Vous êtes sur la page 1sur 42

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/241100203

Solid-state recycling of aluminium alloy swarf


through cold profile extrusion and cold rolling

Article in Journal of Materials Processing Technology November 2011


DOI: 10.1016/j.jmatprotec.2011.06.010

CITATIONS READS

21 194

3 authors, including:

Ryoichi Chiba Mitsutoshi Kuroda


National Institute of Technology, Asahikawa Yamagata University
59 PUBLICATIONS 281 CITATIONS 104 PUBLICATIONS 1,754 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A theoretical study of grain boundary plane rotation for FCC bi-crystal structure by molecular
dynamics simulation View project

All content following this page was uploaded by Ryoichi Chiba on 26 November 2014.

The user has requested enhancement of the downloaded file.


Solid-state recycling of aluminium alloy swarf
through cold profile extrusion and cold rolling

Ryoichi Chiba(a),*
(a)
Department of Mechanical systems Engineering, Asahikawa National College of
Technology,

2-2-1-6 Shunkodai, Asahikawa, Hokkaido 071-8142, JAPAN

E-mail: chiba@asahikawa-nct.ac.jp

Tel: +81-166-55-8003

Fax: +81-166-55-8003

Tamon Nakamura(b)
(b)
Department of Mechanical systems Engineering, Yamagata University,

4-3-16 Jonan, Yonezawa, Yamagata 992-8510, JAPAN

Present affiliation:

Construction Department, Fujita Corporation,

4-25-2 Sendagaya, Shibuya-ku, Tokyo 151-8570, JAPAN

Mitsutoshi Kuroda(c)
(c)
Graduate school of science and engineering, Yamagata University,

4-3-16 Jonan, Yonezawa, Yamagata 992-8510, JAPAN

*
Corresponding author (R. Chiba)

1
Abstract

In this study, the possibility of solid-state recycling of aluminium alloy machining swarf using cold
extrusion and a subsequent cold rolling process is investigated. Cast Al-Si alloy swarf was cold
compacted into billets and successfully profile-extruded into square bars with a rectangular
cross-sectional aspect ratio of 1:1.8 under an extrusion ratio of 4 or more. After annealing, the extruded
bars underwent multi-pass cold rolling into 1-mm thick strips with a total rolling reduction of 85%.
Optical microscopy demonstrated that in material recycled using only an extrusion process, coarse
residual voids existed in regions where insufficient plastic strain was introduced, causing a visible
expansion of the material during heat treatment. However, uniaxial tensile tests showed that
extrusion-recycled material had a higher mechanical strength than the original aluminium alloy, implying
sufficient bonding among the individual pieces of machining swarf. It was also found that the strength
and density of material recycled through extrusion and an additional rolling process were superior to
material recycled using extrusion only. Moreover, it was observed that the ductility of the recycled
materials was inferior to that of the original aluminium alloy.

Key words: solid-state recycling, aluminum alloy, machining swarf, extrusion, rolling,
mechanical property

2
1. Introduction

Currently, the machining swarf discharged from various factory machine tools is
directly discarded or remelted after being retrieved by scrap metal dealers for recycling
into ingots or die-casting products. However, the recycling efficiency of metal scrap,
including machining swarf, is very low in existing recycling processes. For example,
Lazzaro and Atzori (1992) showed that the metal yield rate in the conventional recycling
of aluminium scrap is approximately 55%. A portion of the metal loss, including the
dross formed during the remelting process, is landfilled. Because the commingling of
non-metal inclusions with molten metal cannot be avoided during the remelting stage, the
purity of recycled ingots is reduced, resulting in a degradation of their mechanical
properties. Moreover, the remelting process consumes a large amount of energy, e.g.,
1619 GJ/t for aluminium (Gronostajski et al., 2000), and is therefore unfavourable from
an economic standpoint.
Several studies on solid-state recycling processes, that is, the direct recycling of
machining swarf into bulk materials using severe plastic working without remelting, have
recently been conducted for the purposes of improving recycling efficiency, energy use
and expense. Through the use of plastic working processes such as extrusion and rolling,
which can be used to recycle scrap materials with microstructural control, machining
swarf can be recycled into materials with excellent mechanical properties. In addition,
energy conservation is possible when recycling metals using these processes;
Gronostajski et al. (2000) estimated that for aluminium, a reduction in energy use of
about 70% is possible compared to the existing melting method. Consequently, these
processes have considerably low recycling costs. They are also better for the
environment because their yield rate for recycled products is quite high. For aluminium,
the yield rate was estimated by Lazzaro and Atzori (1992) to be more than 95%.
A great majority of existing studies have been aimed at recycling machining swarf
into round bars or plates using hot/warm forward extrusion at the appropriate below
melting temperatures. Yoshikawa et al. (1980) examined the surface conditions,
densities, dimensional accuracy, mechanical properties and microstructures of round bar
products that were extruded under various extrusion conditions from compacted
aluminium alloy swarf. Subsequently, they carried out a similar investigation with
3
copper alloy swarf, defining the conditions for making good recycled products
(Yoshikawa and Itoh, 1981). Nakanishi et al. (1995) examined the relationship between
the extrusion ratio and mechanical properties of hot extruded machining swarf of AZ91
magnesium alloy. Gronostajski et al. (1997) described a method for the direct
conversion of aluminium alloy swarf into bar-shaped products through granulating,
pre-pressing and hot extrusion. Fogagnolo et al. (2003) studied a method for recycling
aluminium alloy swarf by cold/hot pressing followed by hot extrusion. Sato et al. (2004)
carried out hot profile extrusion tests for recycling magnesium alloy swarf and
discussed the development of a peculiar and wavy surface roughness observed on the
extrudants. Chino et al. (2004) performed solid-state recycling for machining swarf of
pure iron using hot extrusion followed by annealing to improve the ductility. Suzuki et
al. (2005) investigated the effects of the size and cleanliness of machining swarf and the
extrusion conditions on the mechanical properties and corrosion resistance of recycled
strip-shaped materials. Aida et al. (2005), in a recycling process using hot profile
extrusion, significantly improved the surface roughness of extruded magnesium alloy
swarf by changing the exit radius of the die orifice. Moreover, the same research group
(Aida et al., 2006) studied the effect of the ram speed in hot profile extrusion processes
on the surface quality and mechanical properties of extruded AZ31B magnesium alloy
swarf. Chino et al. (2006) recycled AZ31 magnesium alloy swarf into round bars using
hot extrusion and investigated the relationship between the extrusion condition and the
dispersion state of oxide contaminants in the recycled material. Aida et al. (2007)
obtained finely grained recycled material with a high strain-rate superplasticity using a
hot profile extrusion process for AZ91D magnesium alloy swarf. Fudetani et al. (2008)
conducted hot extrusion of Mg96Zn2Y2 alloy swarf and investigated the superplasticity
properties of the round bar-shaped extrudants. Tekkaya et al. (2009) provided the details
for an elaborate solid-state recycling process for 6060 aluminium alloy swarf using hot
profile extrusion and presented not only the mechanical properties but also the drilling
properties of the extruded swarf. Gley et al. (2010) investigated a direct recycling
method for 1050 aluminium alloy in the form of pins mixed with 6060 aluminium alloy
swarf that used hot profile extrusion at 500 and observed the deteriorating effect of
the cooling lubricant on the mechanical properties of the extrudants, which had a

4
rectangular cross section.
In recent years, related studies using backward extrusion (Murakoshi et al., 2007),
forward extrusion followed by normal rolling (Suzuki et al., 2003) or differential speed
rolling (Suzuki et al., 2007), as well as special processing methods that can introduce
large strain such as equal-channel angular pressing (ECAP) (Aida et al., 2004), high
pressure torsion (HPT) (Takahashi et al., 2009), cyclic extrusion compression (CEC)
(Peng et al., 2009) and friction extrusion (Tang and Reynolds, 2010) have been reported.
Although the studies mentioned above have targeted single-phase materials, the
solid-state recycling of machined chips into composite materials has also been attempted.
These have included magnesium-based composites with SiC particles (Aida et al., 2008)
and aluminium-based composites with reinforcing particles such as tungsten, which was
adopted by Gronostajski et al. (1996), and SiC, which was utilised by Tekkaya et al.
(2009). It was recently revealed that the hot extrusion of aluminium alloy machined
chips with pure aluminium powder added as a binder achieves better bonding and
results in improved mechanical properties for recycled products made from two-phase
materials (Sherafat et al., 2009).
However, most recycling processes based on hot working require comparatively
large amounts of thermal energy. As an alternative, the recycling of machining swarf
through cold working has been investigated to reduce processing costs and energy
requirements. This process has the potential for greatly improving material properties and
energy efficiency; it uses only approximately 2% of the energy of existing recycling
processes (Allwood et al., 2005). Furthermore, the process requires only simple
equipment to convert scrap directly into a material product.
There have been a limited number of studies on the recycling of metal scrap using
severe cold working. Allwood et al. (2005) cold-extruded finely chopped fragments of
A1050-H14 aluminium alloy sheets at different extrusion ratios and showed that the
extent of bonding improves with an increased extrusion ratio. However, the study did not
reveal the resultant density, mechanical properties, or microstructure of the extrudants.
The present study investigated the possibility of recycling Al-Si alloy machining
swarf using cold extrusion and a subsequent cold rolling process. After mill-generated
machining swarf was cold-compacted into billet form, square bars were produced

5
through annealing and a subsequent cold profile extrusion process. The bars were then
cold-rolled into strips. A comparison was then made of the surface appearances,
densities, microstructures and mechanical properties at room temperature of the recycled
and virgin/original materials. Finally, the possibility of solid-state recycling of aluminium
alloy machining swarf at room temperature was discussed.

2. Experimental procedure

2.1. Preparation of aluminium-alloy machining swarf

Machining swarf was prepared by side-milling a commercial AC4CH aluminium


alloy ingot (Mitsubishi Plastics, Inc.) using cutting oil. This material was chosen because
it is a cast alloy with a relatively coarse biphasic microstructure, making it easy to
observe the processing state within the material, which is suitable for verifying the
effectiveness of the presented recycling process. The chemical composition of the
AC4CH aluminium alloy is given in Table 1. The milling conditions were as follows: a
cutting rate of 46 m/min, feed rate of 0.1 m/min and cut depth of 1 mm. Figure 1(a) shows
the appearance of the machining swarf. Needle-shaped, spirally-twisted machining swarf
that was about 2530 mm long and 1 mm wide was obtained.
The machining swarf was then degreased with acetone inside an ultrasonic bath for
10 min. It would be financially impractical to completely remove the extraneous matter
remaining on the swarf (e.g., cutting oil), and it has been reported that in the recycling of
6061 aluminium alloy machining swarf through hot extrusion, the effects of the cutting
oil on the mechanical properties of the extrudant are negligible (Suzuki et al., 2003). In
the present study, however, the machining swarf was carefully cleaned to exclude any
factors other than the properties of the target material.

2.2. Pre-compaction of machining swarf

For use as billets in the extrusion process, the cleaned machining swarf was filled
into a container for powder compaction and uniaxially compressed at 303 MPa (Hu et al.,

6
2008) under room temperature into a cylinder 20.5 mm in diameter and approximately 12
mm high. Figure 1(b) shows the appearance of the as-compacted machining swarf.
Machining swarf interfaces were clearly visible to the naked eye. A 500-kN hydraulic
universal testing machine (Takes Group Co., Ltd.) was used for the compaction, as well
as for the subsequent extrusion process described in the next subsection. Because the
densities of the compacts made under various load holding times of 0 to 180 s were
almost equal, the compacts were unloaded immediately after the loading pressure
reached 303 MPa.
Before the subsequent extrusion process, the compacts were annealed for 1 h at
300. This was done because, during the cold extrusion of the as-compacted billets,
some large cracks appeared in the extrudants, and the extrusion load exceeded the
capacity of the testing machine, thereby making successful extrusion impossible.
Alternatively, it would be possible to anneal the swarf before compaction, which would
enhance the recrystallisation in the swarf owing to the mass effect (Sawamura et al.,
2005) and lead to a further softening of the billets. As a result, the billets would be
extruded at a smaller extrusion load.

2.3. Cold profile extrusion

After two of the annealed cylindrical compacts described in subsection 2.2 were
placed in a container for extrusion, they were profile-extruded into square bars at room
temperature. A half-angle of 60 was adopted for the die (Takahashi and Ohuchi, 1980)
to avoid the existence of dead metal in the extrusion and provide a minimum extrusion
pressure. Table 2 shows the dimensions of the die orifices used in our experiments. Each
die included a rectangular-shaped orifice with an aspect ratio of either 1:1.8 or 1:3.8, the
dimensions of which were determined based on extrusion ratio R (for R = 4, see Figs.
2(b) and (d)). A ratio of R = 4 was chosen based on a study on the solid-state recycling of
finely chopped fragments of aluminium alloy sheets using cold extrusion (Allwood et al.,
2005), where the present authors interpreted from macroscopic pictures of the extrudants
that a ratio of R = 4 would be needed for the solid-state bonding of aluminium alloys. For
comparison, the compacts were also extruded into square bars using a die with an orifice

7
having an aspect ratio of 1:1.8 and extrusion ratio of R = 6.
Applying grease on the inner surface of the container, in addition to the conical
surface and land surface of the dies at every extrusion prevented an increase in the
extrusion load as a result of friction. The extrusion load was read directly from the load
meter of the testing machine. The extrusion ram speed was manually maintained at
approximately 2 mm/min using the hydraulic valve control of the testing machine.

2.4. Cold rolling

Rolling was performed to shape the material recycled through the extrusion into a
plate shape and improve its density and mechanical properties. The steady deformation
region (excluding the ends) of the square bars with the 1:1.8 aspect ratio and R = 4
obtained by the extrusion was cut into 25-mm long pieces, which were then rolled
repeatedly along the extrusion direction (ED) into 1-mm thick strips at room temperature.
The cut pieces of the extruded material were annealed at 300 for 1 h only before the
first pass of rolling, because the cold rolling of the as-extruded material tended to cause
cracks on the lateral surfaces. A self-manufactured rolling machine with a pair of
150-mm diameter rolls was used. The rolling conditions were as follows: 20 passes in
total with a 9% rolling reduction per pass at a revolution speed of 1 rad/s under
non-lubricated conditions. That is, the thickness of the cut pieces of the extruded
material was reduced by a total of 85% through the rolling process. As shown in Table 3,
the rolling reduction actually varied at each pass because of the elastic deformation of the
rolls and shafts of the rolling machine.

2.5. Testing of mechanical properties and microstructural analysis

Tensile specimens with a 4-mm gage width and 10-mm gage length were cut out
from an original AC4CH ingot, extruded bars and rolled strips using a milling machine
(the latter two are hereafter called extrusion-recycled material and rolling-recycled
material, respectively). For the recycled materials, the specimens were cut out from each

8
steady deformation region*, and their tensile axis was parallel to the extrusion/rolling
direction. Figure 3 shows the detailed dimensions of each specimen. Because the
extrusion-recycled material had sufficient thickness to prevent the specimen from
fracturing under the maximum load of the testing machine, the specimen was bisected
along the dashed line shown in the figure using a wire electric discharge machine.
Moreover, for the rolling-recycled material, the width of the grip sections and radius of
the shoulder fillets in the specimen were decreased because the available width for
cutting out the specimen was limited. However, all of the specimens conformed in terms
of gage width and gage length. Tensile tests were conducted at room temperature at a
crosshead speed of 2 mm/min using a desktop universal testing machine
(INSTRON-3365). Their strain was measured using an extensometer (10-mm gauge
length).
Additionally, Vickers hardness tests were carried out at an applied load of 1.96 N
and a holding time of 15 s using a microhardness tester (Akashi model-MVK). After
measuring at five points having different z-coordinate values on a line with a certain x
coordinate value in the gray plane in Fig. 4(a), the average of three measured values,
excluding the maximum and minimum, was considered as the Vickers hardness value at
the x-coordinate position.
The microstructures of the recycled materials were observed using optical
microscopy. The observed points were located in planes parallel to the extrusion/rolling
direction, as shown in Fig. 4. After polishing with emery paper and buffing with an
alumina suspension for microstructure observation, the specimens were etched using a
solution of 0.5 ml of 40% hydrofluoric acid and 100 ml of distilled water.

2.6. Finite element analysis of extrusion process

To investigate the strain distribution in the plane perpendicular to the ED for the
extruded bar with an aspect ratio of 1:1.8 and R = 4, an extrusion analysis was performed

* Using a grid method, it was confirmed that a region of the extrusion-recycled material more
than 20-mm away from the front end was subjected to steady deformation.
9
using DYNA3D, which is an explicit nonlinear finite element code. Considering the
symmetry of the deformation, a quarter of a billet with a diameter of 20 mm and length of
20 mm was meshed with eight-node solid elements with a single integration point per
element. The numbers of divisions in each direction were as follows: 31 along the radial
direction, 10 along the circumferential direction and 120 along the axial direction. The
billet was assumed to be an isotropic power-law hardening elastic-plastic body obeying
the isotropic hardening rule and von Mises yield criterion, whereas the tools were defined
to be perfectly rigid bodies; that is, material type 18 was chosen from the DYNA3D
material model for the former and material type 20 for the latter. The experimental flow
stress data obtained in uniaxial compression tests of the annealed AC4CH ingot at an
initial strain rate of 0.005 s1 at room temperature were used as the deformation resistance
properties in the constitutive model. The experimental stress-strain curve is shown in Fig.
5 along with its power-law hardening approximation for parameters given in Table 4.
The friction between the billet and tools was considered to obey the Coulomb rule, with
an assumed constant friction coefficient of 0.1. The extrusion ram speed was fixed at 10
m/s to reduce the CPU cost and improve the performance (speed scaling). It should be
noted that the dynamic effect on the deformation from this technique was verified to be
negligible: the differences in the calculated results were negligible even when the ram
speed was changed to 5 m/s or 1 m/s.

3. Results

3.1. Cold profile extrusion process

The appearances of specimens obtained using cold profile extrusion with R = 4 and
corresponding dies are shown in Fig. 2. The maximum extrusion loads were
approximately 200 kN and 220 kN for the die orifice aspect ratios of 1:1.8 and 1:3.8,
respectively. As shown in Fig. 2, both specimens showed straight extrusion without
warping. Unbonded swarf was observed at the front end of each specimen because
insufficient strain was introduced in this region. The surfaces of the extruded specimens
were considerably hardened according to the degree of work hardening through large

10
plastic deformation at room temperature. The specimens with the 1:1.8 aspect ratio were
completely shiny and smooth square bars except for their front end sections. On the other
hand, large cracks, some of which reached the centre area, were observed in the
specimens with the 1:3.8 ratio. Because it is obvious that the extruded specimens with the
1:3.8 aspect ratio cannot be used as recycled material, the authors do not consider them as
a target for discussion.
When the extruded square bars were annealed for the subsequent rolling process,
visible expansion was observed in some, but not all, of the specimens, as shown in Fig. 6.
In the extrusion-recycled material with an extrusion ratio of R = 4 and the 1:1.8 aspect
ratio, an equal number of expanded and unexpanded specimens were obtained during the
annealing process. In contrast, no extruded specimens with a ratio of R = 6 and the 1:1.8
aspect ratio caused such a shape variation after being annealed under the same conditions
(300-1 h). Note that the maximum extrusion load was approximately 240 kN in the
extrusion for R = 6 and the 1:1.8 aspect ratio.

3.2. Cold rolling process

The extruded material with a ratio of R = 4 and the 1:1.8 aspect ratio that remained
unexpanded during the annealing stage was additionally processed using multi-pass
rolling. Figure 7 shows the appearance of a rolled specimen. Although several cracks
with lengths up to 3 mm are observed at both edges, and slight warping in the normal
direction (ND) can also be seen, a shiny solid strip was obtained. In contrast, after rolling
the extruded specimens that expanded during the annealing process, many large cracks
were observed in the rolled specimens. Some specimens split during the early stages of
the multi-pass rolling because of the enlargement of voids by the annealing process, with
the cracks appearing at the surfaces triggering fractures.

3.3. Microstructures of recycled materials

The extruded- and additionally-rolled specimens were cut using a wire electric
discharge machine, and their heat-affected zones were removed using emery paper prior

11
to microstructure observation. In addition, in advance of our microscopic observation,
several cut planes were macroscopically inspected. While no swarf interfaces were found
in any of the planes, in some extruded specimens for R = 4 and the 1:1.8 aspect ratio,
voids of up to approximately 100 m were observed at locations near the extrusion axis in
the central longitudinal plane (the gray plane in Fig. 4), as shown in Fig. 8.
Figure 9 shows the microstructures at the respective observation points indicated in
Fig. 4. For reference, the microstructure of the original AC4CH ingot is shown in Fig.
9(a). In the original ingot, needle-like eutectic Si phases were observed in coarse primary
Al phases. In the extrusion-recycled material shown in Figs. 9(b)(d), no machining
swarf interfaces were found; however, some voids can be observed in Fig. 9(b). Grain
boundaries were observed irrespective of the observation point. The crystal grains were
comparatively equiaxed at observation point A (Fig. 9(b)), which was located on the
extrusion axis, and were significantly stretched along the ED at point C (Fig. 9(d)), which
was near the external surface.
In general, the plastic strain introduced into billets through an extrusion process is
non-uniformly distributed in the cross section perpendicular to the ED. An elastic-plastic
FE analysis conducted to simulate the present extrusion experiment also demonstrated
that the strain had a pronounced distribution in the cross section (see section 3.6).
Therefore, the number and size of the voids remaining in the extruded specimen per unit
area varied in the cross section depending on the locally introduced plastic strain, as
shown in Figs. 9(b)(d). It can also be seen that, in response to the strain level, the
eutectic Si phases broken into short needle-like particles (average length of 15 m)
through the cutting work were further refined (about 5 m on average) near the external
surface of the specimen (Fig. 9(d)).
In the rolling-recycled material, as shown in Fig. 9(e), the eutectic Si phases were
considerably further refined than in the extrusion-recycled material, dispersing along the
RD in the specimen without aggregation. This microstructure was remarkably similar to
the one observed in an Al-Si alloy (AC4C) that was hot-rolled for a rolling reduction of
80% (Toyama et al., 2010). In Fig. 9(e), the voids are collapsed to the extent that they
cannot be identified; however, because the minimum size for an accurate identification of
voids in our optical microscope was approximately 5 m, the existence of voids smaller

12
than 5 m cannot be denied. It should be noted that, unlike in the extrusion-recycled
material, no grain boundaries were observed in the rolling-recycled material. This is
probably because, owing to severe cold working, the grain sizes were too small to be
identified in the microscope.

3.4. Relative density

The densities of the specimens were measured using the Archimedes method at each
stage of the recycling process. The relative densities at each stage compared to the density
of the original AC4CH ingot are presented in Fig. 10. The 92.7% average value for the
cold compacts (i.e., billets) increased to 96.9% for the extruded specimens for R = 4 and
the 1:1.8 aspect ratio and to 97.7% for the additionally-rolled specimens. As shown in Fig.
9, the severe plastic working processes caused a step-by-step collapsing of voids inside
the material, resulting in an increase in relative density. A relative density of 99.0% was
obtained for the extrusion-recycled material for R = 6 and the 1:1.8 aspect ratio, which
unexpanded with a 100% probability during the annealing process.

3.5. Mechanical properties at room temperature

Figure 11 shows the 0.2% proof stress, ultimate tensile strength and fracture strain of
the respective specimens. The recycled materials had higher strengths than the original
AC4CH ingot. It can also be seen that the strengths were higher in the rolling-recycled
material than in the extrusion-recycled material. In contrast, the recycled materials were
inferior to the original material in terms of fracture strain.

3.6. Numerical prediction of introduced strain

Figure 12 shows the distribution of equivalent strain obtained through an FE


analysis in a square bar extruded for R = 4 and the 1:1.8 aspect ratio. Variations in the
equivalent strain in the cross section perpendicular to the ED in the steady deformation
region are shown, in which the effects of the end section vanish sufficiently. The

13
variation along the long axis of the rectangular cross section from the centre is shown in
the right half of the figure, whereas variation along the short axis is shown in the left
half. The measured values of Vickers hardness at certain locations on the long axis (i.e.,
x-axis in Fig. 4(a)) are also presented in the figure. The distribution tendencies of the
equivalent strain and Vickers hardness are in agreement, and therefore the FE analysis
results shown are generally valid. The FE analysis shows that an equivalent strain of
about 1.5 is introduced at the centre of the cross section, whereas a huge strain
exceeding 4 is introduced at the external surfaces.

4. Discussion

The possibility of using cold working for the solid-state recycling of machining
swarf into plate-shaped bulk material, which is as much in demand as rod material, is
discussed. It has been reported that hot working can directly recycle compacted
machining swarf into plate-shaped bulk material. For example, Suzuki et al. (2005)
successfully obtained a strip with a 545mm rectangular cross-section (an aspect ratio
= 1:9) from 6061 aluminium alloy swarf using hot profile extrusion. However, because
cold working is accompanied by an inferior deformability of the material, the extrusion
into a square bar with an aspect ratio of 1:3.8 and R = 4 at room temperature failed (Fig.
2(c)) in the present study, even though the extrusion into one with an aspect ratio of
1:1.8 and R = 4 succeeded (Fig. 2(a)). This shows that when using cold working for
solid-state recycling into a plate-shaped bulk material, it is very difficult to obtain the
bulk material directly from the extrusion process. Thus, it is necessary to first make a
recycled square bar with an aspect ratio of 1:1.8, followed by the formation of a plate
shape via secondary processing (e.g., rolling).
The cause of the volume expansion observed in the annealing process for the
extrusion-recycled material, as shown in Fig. 6, is next discussed. The relative density
of the extrusion-recycled material for R = 4 and the 1:1.8 aspect ratio was lower than
97% (see Fig. 10), and thus a number of voids existed in the material. This fact was
confirmed by optical microscopy, and both the number and size of the voids were found
to be large in the region with a small introduced strain, which the FE analysis showed
14
was the neighbourhood of the extrusion axis. Therefore, this expansion was probably
caused by heat-expanded air within the relatively large voids remaining inside the
specimen. In the extrusion-recycled material for R = 4 and the 1:1.8 aspect ratio, voids
of up to approximately 100 m were observed, as shown in Fig. 8, and the
extrusion-recycled material that included such coarse voids would correspond to the
specimens that expanded during the annealing process.
Gronostajski et al. (1996) hot-extruded cylindrical compacts of machining swarf
composed of a different kind of aluminium alloy, namely AlCu4, at 500 for R = 4,
and the relative density of the obtained recycled rods was approximately 99%. From a
comparison of this fact with Fig. 10, it is obvious that for an identical extrusion ratio,
hot extruded machining swarf is denser than cold extruded swarf.
Figures 9(b)(d) show that no machining swarf interfaces existed in the
extrusion-recycled material. Thus, the individual pieces of swarf appear to be sufficiently
bonded to each other, even for R = 4, in the cold extrusion. Note that Gronostajski and
Matuszak (1999) stated that for the solid-state bonding of comminute aluminium alloy
chips by hot extrusion, the extrusion ratio should be at least 4. The degree of solid-state
bonding depends mainly on the plastic strain introduced. In the present extrusion process,
the introduced strain was the smallest at the centre of the cross section, as shown in Fig.
12. From the FE result and optical micrograph of Fig. 9(b), it can be inferred that an
introduced equivalent strain of approximately 1.5 achieves solid-state bonding but
allows relatively large voids to remain in the material, which might cause volume
expansion during the annealing process. In other words, for this level of introduced
strain, although the surface oxide films of the machining swarf were broken, and
individual pieces of the swarf were bonded to each other, the voids were not sufficiently
collapsed, leading to a decrease in the relative density of the recycled material.
The distribution of equivalent strain in the square bar extruded for R = 6 and the
1:1.8 aspect ratio was not analysed because a large mesh distortion obstructed a correct
FE computation. However, the equivalent plastic strain introduced at the centre of the
rectangular cross-section, i.e., the minimum value of equivalent plastic strain introduced,
can be predicted to be at least 1.8, which is calculated assuming an equi-biaxial
compressive deformation. Thus, introducing an equivalent plastic strain of over 200% is

15
suggested to exclude any voids that might cause volume expansion during the annealing
process.
Figure 11 shows that the recycled materials had higher strengths than the original
AC4CH ingot. This resulted from the fact that the solid-state bonding among individual
pieces of swarf was sufficiently achieved by the profile extrusion for R = 4 and the 1:1.8
aspect ratio, and additionally, the material was strengthened via grain and Si phase
refinements because of the severe cold working. The reason that such high mechanical
strengths were obtained even though the relative density of the recycled materials was
lower than 98% is discussed here. Mugica et al. (2004) presented the relationship
between the porosity and tensile properties in a cast Al-Si alloy whose chemical
composition differed slightly from AC4CH, reporting that the yield strength (0.2%
proof stress) and tensile strength were insensitive to varying porosity in the range of
volumetric total porosity lower than 3%; even when a porosity of about 3% was
included, the yield strength differed little from that in the case of a porosity lower than
0.1% and the tensile strength decreased by only about 20%. Hence, the high mechanical
strengths of the recycled materials observed in the present study were a consequence of
an increase in the strengths by grain and Si phase refinements as the result of the severe
cold working, which exceeded the slight decline in the strengths depending on the
porosity.
Moreover, Fig. 11 shows that the strengths were higher in the rolling-recycled
material than in the extrusion-recycled material. This was because of the fine dispersion
of Si particles and the improvement in the density by void collapse, as shown in Fig. 9.
During the rolling process, the equivalent plastic strain introduced into a workpiece can
be approximated by considering the constant volume law as follows:

2
p = y2 + z2 + y z
3 , (1)

y = ln(1 r ) , (2)

w1
z = ln , (3)
w0

where r denotes the rolling reduction; w0 and w1 are the width of the workpiece before and
16
after rolling, respectively; y indicates the true thickness strain; and z is the true width
strain. Actually, the workpiece not only stretches/compresses along the RD, TD and ND,
it also undergoes shear deformation at every pass of the multi-pass rolling process.
However, the effects of shear deformation are ignored here for simplicity. Thus, the
equivalent plastic strain calculated in Eq. (1) somewhat underestimates the strain actually
introduced into the workpiece. In our experiment, because the total rolling reduction was
85%, w0 = 12 mm and w1 = 14.5 mm, Eq. (1) yields p = 2.1. That is, improvements in

the density and mechanical properties were achieved to the extent shown in Figs. 10 and
11 by introducing an equivalent plastic strain of at least 2.1 into the annealed
extrusion-recycled material.
Figure 11 also shows that the recycled materials were inferior to the original
material in terms of fracture strain. The possible reasons for this are as follows: (i) the
densities of the recycled materials were lower than that of the original ingot, and (ii)
dislocations were accumulated in the materials through the severe cold working, leading
to a decrease in fracture strain during the tensile tests. In contrast, material recycled by
hot working has a fracture strain that is comparable or superior to that of the original
material from which the machining swarf is prepared (Suzuki et al., 2007) because
dynamic recrystallisation during the hot working and static recrystallisation after the hot
working take place (Tekkaya et al., 2009)

17
5. Conclusions and future work

Recycled materials were obtained from Al-Si alloy (AC4CH) machining swarf by
cold-compaction followed by annealing and severe cold plastic working. Based on an
investigation of their microstructures and mechanical properties at room temperature, the
following conclusions can be drawn:
(1) The cold profile extrusion and additional cold rolling of the machining swarf produce
recycled materials that have mechanical properties comparable to the original ingot.
(2) During solid-state recycling into plate-shaped bulk material via cold working, it is
difficult to obtain the bulk material directly through the extrusion process. It is thus
necessary to first make a recycled square bar with an aspect ratio of 1:1.8, and then
shape it into a plate form using a secondary processing stage (e.g., rolling).
(3) In the profile extrusion of a cylindrical compact into a rectangular bar with a
cross-section aspect ratio of 1:1.8, an extrusion ratio of 4 sufficiently bonds individual
pieces of the machining swarf, producing recycled materials with excellent
mechanical properties better than those of the original ingot. However, coarse voids
remain around the extrusion axis because of the insufficient plastic strain introduced
during the processing.
(4) An extrusion ratio of at least 6, that is, an introduced equivalent strain of about 200%,
is needed if heat treatment of the extrusion-recycled material is intended (e.g., for a
secondary processing stage). The extrusion-recycled material with an extrusion ratio
of 4 has the possibility of visibly expanding during the heat treatment process and
simultaneously breaking at the external surface.
(5) The rolling-recycled material is superior to the extrusion-recycled material in
mechanical strength because of the fine dispersion of Si particles and an increase in
density caused by further void reduction.
It has been reported that in the cold solid-state recycling of 6061-T651 aluminium
alloy machining swarf, the tensile strength of the recycled material increases when a
powder of the identical material is added to the swarf (Fujii et al., 2003). Therefore, the
extrusion of billets made from machining swarf mixed with a certain amount of powder
may allow the powder to fill in the voids included in the extrudant, resulting in an
increase in density and mechanical properties in our experiments as well. Furthermore, it
18
is necessary to investigate the effects of machined chip size on the mechanical properties
and microstructure of solid-state recycled materials because it is known that such effects
are not negligible in other materials: AZ91D magnesium alloy (Hu et al., 2008) and
AZ31B magnesium alloy (Wu et al., 2009).
In the present study, the rolling process was repeated along the ED to form strips.
Appropriate changes in the rolling direction during the multi-pass rolling process may
allow us to control the shape of the plate-shaped recycled material to a certain extent.
Thus, this is worth pursuing in future research.

19
Acknowledgements

This work was financially supported by KAKENHI, a Grant-in-Aid for Encouragement of Scientists
(21919020). The authors thank Mr. Y. Shima for preparing the pre-compaction and extrusion tools. Dr. N.
Horikawa and Prof. T. Sato are gratefully acknowledged for their support of our microstructural
observations.

References
Aida, T., Takatsuji, N., Matsuki, K., Kamado, S., Kojima, Y., 2004. Homogeneous consolidation process
by ECAP for AZ31 cutting chips. Journal of Japan Institute of Light Metals 54, 532-537. (in Japanese)
Aida, T., Takatsuji, N., Matsuki, K., Kawabe, Y., Kamado, S., 2007. Grain refining by hot extrusion of
AZ91D magnesium alloy machined chips and resulting high strain rate superplasticity. Journal of Japan
Institute of Light Metals 57, 391-397. (in Japanese)
Aida, T., Takatsuji, N., Matsuki, K., OHara, T., Kamado, S., 2005. Improvement in surface properties of
extrusions from Mg-Al-Zn based alloy machined chips. Journal of Japan Institute of Light Metals 55,
400-404. (in Japanese)
Aida, T., Takatsuji, N., Matsuki, K., OHara, T., Kamado, S., 2006. Effect of extrusion speed on properties
of the extruded AZ31B magnesium alloy machined chip. Journal of Japan Institute of Light Metals 56,
166-171. (in Japanese)
Aida, T., Takatsuji, N., Matsuki, K., Satou, T., Kamado, S., 2008. Mechanical properties of SiC
particle-AZ31B magnesium alloy machined chips composites prepared by hot extrusion after ECAP.
Journal of Japan Institute of Light Metals 58, 104-110. (in Japanese)
Allwood, J.M., Huang, Y., Barlow, C.Y., 2005. Recycling scrap aluminium by cold-bonding, In:
P.F.Bariani (Ed.), the 8th International Conference on Technology of Plasticity. Edizioni Progetto,
Verona, pp. 311-312.
Chino, Y., Hoshika, T., Lee, J.S., Mabuchi, M., 2006. Mechanical properties of AZ31 Mg alloy recycled
by severe deformation. Journal of Materials Research 21, 754-760.
Chino, Y., Iwasaki, H., Mabuchi, M., 2004. Solid-state recycling for machined chips of iron by hot
extrusion and annealing. Journal of Materials Research 19, 1524-1530.
Fogagnolo, J.B., Ruiz-Navas, E.M., Simon, M.A., Martinez, M.A., 2003. Recycling of aluminium alloy
and aluminium matrix composite chips by pressing and hot extrusion. Journal of Materials Processing
Technology 143-144, 792-795.
Fudetani, S., Itoi, T., Kubo, T., Kawamura, Y., Hirohashi, M., 2008. Microstructure and mechanical
properties of Mg96Zn2Y2 alloy prepared by extrusion of machined chips. Journal of Japan Institute of
Light Metals 58, 54-57. (in Japanese)

20
Fujii, Y., Toda, H., Kobayashi, T., 2003. Creation of aluminum alloy by severe plastic deformation of
cutting chip and its mechanical properties. Journal of Japan Institute of Light Metals 53, 368-372. (in
Japanese)
Gronostajski, J., Marciniak, H., Matuszak, A., 2000. New methods of aluminium and aluminium-alloy
chips recycling. Journal of Materials Processing Technology 106, 34-39.
Gronostajski, J., Matuszak, A., 1999. The recycling of metals by plastic deformation: an example of
recycling of aluminium and its alloys chips. Journal of Materials Processing Technology 92-93, 35-41.
Gronostajski, J.Z., Kaczmar, J.W., Marciniak, H., Matuszak, A., 1997. Direct recycling of aluminium
chips into extruded products. Journal of Materials Processing Technology 64, 149-156.
Gronostajski, J.Z., Marciniak, H., Matuszak, A., 1996. Production of composites on the base of AlCu4
alloy chips. Journal of Materials Processing Technology 60, 719-722.
Gley, V., Khalifa, N. B., Tekkaya, A. E., 2010. Direct recycling of 1050 aluminum alloy scrap material
mixed with 6060 aluminum alloy chips by hot extrusion. International Journal of Material Forming 3,
853-856.
Hu, M., Ji, Z., Chen, X., Zhang, Z., 2008. Effect of chip size on mechanical property and microstructure
of AZ91D magnesium alloy prepared by solid state recycling. Materials Characterization 59, 385-389.
Lazzaro, G., Atzori, C., 1992. Recycling of aluminium trimmings by conform process. Light Metals 3,
1379-1384.
Mugica, G.W., Tovio, D.O., Cuyas, J.C., Gonzalez, A.C., 2004. Effect of porosity on the tensile
properties of low ductility aluminum alloys. Materials Research 7, 221-229.
Murakoshi, Y., Hatsukano, K., Matsuzaki, K., 2007. Consolidation of Mg alloy chips by hot-extrusion
and their mechanical properties. Journal of the Japan Society of Powder and Powder Metallurgy 54,
653-657. (in Japanese)
Nakanishi, M., Mabuchi, M., Kubota, K., Higashi, K., 1995. Relationship between extrusion ratio and
mechanical properties of extruded machined-chips of AZ91 magnesium alloy. Journal of the Japan
Society of Powder and Powder Metallurgy 42, 373-377. (in Japanese)
Peng, T., Wang, Q.D., Lin, J.B., 2009. Microstructure and mechanical properties of Mg-10Gd-2Y-0.5Zr
alloy recycled by cyclic extrusion compression. Materials Science and Engineering: A 516, 23-30.
Sato, H., Aida, T., Takatsuji, N., Matsuki, K., Murotani, K., 2004. Hot extrudability of Mg-Al-Zn based
alloy cutting chips and surface properties of extrusions. Journal of Japan Institute of Light Metals 54,
14-18. (in Japanese)
Sawamura, J., Toda, H., Kobayashi, T., Niinomi, M., Akahori, T., 2005. Effect of residual cutting oil in
heavily deformed in-situ composite produced by cutting chips of dissimilar metals, the 108th
springtime conference of Japan institute light metals, Toyohashi, pp. 323-324. (in Japanese)
Sherafat, Z., Paydar, M.H., Ebrahimi, R., 2009. Fabrication of Al7075/Al, two phase material, by
recycling Al7075 alloy chips using powder metallurgy route. Journal of Alloys and Compounds 487,
395-399.

21
Suzuki, K., Huang, X., Watazu, A., Shigematsu, I., Saito, N., 2007. Recycling of 6061 aluminum alloy
cutting chips using hot extrusion and hot rolling. Materials Science Forum 544-545, 443-446.
Suzuki, K., Shigematsu, I., Imai, T., Saito, N., 2005. Influences of chip characteristics and extrusion
conditions on the properties of a 6061 aluminum alloy recycled from cutting chips. Journal of Japan
Institute of Light Metals 55, 395-399. (in Japanese)
Suzuki, K., Shigematsu, I., Xu, Y., Imai, T., Saito, N., 2003. Mechanical and corrosion properties of 6061
aluminum alloys recycled by hot-extrusion of cutting chips. Journal of Japan Institute of Light Metals
53, 554-560. (in Japanese)
Takahashi, H., Ohuchi, K., 1980. A study of cold extrusion of sintered copper preform. Journal of the
Japan society for technology of plasticity 21, 967-974. (in Japanese)
Takahashi, T., Kume, Y., Kobashi, M., Kanetake, N., 2009. Solid state recycling of aluminum machined
chip wastes by compressive torsion processing. Journal of Japan Institute of Light Metals 59, 354-358.
(in Japanese)
Tang, W., Reynolds, A.P., 2010. Production of wire via friction extrusion of aluminum alloy machining
chips. Journal of Materials Processing Technology 210, 2231-2237.
Tekkaya, A.E., Schikorra, M., Becker, D., Biermann, D., Hammer, N., Pantke, K., 2009. Hot profile
extrusion of AA-6060 aluminum chips. Journal of Materials Processing Technology 209, 3343-3350.
Toyama, K., Matsuura, K., Ohno, M., Sato, T., Nakayama, M., 2010. Changes in microstructure and
mechanical properties of cast Al-Si alloy due to hot rolling. Journal of Japan Institute of Light Metals
60, 7-11. (in Japanese)
Wu, S., Ji, Z., Zhang, T., 2009. Microstructure and mechanical properties of AZ31B magnesium alloy
recycled by solid-state process from different size chips. Journal of Materials Processing Technology
209, 5319-5324.
Yoshikawa, M., Itoh, S., 1981. On the hot-extrusion of machining swarf of copper alloy. Journal of the
Japan society for technology of plasticity 22, 683-688. (in Japanese)
Yoshikawa, M., Itoh, S., Nishimoto, K., 1980. On hot-extrusion of swarf of aluminum alloy. Journal of
the Japan society for technology of plasticity 21, 792-799. (in Japanese)

22
Captions of tables and figures
Table 1 Chemical compositions of AC4CH aluminium alloy.

Table 2 Orifice dimensions of dies.

Table 3 Rolling schedule used in experiments.

Table 4 Parameters for characterising the annealed AC4CH ingot.

Fig. 1 Appearance of (a) AC4CH aluminium alloy machining swarf and (b) as-compacted machining
swarf.

Fig. 2 Surface appearance of extruded bars and corresponding die geometry for R = 4: (a) extruded
specimen and (b) die geometry for aspect ratio of 1:1.8, and (c) extruded specimen and (d) die geometry
for aspect ratio of 1:3.8.

Fig. 3 Dimensions of uniaxial tensile specimens cut out from (a) original AC4CH ingot, and (b) extruded
and (c) rolled specimens.

Fig. 4 Microstructure observation points for (a) extruded and (b) rolled specimens.

Fig. 5 Experimental and approximated stress-strain curves of annealed AC4CH cylinders with 6-mm
diameter and 7-mm length.

Fig. 6 Expansion observed in annealed extrudant for R = 4 and the 1:1.8 aspect ratio.

Fig. 7 External appearances of strip shaped through cold rolling: (a) RD-TD and (b) RD-ND planes.

Fig. 8 Optical micrograph of visible void in extruded specimen for R = 4 and the 1:1.8 aspect ratio.

Fig. 9 Optical micrographs of (a) original AC4CH ingot and recycled specimens: (b) point A, (c) point B
(d) point C in specimen extruded for R = 4 and the 1:1.8 aspect ratio, and (e) point D in specimen
additionally processed by rolling.

Fig. 10 Variations in relative density.

Fig. 11 Mechanical properties of original AC4CH ingot and recycled materials: extruded specimen for R
= 4 and the 1:1.8 aspect ratio, and that additionally processed by rolling.

Fig. 12 Spatial variations in equivalent strain and Vickers hardness in specimen extruded for R = 4 and
the 1:1.8 aspect ratio.

23
10mm

(a)

5mm

(b)

Fig. 1
(a)

10mm

(b)
Cracks

(c)

10mm

(d)

Fig. 2
(a) (b) (c)

Fig. 3
w0 y

Observation points
z (ED)
A w0/2 Extrusion axis
B x
t0
C
t0/2

(a)

Normal direction (ND)

w1/2
w1 Observation point
D Rolling direction (RD)

Transverse direction (TD)


(b)

Fig. 4
300
True compressive stress /MPa

250
Experiment
200

150

Power-law hardening rule


100
= Fn
50

0
0 0.2 0.4 0.6 0.8 1
True compressive strain

Fig. 5
Crack due to expansion

Expansion

10mm

Fig. 6
10mm

(a)

10mm

(b)

Fig. 7
Fig. 8
100m

(a)

Voids

Voids

Extrusion direction 100m

(b)

Extrusion direction 100m

(c)
Extrusion direction 100m

(d)

Rolling direction 100m

(e)

Fig. 9
R=6 Rolling reduction = 85%

R=4

for the 1:1.8 aspect ratio

Fig. 10
300 5
Proof stress
250 Ultimate tensile strength
Fracture strain 4

Fracture strain [%]


200
Strength [MPa]

150

2
100

1
50

0 0
Original Extruded Rolled
AC4CH specimen specimen

Fig. 11
85
HV
5
Measured values
Mean values
80

Vickers hardness, HV
4
Equivalent strain

Equivalent strain

75
3

70
2

1 65
6 5 4 3 2 1 0 1 2 3 4 5 6

Distance from extrusion axis /mm

Fig. 12
Table 1

(mass%)
Cu Si Mg Zn Fe Mn Ti Sb Al
0.01 6.9 0.37 0.02 0.13 0.01 0.13 0.001 Bal.
Table 2

Aspect ratio Extrusion ratio Orifice size of die* /mm


4 6.8 12.0
1:1.8
6 5.6 10.0
1:3.8 4 4.7 18.0
*
1 mm corner radii are set at four corners.
Table 3

Rolling reduction per pass /%


Number of passes
Ave. Std. Min. Max.
20 9.07 1.02 6.72 10.05
Table 4

Parameter Value
Youngs modulus 74 GPa
Poissons ratio 0.34
Density 2700 kg/m3
F-value 220 MPa
n-value 0.12

View publication stats

Vous aimerez peut-être aussi