculated for two values of total heat
content. Temperature profiles along the normal to a shear band (assumed
infinitesimally thin): on the left-hand side for a heat content of H = 0.4 kJ m−2
generated instantaneously at t = 0; on the right-hand side for H = 2.2 kJ m−2 . The
profiles, for successive times t (in nanoseconds) as shown, are calculated using
equation (1) with parameters (given in the text) for the BMG Vitreloy 1. The profiles in
bold are those for the maximum half-width w of melting (0.2 μm on the left-hand
side, 1.0 μm on the right-hand side) of a tin coating for the two H values. For the
coating, melting requires a temperature rise of 207 K above the ambient 298 K.
where a = (2πe)−1/2 , occurring at a time tmax
a2
tmax =
2α
H
ρC
2
1
.
2
Tmax
(3)
For Vitreloy 1 near the melting point of tin, ρ = 6,125 kg m−3
(ref. 27), C = 420 J kg−1 K−1 and α = 3 × 10−6 m2 s−1 (ref. 28). For
melting tin at 232 ◦ C, with an ambient of 25 ◦ C, Tmax = 207 K.
Substituting these values into equations (2) and (3), for the
maximum estimated half-width x of 1 μm, gives H = 2.2 kJ m−2
and tmax = 167 ns; for the minimum x of 200 nm, H = 0.4 kJ m−2
and tmax = 7 ns (corresponding temperature profiles are shown in
Fig. 4). These values of tmax are comfortably longer than the time
resolution of the fusible-coating method.
The tensile strength of Vitreloy 1 (1,900 MPa; ref. 27), assuming
shear at 45◦ to the axis, implies a shear yield strength of 850 MPa.
Assuming that the entire work done in shear is transformed into
heat, the estimated range of H implies shear offsets of 0.5−2.6 μm,
within the range observed (Fig. 2a). Earlier calculations3 of work
done gave H = 2 kJ m−2 , within our range.
The temperature rise at the centre of the band Tcentre when
shear stops (at t = 0) is given by earlier analyses8,29 , adapted here in
terms of H :
H
H
1
1
1
V
=√
,
Tcentre = √
(4)
αδt
αy
π ρC
π ρC
where δt is the duration of shear at relative velocity V giving a shear
offset y = V δt . The maximum V is 0.9ct where ct ≈ 2.47 × 103 m s−1
is the speed of a transverse sound wave3,27 . Taking this maximum
V with y = 0.5−2.6 μm, gives a lower-bound estimate of
δt = 0.2−1.2 ns (much shorter than the estimated tmax of 7–167 ns,
thereby validating the use of equation (1) and the form of
the profiles in Fig. 4). From equation (4), the corresponding
upper-bound estimate of Tcentre is 3,100–8,300 K (that is, a
maximum actual temperature 3,400–8,600 K). In practice, with
slower shear (larger δt ), the temperature rise may be much lower.
Even
√ for the lower-bound range of δt , the thermal diffusion length
2 αt , doubled to give the total width, is 100–240 nm, much
greater than the accepted shear-band thickness of 10–20 nm16,17 .
Thus, shear-band operation cannot be fully adiabatic, as noted
earlier26 , and temperature rise cannot be the factor controlling
shear-band thickness.
The fusible-coating method permits the measurement of local
heating during shear-band operation, with temporal and spatial
resolution far exceeding conventional infrared measurements. As
thermal-diffusion times greatly exceed the shear durations, a simple
calculation can be used to evaluate the heat content in a band and
the resulting temperature profiles. Our measurements suggest that
there can be remarkable temperature excursions (rises of thousands
of degrees for a few nanoseconds) within a shear band during
its operation, and it is therefore not surprising that there can
be structural changes such as the formation of nanocrystals and
nanovoids9,30,31 . Even though the local heating demonstrated in this
work is not the origin of shear localization, the temperature changes
must be taken into account in analysing shear-band operation,
including such phenomena as the apparent onset of ductility at high
strain rates32 .
METHODS
Rectangular cross-section bars of nominal dimensions 3.0 mm × 9.0 mm ×
30 mm were electro-discharge machined from Vitreloy 1, a BMG of nominal
composition (at.%) Zr41.2 Ti13.8 Cu12.5 Ni10 Be22.5 , supplied (ATI, Laguna Niguel,
California) as 3.0-mm-thick plate. Notches of root radius 110 μm were cut in
the bars with a low-speed diamond wire saw. Characterization of the material
(composition, amorphicity and mechanical properties) and further description
of specimen preparation for bend testing can be found elsewhere25 . The vertical
surfaces (Fig. 1a) were polished to a mirror finish using 2,400 SiC grit paper,
followed by 1 μm and 0.3 μm diamond paste, then triple cleaned in acetone.
Deposition of tin (blanket thickness 50 nm) on these surfaces and on glass
microscope slides and polycrystalline aluminium substrates was by UHV
magnetron sputter deposition (Sn target 99.999% purity; base pressure 10−7
Pa). Tin-coated substrates were exposed to a single pulse of duration 10–50 ns
and energy 200–400 mJ from a Lambda Physik (Göttingen, Germany) COMPex
series KrF excimer laser (248 nm). With energy sufficient for local melting, the
changes in morphology that follow melting could be examined.
The tin-coated double-notched Vitreloy beams, subjected to four-point
bend loading under displacement control at 0.0016 mm s−1 using an Instron
servo-hydraulic testing machine, were loaded until failure. In such testing,
catastrophic failure occurs at only one of the notches. The remaining notch
represents a sample unloaded immediately before catastrophic fracture and
captures the shear banding that has evolved during loading. High-resolution
scanning electron microscopy (JEOL 6340 FEGSEM) was used to examine, in
particular, the notch-root regions near which there is extensive shear banding25 .
Complementary experiments were conducted on samples sequentially loaded
to confirm the evolution of shear banding and associated changes in tin-coating
morphology in notch regions. Energy-dispersive analysis confirmed that the
beads observed along shear bands were composed of tin and not of the metallic
glass. Also, it was observed that the regions adjacent to the bands were
depleted in tin.
Received 29 August 2005; accepted 14 October 2005; published 18 December 2005.
References
1. Pampillo, C. A. Review: Flow and fracture in amorphous alloys. J. Mater. Sci. 10, 1194–1227 (1975).
2. Bengus, V. Z. et al. New features of the low temperature ductile shear failure observed in bulk
amorphous alloys. J. Mater. Sci. 35, 4449–4457 (2000).
3. Bengus, V. Z. et al. Some peculiarities of ductile shear failure of amorphous alloy ribbons. Int. J.
Rapid Solid. 8, 21–31 (1993).
4. Bruck, H. A., Rosakis, A. J. & Johnson, W. L. The dynamic compressive behavior of beryllium bearing
bulk metallic glasses. J. Mater. Res. 11, 503–511 (1996).
5. Liu, C. T. et al. Test environments and mechanical properties of Zr-base bulk amorphous alloys.
Metall. Mater. Trans. A 29, 1811–1820 (1998).
17
nature materials VOL 5 JANUARY 2006 www.nature.com/naturematerials
©2006 Nature Publishing Group
LETTERS
6. Flores, K. M. & Dauskardt, R. H. Local heating associated with crack tip plasticity in Zr-Ti-Ni-Cu-Be
bulk amorphous metals. J. Mater. Res. 14, 638–643 (1999).
7. Gilbert, C. J. et al. Light emission during fracture of a Zr-Ti-Ni-Cu-Be bulk metallic glass. Appl. Phys.
Lett. 74, 3809–3811 (1999).
8. Wright, W. J., Schwarz, R. B. & Nix, W. D. Localized heating during serrated plastic flow in bulk
metallic glasses. Mater. Sci. Eng. A 319–321, 229–232 (2001).
9. Kim, J.-J., Choi, Y., Suresh, S. & Argon, A. S. Nanocrystallization during nanoindentation of a bulk
amorphous metal alloy at room temperature. Science 295, 654–657 (2002).
10. Yang, B. et al. In-situ thermographic observation of mechanical damage in bulk-metallic glasses
during fatigue and tensile experiments. Intermetallics 12, 1265–1274 (2004).
11. Yang, B. et al. Dynamic evolution of nanoscale shear bands in a bulk-metallic glass. Appl. Phys. Lett.
86, 141904 (2005).
12. Johnson, W. L. Bulk glass-forming metallic alloys: science and technology. Mater. Res. Soc. Bull. 24,
42–56 (1999).
13. Inoue, A. Stabilization of metallic supercooled liquid and bulk amorphous alloys. Acta Mater. 48,
279–306 (2000).
14. Li, J. C. M. Behavior and properties of shear bands. Polym. Eng. Sci. 24, 750–760 (1984).
15. Xue, Q., Meyers, M. A. & Nesterenko, V. F. Self-organization of shear bands in titanium and
Ti-6Al-4V alloy. Acta Mater. 50, 575–596 (2002).
16. Donovan, P. E. & Stobbs, W. M. The structure of shear bands in metallic glasses. Acta Metall. 29,
1419–1436 (1981).
17. Pekarskaya, E., Kim, C. P. & Johnson, W. L. In situ transmission electron microscopy studies of shear
bands in a bulk metallic glass based composite. J. Mater. Res. 16, 2513–2518 (2001).
18. Polk, D. E. & Turnbull, D. Flow of melt and glass forms of metallic alloys. Acta Metall. 20,
493–498 (1972).
19. Spaepen, F. & Turnbull, D. A mechanism for the flow and fracture of metallic glasses. Scripta Metall.
8, 563–568 (1974).
20. Spaepen, F. A microscopic mechanism for steady state inhomogeneous flow in metallic glasses. Acta
Metall. 25, 407–415 (1977).
21. Steif, P. S., Spaepen, F. & Hutchinson, J. W. Strain localization in amorphous metals. Acta Metall. 30,
447–455 (1982).
22. Pampillo, C. A. Localized shear deformation in a glassy metal. Scripta Metall. 6, 915–918 (1972).
23. Krishnanand, K. D. & Cahn, R. W. Recovery from plastic deformation in a Ni/Nb alloy glass. Scripta
Metall. 9, 1259–1261 (1975).
24. Leamy, H. J., Chen, H. S. & Wang, T. T. Plastic flow and fracture of metallic glass. Metall. Trans. 3,
699–708 (1972).
25. Lowhaphandu, P. & Lewandowski, J. J. Fracture toughness and notched toughness of bulk amorphous
alloy: Zr-Ti-Ni-Cu-Be. Scripta Mater. 38, 1811–1817 (1998).
26. Hufnagel, T. C., Jiao, T., Li, Y., Xing, L.-Q. & Ramesh, K. T. Deformation and failure of
Zr57 Ti5 Cu20 Ni8 Al10 bulk metallic glass under quasi-static and dynamic compression. J. Mater. Res. 17,
1441–1445 (2002).
27. Wang, W. H., Dong, C. & Shek, C. H. Bulk metallic glasses. Mater. Sci. Eng. R 44, 45–89 (2004).
28. Yamasaki, M., Kagao, S. & Kawamura, Y. Thermal diffusivity and conductivity of supercooled liquid
in Zr41 Ti14 Cu12 Ni10 Be23 metallic glass. Appl. Phys. Lett. 84, 4654–4655 (2004).
29. Eshelby, J. D. & Pratt, P. L. Note on the heating effect of moving dislocations. Acta Metall. 4,
560–562 (1956).
30. Chen, H., He, Y., Shiflet, G. J. & Poon, S. J. Deformation-induced nanocrystal formation in shear
bands of amorphous alloys. Nature 367, 541–543 (1994).
31. Jiang, W. H. & Atzmon, M. The effect of compression and tension on shear-band structure and
nanocrystallization in amorphous Al90 Fe5 Gd5 : a high-resolution transmission electron microscopy
study. Acta Mater. 51, 4095–4105 (2003).
32. Sergueeva, A. V. et al. Shear band formation and ductility in bulk metallic glass. Phil. Mag. 85,
2671–2687 (2005).
Acknowledgements
N. Stelmashenko is thanked for assistance with sputter deposition, and M. J. Stowell for useful
discussions. Experimental support for J.J.L. on sabbatical at the University of Cambridge was provided
by the Office of Naval Research, DARPA and Reference Metals. Approved for public release.
Distribution unlimited. A.L.G. acknowledges support from the European Commission RTN
‘Ductilisation of BMG’.
Correspondence and requests for materials should be addressed to J.J.L.
Competing financial interests
The authors declare that they have no competing financial interests.
Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/
18
nature materials VOL 5 JANUARY 2006 www.nature.com/naturematerials
©2006 Nature Publishing Group