Investigation of forming speed and friction on drawability of Al 5182-O using a

servo press with CNC cushion

1,2 2 3 2
Long Ju , Tingting Mao , Shrinivas Patil , Taylan Altan
1
School of Mechanical Engineering, University of Science and Technology Beijing, Beijing 10083,
China
2
Center for Precision Forming, The Ohio State University, Columbus, Ohio 43210, USA
3
Aida-America Corporation, Dayton, Ohio 45424, USA

Abstract
Forming speed and lubrication affect parts quality and productivity in automotive industry. Servo
presses, increasingly used in the automotive stampings, have the capability to provide variable ram
motions for various stamping operations. In the present study, deep drawing tests were conducted in
a servo press to investigate the effects of forming speed on the drawability for Al alloy (5182-O)
sheets at room temperature. Two lubricants, dry film lubricant and commercial mineral oil lubricant,
were evaluated under different ram motions, respectively. A non-isothermal finite element (FE)
model was used to design and analyze the deep drawing tests under various forming speeds.
Experimental results showed that deeper draw depths could be achieved when applying higher
forming speeds. The dry film lubricant was found to provide extremely low friction coefficient,
compared with the oil-based lubricant. However, the dry film lubricant showed more sensitivity to the
changes of forming speed. The punch load was shown to decrease with increasing forming speed in
the tests. The temperature increase, induced by plastic deformation and friction work, was also
estimated by using the thermal-mechanical FE model at each drawing speed. The results and
discussions give insights into the effects of forming speed on deep drawing process of aluminum
alloys at room temperature.
Keywords:
Forming speed; Drawability; Al alloy; Non-isothermal FE model

development time and cost, finite element method

1 INTRODUCTION
(FEM) offers special advantages in sheet metal
Aluminum alloys are playing a significant role in forming analysis. With the FE-based inverse analysis,
automotive manufacturing, due to their high strength a reliable and practical methodology was developed
and light weight. However, forming of aluminum alloys to evaluate the performance of different lubricants
presents new challenges because of aluminum’s low using deep drawing test. The performance of some
formability. Deep drawing is the most common tested lubricants was found to change with contact
process used in sheet metal forming. Considering the pressure and ram speed [6]. By using FE models, the
drawability and quality, drawing of Al parts could be optimal ram speed could be estimated which is
affected by several factors, such as material beneficial to prevent failures [7]. Additional studies
properties, tool geometry, interface conditions and have focused on the effects of deformation and
process parameters [1]. Among various proven friction induced heat on the formability of advanced
results, forming speed was found to have particular high strength steels (AHSS). Maximum temperatures
effects on deep drawability of Al alloys. In a recent 86 ºC in the deformed cup (DP590) and 46 ºC in the
study, it was found that not only a higher forming die corner were predicted by using non-isothermal FE
speed could help enhance the draw depth of an simulations [8]. In drawing aluminum alloys, the small
experimental aluminum alloy door panel, but also the temperature rise (32 ºC) in the die corner was
increase in the forming rate affects drawability [2]. predicted by an analytical model in strip drawing.
Several reasons were suggested to explain the Adhesive wear, as a result of local peak stresses, was
effects of forming speed on the drawability of found in the same location [9].
aluminum alloys at room temperature, such as stain- Considering the practical aspects of forming, the
rate and temperature dependent material properties, application of servo drive presses continues to
heat generation at critical locations and friction force increase in blanking, deep drawing and warm forming
varying with sliding speed [1, 3-5]. Especially, the of light weight alloys [10]. The most impressive
relationship between the coefficient of friction and the advantage of servo press is the free slide motion
sliding speed was investigated by using strip drawing control, such as reverse motion, variable-speed and
tests. Results demonstrated that the punch speed has acceleration/deceleration functions [11]. For deep
a significant effect on drawing force [5]. To reduce
drawn parts such as door panels and fenders, it is 2.2 Illustration of tooling for the deep drawing
verified that the draw depth and productivity have test
been improved by using a servo press line [12]. The The schematic of the deep drawing tooling is
special slide motions, such as pulsating and stepwise illustrated in Figure 2. This die was originally designed
drawing motions, controlled by hydraulic press or to form advanced high strength steels. The die cavity
mechanical servo press, were successfully applied in was assembled by several inserts with different radii
drawing operations to prevent wrinkles and cracks and curvilinear shapes. The clearance between punch
[11, 13]. and die was 1.6 mm. The maximum drawing depth
The main objective of the present study is to allowable in this die is about 80 mm.
investigate the effects of forming speed on drawability
under different friction conditions. By means of deep
drawing tests, dry film lubricant and mineral oil
lubricant were evaluated for forming the aluminum
alloy 5182-O. Different speed profiles under two slide
motions were tested in a servo press. Non-isothermal
finite element analysis was developed to determine
forming parameters for the tests and to investigate
more detailed effects of forming speed on the drawing
process at room temperature. The comprehensive
discussions on material properties, friction and heat
(a)
generation are provided to help to understand the
influence of drawing speed on the practical stamping
operations.

2 EXPERIMENTAL PROCEDURES
2.1 Test material and dry film lubricant
Aluminum alloy 5182-O has an extensive application
in automotive industry, due to its light weight and
good drawability. In the present study, the material
properties were characterized by both tensile test and
viscous pressure bulge (VPB) test. The tensile test
data included 22% uniform elongation, 124 MPa yield (b)
stress and 285 MPa ultimate tensile stress. VPB test, Figure 2: (a) Top and (b) cross-sectional view of deep
developed by CPF, could provide higher strain values drawing tooling, dimensions are in mm (R1=1501.6,
under biaxial state of stress [14]. Fig. 1 shows the R2=1998.4, R3=51.6, R4=55.6, R5=61.6, R6=66.6,
flow stress curves obtained from the two tests. It can R7=20, R8=10).
be seen that, with the VPB test, the flow stress could
be obtained at true strain about 0.5. In a previous 2.3 Servo press
study on Al 5182-O, various lubricants were evaluated In the current experimental study, a 300 ton Aida
using cup drawing test in a hydraulic press. It was mechanical servo press with 25 ton CNC cushion was
shown that dry film lubricant has a better performance used to conduct the deep drawing tests, as shown in
during drawing operation [15]. The Al 5182-O blanks Figure 3. The ram movement can be programmed to
(thickness 1.2 mm) used in the present study were several desired profiles. Using the CNC die cushion,
2
pre-coated with dry film lubricant (1±0.3 g/mm ). the press has the capability to control the forming
500
speed and blank holder force in deep drawing
Tensile test, operations. The die cushion force can be varied
True stress (MPa)

400 1.5 mm through the stroke and the maximum value is 25 ton.
The press cushion also has the pre-acceleration
300 function, which can be controlled to several states
Bulge test,
200 1.2 mm (strong, medium and weak), in order to reduce the
impact of the die upon the blank. The data
100 measurement and storing function can record ram
positions, speeds, as well as punch and die cushion
0
loads in real time.
0 0.1 0.2 0.3 0.4 0.5
True strain

Figure 1: Flow stress curves of Al 5182-O obtained

from tensile test and VPB test.
 ram speed. Thus, according to Eq.1, the equivalent
-1 -1
strain rate range 0.001s and 0.1s during the
drawing process could be estimated. Although the
enhanced formability in Al 5182-O during high strain
rate free-forming has been reported, the material
properties still exhibited very little sensitivity at quasi-
-1
static strain rate range 0.001~0.1s at room
temperature [17].

3 FINITE ELEMENT SIMULATIONS

3.1 FE modeling of deep drawing test
Figure 3: Experimental deep drawing tooling The finite element simulations were carried out to
assembled in the 300 ton servo press design the blank shape and to estimate forming
2.4 Ram speed for the deep drawing test parameters, such as blank holder force and draw
depth. Generally, the heat generation, induced by
Two types of speed profiles are analyzed, including 1)
deformation and friction, is neglected in most FE
mechanical crank motion (1 SPM, 10 SPM and 18
analysis of sheet forming process at room
SPM), and 2) constant speed during deformation (50
temperature. However, it may affect the material
mm/s and 310 mm/s). As seen in Figure. 4, the
behavior, friction conditions and even the
mechanical crank motion is a basic sine curve.
performance of lubricants. To understand the
However, for constant speed profile, the ram speed
complex interactions between those factors during
could not be kept constant during the whole forming
deep drawing of Al alloys at room temperature, it is
stroke, because at a certain stroke position it is
necessary to analyze the process using a non-
necessary for the slide to decelerate to the bottom
isothermal FE model.
dead center (BDC).
In the current study, a 3D model is built for non-
isothermal forming using commercial software
PAMSTAMP (Figure 5a). The punch is fixed in the
press, while upper die and blank holder move
towards the punch. The blanks are designed with
rectangular shape with chamfered corners (Figure
5b). In the FE model, 4-node shell element was used
to present the deformation and heat transfer behavior
of the blank. The mechanical and thermal properties
of the sheet material are summarized in Table 1.
Based on previous study on lubricant using cup
drawing test, the range of coefficient of friction (COF)
0.08~0.12 was selected for dry film lubricant and
mineral oil lubricant [15]. In the current FE
Figure 4: Ram speed profiles with crank motion (1 simulations, the specific COF for each forming
SPM, 10 SPM and 18 SPM) and constant (50 mm/s condition is determined by minimizing the difference
and 310 mm/s) during deformation. between measured and predicted results, including
Since the ram speed changes during the drawing flange length and punch load. The tools are set as
process, the sheet material deforms at variable strain rigid bodies with 6 mm thermal thickness, to enable
rates. In round cup drawing, the relationship between the calculation of temperature gradient. The initial
effective strain rate at the flange and punch speed temperature of blank and tools is 25 °C. Based on
can be expressed as Eq. (1) [16]: previous study on warm drawing of Al 5754 round
cups, in order to emulate the effect of contact
(1) pressure on the heat transfer coefficient (HTC), the
Where, is effective strain rate at the flange region; interface HTC of sheet-punch is selected as 5000
2 2
W/m ·K, while 1000 W/m ·K is for sheet-die and
represents punch speed; is punch
sheet-blankholder [18]. However, the effects of
radius; r is initial blank radius (in round cup drawing). convection and radiation were negligible in the
However, for the rectangular shape punch and sheet current cases.
used in the current case, the approximate equivalent
values were used for the calculation. The dimensions die
can be obtained from Figure 2.
The maximum ram speed 18 strokes per minute
(SPM), corresponding to the contact speed of around Blank
310 mm/s, was suggested in the tests (as shown in Blank holder Punch
Figure 4). The average speed during deformation is
approximately 10 mm/s when using 1 SPM as the (a)
 30 Measurement

Flange length (mm)

25 FE simulation
20
15
10
5
0
L1 L2 L3 L4

(b) Figure 7: Comparison of flange lengths between

experiment and FE simulation (COF 0.1 was applied).
Figure 5: (a) 3D FE model of the deep drawing
process; (b) Designed blank dimension (mm) 300

Punch force (kN)

250
Table 1: Al 5182-O material properties used in the FE
simulations 200
Property Description 150
100
Flow stress curve Bulge test data (Fig. 1) FE simulation (cof=0.1)
50
Young’s modulus (E) 70.6 GPa Experiment
Poisson ratio (υ) 0
0.341
0 20 40 60 80
Coefficient of friction (COF) 0.08~0.14 Draw depth (mm)
Thermal conductivity (λ) 130 W/m·C
Figure 8: Load-stroke curves from experiment and FE
Specific heat capacity (C) 900 J/kg·C simulation (COF 0.1 was applied).
According to the preliminary tests and FE predictions,
3.2 Validation of FE model and determination of the minimum acceptable BHF to prevent heavy
forming parameters wrinkles is 125 kN, which is kept constant during the
Preliminary tests were conducted to set up the 300 whole forming stroke. It is found that, under 125 kN
ton servo press and validate the FE model developed BHF, the critical draw depth is 60.8 mm. All tests
in PAMSTAMP. As can be seen in Figure 6 (a), the under same forming condition were repeated three
part was successfully drawn to 60.8 mm at 310 mm/s, times. In this study, mineral oil lubricant is also used
while the blank holder force (BHF) 125 kN was to investigate the effects of lubrication. The test
applied. Figure 6 (b) shows the simulated part in conditions for the deep drawing tests are summarized
PAMSTAMP. By using COF 0.1, the flange lengths in Table 2.
and load stroke curves were predicted and compared
with measurements. The comparisons show a good Table 2: Process conditions used in deep drawing
match between experiments and simulated results, as tests
illustrated in Figure 7 and Figure 8.
Conditions Description
Sheet material Al 5182-O
Thickness (t) 1.2 mm
Testing speed profile See Figure 4
Blank holder force
125 kN (constant)
(BHF)
Dry film lubricant/Mineral
Lubricant
Oil
(a) (b)
Draw depth (d) 60.8 mm (Max. 80 mm)
Figure 6: (a) Defect free formed part (draw depth 60.8
mm); (b) thinning distribution of simulated part in Cushion pre-
Medium
PAMSTAMP. acceleration

4 RESULTS AND DISCUSSION

4.1 Experimental results of deep drawing tests
Figure 9 and Figure 10 present the test results for the
samples coated with dry film lubricant. It was found
that, the part could be formed without cracks when
applying higher crank motion speed (18 SPM), as approximately 52.9 mm, 55.5 mm and 53.6 mm at 1
shown in Figure 9. Due to the convex shape and the SPM, 10 SPM and 50 mm/s, respectively. It also can
smaller corner radius (R5 shown in Figure 2), the be found from Figure 11 that the maximum punch
cracks always initially occurred at the right corner and load is increased by 19 kN when using lower ram
the localized necking was observed at the left corner speed 1 SPM, while as shown in Figure 12, the
when using 10 SPM. More severe cracks occurred at increase amounts to 13 kN by using 50 mm/s.
both left and right corners when 1 SPM was applied. The punch load curves for drawing with mineral oil
When using constant speeds during deformation, it lubricant are shown in Figure 13. Comparing with
also shown that a better drawing quality was obtained using dry film lubricant (Figure 11), the maximum
at 310 mm/s than at 50 mm/s, as shown in Figure 10. punch load was reduced by around 8% when using
In order to investigate the effects of lubricant on oil lubricant in the deep drawing test. Several factors
maximum draw depth under different forming speed, may have affected the drawing results, including 1)
commercial mineral oil lubricant, originally used for the mineral oil was manually applied on the sheet
AHSS deep drawing, was also tested at 10 SPM and surface while the amount was not precisely
18 SPM. BHF and pre-acceleration were kept the controlled; and 2) the punch and die were cleaned
same as the tests with dry film lubricant. Experimental with slight mineral oil instead of acetone.
results showed that draw depth 80mm, which is the
300
maximum allowable stroke for the die, could be 19 kN
successfully achieved without any defect, at both 10 250

Punch load (kN)

SPM and 18 SPM.
200

10 SPM 150

(b) 100 1 SPM

18 SPM 10 SPM
50
18 SPM
0
Figure 9: Results of deep drawing tests with dry film 0 20 40 60 80
lubricant under different crank motion based forming Draw depth (mm)
speeds, stroke 60.8 mm: (a) 10 SPM, crack at right
Figure 11: Load-stroke curves at forming speeds 1
corner and necking at left corner, (b) 18 SPM, defect
SPM, 10 SPM and 18 SPM, with dry film lubricant
free formed part
(BHF 125 kN, stroke 60.8 mm)
(c) 300
50 mm/s 13 kN
250
Punch load (kN)

200

(d) 150
100
310 mm/s 50 mm/s
50 310 mm/s
0
0 20 40 60 80
Figure 10: Results of deep drawing tests with dry film Draw depth (mm)
lubricant under different constant forming speeds, Figure 12: Load-stroke curves at forming speeds
stroke 60.8 mm: (c) 50 mm/s, crack at right corner, 50mm/s and 310mm/s, with dry film lubricant (BHF
(d) 310 mm/s, defect free formed part 125 kN, stroke 60.8 mm)
4.2 Comparison of punch load and determination 300
of friction coefficient
250
Punch load (kN)

It was reported that a sudden decrease of punch load

during the forming stroke could roughly indicate the 200
fracture in the failed part. Figure 11 and Figure 12
150
illustrate the punch load versus stroke with dry film
lubricant under the speeds mentioned above. The 100
punch load curves at 10 SPM and 18 SPM are found
10 SPM
50 18 SPM
to be so close as a result of the small difference
between their average speeds during deformation. 0
From the results, the load drop could be obviously 0 20 40 60 80
observed at forming speeds 1 SPM, 10 SPM as well Draw depth (mm)
as 50 mm/s. It is seen that the limit draw depths are
 Figure 13: Load-stroke curves at forming speeds 10 location of peak temperature in the punch moved
SPM and 18 SPM, with oil lubricant (BHF 125 kN, from the corner to the wall with forming stroke.
stroke 80 mm)
As mentioned above, by using appropriate COF in A
simulations, the difference in punch forces between
FE predictions and experimental results can be
minimized. Based on punch load curves (Figure 11
and Figure 12) recorded during the deep drawing
tests, the COF were predicted at ram speed 10 mm/s
(approximately average value for 1 SPM), 50 mm/s
and 310 mm/s, as shown in Figure 14. In the FE F
models, the constant drawing speeds during the Figure 15: Temperature distribution in the drawn part
whole stroke were inputted instead of actual data at forming speed 310 mm/s, stroke = 60 mm
from the tests. The predicted coefficient of friction
was found to decrease with increasing ram speed. 80 E
70

Temperature (ºC)
Although the static COF was used in the FE analysis
of deep drawing process, it could also be concluded 60 310 mm/s
that the friction conditions may change with forming 50
speed for dry film lubricant. 50 mm/s
40 A B 10 mm/s
0.14 C D
30 F
Coefficient of Friction

20
0.12
10
0.1 0 200 400 600 800
BHF: 125 kN Curvilinear length (mm)
Stroke: 60.8 mm Figure 16: Predicted temperature distributions in the
0.08
Lubricant: Dry film drawn part alone the selected section A-F at different
0.06 forming speeds, stroke = 60 mm
10 mm/s 50 mm/s 310 mm/s
5 CONCLUSIONS
Figure 14: Predicted coefficient of friction at different Deep drawing tests were utilized to investigate the
forming speeds effects of forming speed on drawability of aluminum
4.3 Prediction of temperature distribution in the alloy. The main conclusions are summarized as
drawn parts follows:
Considering the heat generation during deep drawing 1) The tests were conducted in the servo press
process at room temperature, the proposed thermal- which can provide controllable slide velocity. The
mechanical FE model was able to predict the non-isothermal FE model was developed by
temperature increase in the drawn part. The predicted using commercial software PAMSTAMP to help
temperature distribution in the part is as shown in to determine test parameters and to analyze the
Figure 15. The higher temperature is found round the drawing process. The model was verified by
die-shoulder corners. Based on the predicted COF, comparing the predicted and measured flange
friction work is found to change with forming speed, lengths as well as punch loads.
which affects the temperatures at the interface 2) It was found that higher forming speed could help
between tool and sheet. Also lower forming speed to obtain a deeper draw depth when using dry
provide more time for the heat transfer from the film lubricant. The ductile fracture process
drawn part to the cold die. The section A-F (Figure including localized necking could be observed at
15) was selected to evaluate the effect of forming the part corners during the deep drawing. The
speed on the temperature distributions. As illustrated comparisons between experimental load stroke
in Figure 16, the maximum temperatures at location E curves demonstrated that the punch load was
are 77.3 ºC at 310 mm/s, 52 ºC at 50 mm/s and 34.8 reduced by using a higher forming speed. The
ºC at 10mm/s. According to the observations from the friction coefficient, determined by matching the
drawing tests, the deviation in temperatures on the punch forces between predicted and
drawn part was reasonable and not severe to change experimental results, decreased with increasing
the performance of the lubricant during the tests. forming speed. It could be concluded that the
variable frictional behavior during deformation
Regarding the temperature rises in the tools, that the
highest temperatures 27 ºC and 32 ºC were predicted was the main reason for the forming speed
effects.
in the punch and die, respectively, at forming speed
3) Regarding the heat generation during
310 mm/s. The peak temperature in the die occurred
deformation, the peak temperature 77.24 ºC in
at the corner regions. However, it was found that the
the drawn parts was predicted under forming
speed 310 mm/s and it was found that there
 were only slight temperature rises in the tools. [10] Altan, T., Groseclose, A., 2009,Servo-drive
The results of heat generation were evaluated to presses recent developments.
have no significant effects on the drawing Umformtechnisches Kolloqium Darmstadt 10.
process for only single stroke operation. [11] Osakada, K., Mori K.,, Altan T.,, Groche, P.,
2011, Mechanical servo press technology for
6 ACKNOWLEDGEMENTS
metal forming. CIRP Annals-Manuf. Technol.
The Authors would like to extend special thanks to 60.2: 651-672.
the member companies of the Center of Precision [12] Taoka, H., Mawari, H., Higashi, H., Ikehara, H.,
Forming (CPF) that funded this study. Special thanks Hashimoto, M., Kawano, Y., 2009, Development
are due to Shiloh Industries, Inc. (Cliff Hoschouer) of the world's fastest servo press line for
and Honda Engineering (Dennis O’Connor) for manufacturing automotive body panels. Mater.
supporting this project. Process. Technol. 50.12: 33-38.
7 REFERENCES [13] Mori, K., Maeno T.,, Maki, S., 2007, Mechanism
of improvement of formability in pulsating
[1] Emmens, W.C., 1997, Some frictional aspects of
hydroforming of tubes. Int. J. Mach. Tools
aluminium in deep drawing. International
Manuf. 47.6: 978-984.
congress on Tribology of Manufacturing
Processes. [14] Nasser, A., Yadav, A., Pathak, P., Altan, T.,
2010, Determination of the flow stress of five
[2] Hayashi, H., Nishimura, H., 2009, The
AHSS sheet materials (DP 600, DP 780, DP
Application of Servo Press Machine to Forming
780-CR, DP 780-HY and TRIP 780) using the
of Sheet Metals with Low Formability. The
uniaxial tensile and the biaxial Viscous Pressure
Annals of “Dunarea De Jos” University of Galati
Bulge (VPB) tests. J. Mater. Process. Technol.
Fascicle V, Technologies in Machine Building,
210.3: 429-436.
ISSN1221-4566.
[15] Mao, T., Altan, T., 2013, FE simulation for the
[3] Li, D.M., Ghosh, A.K., 2004, Biaxial warm
design of a test tooling for forming Al 5182-O in
forming behavior of aluminum sheet alloys. J.
a servo drive press-Part II. The Ohio State
Mater. Process. Technol. 145.3: 281-293.
University, CPF-2.2/13/05
[4] N. Abedrabbo, F. Pourboghrat, J. Carsley,
[16] Naka, T., Yoshida, F., 1999, Deep drawability of
Forming of AA5182-O and AA5754-O at
type 5083 aluminium–magnesium alloy sheet
elevated temperatures using coupled thermo-
under various conditions of temperature and
mechanical finite element models. Int. J. Plast.
forming speed. J. Mater. Process. Technol. 89:
23.5 (2007): 841-875.
19-23.
doi:10.1016/j.ijplas.2006.10.005
[17] Lademo, O.G., Engler, O., Benallal, A., 2012,
[5] Groche, P., Möller, N., 2012, Tribological
Effect of strain rate and dynamic strain ageing
Investigation of Deep-Drawing Processes Using
on work-hardening for aluminium alloy AA5182-
Servo Presses. ASME 2012 International
O. Int. J. Mater. Res. 103(8):1035-1041.
Manufacturing Science and Engineering
Conference, American Society of Mechanical [18] Serhat, K., Spampinato, G., Altan, T., 2008, An
Engineers. experimental study on nonisothermal deep
drawing process using aluminum and
[6] Kim, H., Sung, J.H., Sivakumar, R., Altan, T.,
magnesium alloys. J. Manuf. Sci. Eng. 130.6:
2007, Evaluation of stamping lubricants using
061001.
the deep drawing test. Int. J. Mach. Tools Manuf.
47(14), 2120-2132.
[7] Tso, P.L., Liu, H.K., Li C.H., 2007, Experimental
Study of Drawing Process with Servo Press.
ASME 2007 International Mechanical
Engineering Congress and Exposition. American
Society of Mechanical Engineers, pp. 17-21.
[8] Kim, H., Altan, T., Yan, Q., 2009, Evaluation of
stamping lubricants in forming advanced high
strength steels (AHSS) using deep drawing and
ironing tests. J. Mater. Process. Technol. 209.8:
4122-4133.
[9] Groche, P., Nitzsche, G., Elsen, A., 2008,
Adhesive wear in deep drawing of aluminum
sheets. CIRP Annals–Manuf. Technol. 57:295–
298.