flash gap optimization in precision blade forging · design the flash gap design in closed-die...
TRANSCRIPT
Flash Gap Optimization in Precision Blade
Forging
S. Javid Mirahmadi MAPNA Group, R&D Department, Tehran, Iran
Email: [email protected]
Mohsen Hamedi Faculty of Mechanical Engineering, University of Tehran, Tehran, Iran
Email: [email protected]
Abstract—The main process for manufacturing of
compressor blade made of titanium alloys is forging. In the
precision forging, airfoil dimensions are made to the net
shape and formation of a flash around the forging ensures
the complete die filling. Some factors such as flash gap
dimensions affect the accuracy of the forged blade. The
elastic die deflection during the forging results in airfoil
thickness error. In this paper, effects of flash gap
dimensions and forging parameters on elastic die deflection
are studied by an experimentally verified finite element
method (FEM) model. These effects are analyzed by the
response surface method (RSM). The results show that by
selection of appropriate flash gap dimensions as well as the
process temperature and die speed, acceptable forged blade
dimensions can be achieved.
Index Terms—blade forging, elastic die deflection, flash gap
dimensions, precision forging
I. INTRODUCTION
Closed-die forging with flash is a metal forming
process in which usually two halves of dies move toward
each other to deform a billet. At the end of a successful
process, the billet fills the die cavity to make the desired
shape. During the process, the excess material is
squeezed out of the die cavity through a restrictive
narrow gap-named flash gap-that forms a flash around the
forging. The most important function of the flash is to
force the flowing material to completely fill the die cavity.
Although inappropriate design of flash gap may result
in a complete die filling, it will lead to excessive die wear
and a considerable elastic die deflection. In the
conventional close die forging, the elastic die deflection
usually is not a problem; but in the precision forging it
should be considered.
Lee et al. analyzed the forging load, die cavity filling
and strain distribution in the precision forging process
with and without flash by using an upper bound elemental
technique. They reported a good agreement between their
model and experimental results [1]. Ranatunga et al.
proposed an upper bound elemental technique (UBET) to
Manuscript received January 10, 2017; revised April 25, 2017.
design the flash gap design in closed-die forging of
axisymmetric shapes. They reported a good agreement
between finite element method simulations and UBET
results [2]. Samolyk and Pater used the slip-line field
elemental technique (SLFET) method in the closed-die
forging flash design process [3]. Tomov et al. simulated
the closed-die forging process by considering the flash
gap dimensions. They investigated the various rules of
flash gap design and reported a considerable difference
between the results of various flash design rules [4], [5].
Samolyk and Pater used SLFET for flash gap design with
V-shaped notches in the flash land [6]. Fereshteh-Saniee
and Hosseini studied the flash dimensions and bar size
effects on the forging load and metal flow experimentally
[7]. Shahriari et al. optimized the flash gap dimensions in
the superalloy blade forging by finite element method [8].
Zhang et al. investigated a novel flash structure with
resistance wall and its effects on the forging load and die
wear. They concluded great influence of the proper
resistance wall on the forging load and die wear [9].
Recently Langer et al. studied a movable slash gap in hot
forging [10]. Up to now, based on the experimental observation,
several rules are developed to design the flash gap. Some
of them are presented as mathematical equations and the
others are as tables or graphs. Sleeckx and Kruth
reviewed all of the rules and their origins [11]. Most of
the rules are developed for the steel forgings and use the
weight of the forging in their equations. Since some
materials like titanium alloys are approximately half the
density of steel, application of available rules for flash
gap design is inappropriate.
This paper is focused on the flash gap design in
isothermal precision forging of a blade made of a
titanium alloy. As concluded by Tomov et al. [4], the
flash design rules result in a very different flash gap
dimensions that is difficult to choose one of them; so in
this paper the effects of flash gap dimensions, the percent
of the excessive material, process temperature and
forging speed on the forging load and elastic die
deflection are investigated by the finite element method
(FEM) and analyzed by the response surface method
(RSM).
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 200doi: 10.18178/ijmerr.6.3.200-205
II. BLADE FORGING
Due to the mechanical and metallurgical properties of
the material under deformation, forging of a compressor
or turbine blade is a complex metal forming process [8].
In the precision blade forging, the desired dimensional
and geometrical tolerances increase complexity of the
process. The blade forging process starts with locating a
preform with a circular or oval cross-section on the lower
die (Fig. 1). After finishing the process, the airfoil forms
inside the cavity with flash around it according to Fig. 2.
The flash has two main dimensions: thickness and width.
By appropriate determination of the flash gap dimensions,
as well as excessive material to fill the flash gap and
compensate the enlargement of die cavity due to the
elastic die deflection, complete cavity filling and die life
will be achieved.
Figure 1. A perform with a circular cross section located on the lower die.
Figure 2. A forged blade cross-section with flash.
III. MODELING AND SIMULATION
A. FEM Model Development
In order to simulate the isothermal forging process and
study the effect of flash gap dimensions as well as the
process temperature and deformation rate on the elastic
deflection, a coupled thermomechanical modeling based
on the rigid-viscoplastic finite element method was
developed. Deform-2D, a commercially available FEM
package was utilized to perform the simulations.
According to the airfoil geometry, the airfoil forging can
be considered as a plane-strain deformation. The flow
stress of the material was determined as a function of
strain rate, temperature and strain by isothermal hot
compression tests. The flow stress data were corrected for
the effect of the friction and adiabatic temperature rise
and were implemented in the simulations. The friction
factor in the billet-die interface was determined as a
function of temperature and die velocity [12] and used in
the modeling. The geometry of the billet and dies were
modeled in a CAD package and imported to the FEM
software via the standard format IGES. The dies and the
billet were considered to be elastic and rigid-viscoplastic,
respectively. The developed FEM model is shown in Fig.
3.
Figure 3. The developed FEM model.
B. Parameters Selection to Study Their Effects
In order to study the effects of gap dimensions as well
as the process parameters on the accuracy of isothermal
forged blades, the response surface method (RSM) was
utilized to model and analyze of elastic die deflection as a
function of flash gap dimensions. The selected
parameters and their level are presented in Table I.
TABLE I. THE STUDIED FACTORS WITH THEIR LEVELS
Factor Low level High level ‒α +α
A: Flash thickness (mm) 0.5 1.0 0.16 1.34
B: Flash width to thickness ratio 2.0 4.0 0.62 5.38
C: Excessive material (%) 8 15 3.18 19.82
D: Temperature (°C) 875 920 844 950
E: Die speed (mm/s) 1 1.8 0.45 2.35
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 201
IV. FEM MODEL VERIFICATION
In order to verify the developed FEM model, a set of
experiments was done. Two halves of dies were designed
and manufactured by a Ni-base superalloy to withstand
the stresses as well as the high temperature of the tests.
After isothermal forging and trimming the blades, they
were subjected to coordinate measuring according to Fig.
4. The blades were located according to 3-2-1 rule at the
root section. The increasing of the blade thickness
compared to its nominal thickness was considered as
elastic die deflection. The measured dimensions were in a
good agreement with the simulation results. The
isothermal forged blade is shown in Fig. 5.
Figure 4. The locating and measuring points of the blade.
Figure 5. The isothermal forged blade.
V. RESULTS AND DISCUSSION
As stated before in section 1, flash gap dimensions
have a considerable effect on the elastic die deflection
and as a result airfoil thickness error. Other process
parameters, process temperature and forging speed,
which have effects on the elastic die deflection can be
controlled accurately in the isothermal blade forging.
The lower die elastic deflection for flash thickness and
flash width to thickness ratio equal to 0.75 and 3
respectively is shown in Fig. 6. The excessive volume
fraction of material is 3.18% and the maximum elastic die
deflection is approximately equal to 0.068 mm at
temperature 897.5 °C and die speed 1.4 mm/s. The die
filling status in Fig. 7 shows that because of insufficient
excessive material, die cavity is not filled completely. By
decreasing the process temperature to 844 °C, although
the lower die elastic deflection increased to 0.146 mm,
but increasing the excessive material volume fraction to
11.5% resulted in completely die filling (Fig. 8). In both
Fig. 6 and Fig. 8, the lower die geometry before the
forging start is shown by a red profile and the deflected
die geometry is scaled by 50.
Figure 6. Lower die elastic deflection (mm) for flash thickness=0.75
mm, flash width to thickness ratio=3, excessive material=3.18%, temperature=897.5°C, die speed=1.4 mm/s.
In order to evaluate the effects of the factors presented
in Table I and their interaction effects on the elastic die
deflection, analysis of variance (ANOVA) was performed
and the results are presented in Table II. The results show
that the flash thickness, flash width to thickness ratio,
excessive material volume fraction, process temperature
and forging speed have a considerable effect on the
elastic die deflection and as a result airfoil thickness error.
According to P-values, the process temperature and flash
thickness have the most significant effect on the elastic
die deflection. The fitted model to the elastic die
deflection by the coded values is presented in equation
(1). In Fig. 9 and Fig. 10 the model is plotted for 8% and
15% excessive material volume fraction, respectively, for
temperatures 875°C and 920°C and speeds 10 and 63
mm/min.
3 3
3 3 3 2
Deflection 0.23 0.021 5.941 10 5.623 10
0.027 9.920 10 5.166 10 8.314 10
A B C
D E BC C
(1)
TABLE II. ANOVA TABLE FOR ELASTIC DIE DEFLECTION.
Source Squares×10-4 df Square×10-4 F-Value P-Value
Model 619 7 88.5 54.63 > 0.0001
A 192 1 192 118.47 > 0.0001
B 15.3 1 15.3 9.44 0.0041
C 13.7 1 13.7 8.46 0.0063
D 311 1 311 192.12 > 0.0001
E 42.6 1 42.6 26.32 > 0.0001
BC 8.54 1 8.54 5.27 0.0278
C2 36.2 1 36.2 22.36 > 0.0001
Residual 56.7 35 1.62
Cor Total 676 42
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 202
Figure 7. Not completely filled die cavity at the end of the die stroke.
Figure 8. Lower die elastic deflection (mm) for flash thickness=0.75 mm, flash width to thickness ratio=3, excessive material=11.5%,
temperature=844°C, die speed=1.4 mm/s.
In the first comparison of Fig. 9 and Fig. 10 it is
concluded that for 8% excessive material, the flash width
to thickness ratio has no considerable effect on the elastic
die deflection; however for 15% excessive material, the
flash width to thickness ratio effect is significant. For 8%
excessive material the flash land is not fully covered by
the flash; so, increasing the flash width to thickness ratio
has no any effect on the elastic die deflection.
The plots in Fig. 9 and Fig. 10 show that more flash
thickness and less flash width to thickness ratio results in
less forging force and elastic die deflection. So the extent
of elastic die deflection can be controlled by increasing
the flash thickness. However, the trimming process is a
limiting criterion. If the flash thickness exceeds the
appropriate dimension, the elastic die deflection may be
acceptable but the trimming process may results in some
other geometrical and dimensional errors. If the flash
width is not chosen wide enough, it will not have
adequate strength and its wear will results in increasing
the flash thickness. So, according to the desired
dimensional tolerances, selection of appropriate flash
thickness and flash width to thickness ratio, insures
complete die filling as well as minimizing the process
force.
Figure 9. Elastic die deflection for 8% excessive material, A) Temperature=875°C, speed=60 mm/min, B) Temperature=920°C, speed=60 mm/min, C) Temperature=875°C, speed=100 mm/min and D) Temperature=920°C, speed=100 mm/min.
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 203
Figure 10. Elastic die deflection for 15% excessive material, A) Temperature=875°C, speed=60 mm/min, B) Temperature=920°C, speed=60 mm/min, C) Temperature=875°C, speed=100 mm/min and D) Temperature=920°C, speed=100 mm/min
Assessment of the plots reveals that even though
increasing the temperature reduces the die material elastic
modulus, but the forging material flow stress is more
affected by the temperature; so, increasing the
temperature reduces the elastic die deflection
significantly.
More forging speed reduces the friction factor,
however it increases the elastic die deflection; but its
effect is not stronger than temperature. This effect is
because of increasing the material flow stress with raising
the deformation rate.
Assessment of the plots show that by selection of
appropriate flash gap dimensions as well as the process
temperature and forging speed, the desired dimensional
tolerance can be achieved.
VI. CONCLUSION
In this paper the effects of the flash gap dimensions as
well as the process parameters are studied by a verified
FEM model and analyzed by response surface method.
The results show that flash thickness and flash width to
thickness ratio have a considerable effect on the elastic
die deflection and consequently thickness error. The
effect of the flash width to thickness ratio depends on the
amount of the excessive material. Between the process
parameters, the process temperature has a great effect on
the elastic die deflection and thickness error. By
appropriate selection of the flash gap dimensions as well
as the process parameters, the desired dimensional
tolerances can be achieved.
ACKNOWLEDGMENT
This study was supported by MAPNA Group and
performed at Mavadkaran Engineering Company under
the supervision of Mr. Mohammad Cheraghzadeh which
is appreciated.
REFERENCES
[1] J. H. Lee, Y. H. Kim, and W. B. Bae, “A study on flash-and flashless-precision forging by the upper-bound elemental
technique,” Journal of Materials Processing Technology, vol. 72,
no. 3, pp. 371-379, 1997. [2] V. Ranatunga, J. S. Gunasekera, W. G. Frazier, and K. D. Hur,
“Use of UBET for design of flash gap in closed-die forging," Journal of Materials Processing Technology, vol. 111, no. 1, pp.
107-112, 2001.
[3] G. Samolyk and Z. Pater, “Application of the slip-line field method to the analysis of die cavity filling,” Journal of Materials
Processing Technology, vol. 153, pp. 729-735, 2004. [4] B. Tomov, V. Gagov, and R. Radev, “Numerical simulations of
hot die forging processes using finite element method,” Journal of
Materials Processing Technology, vol. 153, pp. 352-358, 2004. [5] B. Tomov, R. Radev, and V. Gagov, “Influence of flash design
upon process parameters of hot die forging,” Journal of Materials Processing Technology, vol. 157, pp. 620-623, 2004.
[6] G. Samołykn and Z. Pater, “Use of SLFET for design of flash gap
with V-notched lands in a closed-die forging,” Journal of
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 204
Materials Processing Technology, vol. 162–163, pp. 558-563, 2005.
[7] F. Fereshteh-Saniee and A. Hosseini, “The effects of flash
allowance and bar size on forming load and metal flow in closed die forging,” Journal of Materials Processing Technology, vol.
177, no. 1, pp. 261-265, 2006. [8] D. Shahriari, M. Sadeghi, M. cheraghzadeh, and M. Taghipour.
“Flash design optimization in blade forging using fem method,” in
Proc. AIP Conference Proceedings, vol. 908, no. 1, pp. 1035-1040, 2007.
[9] Y. Zhang, Z. An, and J. Zhou, “Optimization design of flash structure for forging die based on kriging-pso strategy,” Advances
in Swarm Intelligence, vol. 6145, pp. 373-381, 2010.
[10] J. Langner, M. Stonis, and B. A. Behrens, “Investigation of a moveable flash gap in hot forging,” Journal of Materials
Processing Technology, vol. 231, pp. 199-208, 2016. [11] E. Sleeckx and J. P. Kruth, “Review of flash design rules for
closed-die forgings,” Journal of Materials Processing Technology,
vol. 31, no. 1, pp. 119-134, 1992. [12] S. J. Mirahmadi, M. Hamedi, and M. Cheraghzadeh,
“Investigating friction factor in forging of Ti-6Al-4V through isothermal ring compression test,” Tribology Transactions, vol. 58,
no. 5, pp. 778-785, 2015.
S. Javid Mirahmadi was born in Isfahan, a
historical city in Iran. He received his B.Sc. from University of Science and Technology
(2007), M.Sc. from KN Toosi University of
Technology (2009) and Ph.D. from University of Tehran (2015) all in manufacturing
engineering. Dr. Mirahmadi was with Mavadkaran
Engineering Company, a subsidiary of
MAPNA Group, for 5 years as a researcher in the field of metal forming processes. Then he
joined MAPNA Group R&D, active in developing advanced manufacturing technologies. He has published several papers in the field
of titanium and superalloy forming.
Mohsen Hamedi is professor of
manufacturing in School of Mechanical Engineering, University of Tehran, Iran where
he teaches graduate and undergraduate
courses. He obtained his Ph.D. (1995) from UNB, Canada and M.Sc. and B.Sc. from
University of Tehran respectively. Dr. Hamedi has supervised more than 60
Ph.D. and M.Sc. dissertations. He has
published more than 120 papers in the refereed international journals and conferences. His research area of
interests are micro-sensor and micro-actuator design and fabrication, mechanical energy harvesting and manufacturing process optimization.
International Journal of Mechanical Engineering and Robotics Research Vol. 6, No. 3, May 2017
© 2017 Int. J. Mech. Eng. Rob. Res. 205