Sie sind auf Seite 1von 9

MAT ER IA LS CH A R A CTE R IZ A TI O N 5 9 (2 0 0 8) 9 17

Effect of silver addition on the creep parameters of Sn7 wt.%


Bi alloy during transformation
F. Abd El-Salam, A.M. Abd El-Khalek, R.H. Nada, A. Fawzy
Physics Department, Faculty of Education, Ain Shams University, Cairo, Egypt

AR TIC LE D ATA

ABSTR ACT

Article history:

The straintime relations of Sn7 wt.% Bi and Sn7 wt.% Bi1 wt.% Ag alloys have been

Received 27 June 2006

obtained under different constant stresses (8.719.1 MPa) in the temperature range (313

Received in revised form

373 K). The results showed that the transient and the steady state creep parameters n, and

13 October 2006

st depend on both the deformation temperature and the applied stress. The temperature

Accepted 17 October 2006

dependence of these parameters indicates a transition temperature at 343 K. The Ag-free


samples are more hard than those with Ag addition. This is explained by precipitation of Bi

PACS:

atoms which segregate to form a non-coherent cubic -phase Sn7 wt.% Bi samples. While,

62.20Fe

Ag addition prevents precipitation segregate on the grain boundaries. The analysis of the X-

62.20Hg

ray diffraction patterns shows that, Ag addition increases the Sn crystallite size but

81.40Lm

decreases Bi crystallite size. The energy activating the steady state creep in the two alloys in
the low and high temperature ranges (before and after transformation temperature) were
found to be 51.2 1.6 and 88.1 1.4 kJ/mol for Sn7 wt.% Bi alloy and 34.3 1.9 and 57 1.8 kJ/
mol for Sn7 wt.% Bi1 wt.% Ag alloy, respectively.
2006 Published by Elsevier Inc.

Keywords:
Creep parameters
Ag addition
Crystallite size
Activation volume
Grain boundary

1.

Introduction

In view of the environmental and health issues concerning the


toxicity of Pb present in the most prominent PbSn solder [1],
much attention has been paid to the development of lead-free
solder alloys which avoid the environmental problem of lead
[2,3].
The choice of alloy used depends upon factors as the
strength and corrosion resistance. Most of these solders are tin
containing binary and ternary alloys [4]. The low hardness and
consequently the low mechanical strength of pure tin make it

unsuitable for use as a bulk material for construction unless


strengthened by hardening alloying elements [5].
Pewters (alloys of more than 90% Sn) are used mostly as
solders for packing and interconnection in the electronic,
electrical, and engineering industries because of their ease
of fabrication into any required shape [5]. They are widely
used for utensils and have to some extent replaced the
traditional silverware which was formerly part of the domestic scene.
The determination of the mechanical properties of materials is of great importance in defining their endurance limit

Corresponding author. Physics Department, Faculty of Education, Ain-Shams University, Heliopolis, Roxy, Cairo, Egypt.
E-mail address: asaad_abdelkahlk@yahoo.com (A.M. Abd El-Khalek).
1044-5803/$ see front matter 2006 Published by Elsevier Inc.
doi:10.1016/j.matchar.2006.10.006

10

MAT ER IA LS CH A R A CTE RI ZA TI O N 5 9 (2 0 0 8) 917

sensitive deformation at room temperature and is superplastic [14]. Mechanical and microstructural measurements indicated that slip is the deformation mechanism which operates
at high strain rates, while grain boundary sliding (GBS)
dominates at low strain rates. The mechanisms involved
were controlled by the properties and behaviour of the grain
boundaries during superplastic deformation [15].
As grain boundary sliding would promote grain boundary
cracking and lead to limited ductility, it was proposed [16] that
diffusional creep is the principal mode of strain in the
superplastic range. Strain rates less than 1, are associated
with a contribution by rate insensitive climb-controlled slip
which becomes more important at high rates of strain.
In the present work, structure and creep characteristics of
the Sn7 wt.% Bi alloy, were investigated around phase
transformation. The effect of 1 wt.% Ag addition on these
characteristics has been also investigated.

2.

Fig. 1 a) The equilibrium phase diagram of Tin-Bismuth, b)


X-rays diffraction pattern of Tin-Bismuth.

under definite working conditions and so comes their


optimum suitability in technological industrial applications.
Changes in the mechanical properties of materials were
observed around and during phase transformation [2,6].
SnBi alloys, which outperform the SnPb alloys [7]
attracted attention because of their good properties as solders
for some delicate electronic tools [8]. The transient creep
behaviour of Sn0.2 wt.% Bi alloy predeformed by twisting was
studied at different temperatures [9]. Resoftening was observed with higher degrees of predeformation associated with
increasing energies of activation. In the same alloy with
different grain size, increasing grain size caused a decrease in
the parameter and increased n. Increased predeformation to
a critical value caused a decrease in both and n to minima,
depending on grain size, after which an increase in both and
n was observed [10]. In Sn0.5 at.% Bi alloy the steady state
creep rate was found [11] to decrease on increasing grain size
while the activation volume increased.
A Sn5 wt.% Bi master alloy and ternary alloys with
different Pb content were prepared by melt spinning technique and then irradiated to 1.2 MGy -radiation at room
temperature [12]. The changes in the measured values were
attributed to composition inhomogeneity caused by the
distribution mode of the third element in the matrix and the
defects induced by irradiation. The ductility of the BiSn
eutectic alloy significantly improved [13] by adding small
amount of Ag (less than 0.5 wt.%).
The microstructure of the Sn5 wt.% Bi alloy detected as a
dispersion of Bi particles in equiaxed Sn grains shows rate

Experimental

The materials investigated, Sn7 wt.% Bi and Sn7 wt.% Bi


1 wt.% Ag (hereafter termed alloy A and alloy B, respectively)
solder alloys were prepared from pure elements 99.99%, by
melting the constituents in a Pyrex tube under a fluxing agent
(colophony) to prevent the sample oxidation in air. The melt
was heated on a benzene flame for more than 15 min while the
tube was shaken to ensure homogenization of the melt. After
solidification the ingot was swaged in the form of wires of
diameter 1.2 mm for creep study and sheets of 0.2 mm in
thickness for X-rays diffraction patterns.
Following previous work the primary mean of microstructural control was annealing at room temperature [14].
This procedure avoids the complication of the solution of Bi
during elevated temperature annealing and reprecipitation
on cooling prior to room temperature testing. The wire samples were annealed for 2 h at 408 K to eliminate the cold work
acquired during swaging and were then quenched in cold
water at 273 K to have an identical thermal history. The
specimen compositions (wires or sheets) determined by
chemical analysis were very close to the alloy composition
of the melt.
The experiments were carried out in a convenient creep
machine equipped with a strain resolution equal to 10 4. The
average of three creep runs were performed under constant
loads corresponding to stresses, , of 8.7, 10.4, 12.1, 15.6,
19.1 MPa and at testing temperatures ranging from 313 K to
373 K in steps of 10 K. The accuracy of temperature and

Table 1 Average crystallite size from analysis of the Xrays diffraction patterns
Samples

Crystallite size () of
Sn

Sn7 wt.% Bi
Sn7 wt.% Bi1 wt.% Ag

Bi

323 K

353 K

323 K

353 K

681.6
1006

849.8
909.2

245
176

480
260

MAT ER IA LS CH A R A CTE R IZ A TI O N 5 9 (2 0 0 8) 9 17

11

Fig. 2 Representative straintime curves for Sn7 wt.% Bi (alloy A), and Sn7 wt.% Bi1 wt.% Ag (alloy B) at different stresses and
different deformation temperatures.
elongation measurements are of the order of 1 K and 10 5 m,
respectively.

3.

Results

The SnBi phase diagram [17] is given in Fig. 1a. A representative example for X-rays diffraction patterns is shown in
Fig. 1b for SnBi (323 K). From the analysis of the X-rays
diffraction patterns, the average crystallite size of both Sn and
Bi are given in Table 1 for different temperatures.
Representative creep curves of alloys A and B are given in
Fig. 2. Fig. 2 (a, b) shows the creep curves for both alloys at
certain temperature (363 K) under different stresses and Fig. 2

(c, d) for both alloys under a constant stress of 12.1 MPa at


different temperatures. It is clear that these isothermal creep
curves are sensitive to both the deformation temperature, T,
and the applied stress, , and show a monotonic shift towards
higher strains and lower fracture times with increasing the
deformation temperature. The level of creep strain for alloy B
is generally higher than that of alloy A under the same testing
conditions.
The transient creep strain, tr was found to obey a relation
of the form [18],
etr btn

where t is the transient creep time, and , n are constants


depending on the test conditions.

Fig. 3 The relation between lntr and lnt for different deformation temperatures for representative samples of both alloys
A and B.

12

MAT ER IA LS CH A R A CTE RI ZA TI O N 5 9 (2 0 0 8) 917

Fig. 4 The temperature dependence of the creep parameters n, and st; (a, c, e) for alloy A and (b, d, f) for alloy B.

Fig. 5 The relation between both ln and lnst and 1000/T: (a, c) for alloy A and (b, d) for alloy B.

MAT ER IA LS CH A R A CTE R IZ A TI O N 5 9 (2 0 0 8) 9 17

13

From the relation between lntr and lnt given in Fig. 3, as


representative examples at different temperatures under a
certain constant stress of 12.1 MPa, the parameter n is
obtained as the slope of the straight lines and is deduced
from the intercept at the lntr axis.
The temperature dependence of the parameters n under
different stresses is given for both alloys in Fig. 4 (a, b) and for
the parameter Fig. 4 (c, d). Fig. 4 shows two straight lines for n
and relations with increasing stress and temperature. The
behaviour of variation changes to higher values at temperatures above 343 K.
The steady state strain rate (st) of the tested samples is
calculated from the slopes of the linear parts of the creep
curves of Fig. 2. It increases with increasing both the
deformation temperature and the applied stress. Fig. 4 (e, f)
shows that st increases slowly in the low deformation

Fig. 7 (a, b) Relation between lnst and at different


deformation temperatures, T, for alloy A and alloy B. (c) The
dependence of the activation volume, V, on the working
temperature, T, for both alloys A and B.

temperature range 313333 K, while a rapid increase occurs


starting from 343 K in both alloys. It is seen that under the
same test conditions alloy B yields creep rates so high
compared with those of alloy A.
The activation energy of the transient creep or the steady
state creep, E, for both alloys was calculated using an
Arrhenius equation of the form [19],
b b VexpE=KT

and the reported equation:


est const:expE=KT
Fig. 6 (a, b) Relation between lnst and ln at different
deformation temperatures, T, for alloy A and alloy B. (c) The
dependence of the stress sensitivity parameter, m, on the
deformation temperature, T, for both alloys A and B.

where is a constant depending on the test conditions, K is


the Boltzmann constant, T is the absolute temperature. The
relation between ln and 1000/T, Fig. 5 (a, b) and lnst and 1000/
T, Fig. 5 (c, d) gave straight lines for the different applied

14

MAT ER IA LS CH A R A CTE RI ZA TI O N 5 9 (2 0 0 8) 917

temperatures. The temperature dependence of m values given


in Fig. 6c shows that m values are higher in alloy B than in
alloy A, indicating that alloy B is more ductile than alloy A.
The value of the activation volume, V = lnst / , as derived
from the slopes of the straight lines of Fig. 7 (a, b) relating lnst
and the applied stress, , was found to increase with
increasing deformation temperature, as shown in Fig. 7c.
As creep is a continuous thermally activated process, the
parameter was reported to correlate with the steady state
creep rate st through the relation [22]:
b bo est g

where o is a constant, and the exponent (= ln / lnst)


relates both the transient and the steady creep stages. is the
slope of the straight lines relating ln and lnst given in Fig. 8
which shows that is a stress independent ratio for both
alloys A and B. For alloy A, assumed the value 0.63, and for
alloy B it was 0.69.

4.

Fig. 8 The relation between ln and lnst for both alloys A


and B.

stresses. Each line is formed of two parts supposed to be


separated at the phase transformation temperature of each
alloy. The slope of the low temperature part characterizes the
activation energy before transformation and the second part
in the high temperature gives the activation energy after
transformation. For alloy A, the average activation energy for
transient creep, Fig. 5 (a, b) before and after transformation
were 23.4 and 47.2 kJ/mol, and for alloy B, 24.5 and 44.7 kJ/mol,
respectively. The close values of these activation energies
consist with conclusion that the activation energy for creep is
not measurably affected by minor impurities, (1 wt.% Ag in
alloy B) [20].
The calculated activation energies from Fig. 5 (c, d), for the
steady state creep were found independent on the applied
stress, but showed dependence on the range of the deformation temperatures being below or above 343 K. Below 343 K, the
calculated energies assumed the value 51.2 1.6 kJ/mol for
alloy A and 34.3 1.9 kJ/mol for alloy B. Above 343 K, a value of
88.1 1.4 kJ/mol was obtained for alloy A and a value of 57
1.8 kJ/mol was obtained for alloy B.
It is well known that the steady state creep rate (st) is
related to the applied stress () according to the relation [21]:
est crm

where c is a constant and m is the stress sensitivity parameter,


and both parameters were found to depend on the testing
conditions. The relation between lnst and ln was plotted and
straight lines for both alloys were obtained as shown in Fig. 6
(a, b). The slopes of the straight lines in Fig. 6 yield the values
of the stress sensitivity parameter m, at different deformation

Discussion

The phase diagram of SnBi system [17] shows that Sn


dissolves about 1% Bi at room temperature. The remaining Bi
is present as essentially pure Bi. Under equilibrium states the
Sn7 wt.% Bi alloy consists of two phases: a body centred
tetragonal -Sn matrix phase and a rhombohedral Bi second
phase. During preparation, the intrinsic point defects in
thermal equilibrium and concentrations determined by Boltzmann's factor, exp( E/KT), are expected to form in ratios
depending on the heat treatment and cooling regime. This will
cause a lack of homogeneity in the prepared samples such
that they become out of equilibrium. At 343 K and above the
single Sn-rich phase, -Sn, dominates. It appears from the
phase diagram that the transformation temperature is that
where the alloy decomposes into a primary solid solution of Bi
in Sn and a Bi phase with very low Sn content. The operating
deformation mechanism should therefore change at this
temperature.
The morphology of the Bi in SnBi alloy was difficult to
determine but the results can be interpreted on the basis that
the Bi is in solution in the ingot alloy obtained at the solution
temperature (343 K) and that Bi precipitates and coarsens at
room temperature. After 1 h annealing at room temperature,
the alloy was found to be fully recrystallized with an equiaxed
microstructure in cross or longitudinal sections with very fine
Sn grains. Further annealing at room temperature causes
grain growth. The grains remain of uniform size and equiaxed
[14].
It is inferred that segregation of Bi and Ag to or near the Sn
boundaries may occur [14]. In SnBi alloys the analysis of Xray diffraction patterns like that represented in Fig. 1b, free Bi
exists in their binary and ternary systems (015), (211) and (128).
Besides, in the ternary SnBiAg alloy-free Ag, (200) with
relative intensity (RI) 43.67%, also exist. The solute atoms
distort the solvent lattice and change the average size of its
elementary cell. In the solid solution, the solvent retains its
original crystal lattice, but there exists a relative strengthening. The increased strengthening, which results due to a
decrease in lattice constant is stronger than that due to an

15

MAT ER IA LS CH A R A CTE R IZ A TI O N 5 9 (2 0 0 8) 9 17

increase in lattice constant. Accordingly, the strengthening


due to Bi (radius of 1.56 ) in Sn (1.51 ) will be smaller than
that due to Ag (1.44 ).
In view of the surface activity theory [23] two mechanisms
for the effect of impurity atoms on the growth and properties
of crystals are possible. Accordingly, Bi atoms in Sn matrix
should migrate away from grain boundaries to the interior of
the grains leaving more mobile boundaries. On the other hand
Ag atoms in Sn matrix migrate towards grain boundaries thus
limit their mobility and inhibits grain growth.
However, a redistribution process of the solute atoms (Bi or
Ag) in the matrix Sn may take place depending on the existing
medium and the external factors such as temperature and
stress, leading to probable deviation from this theory. The
lower value 681.6 of the crystallite size (C.S.) of the Sn matrix
in alloy A at 323 K, Table 1, may be due to the Bi atoms being
less activated at the quenching temperature and the low
temperature 323 K, to reach their expected positions at the
grain cores as states the surface activity theory [23].
Therefore, some of the Bi atoms will disperse near the Sn
grain boundary regions and pin mobile dislocations. This,
increases strengthening and limits grain boundary motion.
This consists with the common reported [14] effect of alloying
elements which act principally to refine and stabilize the Sn
grains size and are more effective in large concentrations. The
above explanation is supported by the increased crystallite
size of Sn in the same sample at 353 K to 849.8 , Table 1,
where many Bi atoms were activated to reach grain cores
which facilitated grain boundary motion and enabled grain
growth. At the same time, the increased migration of Bi to the
grain cores at 353 K may account for the increased Bi grain size
from 245 at 323 K to 480 at 353 K, Table 1. This shows that
increasing temperature increased both Sn and Bi grains [14].
The general moment [23] of Ag is 32.6 1010e, of Sn is
33.1 1010e and for Bi is 52.2 1010e where e is the electronic
charge. Therefore, Ag atoms should go to the grain boundaries
of either Bi or Sn. As the moment difference between Sn and
Ag is very small, +0.5 1010e, the behaviour of Ag atoms in Sn
under the working conditions may be of neutral tendency to
move to either grain boundaries or to grain cores, but high
probably go to Bi grain boundaries due to their large moment
difference +20.6 1010e. Taking the molecular volumes into
account, Ag (10.28 10 6 m3) can be easily accommodated in Bi
(21.3 10 6 m3) rather than in Sn (16.26 10 6 m3). It can
therefore be accepted that most of the Ag atoms prefer
migrating towards Bi grain boundaries leaving Sn boundaries
more free to move and increase from the value 681.6 for the
binary alloy to 1006 for the ternary alloy, Table 1, while Bi
grain boundaries show the observed decrease in their crystallite size from 245 to 176 in the ternary alloy (Table 1).
Increasing temperature from 323 K to 353 K may have
activated some Ag atoms to disperse towards Sn grain
boundaries thus freeing some Bi boundaries from their
pinning points to reach larger crystallite size, 260 , while
pinning some Sn boundaries to minimize Sn crystallite size to
909.2 , Table 1. The addition of 1% Ag effectively inhibits the
uncontrolled growth of the high strength large structure
during the solidification of the binary SnBi alloy [1]. It is
believed that such small addition of Ag goes primarily into
solution within the -Sn phase [14]. It appears that Ag

promotes a higher nucleation density of solid phases during


solidification. The high density and apparent random relative
positioning of these solid phase nuclei cause them to
immediately impinge upon one another and prevent the
formation of the high strength large plate-like structure. Thus,
the ternary alloy samples (alloy B) will be more soft than the
binary alloy A, as clear from Fig. 2, which shows higher strain
levels for alloy B than those for alloy A. This shows that the
addition of 1% Ag to the SnBi alloy improves the tensile
ductility as was found previously [1].
Although the presence of silver is expected to harden the
alloy, the observations in Figs. 2, 4 (e, f), 6c and 7c), show
continuous softening behaviour for the samples of alloy B. It
may be that the Ag precipitate particles which are more
harder, Y(Ag) [Young's modulus of Ag] = 80.5 GN m 2, than the
matrix, Y (Sn) = 52 GN m 2, and the second phase, Y (Bi) = 34 GN
m 2, and the difference in ductility makes the whole structure
similar to that of a non-Newtonian fluid containing rigid
particles [24] which rotate practically without deformation
while the soft phases, Sn and Bi, undergo plastic deformation
performing accommodation between both by sliding along the
interfaces [25].
The effect of Ag addition on mechanical properties can be
seen from comparing the tensile test curves of both alloys A
and B shown in Figs. 2, 4, 6c and 7c). Significant improved
ductility and softening in the Ag-doped alloy compared to the
Ag-free binary alloy is evident. The temperature dependence
of the parameters n, and st (Fig. 4) shows nonlinear relations
reflecting the existence of two stages. The first stage at lower
temperatures shows slow rate of strain variations with low
strain levels depending on the applied stress. The second
stage at higher temperatures, where many of the blocked
dislocations are freed from their pinning points shows marked
increased strain levels. It is therefore considered that there
exists two strain rate stages separated by a narrow temperature range centred at the transformation temperature 343 K.
Below 343 K the grains are relatively small (681.6 for Sn at
323 K) depending on the deformation temperature, while
above 343 K the grains have grown to higher values (849.8 for
Sn at 353 K), Table 1 for alloy A.
On the basis of evidence and interpretation accepted in
literature about SnBi alloy [14], the present data, Fig. 4,
establish that distinct mechanisms of deformation are dominant at low and high levels of strain. Because of the absence
of twinning in this alloy [14] the dominant low rate mechanism is the thermally activated grain boundary sliding (GBS)
and the high rate mechanism is slip. The region of transition

Table 2 Average activation energy for both alloys A and B


at transient and steady state creep stages
Creep
stage

Transient
creep
Steady state
creep

Activation energy for


alloy A (kJ/mol)

Activation energy for


alloy B (kJ/mol)

Below
(343 K)

Above
(343 K)

Below
(343 K)

Above
(343 K)

23.4

47.2

24.5

44.7

51.2

88.1

34.7

57

16

MAT ER IA LS CH A R A CTE RI ZA TI O N 5 9 (2 0 0 8) 917

in which GBS is becoming an important deformation mode, is


the region of superplasticity for the Sn7 wt.% Bi alloy. The
transition from GBS to slip as dominant deformation mode
occurs in a relatively narrow range of strain rates [26].
The average activation energy for transient creep below
343 K is 23.8 kJ/mol, and above 343 K is 45.95 kJ/mol, Fig. 5 (a, b).
For the steady state creep, Fig. 5 (c, d), it was 42.95 kJ/mol below
343 K, and 72.55 kJ/mol above 343 K. The activation energies at
high temperatures for the steady state creep, Table 2, show
that dislocation creep is the rate controlling mechanism [27].
From Fig. 5 and Table 2 it is clear that the activation energies
for alloy A are higher than those of alloy B. This might be
rendered to the refining effect of Ag which is associated by the
prevention of the formation of the high strength large
structure which forms in the alloy A and needs higher energies
than alloy B.
It is to be noted that grain boundary sliding (GBS) as the
mechanism of plastic deformation in which the most evident
metallographic feature is shearing at grain boundaries, would
promote grain boundary cracking and lead to limited ductility
[16]. It is therefore recognized that for the high strain
accommodation processes such as grain boundary migration
which discourage cracking [28] and local slip are necessary
[29]. The SnBi alloy is expected to be superplastic at room
temperature [15]. It was proposed that the elasticity in
superplastic alloys is determined by the behaviour of grain
boundaries [30]. Superplastic alloys subjected to mechanical
test behave in a different manner than coarse-grained
materials. The major contribution to the nonlinearly elastic
[31] strain, is from an elastic deformation along grain
boundaries due to the reversible motion of grain boundary
dislocations.
The elastic properties of grain boundaries are influenced by
their imperfections and, in particular, by the occurrence of
grain boundary dislocations [32]. This type of imperfections is
usually associated with grain boundary ledges, and consequently the curvature of the grain boundary plane increases
with increasing density of grain boundary dislocations. The
elastic strain along the boundary is thus reduced at a given
shear stress level. The linear density of grain boundary
dislocations depends on the ratio between the rate at which
they are generated on dissociation of the lattice dislocations
that have entered the boundary plane and the rate of their
subsequent annihilation [15].
Thus, in fact, the reason for elastic recovery in the
superplastic alloy may be that the grain-boundary structure
reverts with decreased, density of grain-boundary dislocations. As the strain rate is increased, the conditions of
deformation change to those of normal hot deformation.
According to the network model of Mclean [33] creep
deformation occurs by bowing out of the longest link in the
network which decreases the average length of the remaining
links and rises the flow stress expressed here as the observed
increase in the steady strain rate with increasing the applied
stress at the different deformation temperatures, Fig. 5.
Besides, the formation of subgrain structure during primary
creep leads to an increase in the dislocation density of the
subgrain network to a level that remains essentially constant
during steady state creep [9]. Due to this high dislocation
density, a lower density of dislocations pinned by the

migrating Bi atoms (and/or Ag atoms) exist and an increased


density of the mobile dislocations is expected with values
increasing with increasing the applied stress, as clear from the
increased strain rate in Fig. 4 (e, f).
The increased creep strain by increasing deformation
temperature and the applied stress, Fig. 2, may be due to the
formation of new dislocation sources in the early stage of
creep and the role of the applied stress which increases the
driving force for the mobile dislocations and increases their
density, which in turn contribute to the creep rate, Fig. 4 (cf).
This can be observed from the increasing stress sensitivity
parameter m with increasing temperature, Fig. 6c. The
softening of alloy B over that of alloy A is observed from the
higher level of m for alloy B, Fig. 6c.
The contribution of the deformation temperature to st
proves to be the same as that for the transient creep parameter
, Fig. 4 (cf) and this repeated behaviour satisfies the concept
that creep is a continuous process. The value of (= ln /
lnst) shows to what extent the transient stage characteristics
extend to the steady state creep. The obtained values of from
Fig. 8 were 0.63 for alloy A, and 0.69 for alloy B. The density of
dislocations at the end of the so-called substeady state creep
in the middle of the primary creep state [34] increases within
grains and near grain boundaries. These dislocations get into
uniform distribution where the steady state creep stage starts.
This shows why high values are obtained from Fig. 8. The
obtained m values in the range (1.11.75) and (1.353.4) for
alloys A and B, respectively, Fig. 6 (a, b) are generally smaller
than those suggested by creep theories based on a dislocation
glide along the grain boundaries. These m values exceed the
value 0.5 suggested for creep theories based on dislocation
climb mechanism along the grain boundaries in the investigated temperature range [35].
The activation energy values in the low temperature range
before transformation for both creep stages in both alloys
lying in the range 23.451.2 kJ/mol, Fig. 5, suggest that the
creep process in this stage is controlled by grain boundary
sliding or migration mechanisms, which agree with previous
finding [14]. Such mechanisms might be more predominant
since they are thermally activated processes. Increasing the
deformation temperature of the tested samples a dissolution
process takes place leading to marked variations in composition of the two phases (Sn-rich phase and Bi-rich phase)
associated with diffusion activity. This gives rise to a solution
process at one of the interphase boundaries and a precipitation process at the other interphase boundary. These changes
in composition strongly affect the movements of dislocations
at the interphase boundaries.
At elevated temperatures, if a gliding dislocation was held
by an obstacle, a small amount of climb might permit it to
surmount the obstacle, allowing it to glide to the next set of
obstacles where the process was repeated. The glide step
produces almost all of the strain. The climb process of
dislocations requires diffusion of vacancies or interstitials,
the rate controlling step is therefore atomic diffusion. As
diffusional processes are more difficult in the ordered lattice
[36] (here it is the binary alloy A), thus the activation energies
for the formation and migration of vacancy increase with
order, resulting in higher activation energies for diffusion and
for creep, Fig. 5c (88.1 kJ/mol) compared with Fig. 5d for alloy B

MAT ER IA LS CH A R A CTE R IZ A TI O N 5 9 (2 0 0 8) 9 17

(57 kJ/mol), and therefore lower creep strain [37], Fig. 2 (a, c), as
compared with Fig. 2 (b, d).
On the application of stress the steady creep rate of the
structure existing at the deformation temperature depends
mainly on the interaction between the solute atoms and
dislocations which leads to a viscous motion of dislocations
on their slip planes with increasing dependence on stress and
temperature, Fig. 7a. This consists with the values close to 3
for the stress sensitivity parameter m in Fig. 6c which shows
also that the binary alloy A is harder than alloy B.
The increase in the activation volume by increasing
temperature, Fig. 7c, can be explained on the basis of the
dislocation density because the activation volume is directly
proportional to the average spacing of dislocations [38]. As
thermal energy is provided, the dislocation density decreases
and the average distance between dislocations increases. Fig.
2 shows that the strain in alloy B samples is higher than that
for alloy A under the same conditions. This indicates that alloy
B is more soft than alloy A, which implies that the dislocation
density is lower in alloy B. Consequently, the average distance
between dislocations increases in alloy B than in alloy A. This
leads to an increase in the activation volume by increasing
temperature. This consists with Fig. 7c which shows that the
level of the activation volume values is higher for alloy B.

5.

Conclusions

1. The phase diagram of SnBi shows that Sn dissolves about


1 wt.% Bi. The remaining Bi is presented as free pure Bi
which was detected in X-rays patterns in the binary and
ternary alloys.
2. The results were interpreted on the basis that Bi is in
solution in the ingot alloy and precipitates and coarsens at
room temperature.
3. The increased migration of Bi to Sn grain cores accounts for
the increased Bi grain size at high temperature.
4. Although Ag, which is the more harder, Y(Ag) = 80.5 GN
m 2, is expected to harden the alloy, the results showed
continuous softening in the matrix, Y(Sn) = 52 GN m 2 and
the second phase, Y(Bi) = 34 GN m 2 as the whole structure
can therefore be similar to a non-Newtonian fluid containing rigid particles, which rotate practically without deformation while the soft phases undergo plastic deformation
performing accommodation between both by sliding along
the interfaces.
5. At the transformation temperature 343 K the alloy decomposes into a primary solid solution of Bi in Sn and a Bi
phase with very low Sn content. The operating mechanism
which is GBS changes after transformation into the slip
mechanism.
6. The refining effect of Ag which is associated by the
prevention of the formation of the high strength large
structure lower the activation energies and increases the
activation volume of the Ag containing alloy.

17

REFERENCES
[1] Mc Cormak M, Kammlott GW, Chen HS, Jin S. Appl Phys Lett
1994;65(9):1100.
[2] Wade N, Wu K, Kunii J, Yamada S, Miyahara KK. J Electron
Mater 2001;30(9):1228.
[3] Lang F, Tanaka H, Munegata O, Taguchi T, Narita T. Mater
Charact 2005;24:223.
[4] El-Bahay MM, El-Mossalamy ME, Mahdy M, Bahgat AA. Phys
Status Solidi A Appl Res 2003;198(1):76.
[5] Barry BTK, Thwaites CJ. Tin and its alloys and compounds.
New York: Wiley; 1983 (chaps. 3, 4).
[6] Abd El-Salam F, Mostafa MT, Nada RH, Abd El-Khalek AM.
Phys Status Solidi A Appl Res 2001;185:331.
[7] Glazer J. Almit Tech J 1996;29:1.
[8] Dogra K.S.; Ins. Electronics Packaging. Glen Ellyn (IL), USA,
1986; p. 631.
[9] Abd El-Salam F. Egypt J Phys 1985;16(2):337.
[10] Abd El-Salam F. Egypt J Solids 1984;6(1):101.
[11] Saad G, Abd El-Salam F, Mostafa MT. Surf Technol 1984;22:73.
[12] Nada RH. Radiat Eff Defects Solids 2002;157:521.
[13] Mc Cormak, Chen M, Kammlatt GW, Jin. J Electron Mater (USA)
1997;26(8):954.
[14] Alden TH. Acta Met 1967;15:469.
[15] Valiyev RZ, Emaletdinov AK, KaiByshev OA, Tsanev NK. Phys
Metals 1985;6(5):1008.
[16] Avery DH, Backofen WA. Trans Am Soc Met 1965;58:551.
[17] Hansen M, Anderko K. Constitution of binary alloys. New
York: Mc Grew Hill Book Co. Inc.; 1958. p. 337.
[18] Friedel J. Dislocations. London: Pergamon Press; 1964. p. 304.
[19] Abd El-Khalek AM. Physica B 2002;315:7.
[20] Ashby OD, Dorn JR. Trans AIME 1952;194:959.
[21] Hollomon GH. Trans AIMe 1954;162:268.
[22] El-Sayed MM, Abd El-Haseeb R, Abd El-Salam F, Nagy MR.
Phys Status Solidi 1994;A143:K83.
[23] Semenchenko VK. Surface phenomena in metals and alloys.
London: Pergamon; 1961.
[24] Suery M, Baudelet BB. Phila Mag 1980;41:41.
[25] Toscno EH. Scr Metall 1983;17:309.
[26] Davies PW, Stevens RN, Wilshire B. J Inst Met 1966;94:49.
[27] Rowtherham L, Smith A, Greenough G. J Inst Met 1951;79:439.
[28] Gifkins RC. In: Averbach, et al, editor. Fracture. N.Y.: John
Wiley; 1959. p. 579.
[29] Brunner H, Grant NJ. Trans Am Inst Min Metall Eng
1959;215:48.
[30] Schneibel JH, Mozzlidine PM. Proc. 5th Int. Conf. on Strength
of Metals and Alloys, Aahen; 1979. p. 259.
[31] Schneibel JH. Acta Met 1980;28(10):1527.
[32] Ashby MF. Surf Sci 1972;31:498.
[33] Mclean D. Trans Soc AIME 1968;242:1195.
[34] Bo Wang Liu, Tang C, Qiu W, Tien F. Mat Sci Forum
1995;175:443.
[35] Ball A, Hutchison MM. Met Sci 1969;3(3):1.
[36] Sharma G, Ramanujan RV, Kutty TRG, Tiwari GP. Mater Sci
Eng 2000;A278:106.
[37] Mc Kamey CG, Maziasz PJ, Jones JW. J Mater Res 1992;7:2089.
[38] Garofalo F. Fundamentals of Creep and Creep Rapture in
Metals; 1965. p. 29. New York and London.

Das könnte Ihnen auch gefallen