Among Severe Plastic Deformation (SPD) techniques, such as High Pressure Torsion
(HPT), Twist Extrusion (TE) and Multi Directional Forging (MDF), Equal Channel
Angular Pressing (ECAP) is a novel technique, which is widely used for fabrication
of bulk nanostructure materials (Kim, 2001). The fundamental
of ECAP, developed by Segal and co-workers (Dumoulin et
al., 2005), is schematically shown in Fig. 1. The
well lubricated sample is placed in the entry channel and extruded into the
exit channel by a punch without any changing in the original cross-section (Zhang
et al., 2007). Simple shear, which exist in the intersecting of the
two channels, provides a high strain in the ECAPed material (Yang
et al., 2005). For the ECAP die shown in Fig. 1,
there are two important parameters that play prominent role in the magnitude
of strain. These are die channel angle (Φ) and outer corner angle (Ψ).
Xu et al. (2006b) have suggested a relationship
between the magnitude of effective strain (εeq) and these angles
as shown in Eq. 1:
This process can be repeated several times to produce large strain in the sample.
The magnitude of effective strain is achieved by multiplying the magnitude of
effective strain in the first pass by the number of passes (Xu
et al., 2006a).
|| The schematic illustration of ECAP process
Using classical ECAP die to provide high strain in the sample by repeating
the process, with or without intermediate rotation, is time consuming because
the sample must be removed from the exit channel die and reinserted in order
to achieve large numbers of passes and high imposed strain. To avoid these problems,
ECAP with parallel channels (Rosochowski and Olejnik, 2002)
and ECAP using rotary-die (Ma et al., 2005) have
been proposed. The effect of ECAP process on aluminum 6061 alloy had been investigated
by Mosneaga et al. (2001) and demonstrated the
efficiency of this method to form ultra fine grained materials.
So far, many FEM studies were carried out on the deformation behavior on the
sample in the classical ECAP die (Kim, 2002; Figueiredo
et al., 2006). Djavanroodi and Ebrahimi (2010)
investigated the effects of various values of die channel angle and back punch
pressure on the strain behavior and pressing pressure in two friction conditions.
In general, increasing in the magnitude of die channel angle causes to decrease
in the magnitude of effective strain but increase in the homogeneity of strain
distribution on the cross-section of sample. Increasing in the uniform strain
behavior with using back pressure is another outcome. Zhao
et al. (2005) studied the deformation behavior of pure aluminum during
ECAP with 2D simulation and indicated the in-homogeneity of strain distribution
increases in the case of friction and sharp outer corner die. Basavaraj
et al. (2009) investigated the strain behavior and pressing pressure
on the multi pass ECAP and revealed the prominent of die channel angle in the
strain distribution. Oruganti et al. (2005) reported
the effects of back pressure, friction coefficient and temperature on the material
flow and in-homogeneity of strain and strain rate behavior in the cross-section
of the sample in the ECAP process by 2D FEM. It was demonstrated; high level
of back pressure with reduced friction is required for high magnitude of effective
strain and uniformly strain distribution. Raab (2005) studied
experimentally and analytically the strain distribution in the ECAP with parallel
channels. He has shown that, using of Φ = 100° with k = d (k is the
distance between two parallel channels) provides condition for more homogeneity
of strain distribution. Suo et al. (2007) simulated
the influence of various routes on the material flow during ECAP. They have
shown that, routes BC and C are the best routes to achieve high magnitude
of effective strain with more homogeneity of strain distribution in comparison
with route A. Karpuz et al. (2009) investigated
the effect of strength coefficient and strain hardening exponent on the size
of dead zone in the ECAP process. They showed that, to reach the lowest outer
corner angle, the lowest strain hardening exponent and strength coefficient
ECAP process using rotary-die is shown in Fig. 2. In this
technique, removal of sample from the die channel for each pass is eliminated.
Three equal length punches are placed in the channels as shown in Fig.
2a. The sample is inserted in the die. At the end of first pass (Fig.
2b), the ECAP set-up is rotated through 90° and the sample is pressed
again as shown in Fig. 2c (Ma et al.,
2005). The route of process that is used in this method is A where the sample
has not any rotation between consecutive passes. This route and route BA
have the highest in-homogeneity among of four fundamental routes in the ECAP
process (Kim and Namkung, 2005).
In this study two modified features was studied using 3D FEM: First an ECAP
process that eliminates ejection of the sample from the channels with routes
C (for more uniformly strain distribution) in consecutive passes. The schematically
representation of ECAP process modification is shown in Fig. 3a-d.
In this method two punches are used instead of four punches in the rotary-die
ECAP to solve in-homogeneity of the strain distribution by using route C. After
first pass (Fig. 3c), the ECAP set-up is rotated through 90°
and sample is pressed for the second time. This process is repeated several
times to achieve high strain in the sample. Secondly, an ECAP process in which
the magnitude of effective strain and homogeneity of effective strain in the
cross-section of the sample is increased by modifying the design of ECAP dies
shape as shown in Fig. 4. The effective strain behavior and
in-homogeneity of strain distribution in the cross-section of the sample were
investigated. Also, pressing force required for deforming the samples in simple
and modified ECAP die have been obtained. Four passes are carried out for each
of ECAP dies. Experimental data is used for validation of simulation results.
Finite element methods: All of the simulations were carried out using
commercial FEM code DEFORM-3DTM V5.0. Two distinct models for the
ECAP die were considered that is shown in Fig. 4.
|| The rotary-die ECAP process: (a) initial state, (b) after
one pass and (c) after 90° die rotation
||The modification of ECAP process for achieving uniform strain
distribution without removing the sample from the die: (a) initial state,
(b) start pressing the sample, (c) end of first pass and rotate the ECAP
set-up and (d) start second pass
|| The two ECAP die shapes (simple and modification) simulated
for this study
||The ECAP process using two punches: (a) press with punch A,
(b) end of first pass, (c) press with punch B and (d) end of second pass
In both cases, the dies, punches were assumed to be rigid and the material
properties follow the relationship where,
74 MPa is strength coefficient and 0.17 is strain hardening exponent. The die
channel angle of 90° and the outer corner angle of 20° were used in
both ECAP dies. The speed of 1 mm sec-1 was applied for punches A
and B. The numbers of elements were 3200 and automatic re-meshing was applied
to accommodate large deformation during all of the simulations. All of the simulation
conditions were the same except shape of die models.
At the first, sample up to four passes was simulated, using the simple ECAP die. Magnitude of effective strain and in-homogeneity of strain distribution in the cross-section of the sample was studied. Then, using the modified ECAP die, the distribution of strain and homogeneity of strain distribution was compared with the results of simple ECAP die shape. At the end, the pressing force needed for the simple and modified ECAP dies in the first pass were contrasted.
For simplifying the simulation, two punches system for eliminating the rotation
of ECAP set-up is used with punch A in entrance channel and punch B in existence
channel as shown in Fig. 5a-d. By using
this system, the rotation of die and sample between consecutive passes is eliminated
and simulated of process becomes easier. The sample is placed in vertical channel
(Fig. 5a) and it is pressed by punch A until the sample is
completely driven to the exit channel (Fig. 5b). Until this
time, punch B is free and do not have any contact with the sample. Now, punch
A is become free and then, punch B starts pushing the sample to the vertical
channel (Fig. 5d).
RESULTS AND DISCUSSION
Validation of simulation analysis: To validate simulation results, ECAP
die with the die channel angle of 90° and outer corner of 20° was used.
The commercial pure aluminum was used as sample that homogenized at 670°C
for 0.5 h. The chemical composition of pure Al is represented in Table
1. The speed of punch was 1 mm sec-1 and MoS2 was
used as lubrication. The diameter of sample was 20 mm with a length of 100 mm.
Also, tensile test were performed to obtain true strain-stress relationship
according to ASTM B557M, 2002. The magnitude of friction coefficient in the
simulation condition was assumed 0.1. The hydraulic press, ECAP die set up and
deformed sample are shown in Fig. 6 and 7,
After one pass, the pressing pressure magnitudes obtained from experimental
work and simulation results are 57 and 54 MPa, respectively. There is a good
agreement between experimental and simulation has been achieved.
|| The chemical composition of commercial pure aluminum that
is used as sample during ECAP process
|| The ECAP set-up for experimental work: hydraulic press and
ECAP die set-up
|| The ECAPed samples after one pass in experimental work and
|| Comparison of the magnitude of effective strain results
|| The contours of ECAPed sample after first and third passes
in the simple and modified ECAP dies
Also, magnitudes of simulated effective strain are compared with the Eq.
1 in Table 2. As can be viewed, there are good agreements
between them. The simulation running time for the sample with this geometry
(d = 20 mm and l = 100 mm) was about 13 h. In order to reduce the computational
running time the dimension of sample was reduced to (d = 8 mm and l = 50) for
all of the analyses, with this new dimension, the running time for each simulation
was about 3 h.
Strain behavior of samples during ECAP in simple and modified dies:
Finite element analysis was carried out for the ECAP process of pure aluminum
up to four passes. Figure 8 and 9 compare
strain behaviors in simple and modified ECAP dies. As it can be seen, the number
of passes increases the difference in the effective strain between simple and
modified dies increases. For the first and forth passes this difference is approximately
27 and 38%, respectively. Using the modified ECAP die causes rough surface in
the first, third pass and smoother surface in the second and fourth pass as
shown in Fig. 10. Also, the diameter of the deformed sample
with the modified ECAP die in the first and third passes reduces to 7 mm as
shown in Fig. 8. This decrease in the diameter of the deformed
sample is compensated in the second and fourth passes.
There are three distinct regions in the each ECAPed sample. These regions are
head, steady-state and tail. Between the unavoidable non-uniformly deformed
head and tail regions, there is a steady-state region where the strain distribution
is homogeneous along the deformation axis. Previous simulation studies have
good agreement with these results (Basavaraj et al,
2009; Xu et al., 2006a).
||The contours of ECAPed sample after second and fourth passes
in the simple and modified ECAP dies
|| The 3D shape of sample during ECAP process with simple and
modified ECAP die up to four passes
||The magnitude of effective strain for different number of
pressing passes in two ECAP die conditions
The magnitude of effective strain for each pass of the ECAPed sample which
is represented in Fig. 11 shows higher strain value imposes
to the sample in the modified ECAP die. The magnitude of effective strain increases
34% in the first pass and 38% in the fourth pass, respectively. So, with increasing
in the number of passes, the slope of the growth rises gradually.
||Effective strain distributions in the cross-section of the
ECAPed samples for different passes in simple and modified ECAP die
The effective strain distributions in the cross-section of the ECAPed samples
in the steady state zone for different passes in simple and modified ECAP die
are shown in Fig. 12. As can be viewed, using the modified
ECAP die has several advantages in comparison to simple ECAP die.
|| The magnitude of inhomogeneity index for simple and modified
There is much less in-homogeneity strain in the cross-section of the sample
in the modified ECAP die. Also, the magnitude of effective strain increases
with increasing the number of passes. One disadvantage of modified die is the
high surface roughness of the sample; this can be removed by machining from
the surface of the samples. For simple ECAP die, the level of in-homogeneity
of strain distribution decreases as the number of passes increases. So, increasing
the number of passes, the level of the uniformly strain increases. The result
from the simple ECAP die in the first pass has good agreement with the studies
obtained by Nagasekhar et al. (2005).
The degree of inhomogeneity is defined as (Basavaraj et
maximum, minimum and average values of plastic strain, respectively. The magnitude
of inhomogeneity index ( ) for simple die after one pass ECAP in the steady
state zone is approximately 1.02 which is in good agreement with Basavaraj
et al. (2009) findings. Table 3 shows the magnitudes
of inhomogeneity index for all simulated conditions in the samples (head, steady
state and tail zones). As can be seen, the homogeneity and distribution of the
strain in the modified die is more uniform than simple die. Also, the inhomogeneity
index reduces with increasing the number of the passes.
Pressing pressure requirement in simple and modified ECAP dies: Using
the modified ECAP die needs excess load for pressing sample. Figure
13 has been brought to compare the magnitude of force requirement for the
first pass in the simple and modified ECAP dies. From these curves, it can be
found that, 57% increase in load in the first pass by using the modified ECAP
die is required in comparison to the load requirement for pressing sample in
the simple ECAP die. For the case of modified ECAP die, the force curve is reached
to the load equal with the simple dies (point A) and after this point,
the force increases with a steeper slope to the peak position. At point A, the
sample, reaches the extruded part in the modified ECAP die. For the extruded
part in the modified ECAP die, the diameter of the channel reduces from 8 to
||Punch force versus ram displacement in the first pass for
simple and modified ECAP dies
After the peak position, the requirement force reduces gradually to the 4 kN.
Two modifications are carried out in this study: ECAP process modification
and ECAP die shape modification. The first modification enable the process to
be continues for several passes, resulting high homogeneity of strain distribution
by using route C and the second modification enables us to obtain high magnitude
of effective strain and high uniform strain distribution during ECAP process
respectively. The simple and modified ECAP dies are used to press pure aluminum
up to four passes for comprising magnitude of effective strain and homogeneity
of strain distribution in the cross-section of the sample. From the results
of the simulations, the following conclusions are made:
||About 40% increase in the effective strain magnitude is achieved by using
of modified ECAP die instead of simple ECAP die
||The in-homogeneity of strain distribution that is seen in the deformed
sample with the simple ECAP die is considerably less by applying the modified
||For the deformed sample with the simple ECAP die, the level of the non-uniformly
strain distribution is decreased with increasing the number of the passes
Also, continues of pressing passes is carried out without ejecting the sample form the channels. For validating of FEM results, experimental work, theoretical relationship and previous studies are employed.
We express our thanks to Manufacturing Technology Center (MTC) at Mechanical Engineering Department in Iran University of Science and Technology for assisting in materials testing.