Journal of Chemical

Journal Technology
of Chemical and Metallurgy,
Technology 56, 2,56,
and Metallurgy, 2021, 410-416
2, 2021

FEM-STUDY OF BIMETALLIC WIRE DEFORMATION

DURING COMBINED ECAP-DRAWING

Irina Volokitina1, Andrey Volokitin1, Abdrakhman Naizabekov1, Evgeniy Panin2

1
Rudny Industrial Institute, 111500, 50 let Oktyabrya 38, Rudny, Kazakhstan Received 14 January 2020
2
Karaganda State Industrial University, 101400, Republic av. 30, Temirtau, Kazakhstan Accepted 10 March 2020
E-mail: [email protected]

ABSTRACT

The method for producing blanks for tools made from high-speed steel grades is presented. The method consists
of sintering the crushed chips of steel grades with certain titanium carbide content. The methodology for the steel
powder preparation, for the pressing process and for sintering mixtures is presented. The heat treatment of the
obtained blanks was carried out according to the necessary conditions. The number of indicators of the obtained
mechanical billets, which are decisive for high-speed steel grades, such as hardness, toughness and resistance to
cutting, are investigated.
Keywords: high-speed steels, titanium carbide, powder metallurgy, sintering, mechanical characteristics.

INTRODUCTION wire and bars from non-ferrous metals and alloys with
an ultra-fine-grained structure and an increased level of
The use of severe plastic deformation to form an mechanical properties.
ultrafine-grained (UFG) structure in metals and al- Due to this unique deformation scheme, a sufficiently
loys that provides a high level of strength and plastic high level of tensile stress develops in the section of the
characteristics of these metals and alloys is one of the workpiece, while the wrong selection of technological
promising directions of modern materials science [1]. parameters will lead to the breakage of the deformable
Among the large number of already developed methods wire. In works [10 - 13], theoretical and experimental
for implementing severe plastic deformation, the equal- results of studies of this process in the deformation of
channel angular pressing (ECAP) is still of particular steel, aluminum and copper wire are presented.
interest [2]. This interest is related to the possibility of At the moment, the development of this combined
obtaining rather massive, but unfortunately not long- process only for homogeneous materials was conducted.
length blanks, with a uniform UFG structure of various At the same time, bimetallic wire is increasingly used
metals and alloys without changing their original cross- as the starting material for the production of wires. It
section. Therefore, this method has not found widespread is a long - length product of composite steel products
industrial use. In recent years, scientists from different consisting of metals and alloys with different chemical
countries have developed a number of combined pro- and physical characteristics. The most commonly used
cesses based on the ECAP principle [3 - 9] to solve connection types are steel-aluminum and steel-copper.
this problem, i.e. in order to obtain long products with The construction of a bimetallic wire consists of a core
an ultrafine-grained structure. Each of these processes and a shell. The core is made of high-strength steel; the
allows to increase significantly the productivity of the shell is a different metal or alloy in properties.
deformation process by annihilating certain ECAP When implementing the deformation scheme shown
disadvantages. These processes include the combined in Fig. 1, a high level of tensile stresses develops in
process “ECAP-drawing” (Fig. 1). Implementation of the workpiece section, so an incorrect selection of
this process makes it possible to obtain steel wire or technological and geometric parameters of this process
410
 Irina Volokitina, Andrey Volokitin, Abdrakhman Naizabekov, Evgeniy Panin

Fig. 1. Scheme of the “ECAP-drawing” process.

will lead to the breakage of the deformable wire. These of the steel core was equal to 8 mm. The core material
parameters are often not the same even for different was chosen AISI-1010 steel (analogous to steel 10).
homogeneous materials. When drawing a bimetallic Aluminum alloy 1100 was chosen as the shell material.
wire or bar, they will be even more unique. Therefore, The deformation was performed at room temperature.
this work is devoted to the study of the bimetallic wire The angle of junction of channels in the ECA-matrix
deformation process of «steel-aluminum» type using the was equal to 145°. At the drawing stage, 5 % compres-
«ECAP-drawing» combined process. sion was provided, up to a diameter of 9.5 mm. The
elastic-plastic type was chosen as the material model
FEM-modeling for the core and shell. Since both materials are station-
One of the most effective methods of theoretical ary relative to each other in a bimetallic wire, a rigid
analysis of any technological process currently is com- unbreakable contact has been established between them.
puter simulation using the finite element method. This On the contact of the aluminum shell and both tools
method of research has several undeniable advantages: (ECA-matrix and drawing), a coefficient of friction of
- possibility to visualize the studied process, even 0.1 was set, which corresponds to the polished surface
inside the workpiece or tool, that is impossible in real with the use of grease.
conditions; When developing this combined process, it was
- complex analysis of several parameters at once found that to prevent wire breakage in the area between
at any point of the studied object, that is also often im- matrix and drawing die, as well as to avoid the forma-
possible with traditional methods, for example, in the tion of defects at the front and rear ends of the wire, it
absence of necessary empirical equations; is necessary to coordinate the pulling speed applied to
- ability to optimize the process by varying the the front end of the workpiece and the pushing speed
values of certain geometric or technological parameters. applied to the back end. Incorrect selection of kinematic
All these advantages have made FEM simulation a parameters will lead to either a wire break or stratifica-
fairly common method of theoretical study. tion of layers, as shown in Fig. 2.
If we consider FEM modeling from the point of view For a stable flow of this process, it is necessary to
of metal forming, then the leading position is occupied use the law of constancy of second volumes. Thus, for
by the Deform program, which allows to simulating a given back speed of 10 mm/s and a reduction in the
almost any deformation process. cross section of the wire from 10 to 9.5 mm, the front
During modeling the deformation of a bimetallic speed should be equal to 11.08 mm/s.
wire using the “ECAP-drawing” combined process, it Before starting the stress-strain analysis, it is also
is necessary to solve two problems at once: necessary to decide which parameters will be studied. To
1) study the stress-strain state under complex load- assess the strain state, it is most convenient to consider
ing, due to the combining two operations, that leads to the criterion «equivalent strain» [14], the value of which
the appearance of two deformation zones; depends on the values of the main deformations and is
2) conduct a study of the stress-strain state for each determined by the formula:
material separately.
2
( ε1 − ε 2 ) + ( ε 2 − ε 3 ) + ( ε 3 − ε1 )
2 2 2
A bimetallic wire of “steel-aluminum” type with a ε=
EQV
(1)
diameter of 10 mm was used as the initial billet, diameter
3

411
 Journal of Chemical Technology and Metallurgy, 56, 2, 2021

related to its chemical composition and mechanical

properties after deformation or heat treatment. Also, the
microstructure, namely the size and shape of the result-
ing metal grain determines the overall quality level of
the product. Predicting and managing microstructures
is a top priority when developing a successful process
and producing a product with the desired properties.
a) In the Deform system, there are 2 approaches to mod-
eling the evolution of microstructure during the processes
of deformation and heat treatment. The first approach
is the traditional Johnson-Mehl-Avrami-Kolmogorov
method (JMAK). The second approach to microstruc-
ture modeling in Deform is the discrete lattice method
implemented using the Cellular Automata algorithm.
In the JMAK method, the final average grain size
is determined based on the initial average grain size,
b)
some constant material values, and fields of variable
Fig. 2. Stratifications at the ends of a bimetallic wire as a data such as temperature, strain, strain rate, and time.
result of misalignment of kinematic parameters: a) - front Grain growth and recrystallization kinetics (in particular,
end; b) - back end. dynamic, meta-dynamic, and static) are also modeled.
To predict the development of the average grain size
When considering the stress state, the “equivalent and the percentage of recrystallization of the material
stress” [14] criterion is often used to estimate the average of interest, it is necessary to determine the initial aver-
level of stress that occurs, and this is often sufficient. age grain size and develop equations for the kinetics of
This criterion depends on the values of the main stresses recrystallization. The development of such equations
and is determined by the formula: involves conducting characteristic material studies to
study the development of grain size as a function of
1
(σ 1 − σ 2 ) + (σ 2 − σ 3 ) + (σ 3 − σ 1 )
2 2 2
σ=
EQV
(2) strain, temperature, and strain rate that occur during the
2
deformation and heat treatment processes.
However, in this combined process, the deformation JMAK expressions use coefficients and exponents
scheme is such that if the initial parameters are incor- defined for constant conditions (temperature, strain
rectly selected, the deformation may be unstable, leading rate, etc.). In real conditions, these values change dur-
to a wire break. Considering this fact, consideration of ing the processing of research objects. DEFORM uses
the equivalent stress in the ECAP-drawing is insufficient. these coefficients and exponents, which are functions
Therefore, for a full assessment of the resulting stresses, of temperature, strain, and strain rate. Since the partial
it is also necessary to study the criterion “average hydro- values of these variables can be calculated by means of
static pressure” [14], which allows you to estimate the finite element modeling, this leads to an increase in the
value of the stress taking into account the sign, i.e., to accuracy of JMAK model.
estimate the value of tensile and compressive stresses. However, the JMAK method does not take into
This criterion is derived from the values of the main account the formation of non-metallic inclusions and
stresses and is determined by the formula: their influence on grain growth. Also, this method does
σ1 + σ 2 + σ 3 not take into account the different shape of the grain.
σ AV = (3) The initial microstructure with the same average size,
3
but with different grain shape has different kinetics of
Simulation of microstructure evolution microstructure development. The classic JMAK method
The microstructure of the material provides the re- is insensitive to the grain shape and thus is limited for
searcher with data about the resulting material, directly use in production.

412
 Irina Volokitina, Andrey Volokitin, Abdrakhman Naizabekov, Evgeniy Panin

Fig. 3. Interaction mechanisms in the Cellular Automata model.

To eliminate the disadvantages of the JMAK meth- By default, the preferred mechanism is the Moore
od, Deform implemented the Cellular Automata (CA) mechanism, because in this case, the number of cells
method [15], based on the use of discrete lattices. The remains constant both in the horizontal and vertical di-
Cellular Automata model represents a microstructure as rections, and also in the diagonal. Taking into account the
a discrete lattice of regularly ordered points. Each point additional features of the Cellular Automata method, we
represents the microstructure as a function of the crystal- will use this method when calculating the microstructure.
lographic orientation, dislocation density, strain distribu- In the work of this module cannot accommodate the
tion, etc. This model, which explicitly represents grains possibility of complex input data, so the modeling of the
and grain boundaries not as an «average» value, is more microstructure for each material will be held separately.
sensitive to local changes in geometry and morphology.
Thus, in the CA method, data on recrystallization and RESULTS AND DISCUSSION
grain development are more accurate and correspond to Strain state
reality than when using the classic JMAK method. The When considering the ECAP stage of this combined
CA method allows modeling microstructure phenom- process, it was found that passing through the channels of
ena such as hardening, property restoration, dynamic the ECA-matrix, the shell and core receive different levels
recrystallization, metadynamic recrystallization, static of strain (Fig. 4(a)). The largest amount of equivalent
recrystallization, and grain growth. The CA method takes strain, reaching ε = 1.5 in some areas, the aluminum shell
into account the effects of strain, strain rate, temperature, receives in the zones of the junction of channels - when
crystallographic orientation, grain boundary misorienta- moving and rubbing against the rounded zones of the cor-
tion, and dislocation density. ners of the joint. The core gets a much smaller deforma-
When calculating the grain size, the equations of tion - the central zone of the core is processed to ε = 0.4,
the JMAK method are used [16]. However, additional the surface layers of the core get a higher deformation, to
calculation modules are used to calculate the grain shape. ε = 0.6. Despite the fact that both materials are in a rigid
The key feature is the choice of interaction between engagement with each other and must deform the same
neighboring calculated grid cells (grains). The program way, such a significant difference in the development of
provides two different mechanisms of interaction: Moore strain is due to the different amount of deformation resist-
and Von Neumann. Fig. 3 clearly demonstrates their ance of both materials. In other words, in this case, one
differences when using the same values of the interac- material is much softer and more pliable than the other.
tion radius. After the drawing stage (Fig. 4(b)), a small increase in

413
 Journal of Chemical Technology and Metallurgy, 56, 2, 2021

a)

b)

Fig. 4. Equivalent strain: a) - ECAP stage; b) - workpiece Fig. 5. Equivalent stress: a) - ECAP stage; b) - drawing
after the drawing stage. stage.

strain is observed only in the shell to ε = 1,7, the level of this parameter is similar to the distribution of the equiva-
strain in the core almost does not change. lent stress (Fig. 6(a)). In the aluminum shell, in areas
without contact with the tool, there are tensile stresses
Stress state of 100 - 110 MPa. In straight sections with contact with
When studying the equivalent stress at the ECAP the matrix, the shell experiences back pressure from
stage, it was found that when passing through the chan- the matrix, which leads to the creation of compressive
nels of the ECA-matrix, different stress levels develop stresses from -80 to -90 MPa. The maximum level of
in the shell and core (Fig. 5(a)), covering the entire zone back pressure is created directly at the channel junc-
of the junction of the channels. In an aluminum shell, tions - here the value of compressive stresses reaches a
the average value of the equivalent stress is 110 - 120 value of -350 MPa.
MPa. In a steel core, due to the simultaneous action of In a steel core, the distribution of tensile and com-
the pulling speed at the front end, which causes stretch- pressive stresses can be called identical to the distribu-
ing, and the backing speed at the rear end, which causes tion in the shell. On the upper sections of the input
compression, there is an alternating distribution of stress. and output channels, as well as on the lower section of
When passing the intermediate channel, the highest the intermediate channel, there are tensile stresses, the
stress occurs in the lower part of the core, reaching 480 value of which is about 230 - 240 MPa. In the opposite
MPa. But, in the output channel, when the influence of sections, compressive stresses reaching -360 MPa are
the back speed weakens, the maximum stresses are real- implemented in all channels.
ized already on the upper part of the core (490 MPa), When drawing (Fig. 6(b)), the deformation zone,
generally balancing the entire deformation zone. as well as when considering the equivalent stress, is
When drawing (Fig. 5(b)), the deformation zone is completely symmetrical. Compressive stresses of ap-
completely symmetrical. In the shell, the average value proximately -160 MPa occur in the shell. In this case,
of the equivalent stress is 130 MPa. The average stress the entire core creates tensile stresses at the level of 90
level in the core is significantly reduced to 280 MPa. MPa. The reduction of these tensile stresses in the core
When considering the average hydrostatic pressure is observed directly in the deformation zone, where their
at the ECAP stage, it was found that the distribution of value is reduced to 40 MPa.
414
 Irina Volokitina, Andrey Volokitin, Abdrakhman Naizabekov, Evgeniy Panin

a)

b)

Fig. 7. Grain size change in the aluminum shell.

the steel core (Fig. 8), it was found that after passing
the channels of the ECA-matrix, the initial grain size of
18 microns practically does not change, only individual
grains are slightly reduced. In the intermediate zone
Fig. 6. Average hydrostatic pressure: a) - ECAP stage;
between matrix and drawing die, the grain size also does
b) - drawing stage.
not change much. However, there is a certain elongation
of the grains in the longitudinal direction.
Microstructure evolution Due to the fact that the core is much thicker than the
When creating the model, the initial average grain shell, the microstructure was considered at the exit from
size was set to 20 microns for the aluminum shell and the drawing die at two points - in the axial and surface
18 microns for the steel core. During the calculation of zones. In the surface zone, the grain size decreases about
the models, changes in grain size in the following zones to 14 microns, while the grain shape is more equiaxed,
were analyzed: with approximately the same areas. In the axial zone,
1) in the output channel of the matrix after the com- the grain size decreases significantly less, to 17 microns,
pletion of the ECAP stage; while the grain shape is strongly elongated, which is a
2) in the intermediate zone between the matrix and consequence of the drawing stage.
drawing die;
3) at the exit of the drawing die after the completion
of the drawing stage.
Considering the microstructure in the shell of a
bimetallic wire (Fig. 7), it was found that after passing
the channels of the ECA-matrix, the initial grain size
decreases to about 13 microns, while the shape of the
grains becomes equiaxed, which corresponds to the
ECAP effect.
In the intermediate zone between matrix and draw-
ing die, the grain size does not change much. However,
there is a certain elongation of the grains in the longitu-
dinal direction, which is the result of the action of tensile
stresses occurring in this area. At the end of the draw-
ing stage, the grain size is reduced to 9 microns, while
elongated grains are also clearly visible in the direction
of deformation. When studying the microstructure in Fig. 8. Grain size change in the steel core.

415
 Journal of Chemical Technology and Metallurgy, 56, 2, 2021

CONCLUSIONS of Metal Flow during Processing by Multi-ECAP-

The considered combined process «ECAP-drawing» Conform, Advanced Engineering Materials, 17, 12,
is quite interesting from the point of view of the pos- 2015, 1723-1727.
sibility of deforming such a material as bimetallic wire. 5. A.R. Fastykovskii, Region for extrolling feasibility
When analyzing the strain state, it was found that each and effective deformation modes, Russian Journal
material in a bimetallic wire receives different values of of Non-Ferrous Metals, 52, 3, 2011, 230-233.
strain in a single cycle, which is due to the difference in 6. A. Naizabekov, S. Lezhnev, E. Panin, I. Volokitina,
the strength and plasticity levels of these materials. The A. Arbuz, T. Koinov, I. Mazur, Effect of Combined
stress state of both materials is different in both zones of Rolling-ECAP on Ultrafine-Grained Structure and
deformation - in the ECAP zone, the deformation zone Properties in 6063 Al Alloy, Journal of Materials
is divided into sections of tension and compression, di- Engineering and Performance, 28, 1, 2019, 200-210.
vided diagonally. In the drawing zone, the deformation 7. A. Naizabekov, S. Lezhnev, E. Panin, A. Arbuz, Com-
zone is completely symmetrical. At all stages of defor- bined process “helical rolling-pressing” and its effect
mation, the level of compressive stress is significantly on the microstructure of ferrous and non-ferrous
higher than the tensile stress. materials, Metall. Res. Technol., 115, 2018, 213.
In the course of FEM modeling of microstructure 8. C.J. Luis-Perez, P.A. Gonzalez, J. Gil, J. Alkorta,
evolution in layers of bimetallic wire of the “steel- Procesado continuo de materiales metalicos mediante
aluminum” type with the combined method of deforma- deformacion plastica en canal poliangular, Spain
tion “ECAP-drawing”, it was found that both layers of patent No 2224787, 2006, (in Spanish).
bimetallic wire are processed unevenly. The surface shell 9. A.A. Ershov, Y.N. Loginov, N.N. Zagirov, E.V. Ivanov,
is processed to a grain size of 9 microns. In this case, the Simulation of the Consolidation of a Porous Material
shape of the grains turns out to be fairly equiaxed. The in a Combined Rolling-Extrusion Process, Metallur-
steel core has different grain sizes in cross-section – in gist, 60, 5-6, 2016, 561-565.
the surface layers, the grain structure has an equiaxed 10. S. Lezhnev, I. Volokitina, A. Volokitin, Evolution of
shape, similar to the shape of the grains in the shell, Microstructure and Mechanical Properties of Steel
with a size of 14 microns. In the axial zone, the grains during Pressing–Drawing, Physics of Metals and
have a size of 17 microns, and after the drawing stage, Metallography, 118, 11, 2017, 1167-1170.
the grain shape is strongly elongated due to the action 11. I. Volokitina, A. Volokitin, Evolution of the Mi-
of tensile stresses. In other words, with relatively equal crostructure and Mechanical Properties of Copper
initial grain sizes in the shell and core, after deformation during the Pressing-Drawing Process, Physics of
according to the “ECAP-drawing” scheme, you can get Metals and Metallography, 119, 9, 2018, 917-921.
a bimetallic wire with a “gradient “ structure, where the 12. I.E. Volokitina, A.V. Volokitin, A.B. Naizabekov,
grain size in the surface area is reduced by more than 2 S.N. Lezhnev, Change in structure and mechanical
times, and the axial zone is almost unchanged. properties of grade А0 aluminum during implemen-
tation of a combined method of ECAE–drawing
Acknowledgements deformation, Metallurgist, 63, 9-10, 2020, 978-983.
This research was funded by the Science Committee 13. S. Lezhnev, A. Naizabekov, A. Volokitin, I. Vo-
of the Ministry of Education and Science of the Republic lokitina, E. Panin, M. Knapinski, Development and
of Kazakhstan (Grant No. AP08052852) research of combined process of “equal channel an-
gular pressing - drawing”, J. Chem. Technol. Metall.,
REFERENCES 52, 2, 2017, 172-179.
1. R.Z. Valiev, I.V. Alexandrov, Nanostructured materi- 14. G.E. Arkulis, V.G. Dorogobid, Plasticity theory,
als obtained by severe plastic deformation, Moscow, Moscow, Metallurgy, 1987,e(in Russian).
Logos, 2000, (in Russian). 15. H.W. Hesselbarth, I.R. Gobel, Simulation of Recrys-
2. V.M. Segal, Method of pressing, USSR patent No tallization by Cellular Automata, Acta Metallurgica
575892, 1977, (in Russian). et Materialia, 39, 9, 1991, 2135-2143.
3. J. Kliber, Advanced Forming Technology, Metalurgija, 16. T. Furu, K. Marthinsen, E. Nes, Modeling Recrystal-
55, 4, 2016, 835-838. lization, Materials Science and Technology, 6, 11,
4. E.I. Fakhretdinova, G.I. Raab, R.Z. Valiev, Modeling 1990, 1093-1102.
416