Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 81 (2014) 2415 – 2420

11th International Conference on Technology of Plasticity, ICTP 2014, 19-24 October 2014,
Nagoya Congress Center, Nagoya, Japan

Effect of forming speed in precision forging process evaluated

using CAE technology and high performance servo-press machine
Soo-Young Kim*, Kaoru Tsuruoka, Tadashi Yamamoto
Yamanaka Eng Co. Ltd., 2-11-2 Ohsaku, Sakura Chiba 285-0802, Japan

Abstract

This study investigates the process design method for precision forging combined with complex and flexible motions realized
using servo-press machine with the aid of computer-aided engineering technology and design of experiment study. The
evaluation results for the simplified unsetting process of a CVT pulley and precision spur gear forging process indicate that the
design method proposed in this study could provide useful information on the effects of press motion for important process
outputs, such as the forming load and deformation behavior of the workpiece.

©© 2014
2014TheTheAuthors.
Authors.Published
Publishedby
byElsevier
ElsevierLtd.
Ltd.This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of Nagoya University and Toyohashi University of Technology.
Selection and peer-review under responsibility of the Department of Materials Science and Engineering, Nagoya University
Keywords: Precision forging; Servo-press machine; Multi-axis press motion; CAE technology; DOE study

1. Introduction

Ever since the invention of the servo-press machine in 1970s, it has been successfully applied in the sheet
forming and fine blanking industries. Recently, there has been a growing demand for developing cost-effective and
high value added forging processes using the servo-press for the forging industry across the world. Consequently,
the development of high performance servo-press is increasingly pursued by major press manufacturing companies,
as indicated by Osakada (2010), Nakano et al. (2010), Kawamoto et al. (2012), and Osakada et al. (2011). In

* Corresponding author. Tel.: +81-43-498-3446; fax: +81-43-481-5100.

E-mail address: sykim@yamanaka-eng.co.jp

1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Department of Materials Science and Engineering, Nagoya University
doi:10.1016/j.proeng.2014.10.343
2416 Soo-Young Kim et al. / Procedia Engineering 81 (2014) 2415 – 2420

particular, the ability to realize complex and flexible press motion is one of the most important aspects desired in
servo-press machine.
In the design of the precision forging process, computer-aided engineering (CAE) is a key technology that is
being adopted for improving the design efficiency and quality of the forged product. Especially while using the
servo-press machine for the forging process, the application of numerical simulation is essential to achieve reliable
process design, as it can provide useful information to overcome the difficulties associated with the increase in
flexibility and complexity of servo-press motion control. The effect of CAE technology can be further improved by
combining it with the design of experiment (DOE) study, considering the increasing number of design parameters
due to flexible motion definition in the forging process design using servo-press machine.
In the present study, we have investigated the possibility of process improvement by combining the CAE
technology, DOE study, and flexible servo-press motion for two forging examples, namely, the simplified upsetting
process of a CVT pulley and the divided flow forging process of a precision spur gear. The commercial forming
simulation software DEFORMTM was used for the numerical simulation and the DOE study. The variations of the
forming load and deformation behavior with respect to the design variables related with the press motion definition
were investigated to determine the adequate conditions for each process.

Nomenclature

SH Range of stroke for high forming velocity (200 mm/s) in pulley upsetting process
VL Forming velocity during the stroke after S H in the pulley upsetting process
S1 Range of stroke for 1st motion in spur gear forging process
V2 Forming velocity during 2nd motion in spur gear forging process

2. Development of servo-press for precision forging process

The high performance servo-press machine FPS-1200, co-developed by Sumitomo Heavy Industries and
Yamanaka Eng Co., was used in this study. The exterior of FPS-1200 is shown in Fig. 1 and the detailed
specifications are listed in Table 1. The FPS-1200 system, used with a 500 kW servo motor, can preserve a
maximum forming load of 1,200 ton and a maximum forming speed of 135.2 mm/sec during a 5.5 mm stroke from
the bottom dead center. In addition to the main slide driven by the servo motor, FPS-1200 has three hydraulic servo
driven cushions, namely, one upper die cushion and two lower die cushions. This facilitates a four axis motion
control, with flexible motion definition for each axis. The capability of handling such complex and flexible motion
definition is one of the unique and strong features of the servo-press, distinguishing it from the conventional press
machines.
Table 1. Detailed specifications of FPS-1200.
Main slide (Dynamic accuracy: JIS 1st class)
Capacity (kN) 12,000
Stroke (mm) 300
Speed (mm/s) Max. 135.2
Die cushions (Hydraulic servo driven)
Capacity (kN) 1,500
Upper
Stroke (mm) 70
Capacity (kN) 3,500
Lower main
Stroke (mm) 70
Capacity (kN) 1,000
Lower sub
Stroke (mm) 220
Main Slide [Upper side]
Upper die cushion
Main die cushion Sub die cushion
[Lower side]
Fig. 1. Exterior of FPS-1200. Fig. 2. Numerical procedure adapted in this study.
 Soo-Young Kim et al. / Procedia Engineering 81 (2014) 2415 – 2420 2417

3. Evaluation of press motion using the CAE and DOE study

Fig. 2 illustrates the protocol adopted to evaluate the effect of forming speed on deformation behavior and the
forming load in the cold forging process. Considering the heat generation due to plastic deformation and friction,
the non-isothermal rigid-viscoplastic finite element simulations were performed for various motion conditions. The
flow stress data of the workpiece was defined as a function of temperature, effective strain, and effective strain rate.
The DOE study was adopted with the numerical simulations for sampling the motion conditions and for evaluation
of the effect of motion parameters on the simulation results.
The DOE study has been successfully applied for designing of forging process and also for the quality control of
the forged product. According to the study reported by Fujikawa et al. (2000) and Yang et al. (2011), the usefulness
of the DOE study can be further improved by combining it with the CAE technology in the process design
procedure. A typical example of such a design improvement realized by synergy of the CAE and DOE study is
shown in Fig. 3. The main problem in this sheet forging process is the existence of a folding defect and under-fill
regions in the final product. In order to circumvent this issue, spring constant and corner radius of the upper die
cushion, spring constant of the mandrel, and friction coefficient were selected as the design variables. Absence of
folding and under-fill in boxed region A were set as the constraint conditions. In addition, minimizing the effective
strain in boxed region B was defined as the objective function of this problem. On the basis of the results obtained
from a total of 122 simulations, the design variables were determined to get the optimal deformation behavior for
this problem.
In the present study, the commercial forming simulation software DEFORMTM was used for the numerical
procedure including the DOE study, to investigate the effect of forming speed on the material behavior of
workpiece during cold forging process.
Sleeve Obj. Response Surface Prediction of defect in initial condition
Region A
v Design Variables 1200 Folding Folding
Region B
Min. Max.
600
FU
50 500
(kN/mm) Result of optimal condition
FU 0
Upper FL
5 40 500 FU FL R P
R Cushion (kN/mm)
40 (kN/mm) (kN/mm) (mm)
R 275
1 5 FU 378.5 20.75 3.84 0.052
FL (mm) 22.5
(kN / mm) FL
Friction coeff. P Lower P 0.05 0.1 No Folding and under-fill
50 5 (kN / mm)
Mandrel Cushion

Fig. 3. Improvement of process design using the synergy of CAE and DOE study in during the sheet forging process.

4. Examples

4.1. Forming load reduction in pulley upsetting process

The motivation of this example is to reduce the forming load in the upsetting process of the CVT pulley product.
The final upsetting stage of the CVT forging process usually requires a higher forming load when compared to the
preceding forging stages. To reduce the forming load, this process is usually performed under hot or warm forging
conditions, which often leads to additional processing time and increased energy costs associated with the heating
and cooling processes.
In this study, we have investigated a simple upsetting process, similar to the final upsetting process of the CVT
pulley, for the press motion pattern shown in Fig. 4. Just before contacting the upper side of the billet, the speed of
the upper die was reduced to 5 mm/s in order to prevent the tool damage caused by the strong impact between the
tool and workpiece. After the initial contact, the tool speed was suddenly raised to 200 mm/s in order to reduce the
cycle time and to increase the workpiece temperature due to the heat generated from the plastic deformation
process. The die speed becomes slower around the final stroke, so as to utilize the effect of flow stress reduction
associated with the lower strain rate under the high temperature condition of the workpiece. Fig. 5 shows the
2418 Soo-Young Kim et al. / Procedia Engineering 81 (2014) 2415 – 2420

simulation model, namely, the two dimensional axi-symmetric non-isothermal rigid-viscoplastic FE model, used in
this study. The flow stress data was selected from the in-house reference data, so as to consider the effects of
temperature and effective strain rate as shown in Fig. 6. In this example, the range of applied stroke S H , forming
speed of 200 mm/s, and the reduced forming speed V L in the final deformation stage, were selected as the design
parameters for the DOE study, as listed in Table 2. The initial workpiece temperature was assumed as the room
temperature.
800 800
[B.C.] 700 700

Effective stress (MPa)

Deformation

Effective stress (MPa)

mf = 0.1 600 600
Start 5 mm/s
Position of slide (mm)

hc = 5 N/s/mm/K
Workpiece 500 500
Eff. strain rate
Material 400 T(0C) 400
99.3 20 0.1
SCM420 300 100 300 1
1 mm 200 10
200 mm/s 200 200 50
SH 50 300
100 400 100 100
ij 500 1000
VL mf = shear friction factor 0 0
hc = heat transfer coeff. 0 0.5 1 1.5 0 0.5 1 1.5
Time (sec) Effective strain Effective strain
Fig. 4. Press motion for pulley upsetting. Fig. 5. FE model used for simulations. Fig. 6. Flow stress data used for simulations.
700 C1

Forming load (ton)

Table 2. Design variables used in the DOE study for pulley
upsetting. 692

Variables Min. Max. N Conditions 685

SH (mm) 40 48 5 40, 42, 44, 46, 48
677
VL (mm/s) 1 20 5 1, 5.75, 10.5, 15.25, 20
670
Table 3. Experimental conditions used for pulley upsetting. 40 C2
44
Variables Min. Max. N Conditions SH (mm) 48 C3
SH (mm) 42 48 3 42, 45, 48
Fig. 7. Response surface plot of forming load with respect to the press
VL (mm/s) 2 14 4 2, 6, 10, 14
motion conditions.

The response surface of the forming load obtained from the results of the DOE study is depicted in Fig. 7. As
shown in this figure, the effect of S H on the forming load became stronger with a decrease in the final forming
velocity V L . The variation in the forming load due to the variation of S H could be ignored for V L exceeding 10
mm/s. This could be attributed to the temperature drop caused by an increase in contact time at the final forming
stage. The forming load reduced steadily for V L ranging from 1 mm/s to 5 mm/s, and became stable for V L
exceeding 10 mm/s. The forming load was minimum for S H of 48 mm and V L of 5.75 mm/s (condition C3). The
temperature and effective stress distributions in the workpiece at three selected conditions are shown in Fig. 8. The
temperature distribution for the condition C2 is the highest, given the fact that the time span of contact between the
workpiece and die is shortened when compared to the other conditions. However, the effective stress distribution
becomes lowest for the condition C3 due to the synergistic effects of increasing temperature and decreasing
effective strain rate around the final stroke.
On the basis of the information obtained from the DOE study, we performed the forging experiment of the
pulley upsetting process. The detailed experimental conditions are summarized in Table 3. The range of S H and V L
was adjusted considering the result of the DOE study, so as to reduce the total number of experiments. The
photographs of the forged product and the surface plot of the forming load variation with respect to the
experimental conditions are depicted in Fig. 9 and Fig. 10, respectively. As shown in Fig. 10, the tendency of the
forming load variation with respect to the design variables was similar to that predicted by the DOE study,
although the absolute values of measured forming load were about 20% - 30% higher than the simulation result.
The main reason for the observed difference is the inaccuracy of the in-house flow stress data used in the
simulations. More reliable material data obtained from carefully treated material test should be applied to improve
the simulation accuracy up to the acceptable level for quantitative evaluation.
 Soo-Young Kim et al. / Procedia Engineering 81 (2014) 2415 – 2420 2419

Temperature 910-915
300 0 50 mm 905-910
915
150 900-905
0 910 895-900

Forming load (ton)

890-895
[0C] 905
Effective stress 900
700 0 50 mm
895

500 890
42 2
[MPa] 1414 10 6
SH (mm) 4848
(a) (b) (c) VL (mm/s)
Fig. 8. Temperature and effective stress distributions Fig. 9. Forged product. Fig. 10. Surface plot of the forming load with respect
for conditions (a) C1, (b) C2, and (c) C3 in Fig. 6. to press motion.

4.2. Flow control in spur gear forging process

In the present study, we investigated a precision forging process of the spur gear having 14 teeth, outer diameter
of 48 mm, and height of 25 mm, in order to evaluate the effect of press motion on the teeth filling behavior and
forming load variation. To effectively reduce the forming load by completely filling the tooth region, we applied
the divided flow forging method proposed by Kondo (1999). Fig. 11 shows the FE simulation model and the DOE
study conditions used for the spur gear forging process. Considering the geometrical symmetry, the upper half
region of one teeth part was used in the simulation to reduce the calculation time for each simulation. The same
flow stress data shown in Fig. 6 was applied in the current simulations. During the 1st motion of normal upsetting
stage up to the stroke S 1 , the forming speed of the upper/lower pins and sleeves were set to 20 mm/s. Before
starting the 2nd motion, the upper and lower pins were released to allow free movement to the backward direction.
In the 2nd motion, the forming speed of upper and lower sleeves was set to V 2 until the end of the stroke. The
stroke of 1st motion S 1 and the sleeve velocity V 2 of 2nd motion were selected as the design variables of the DOE
study for evaluating the teeth filling status and forming load. The region M shown in Fig. 10 was defined as the
region of interest for investigating the contact status. A total of 16 simulations were performed, which took about
16 hours for implementing on a machine with Core i7 4.20GHz CPU and 24GB RAM (three node parallelization
and two jobs per run).
The effect of press motion on the teeth filling behavior and forming load variation of spur gear forging process,
as predicted by the DOE study, is shown in Fig. 12. As shown in the contour plot of the forming load according to
the design variables, the variation of the forming load highly depends on the stroke S 1 . More specially, under-fill
region exists for S 1 less than 11.6 mm. Among the complete filling cases, the minimum forming load can be found
around the press motion condition of S 1 = 11.6 mm and V 2 = 5 mm/s (run #10).
[1st Motion]
Sleeve Pin Region M 1st
Motion 2nd Design S1 V2
Motion Variables (mm) (mm/s)
Forming Speed (mm/s)

Max. 11.8 20
20 Min. 11.2 1
Workpiece
SCM420 [2nd Motion] N 4 4
Die
100kTetra. Ele. V2 11.2 1
11.4 5
0 Conditions
11.6 10
0 S1 12.85
Symmetry plane 11.8 20
Stroke (mm)

Fig. 11. FE model and the DOE study conditions used for spur gear forging.
2420 Soo-Young Kim et al. / Procedia Engineering 81 (2014) 2415 – 2420

[Response contour plot of forming load] [Pressure] 2,500

Under fill conditions Run#2
n#2 Run#10
20 2,000
1,500
148ton 1,000
145ton 500
142ton Under 0
V2 (mm/s)

fill [MPa]
10 140ton [Temperature]
mperature]
200
137ton Run#10 Run#2
n#2 Run#10
n#10
160
5 134ton 120
132ton 80
Run#2
1 40
11.2 11.4 11.6 11.8 0
S1 (mm) [0C]

Fig. 12. Effect of press motion on spur gear forging process during teeth filling behavior and forming load variation.

5. Conclusions

This study investigates the possibility of combining the CAE technology and DOE study for determining the
adequate press motion design for the forging process design using servo-press machine. On the basis of the results
of the DOE study for pulley upsetting simulation, it is possible to effectively predict the adequate condition of
press motion necessary for reducing the forming load. Experimental results showed similar tendency of the
forming load variation as that of the simulation result. However, there are still some substantial differences
between the predicted and measured forming loads. For the precision spur gear forging, the motion condition for
complete filling of the tooth region, in order to minimize the forming load, could be predicted using the DOE study
of 3-dimensional numerical simulations under various motion conditions.
On the basis of the results obtained, it can be realized that the process design method introduced in this study could
provide useful information on the forging process using servo-press with flexible motion definition. For more
practical use to satisfy the industrial needs, the following issues should be addressed in the future:
(1) Collecting accurate flow stress data considering the effects of temperature and strain rate.
(2) Improving the accuracy of non-isothermal simulation handling for cold and warm forging processes.
(3) Improving the robustness of the DOE study to handle complex press motion definition.
(4) Improving the simulation efficiency for reducing the total calculation time required for multiple simulations.

References

Osakada, K., 2010. Application of Servo Presses to Metal Forming Processes. steel research int. 81, 9-16.
Nakano, T., 2010. Press Machine Trends and Servo Press Forming Examples. steel research int. 81, 682-685.
Kawamoto, K., Klumb, D., 2012. Future Application of AC Servo Press Focusing on Forging Process. Proceedings of the 45th ICFG Plenary
Meeting, 113-116.
Osakada, K., Mori, K., Altan, T., Groche, P., 2011. Mechanical Servo Press Technology for Metal Forming. CIRP Annals - Manufacturing
Technology 60, 651-672.
Fujikawa, S., 2000. Application of CAE for Hot-forging of Automotive Components. Journal of Materials Processing Technology 98, 176-181.
Yang, J.B., Oh, J.Y., Polisetty, H., Wu, Z., Wu, W.T., 2011. Shape Optimization in 3D Forging Process. Proceedings of the 10th ICTP, 159-164.
Kondo, K., 1999. Improvement of Product Accuracy in Cold Die Forging. Proceedings of the 6th ICTP, 41-48.