Available online at www.sciencedirect.

com ScienceDirect
2214-7853 © 2015 Elsevier Ltd. All rights reserved. Selection and Peer-review under responsibility of Conference Committee of
Aluminium Two Thousand World Congress and International Conference on Extrusion and Benchmark ICEB 2015 doi:
10.1016/j.matpr.2015.10.007
Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

Aluminium Two Thousand World Congress and International Conference on Extrusion and
Benchmark ICEB 2015

Analysis of aluminium hot extrusion by finite volume method

José D. Bressana,*, Marcelo M. Martinsb, Sérgio T. Buttonc
aDepartment of Mechanical Engineering, Centre for Technological Science - University of Santa Catarina State(UDESC), CEP
89219-710 - Joinville - SC, Brazil. bCentre University Catholic of Santa Catarina, CEP 89203-005 - Joinville-SC, Brazil.
cDepartment of Materials Engineering, Faculty of Mechanical Engineering - State University of Campinas(UNICAMP)- Campus
Cidade Universitário “Zeferino Vaz”, Mendeleyev 200, CEP 13083-860 - Campinas - SP, Brazil.
Abstract
A novel numerical approach to calculate strain rate, stress and velocity fields of axisymmetric hot extrusion process is presented
and discussed. Traditionally, metal forming analysis has been performed by upper-bound, slab, slip-line and Finite Element
(FEM) methods. However, Finite Volume Method (FVM) has been recently developed for metal plastic flow analysis. Hence,
metal extrusion was modelled by flow formulation. Governing equations were discretized by FVM, using Explicit MacCormack
Method in structured and collocated mesh. SIMPLE method was applied to attain necessary pressure-velocity coupling. This
scheme was applied to hot extrusion of aluminium alloy 6351. Numerical results of velocity field from present FVM produced
consistent and good agreement with FEM results, employing FORGE 2008. Present results also suggest that MacCormack and
SIMPLE methods can be applied in the solution of metal forming processes.
© © 2014 2015 Elsevier Elsevier Ltd. Ltd. All All rights rights reserved.
reserved. Selection Selection and and Peer-review Peer-review under under responsibility responsibility of Conference of
Conference Committee Committee of Aluminium of Aluminium Two Thousand Two Thousand World Congress World Congress
and International
and
International Conference on Conference Extrusion and on Benchmark Extrusion and ICEB Benchmark 2015
ICEB 2015.
Keywords: Hot extrusion; finite element method; finite volume method; aluminum alloy 6351.
* Corresponding author. Tel.: +55-47-4009-7958; fax: +55-47-4009-7940. E-mail address: josedivo.bressan@udesc.br
4741 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

1. Introduction
Nowadays, industrial challenges and competitions have been increased and forced the reengineering of all
activities at factories and academy in order to increase productivity, reduce production costs, part weight, part energy
content, energetic efficiency, efficient and rational use of natural resources and design for recyclability of materials.
Hence, the globalization of economy, market dynamics and growing environmental requirements for sustainable
development, have faced the industries, academy and governments with new manufacturing and material challenges.
Nomenclature
σ σ ρ ε rr rz

,ε ,

σ
θθ

density
, shear σ
zz

component normal components of Cauchy of stress Cauchy tensor stress σ true strain, strain rate
tensor σ
T temperature c specific heat σ qx v r r
,
,,
qx v z z

components of heat flow vector qx components of velocity vector effective or equivalent stress,
σ
= ( 2/3 )
ss ijij s
ij

s=σ-σδm
deviatoric stress,
ij ij ijm η σ ε λ x x

hydrostatic pressure, σ m = ( σ rr + σ θθ
+

σ zz ) / 3 metal equivalent viscosity plastic multiplier effective or equivalent

plastic strain rate, ε x
=
( 3/2 )

ε x ijij ε x t time t∆ virtual time step r radial coordinate in

cylindrical coordinate system (r,θ,z) θ angular coordinate in cylindrical coordinate system (r,θ,z) z axial coordinate
in cylindrical coordinate system (r,θ,z) Q ,
r

F , S flow vectors
mn F V ,

control z
volume area of node mn
Q mn 1+t
, 1+t
mn
Q mn
1+t

predictor step, corrector step and current step respectively A, m

1

Q,,m
2

,m
3

,m
4

,m
5

,m
7

,m
8

,m
9

material parameters of Hensel-Spittel equation

In this new national and international scenario, there is the formation of work teams inside the industries,
academy and mixed teams to study and improve technological processes with the aim at advancing the product
quality, durability, sustainability and equipment productivity, as well as to reduce production and maintenance costs.
In a broad sense, the industries (automotive, aerospace, stamping, steel, etc.) should continue to attain these
objectives in addition to introduce High Strength Low Alloy steels - HSLAS, Advanced High Strength steels -
AHSS, to increase the use of lightweight aluminium alloys, magnesium alloys and composites. I addition, to produce
precision parts with zero-defects, to advance the use of near net shape processes, advance additive manufacturing,
and design for recycling, durability or sustainability.
Consequently, nowadays in the actual engineering practice, industrial manufacturing processes are being
investigated and modelled increasingly by interdisciplinary groups of engineers and academics, largely with the
support of software and computers. Numerical and analytical modelling of manufacturing processes together with
microstructure investigations have great potential to increase process productivity and product quality, as well as to
reduce development and production costs, through the following factors [1]:
• reduction of the iterations numbers in the experimental method of try-and-error,
• allows a quick construction of modelling or prototype,
• improve the visualization of manufacturing process.
4742 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

• provides a knowledge of the essential physical mechanisms necessary to control the process in real time,
description of the evolution of material internal damage and avoid undesirable defects, However, mathematical
modelling of thermo-mechanical process of metal forming requires an appropriate mathematical formulation of
material behaviour, boundary conditions and low numerical error. In general, the mathematical techniques employed
to solve metal forming problems are: analytical methods (slab method, upper bound method, slip line field method)
and numerical simulations with mesh (finite element, finite difference, finite volume). On the other hand, the
experimental methods available for solving manufacturing process modelling can be classified into the followings
techniques: material simulation (plasticine, lead, aluminium: stripe pattern grid technique) and material
characterization (simple tension test, compression test, plastic torsion test).
Aluminium hot extrusion is one main metal forming process largely applied in manufacturing bars and products
with complex cross section shape. Traditionally, metal forming analysis has been performed by upper-bound, slab,
slip-line methods and also, more recently, through computational methods such as the finite element method.
In cold and hot extrusion processes, the knowledge of metal flow behavior through die and variables such as the
flow lines, velocity field, deformations, stresses, friction, forces acting on tool and temperature can help to predict
internal defects, failure and geometrical distortions in manufactured components or products. One approach to
obtain such information are employing numerical methods, that are fast and accurate and by this manner, helping to
reduce the development time and tooling costs and optimization of operational conditions in the experimental tryout
process. There are three principal macroscopic defects to be avoided in metal extrusion: surface cracking, internal
cracking or internal burst and geometrical distortions of extruded part.
Numerous materials can be extruded to a wide variety of cross-sectional shapes and dimensions. Considerable
research effort in metal forming of long products with complex cross section such as plane bars, circular cross
section bars, hollow bars, T and H-shape cross section bars have been also carried out experimentally and
theoretically due to its relevance to tool design, understanding of metal flow and to obtain information on geometric
changes, imposed straining, stresses, friction and load distributions. To achieve these process insights, material
simulation by lead and plasticine [2, 3], visioplastic strain analysis by split billets [4], stripe pattern grid technique
[5], analytical models [2, 6] and computational numerical simulations [5, 7] have been employed extensively. FEM
is nowadays very popular and widely employed to metal forming process analysis by commercial codes such as
ABAQUS, DEFORM and FORGE.
In metal forming process analysis, two different classical mathematical approaches can be employed: the flow
approach and the solid approach [8, 9].
In the flow approach formulation, an Eulerian reference coordinate system with fixed mesh in space is employed:
the discrete elements are fixed in space and not in the material, and the material is considered non-linear viscous
fluid or visco-plastic incompressible fluid. Thus, the numerical grid is constant and keeps its initial geometrical
features during numerical simulation. The flow formulation is effective to calculate material internal variables but is
not for moving boundaries [8].
In the solid formulation, a Lagrangian reference coordinate system attached to the moving workpiece or material
is used: the discrete elements remain attached to the workpiece and deforms with it. Hence, mesh moves and distorts
with material deformation or plastic flow. Material is considered elastic-plastic or viscoplastic solid [8]. The solid
approach is very popular and has been effective for large plastic deformation analysis of metal forming by the finite
element method. However, has the drawback of occurrence of mesh severe distortion at some critical points which
causes error in the calculations of field variables. The solution for this problem is to restart the analysis and use a
remeshing technique to increase the accuracy, but with higher computation time [10].
Alternatively, a mixed numerical method has been more recently developed which combine both Lagrangian and
Eulerian approaches: the Arbitray Lagrangian Eulerian-ALE finite element formulation method [11]. In situations
where the mesh undergoes large deformations and becomes unacceptably distorted, ALE formulation gained interest
for its capability to control mesh geometry independently from material geometry. ALE or remeshing methods are
used to create a new undistorted mesh of arbitrary domain which allows the calculations to continue, but again their
implementation is highly difficult and time consuming because of tedious geometrical calculations [11].
Finite volume method, FVM, has been developed since the seventies and applied mainly to computational fluid
dynamic, CFD. However, in the nineties, there was an increase of interests to use FVM in solid mechanics [12, 13],
initially for problems involving fluid-structure interactions [14]. In recent years, FVM has been proposed for solid
4743 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

structure elastic stress analysis to calculate stress, strain and displacements [15, 16]. Bressan et al. [17] have also
presented a numerical approach for calculating transient thermal stress due to a plate thermal shock, using FVM.
In the nineties at Imperial College of London, a free code package named OpenFoam was developed by Jasak’s
[18, 19] for computational fluid dynamic, using the FVM. However, nowadays it is also being used to solve solids
mechanic and metal forming problems [20].
The goal of present work is to propose a new numerical scheme to calculate stress and velocity fields of metal
flow in the axisymmetric extrusion process in steady state of aluminium alloy 6351. The governing equations were
discretized by FVM, using the Explicit MacCormack Method in structured and collocated mesh [21]. A velocity-
pressure coupling method was necessary to obtain a consistent velocity and pressure fields because metal plastic
flow does not present state equations for the evolution of pressure. The SIMPLE Method [21] was applied to attain
pressure-velocity coupling. Present numerical FVM was applied to direct hot extrusion through a conical die of
aluminium alloy 6351 and validated by the numerical results from FEM, employing FORGE 2008.
2. Extrusion modelling by Finite Volume Method
Metal plastic flow in direct hot extrusion process was modelled by FVM, using the flow formulation with a fixed
Eulerian reference coordinate system: metal plastic flow was considered similar to a non-linear viscous fluid flow.
Tannehill et al. [21] suggests that viscous fluid flow can be numerically simulated by FVM, satisfying the mass,
momentum and energy conservation equations and boundary conditions.
2.1. Governing equations of flow formulation
According to the work presented by the authors [1], the governing conservation equations (mass, momentum and
energy) in cylindrical coordinates system (r,θ,z) for axisymmetric case, can be unified in compact matrix structure in
the following form:
∂∂
Q rt +
1
∂

(r∂
FF r

r)
+
∂
z∂
z=
S
(1)
where Q ,
Fr
,
Fz
and S are flow vectors which assume the following format:
Q
=
⎧││⎨││
ρρ
v
rρ

v
z cT
⎫││⎬││
⎧
ρ
v
⎫
⎧
ρ
v
⎫
2
r

z2│
ρ
+
x│
│
ρ
+
x0
S (2)
2.2. Plastic Flow Constitutive Equations
Present metal plasticity constitutive equation for axisymmetric plastic flow considers the material to be
incompressible, isotropic and rigid-viscous-plastic. The relation between the stress tensor and strain rate tensor can
be written in the following form:
-=
σ m 2+ εη x F =
││⎨│
ρ
ρ vv v
r
r-
σ rr -

σ│⎬│
F
=
││⎨
ρ
vv
rz
z-

σ rz │ │ ⎬
=│
│
ρ
v
-
σ⎧││⎨
-

σ
⎫

⎩
ρ
⎭
⎩
cTv
r
qr⎭

⎩
cTv
z
qz│

σε││⎭
│ zr
rz z
zz ⎩
θθ
0
r
.
x
││⎬││⎭

σ (3)
where σ is Cauchy stress tensor. Hydrostatic pressure inside the solid is calculated by
σ
m
= ( σ rr σ+ θθ
+ σ zz )
3/

.
Metal plastic flow equivalent viscosity η is calculated from the Associated Plastic Flow Rule [1]:
η
=13
σεx
(4)
4744 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

the effective strain rate ε x

must not assume values less than 10-3 to avoid numerical instability.
2.3. The Finite Volume Method
Integrating Eq. (1) over the control volume at node (m,n), employing Gauss' Theorem, results in the following
differential equation [1]:
∂
Q
11
⎧
mn

(r
) n,m
-

( r ) n,m + ( z ) n,m - ( z )
n,m + 2

⎫│⌋
mn (5)
where
mn ∂ t

=
- rV ⎨
⌈│⌊
r
sF
∙1
+ r sF ∙ 2
12

⌉│⌋
+
⌈│⌊
sF ∙ 1
+ sF ∙ 2
1

⌉
⎬⎭

+Ss
is the outward surface vector and
V mn

is the control volume area at node (m,n). Explicit MacCormack scheme

was applied in the Eq. (5). This scheme use explicit Euler scheme on time. Thus, Eq. (6) below represents the
predictor step, Eq. (7) represents the corrector step and Eq. (8) represents the final current step. MacCormack
numerical scheme is a pseudo-transient process, where ∆t is a virtual time increment to obtain the final converged
numerical solution.
QQ mn
t
+
1
=
t mn
-∆
t
⎧│⎨│⎩
1
⌈
│
⎛││⎝
F∙
s⎞││⎠
-
⎛││⎝
sF ∙ ⎞ │ │ ⎠
⌉
│
+
⌈
│
⎛││⎝
t
t
n,m
mn ⎫ F ∙ s +
⎞││⎠
-
⎛││⎝
sF ∙ ⎞
⌉
-2
│ rV
mn
││⎠│
r
n,m
+
1⌊
n,m
+
12
│ ⎬ │ ⎭ r n,m
n,m
-12
+

∆ t S (6)
t mn │
│⌋
│
n,m
+
1⌊
n,m 1 2
n,m
1
⌋
QQ mn
t
+
1
=
t mn
-∆
t
⎧
1
⌈
⎛
│⎩
│
│
sF ∙
⎞│
-
⎛│
F∙s⎞│
⌉
│
+
⌈
│
⎛│
sF ∙ ⎞ │
-
⎛│
F ∙ s n,m
-
⎞
⌉
⎫
t
│
│
│
+∆S│⎭

(7)
The MacCormack final current step is calculated by the mean between the predictor and corrector steps,
│⎠
│⎨
│ ⎝ rV
mn
│
│⎠
⎝│
r
n,m ⌊
n,m
+
12
│
│⎠
⎝ r n,m
-
1
n,m
-
12
│
│⎠
⎝
2│⌋
│⎠
│
⎬t│
n,m ⌊
n,m + 1 2
n,m -
1
1⌋
Q mn t
+1
=12
⎛│⎝
QQ mn
t
+1+t
+ 1 ⎞ mn

(8)
where,

t
is current time, t 1+ is next time step, t∆ is a virtual time step and Q mn
1+t

is the final current step. The flow

vector F , inside Eq. (6) and Eq. (7), must be discretized in correct way to ensure the main feature of MacCormack
method: a numerical method of second order accuracy in time and space [21]. MacCormack scheme is explicit,
hence, it is a conditionally stable method. However, certain numerical conditions must be satisfied in order to attain
numerical convergence. These numerical conditions are represented by the CFL number (Courant-Friedrichs-Lewy
number) which depend mainly on the mesh size and time step [21].
The SIMPLE Method was applied to obtain coupling of velocity-pressure variables in the numerical scheme
developed in present work [1].
The scheme called "Ghost Volume" was used at the boundary conditions of control volumes situated at the
external faces at boundaries [21]. In these ghost volumes it can be applied boundary conditions of "Dirichlet Kind"
and "Neumann kind" [21]. Boundary conditions for physical consistency applied for metal flow in direct extrusion
analysis should be such as [19]:
• Inlet: the inlet velocity field should be prescribed (Dirichlet) and pressure should have zero gradient.
• Outlet: the outlet velocity field should have zero gradient (Neumann) and pressure should be prescribed.
• Symmetry line: should be prescribed zero gradients for normal surface gradient and parallel surface
components should use domain values.
• Solid Wall: at solid wall interface between material and die, consider friction model. The friction model used
was Friction Factor represented by nt
mk
τ is the tangential friction stress, k is
yield shear stress and m is friction factor. Considering isotropic material, von Mises yield criteria and shear strain
rate
nt

τ = , where
nt

γx , the shear friction stress is given by:

τ nt

= )3/2( σ ε x γ x nt .
4745 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

3. Material
Standard tensile test specimens were machined from an aluminium 6351 T5 bar [22]. Before machining, the
specimens were solubilized at 515 oC for 60 minutes and air cooled. Strain ε and strain rate εx
hardening behaviour of
material was obtained by fitting Hensel-Spittel law [23], Eq. (9), from tensile tests at temperature T=450oC. Fitted
parameters of Eq. (9) for tensile test curve are available in Forge 2008 and seen in Table 1. The curve is fairly linear
and represent a rigid-perfectly-plastic material, yield stress 255 MPa. Table 2 present other material properties.

σ=A
exp mmTm 1
T 9 ε 2 exp ( m 4 ε )( 1 + ε )
Tm 5 exp m 7 ε ε x Tmm 3 ε x 8 (9)
Table 1 – Hensel-Spittel law parameters obtained to model the flow stress of aluminium alloy 6351 T5 at 450oC.
Material
A (MPa)
m
1 (1/ oC)
m
2
m
3
m
4
m
5
m
7
m
8
m
9 (1/ oC)
(1/ oC)
AA 6351 953.65542 -0.00524 -0.01407 0.10998 -0.00913 0 0 0 0
Table 2 - Material, extrusion process and simulation parameters used in the
numerical simulations of hot extrusion of aluminum alloy Al 6351.
Parameters Values
Density (
ρ
) 2710 (kg.m-3)
Yield stress (
σY
) 255 (MPa)
Area reduction 89 %
Inlet extrusion velocity (V
o
) 10 (mm.s-1)
Quantity of control volumes 1110
Time step (∆t) 10-15 (s)
Die semi-angle (θ) 32.3°
Friction factor parameter (m) 0.5
Material hardening model in FVM: rigid-perfectly-plastic
Material hardening model in Forge 2008: Hensel-Spittel law

4. Extrusion modelling by Finite Element Method

For comparisons with present FVM analysis, the Forge 2008 software based on FEM was used to simulate the
extrusion process with the same conditions and material which were applied in experimental extrusion tests [22].
In FEM, it was employed the usual Lagragian approach. The initial undeformed billet was modeled by a mesh of
tetrahedral elements with constant size of 2 mm and die geometry is seen in Fig. 1. Table 1 presents the coefficients
used to model the billet material by the Hensel-Spittel law [22] employed in Forge 2008 extrusion simulation.
Extrusion process parameters used in FEM, which was the same for FVM, are seen in Table 2. The tools were
modeled as rigid. The viscoplastic friction model was adopted with the friction factor equal to 0.5.
5. Results and discussions
In Fig.1, mesh geometry with 1110 volumes employed in the present numerical simulations by FVM of hot direct
extrusion aluminium alloy Al 6351 is shown. The calculated average material parameters, extrusion process and
simulation parameters employed in FVM simulations of extrusion of aluminium 6351 are shown in Table 2. The
material was assumed rigid-perfectly-plastic with average yield stress of 255 MPa. The number of optimized
volumes was 1110 and about 50,000 iterations was necessary to attain numerical convergence in present FVM.
In Fig.2, the comparison of axial velocity V
z

and radial velocity V

r

simulation results from FEM by Forge 2008

software and FVM present work are presented. In general, the results are in good agreement. However, the accuracy
4746 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

depended on mesh size or refinement. Major discrepancies between simulations are at die exit region. In FEM,
distorted mesh at die exit region can affect accuracy. For higher extrusion ratio or area reduction, larger error are
expected in FEM simulations. Also, simulation running time to attain stead state in FEM is higher than in FVM.
In Fig.3, the influence of mesh refinement on FVM numerical results for equivalent strain rate distribution is
presented for two different number of volumes: 297 and 1110 volumes. Predicted maximum strain rate increased
from 2/s to 5/s near the die wall exit. For higher number of volumes, the numerical results are essentially the same.
r
32.3°
0.03
0.01
Fig. 1. Mesh with 1110 volumes employed in the numerical simulations by FVM of Al 6351.

a)
b)
c) d)
Fig. 2. Comparison of simulation results of extrusion of aluminum 6351. Axial velocity V
z
: a) FEM, Forge 2008 and b) FVM, present work. Comparison of radial
velocity V
r
: c) FEM, Forge 2008 and d) FVM, present work.
Fig. 3. Equivalent strain a) rate (1/s) distributions of direct extrusion b)
of Al 6351 from present work by FVM. Influence of mesh refinement: a) mesh
with 297 volumes and b) mesh with 1110 volumes.
4747 José D. Bressan et al. / Materials Today: Proceedings 2 ( 2015 ) 4740 – 4747

6. Conclusions
From the numerical simulation results for velocity and strain rate fields obtained in axisymmetric direct hot
extrusion, 89% of area reduction, of aluminum alloy 6351 billet from present FVM approach and validated by FEM
with FORGE 2008 software, the following conclusions can be drawn,
• Simulation results from FEM by Forge 2008 and FVM of present work are in good agreement. Thus, Present FVM
approach together with metal flow formulation has produced good and encouraging results.
• Present numerical scheme is explicit, hence, it is a conditionally stable method. In order to satisfies the CFL
condition for convergence it was necessary to use a minimum virtual time step of 10-15. This is an extremely low
value and, thus, the convergence time was influenced.
• It was necessary about 50,000 iterations to attain numerical convergence which is reasonable. Present FVM
numerical scheme together with fixed Eulerian reference coordinate system for metal flow analysis did not employed
the artificial viscosity for convergence and stability as required by incompressible viscous fluid.
• MacCormack and SIMPLE methods can also be extended for modelling and analysis of metal plastic flow, as
suggested by present work, besides the classical application to incompressible fluid flow analysis.
Acknowledgements
The authors would like to gratefully thank the financial support given by the PVS scholarship of CAPES-ITA of
Brazil, CNPq of Brazil, University of Santa Catarina State - UDESC, Brazil and UNICAMP/SP, Brazil.
References
[1] Martins MM, Bressan JD and Button ST, Lead extrusion analysis by finite volume method. In: Oñate E, Owen DRJ, Peric D
and Suárez B
(Eds), XII International Conference on Computational Plasticity. Fundamentals and Applications. COMPLAS 2013,
Barcelona, 2013. [2] Johnson W, Sowerby R and Venter RD, Plaine Strain Slip-Line Fields for Metal Deformation Processes.
Pergamon Press, Oxford, 1982. [3] Mellor PB and Johnson W, Engineering Plasticity. Van Nostrand Reinhold, 1985. [4]
Thomsen EG and Frisch J, Stresses and Strains in Cold-Extruding 2S-O Aluminum, Trans. ASME. 77, (1955) 1343-1353. [5]
Valberg HS, Applied Metal Forming: inclunding FEM analysis. Cambridge University Press, London, UK, 2010. [6] Avitzur B,
Metal Forming, the Application of Limit Analysis. Marcel Dekker, N.Y., 1980. [7] Kobayashi S, OH S and Altan T, Metal
Forming and the Finite-Element Method. Oxford University Press, 1989. [8] Ghosh S, A new finite element description for
simulation of metal forming processes. In: Thompson EG, Wood RD, Zienkiewicz OC and
Samuelson A (Eds), Numerical Methods in Industrial Forming Processes - Numiform 89, A.A. Balkema, (1989) 159-164. [9]
Zienkiewicz OC, Jain PC and Onate E, Flow of solids during forming and extrusion: some aspects of numerical solutions. Int. J.
Solids Struct.
14 (1978) 15-38. [10] Zienkiewicz OC and Huang GC, Adaptive modelling of transient coupled metal forming processes. In:
Thompson EG, Wood RD,
Zienkiewicz OC and Samuelson A (Eds), Numerical Methods in Industrial Forming Processes- Numiform 89, A.A. Balkema,
(1989) 3-10. [11] Souli M, Ouahsine A and Lewin L, ALE formulation for fluid-structure interaction problems. Comput. Methods
Appl. Mech. Engrg., 190
(2000) 659-675. [12] Oñate E, Cervera M and Zienkiewicz OC, A Finite-volume format for structural mechanics. Int. J. Num.
Meth. Engng.,37 (1994) 181-201. [13] Demirdzic I and Martinovic D, Finite volume method for thermo-elasto-plastic stress
analysis. Computer Meth. Applied Mech. and Engng.,
109 (1993) 331-349. [14] Greenshields CJ, Weller HG and Ivankovic A, The finite volume method for coupled fluid flow and
stress analysis. Comput. Model Simul.
Eng.,4 (1999) 213–218. [15] Bailey C and Cross M, A finite volume procedure to solve elastic solid mechanics problems in
three dimensions on an unstructured mesh.
Int. J. Num. Meth. Engng., 38 (1995) 1757–1776. [16] Wenke P and Whell MA, A finite volume method for solid mechanics
incorporating rotational degrees of freedom. Computers and
Structures, 81(2003) 321–329. [17] Bressan JD, Martins MM and Vaz Jr. M, Stress evolution and thermal shock computation
using the finite volume method. Journal of
Thermal Stresses, 33 (2010) 533 - 558. [18] OPENFOAM – The Open Source CFD Toolbox, “Programmer’s Guide; Versão
1.6”, http://www.openfoam.com/docs/. [19] Jasak H, Error analysis and estimation for the Finite Volume Method with
applications to fluid flows, Ph.D. Thesis, Imperial College,
London, 1996. [20] Basic H, Demirdzic I, Muzaferija S, Finite volume method for simulation of extrusion processes. Int. J.
Num. Meth. Engng. 62(2005)475-494 [21] Tannehill JC; Anderson DA and Pletcher RH, Computational Fluid Mechanics and
Heat Transfer. Taylor&Francis, London, 1997. [22] Bressan JD, Martins MM and Button ST, Aluminium extrusion analysis by
the finite volume method. In: Oñate E, Owen DRJ, Peric D and Suárez B (Eds), XII International Conference on Computational
Plasticity. Fundamentals and Applications. COMPLAS 2013, Barcelona, 2013. [23] Hensel A and Spittel T, Kraft-und
Arbeitsbedarf bildsamer Formgebungsverfahren. VEBDeutscher Verlag fur Grundstoffindustrie, Leipzig,
1978.