Numerical Simulation in Bulk Forming Process: Journal For Technology of Plasticity, Vol. 32 (2007), Number 1-2

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Journal for Technology of Plasticity, Vol.

32 (2007), Number 1-2


NUMERICAL SIMULATION IN BULK FORMING
PROCESS
Vukevi Milan, Janji Mileta, ibali Nikola
University of Montenegro, Faculty of Mechanical Engineering, Podgorica, Montenegro
ABSTRACT
Investigation of stress-strain state of bulk forming in open dies has been presented in the paper. By
numerical simulations it is possible, by discretization of the process per deformation phases, to
determine deformation parameters in all phases. Logarithmic strain tensor components are
determined by meridial point cross-section per deformation phase displacements, where a total
strain is calculated as a sum of strains obtained in deformation phases. Stress tensor component is
determined by the method of higher theory of plasticity.
For numerical simulation, DEFORM-2D software package is used, meant to analyze plane and
axi-symmetrical deformations. A stepped concave die shape is adopted, as well as working-piece
dimensions and node point distribution, whose displacements are fallowed during deformation
process. Numerical simulation consists of three phases.
In the first phase numerical experiment is carried out, where, for adopted point coordinates in
non-deformed state, point displacements at the end of deformation process, are obtained. Stress-
strain state parameters are determined on the base of these data at displacement points. The
second phase relates to numerical experiment that is carried out for the same initial conditions,
but bulk deformation process is observed in deformation phases. In each phase, stress-strain
parameters are determined, based on which parameters at the end of deformation process are
obtained. The third phase concerns stress state parameters, determined by DEFORM-2D, and they
are related to the same initial conditions.
Key words: bulk forming, numerical simulation, DEFORM-2D
1. INTRODUCTION
As a result of a rapid development of computer engineering in the field of metal forming a number
of different numerical methods has been developed and used so far. Finite Elements Method -
FEM has been used as the most powerful numerical method. Based on this method, several
commercial software packages for numerical process simulations have been developed. Among
them, one of the most famous software package is DEFORM, produced by Scientific Forming
Technologies Corporation (SFTC), which has been used in this work.
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
48
2. INPUT PARAMETERS OF NUMERICAL INVESTIGATIONS
Bulk forming process in open dies includes a wide class of problems, both from the aspect of
geometry and technological condition. The elements given in Fig.1 have been analysed in the
paper.
- Experimental material is aluminium alloy AlMgSi0,5.
- Investigation is carried out at temperature of hot treatment T=440 [
o
C].
- Deformation is realized at low constant deformation velocity, v=2 [mm/s].
- Hardening curve parameters are c=30.34434 and n=0.097808 for AlMgSi0,5 aluminium
alloy and temperature T=440 [
o
C].
- Friction factor is m=0.114.
- Tool is of stepped concave shape (Fig. 1.). It consists of two pieces, upper and lower ones.
The upper die piece consists of two height degrees where one is with degree, whereas lower
one consists of one height degree.
Working-pieces are cylindrical, of diameter d
0
=33.56 [mm]. The height h
0
is determined out of the
working-piece constant volume before and after pressing process for adopted die dimensions given
in Fig. 1, and it is h
0
=29.58 [mm].
Point coordinates, whose displacement will be followed in numerical experiment and whose stress-
strain state parameters will be determined, are given in Fig. 5. The adopted network of the half of
the axi-symmetrical working-piece is given in Fig. 2. Total node point number is 140.
Figure 1 - Working-piece within dies Figure 2 - Adopted network of the halt of
axi-symmetrical working-piece
3. NUMERICAL SIMULATION FOR CONCAVE DIE SHAPE
For the adopted tool geometry (Fig. 1.) and adopted working-piece geometry (Fig. 2.), as well as
for known friction and hardening curve factors, numerical simulation has been carried out, where
the data input into DEFORM's module Pre Processor, are given in Table 1. Default values of
DEFORM-2D software package are taken for other values.
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
49
Table 1. Input data of DEFORM simulation for concave die shape
Units UNIT ;SI
Simulation Parameters
Geometry GEOTYP ;Axisymmetric
Number of simulation steps NSTEP=1000
Step increment to save STPINC=10
Primary die PDIE(1)=1
Steps by ;Stroke
Step Controls
Stroke per step DSMAX=0.1 [mm]
Stopping Controls Primary die displacement SMAX=0, 11.58 [mm]
X [mm] Y [mm] R [mm]
1 0 29.58 0
2 9.7202 29.58 1
3 10 32.58 1
4 19.300732 32.58 1
5 20 22.58 1
6 35 22.58 0
G
e
o
m
e
t
r
y

7 35 32.58 0
Speed 2 [mm/s]
Name: Upper die
;Rigid
Movement controls
Angle -90
X [mm] Y [mm] R [mm]
1 35 0 0
2 35 10 0
3 20 10 1
4 19.300732 0 1
Name: Lower die
;Rigid
G
e
o
m
e
t
r
y

5 0 0 0
X [mm] Y [mm] R [mm]
1 0 0 0
2 16.78 0 0
3 16.78 29.58 0
G
e
o
m
e
t
r
y

4 0 29.58 0
Name: Working-piece
;Plastic
Mesh Number of mesh elements MGNELM=1000
Contact relation CNTACT ;Master-Slave
Friction model FRCFAC Shear
Upper die -
- Working-piece
Friction FRCFAC=0.114
Contact relation CNTACT ;Master-Slave
Friction model FRCFAC Shear
Lower die -
- Working-piece
Friction FRCFAC=0.114
Finite element network forming process and data base generation are valid for the initial step "-1",
(Fig. 3.). During deformation process, remeshing was done four times, and the process ended in
12
th
phase. After simulation process starting, numerical calculations for DSMAX stroke per step
value are done, and each tenth step is saved in data base. After the numerical simulation is done, a
final working-piece look is obtained in 12
th
phase with a deformed mesh (Fig. 4.).
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
50
Figure 3 - Working-piece with a deformed
network in mesh with dies
Figure 4 -. Working-piece at the end of
deformation process
To compare the results, it is necessary to input the adopted node points r
p0
and z
p0
, into
Point Tracking sub-module, and their distribution is given in Fig. 5. After inputting node point
coordinates of non-deformed state into Point Tracking sub-module, node point coordinates per
phases are generated. The distribution of point in 12
th
or final phase is given in Fig. 6. Using Data
Extract order, node point coordinates per phases are also obtained at the end of deformation
process, namely in 12
th
phase r
p12
and z
p12
.
Figure 5 - Adopted distribution of node point in
non-deformed state
Figure 6 - Point distribution in 12
th
phase
obtained by DEFORM simulation
3.1. Numerical experiment for process continuity
Deformation parameters have been determined by the obtained node points at the end of
deformation process. Data have been processed in MATLAB (v. 7.0). Input data are point
coordinates at the beginning r
p0
, z
p0
and at the end of deform process r
p12
, z
p12
, (Fig. 5. and Fig. 6.).
Point displacements are expressed by:
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
51

=
=
0 12 12
0 12 12
p p z
p p r
z z u
r r u
(1)
Using displacements point u
r12
and u
z12
, partial displacement statements per radius and height
are calculated: cu
r
/cr, cu
r
/cz, cu
z
/cz and cu
z
/cr. Based on these statements, components of
small strains may be determined [6]. Using relations of relative and logarithmic strains (2)
logarithmic strain values are obtained to be compared to the values of numerical FEM
simulation.
) 1 ln( c + = (2)
Effective logarithmic strain values are given in the form of three-dimensional diagram in Fig. 7.
-10
0
10
20
30
-10
0
10
20
30
0
0.5
1
1.5
2
Radius r [mm]
NUMERICAL EXPERIMENT FOR PROCESS CONTINUTY
Height z [mm]
E
f
f
e
c
t
iv
e

lo
g
a
r
it
h
m

d
e
f
o
r
m
a
t
io
n

|
e

Figure 7 - Effective logarithm strain


e
To determine strain rate components, it is necessary to determine displacement speeds, based on
point displacements at the beginning and at the end of final deformation interval, provided strain
rate is constant. For tool stroke increment Az=2 [mm] and deformation velocity v=2 [mm/s], time
increment is At = 1 [s]. For the adopted increment, node point coordinate values are obtained from
the Point Tracking sub-module in 10
th
phase, representing the beginning of the observed interval.
Point coordinates at the end of deformation process, namely at the end of the observed interval are
marked by "12" in index.
Input data are taken from working-piece deform analysis, namely: r
p0
and z
p0
, r
p12
and z
p12
, as well
as the obtained node points r
p10
and z
p10
.
Thus, displacements are:

=
=
0 10 10
0 10 10
p p z
p p r
z z u
r r u
(3)
Displacement increment is:
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
52

A = = = A
A = = = A
z z z u u u
r r r u u u
p p z z z
p p r r r
10 12 10 12
10 12 10 12
(4)
Further, displacement velocities in node points of the deformed network are determined (phase 12)
where partial displacement speeds per radius cv
r
/cr, cv
z
/cr and height cv
z
/cz, cv
r
/cz, are
determined, based on which strain rates are obtained. In this way we come to radial, axial and
angular deformation velocity components, whereas rim velocity is determined as a ratio between
radial displacement velocity and radius.
Effective strain rate values are given in the form of three-dimensional diagram in Fig. 8.
Stress is determined by using the method of higher theory of plasticity. The method is based on
obtaining axial stress component
z
, by solving the basic equation of plasticity [6]. Input data are:
node point coordinates at the beginning r
p0
and z
p0
and at the end of the process r
p12
and z
p12
,
hardening curve parameters c and n, as well as the results out of deformation
r
,
z
,
u
,
rz
and
e
and kinematic analyses
r
c ,
z
c ,
u
c ,
rz
and
e
c .
Effective stress values are given in the form of three dimensional diagrams on Fig. 9.
-10
0
10
20
30
-10
0
10
20
30
0
0.5
1
1.5
Radius r [mm]
NUMERICAL EXPERIMENT FOR PROCESS CONTINUITY
Height z [mm]
E
f
f
e
c
t
i
v
e

d
e
f
.

s
p
e
e
d


c
e
0
10
20
30
0
5
10
15
20
25
2.6
2.8
3
3.2
3.4
Radius r [mm]
NUMERICAL EXPERIMENT FOR PROCESS CONTINUITY
Height z [mm]
E
f
f
e
c
t
iv
e

s
t
r
e
s
s

o
e
Figure 8 - Effective strain rate ] [
1
s
e
c Figure 9 - Effective stress
e
[daN/mm
2
]
3.2. Numerical experiment per phases
With numerical experiment per phases, displacements per phases are determined first, after that
final value of strain, strain rate and stress are obtained by known procedures and methods. Based
on the obtained node point coordinates per phases, data are processed by a programme made
within a programme language MATLAB (v. 7.0). Input data are: node point coordinates at the
beginning r
p0
and z
p0
and per phases: r
pi
and z
pi
, 12 ,..., 2 , 1 = i . Point displacements are expressed
by.

=
=

1
1
pi pi zi
pi pi ri
z z u
r r u
12 ,..., 2 , 1 = i (5)
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
53
Where logarithmic strain are obtained
ri
,
zi
,
i u
,
rzi
and
ei
, 12 ,..., 2 , 1 = i at the end of
each phase of bulk forming.
Logarithmic strain values in 12
th
phase (at the end of bulk deformation process) are calculated by
(6).
_
=
=
12
1 i
ri r
,
_
=
=
12
1 i
zi z
,
_
=
=
12
1 i
i u u
,
_
=
=
12
1 i
rzi rz
i
_
=
=
12
1 i
ei e
(6)
Effective logarithmic strain values obtained per phases are given in the form of a three-
dimensional diagram in Fig. 10.
-10
0
10
20
30
-10
0
10
20
30
0
0.5
1
1.5
2
2.5
Radius r [mm]
NUMERICAL EXPERIMENT PER PHASES
Height z [mm]
E
f
f
e
c
t
i
v
e

l
o
g
a
r
it
h
m

d
e
f
o
r
m
a
t
io
n

|
e

Figure 10 - Effective logarithmic strain


e
To determine strain rate tensor components by numerical experiment per phases, it is necessary to
know logarithmic strain values in each deformation phase. By interpolation to cube polynom, it is
possible to determine logarithmic strain increment for a given time increment, namely logarithmic
strain rate.
By differentiating the expression (2) per time dt, a relation of relative and logarithmic strain rate is
obtained.
c

= e (7)
For adopted deformation velocity v=2 [mm/s], in 12
th
phase for time increment dt=0.1, relative
strain rates are determined.
Effective strain rate values in the observed points of meridial working-piece cross-section are
given in the form of a three-dimensional diagram in Fig. 11.
Stress is determined by using the method of high plasticity. Input data are: node point coordinates
at the beginning r
p0
and z
p0
and at the end of deformation process r
p12
and z
p12
, hardening curve
parameters c and n, as well as the results obtained by deformation
r
,
z
,
u
,
rz
and
e
and
kinematic
r
c ,
z
c ,
u
c ,
rz
and
e
c analyses.
Effective stress values are given in the form of a three-dimensional diagram in Fig. 12.
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
54
-10
0
10
20
30
-10
0
10
20
30
0
1
2
3
4
Radius r [mm]
NUMERICAL. EXPERIMENT PER PHASES
Haight z [mm]
E
f
f
e
c
t
iv
e

d
e
f
.

s
p
e
e
d


c
e
0
10
20
30
0
10
20
30
2.6
2.8
3
3.2
3.4
3.6
Radius r [mm]
NUMERICAL EXPERIMENT PER PHASES
Height z [mm]
E
f
f
e
c
t
iv
e

s
t
r
e
s
s

o
e
Figure 11 - Effective strain rate ] [
1
s
e
c Figure 12 - Effective stress
e
[daN/mm
2
]
3.3. DEFORM results
For concave die shape deformation change, speed and stress values in each phase of the observed
process for the adopted node points are obtained directly out of DEFORM-2D software package.
Values of these components at the end of deform process, effective logarithmic strains, effective
strain rates and effective stress are given in Fig. 13. to Fig. 15.
-10
0
10
20
30
-10
0
10
20
30
0
1
2
3
4
Radius r [mm]
DEFORM RESULTS
Height z [mm]
E
f
f
e
c
t
iv
e

lo
g
a
r
it
h
m

d
e
f
o
r
m
a
t
io
n

|
e

-10
0
10
20
30
-10
0
10
20
30
0
1
2
3
4
5
6
Radius r [mm]
DEFORM RESULTS
Height z [mm]
E
f
f
e
c
t
iv
e

lo
g
.

s
p
e
e
d

c
e
Fig. 13. DEFORM effective logarithm strain
e
Figure 14 - DEFORM effective strain rate
] [
1
s
e
c
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
55
-10
0
10
20
30
-10
0
10
20
30
0.5
1
1.5
2
2.5
3
3.5
Radius r [mm]
DEFORM RESULTS
Height z [mm]
E
f
f
e
c
t
iv
e

s
t
r
e
s
s

o
e
Figure 15 - DEFORM effective stress
e
[daN/mm
2
]
4. CONCLUSIONS
Numerical simulations of bulk deformation process have been carried out in the current paper, by
using finite element method, as well as the programmes based on plasticity theory. The numerical
experiments carried out simulate the conditions of a real object. Stress-strain deformation state
parameters of a working-piece in open dies of axi-symmetrical elements have been determined.
Numerical simulation is performed for a stepped concave die shape and it consists of three phases:
numerical experiment for process, numerical experiment per phases and DEFORM results.
REFERENCES
[1] auevi M.: Theory of Plastic Metal-Working. Svjetlost, Sarajevo, 1979.
[2] DEFORM Mannual - SFTC.
[3] Janji M., Vukevi M., Domazetovi V.: Application of Discretization Method in
Deformation Analysis of odpreska. XXIV JUPITER Conference, Zlatibor, 1998.
[4] Janji M., Domazetovi V., Savievi S., Vukevi M.: Determining hardening curves and
their choice in numerical simulations. XXXI JUPITER Conference, Zlatibor, 2005.
[5] Janji M.: Investigation strain-stress parameters in bulk-forming processes - Doctor's thesis,
Faculty of Mechanical Engineering, Podgorica, 2005.
[6] Musafija B.: Applied Plasticity Theory, University in Sarajevo, Sarajevo, 1973.
[7] Musafija B.: Plastic deformation Metal Working. Svjetlost, Sarajevo, 1972.
[8] Plancak M.: Strain-stress state in processes of cold steel extrusion. Faculty of Engineering
Sciences, Novi Sad, 1984.
[9] Vukevi M., Janji M., Domazetovi V.: Influence of geometrical parameters at deforming
axi-symmetrical samples in open-dies. XXVI JUPITER Conference, Beograd, 2000.
[10] ibali N.: Modeling of bulk metal forming processes using methods of physical
discretization and numerical simulation - Master thesis, Faculty of mechanical Enginering,
Podgorica, 2007.
Journal for Technology of Plasticity, Vol. 32 (2007), Number 1-2
56
NUMERIKE SIMULACIJE KOD PROCESA
ZAPREMINSKOG DEFORMISANJA
Vukevi Milan, Janji Mileta, ibali Nikola
Univerzitet Crne Gore, Mainski fakultet, Podgorica, Crna Gora
REZIME
U radu je data simulacija naponsko deformacionog stanja zapreminskog deformisanja u
otvorenim kalupima, po procedurama koje predstavljaju odreene metode. U numerikim
simulacijama mogue je diskretizacijom procesa po fazama deformisanja, odrediti deformacione
parametre u svim fazama. Komponente tenzora logaritamskih deformacija odreuju se na osnovu
pomjeranja taaka meridijalnog presjeka po fazama deformisanja, a ukupna deformacija se
izraunava kao suma deformacija dobijenih u fazama deformisanja. Dok se komponente tenzora
napona odreuju metodom visioplastinost.
Za numeriku simulaciju koristi se softverski paket DEFORM-2D, koji je namijenjen za analizu
ravanskih i osnosimetrinih deformisanja. Usvojen je stepenasto konkavni oblik kalupa, dimenzije
pripremka i raspored vornih taaka ija se pomjeranja prate tokom procesa deformisanja.
Numerika simulacija se sastoji iz tri etape:
U prvoj etapi se izvodi numeriki eksperiment gdje se za usvojene koordinate taaka u
nedeformisanom stanju dobijaju pomjeranja taaka na kraju procesa deformisanja. Na osnovu
ovih podataka o pomjeranju taaka odreeni su parametri naponsko deformacionog stanja. Ovaj
numeriki eksperiment se odnosi na kontinuitet procesa.
Druga etapa se odnosi na numeriki eksperiment, koji se realizuje za iste poetne uslove, ali se
proces zapreminskog deformisanja posmatra u fazama deformisanja i u svakoj fazi se posebno
odreuju naponsko deformacioni parametri, na osnovu kojih se dobijaju parametri na kraju
procesa deformisanja.
Trea etapa se odnosi na parametre naponsko deformacionog stanja koje DEFORM-2D
odereuje, a odnose se na iste poetne uslove.

You might also like