- Email: [email protected]
Elastic deflections of the blank holder in deep drawing of sheet metal
Journalof
Materials Processing Technology Journalof MaterialsProcessingTechnology59 (1996)34-40
ELSEVIER
Elastic deflections of the blank holder in deep drawing of sheet metal L e o n i d S h u l k i n a, S t e v e n W . J a n s e n b, M u s t a f a A. A h m e t o g l u a, G a r y L. K i n z e l c, T a y l a n A l t a n a'*
a ERCfor Net Shape Manufacturing, Ohio State University, Columbus, OH 43210, USA b Utilase, Inc., 20530 Hoover Road, Detroit, M148205, USA c Department of Mechanical Engineering, Ohio State University, Columbus, OH 43210, USA
Industrial Summary
In general, elastic deformations of deep drawing tools influence the process negatively; thus, they are undesirable. At the same time, an elastically deforming blank holder that adjusts itself to the changing process conditions can be instrumental in achieving uniform distribution of the blank holder pressure over the flange area. This will result in uniform material flow during deep drawing and ultimately in improved quality of the parts. A series of experiments and FEM simulations have been conducted in the ERC for Net Shape Manufacturing in order to investigate the influence of the elastic deflections of the blank holder on the pressure distribution between the blank holder and the sheet.
1. I n t r o d u c t i o n
In deep drawin~ a punch is used to force a sheet metal blank into the cavity of a die. Thus, the clearance between the die and the punch gives the blank the desired shape. The sheet metal is held in place b y means of a blank holder. The blank holder m a y either transmit a certain a m o u n t of predetermined force to the blank or it may be used to create a gap for sheet metal flow which is slightly larger than the thickness of the blank. There are many variables, such as blank holder pressure (BHP) or force (BHF), lubrication, die surface quality, blank shape, etc., which affect the quality of formed parts. All of these variables need to be carefully considered when designing a part and tooling for deep drawing. In most studies of deep drawing, the blank holder is assumed to be nearly "rigid" and deforms little. Thus, it does not adjust to the thickness changes that occur in the sheet metal during deformation. The blank holder touches the sheet metal where the thickness is a maximum and applies pressure only at these points. A uniform pressure distribution may be obtained b y first determining the locations of the increased thickness and then grinding (spotting) the
* Corresponding author. 0924-0136/96/$15.00© 1996ElsevierScience S.A.All rightsreservod Pl10924-0136 (96)02284-4
blank holder at these areas. BHF m a y be applied to the blank holder b y m e a n s of a die cushion, mechanical springs, or nitrogen or hydraulic cylinders. Many studies have shown that the combined rigidity of the dies and press has a great influence on the accuracy in precision blanking and deep drawing [2]. Elastic deflections of press and tooling are difficult to predict and control. Thus, they affect the quality of the parts and are undesirable. These deflections are caused by eccentric forming and cutting forces and result in u n e q u a l distances between the comers of the blank holder and the die. T. Jimma and F. Sekine analyzed deformations of the dies subjected to various loads and p r o p o s e d a method to select a press suitable for a given die set and a given off-center loading [5]. K. Siegert p r o p o s e d a closed loop control system using hydraulic cylinders and proportional valves for c o m p e n s a t i o n of the tilting and h o r i z o n t a l displacement of the upper die relative to the lower die [6]. An elastically deforming blank holder that adjusts itself to the sheet thickness can provide a uniform distribution of the BHP over the flange. This will result in uniform material flow during deep drawing and ultimately in improved quality of the parts. Also, spotting of the blank holder is not required in
L. Shulkinet al. /Journal of Materials Processing Technology 59 (1996) 34-40
this case. E. Doege and G. Stock had some success using a rubber elastic blank holder for forming flat sheets into rectangular pans. They observed several a d v a n t a g e s such as elimination of wrinkles, reduction of local BHP, and easier die try-outs and process adjustments [3]. The elastic deflections imposed intentionally on a blank holder can be roughly d i v i d e d into two categories: "small" and "large." "Small" elastic deflections are n e e d e d w h e n the blank holder is initially in full contact with the sheet, and a very small deformation of the blank holder is enough to cause significant changes in the BHP distribution. "Large" elastic deflections, on the other hand, are n e e d e d w h e n the blank holder first needs to be brought in contact with the sheet surface as in the case with the multi-gage w e l d e d blanks. The deflections must be large enough to change the blank holder surface geometry in order to compensate for the initial gap(s) due to the different thickness of the blank. This m a y require a large BHF and a good understanding of h o w the blank holder contacts the sheet surface and h o w the BHP is distributed. A series of experiments have been conducted at the E R C / N S M in order to investigate the influence of the elastic deflections of the blank holder on the p r e s s u r e distribution at the blank holder-sheet interface.
2. "Small"ElasticDeflections A p h o t o g r a p h of the tooling u s e d for the experiments is shown in Fig. 1 [1]. The experiments are conducted in a 160 ton single action hydraulic MINSTER/Tranemo DPA-160-10 press. The die is m o u n t e d to the u p p e r die shoe. The six nitrogen cylinders as well as a stationary punch are mounted on the lower die shoe. The press brings the die in contact with the sheet which allows the nitrogen cylinders to apply the BHF. A 2.5 inch thick load cell plate is located on top of the nitrogen cylinders. It is provided with 6 pockets for placing the load cells. A 2 inch thick blank holder plate with an inserted blank holder ring is placed on top of the load cells. The inserted ring is divided in two halves. Each half of the inserted ring can be adjusted in its height by shimming. Thus, the tooling can be used for forming axisymmetric n o n - w e l d e d blanks, w e l d e d blanks with same sheet thicknesses on both sides of the weld, or welded blanks of different thicknesses with the weld line in the middle. The main dimensions of the tooling are given in Fig. 2.
35
WeldLine Position / ~ ~.--Nitrogen Blank~ f ~ , ~ C y l i n d e r Holder / V
Plate ~
kd
\
Fig. 1. Phetograph of the experimental tooling and location of nitrogen cylinders
~ 158.2
SHEET
Fig. 2. Main dimensions of the experimental tooling (in mm) The six nitrogen cylinders are arranged into two individually controlled pressure groups consisting of three cylinders each as shown in Fig. 1. By setting a different nitrogen gas pressure for each pressure
L. Shulktn et al. / Journal of Materials Processing Technology 59 (1996) 34-40
36
group, separately controlled values of the BHF can be obtained on each side of the blank holder. This is possible because the blank holder plate deflects elastically under the force produced by the nitrogen cylinders. Laser-beam welded circular blanks from AKDQ steel are used to form the round cups. A 1.8 m m thick sheet is welded to a 0.8 m m thick sheet along the center line. Each half of the inserted blank holder ring is adjusted in its height to compensate for the 1 m m thickness difference between the thick and thin sides of the blank. Thus, the welded blank placed on top of the blank holder is in full contact with the blank holder ring inserts and the BHP is applied to both thick and thin sides of the blank. Fig. 3 s h o w s h o w d i f f e r e n t forces were transmitted from the six nitrogen cylinders through the 2.5 in. thick load cell holding plate to the six load cells. The force values read by the load cells are almost equal to the values of force set by nitrogen cylinders (2.2 tons on the thin side of a sheet metal blank and 0.5 tons on the thick side). It can be concluded that the blank holder acts not as a rigid plate but rather deforms elastically at the areas where the forces of the nitrogen cylinders are applied. This indicates that it m a y be possible to develop an "elastic" blank holder which can adapt its shape to the profile of a sheet metal blank making the deep drawing process more flexible.
kept at 1 ton while on the thin side it varied as given in Table 1. The lubricant was a heavy duty, lightly pigmented, synthetic d r a w i n g c o m p o u n d . The quality of the cups in these examples improves from a fractured cup with significantly distorted weld line to a somewhat non-symmetric cup and finally to a completely symmetric cup w i t h the weld line remaining centered. As seen in Table 1, the 13 inch blanks could not be drawn without failure. Even with very low BHF tearing could not be avoided on the thin side.
a)
b)
Tailor Welded Blank 0.8 / 1.8 mm, Exp. 58 Thin side - Load Cells 1,2,3 Thick side - Load Cells 4,5,6 20
.
.
.
.
.
.
.
.
.
.
Punch Force
c)
- - - 0 - - - Load Ceil#1
r:O ,-.
15
"-'On
Lead Cell#2
+
Load Cell#3
42_
...... i ........... ~ - - " ' " ' i - ~
-~--Load 0o,.
i
× L0adCe,.S
i
/ /
i
i
i
i
iI . , iiiiiiii
uO . ~
.1:2 " ~
f...
,, .................!........
...........
.
Fig. 4. Photographs of welded cups deformed under different BHF ratios. Process parameters are given in Table 1.
Table 1 Experimental conditions for the examples in Fig. 4
rn
-20
0
20
40
60
80
Punch Displacement [mm]
Example
Blank BHF, ton Diameter, in thick side thin side
Fig. 3. BHF variation: load cell readings a) fractured cup Fig. 4 shows the photographs of welded cups deformed u n d e r different BHF ratios between the thick (1.8 mm) and the thin (0.8 mm) side of the blank. The BHF on the thick side of the blank was
13
1
15
b) non-symmetric cup 11
1
5
c) symmetric cup
1
12
11
37
L Shulkin et al./Journal of Materials Processing Technology 59 (1996) 34-40
3. "Large" Elastic Deflections 3.1. Flexible blank holder experiments In deep drawing of tailor welded blanks with a single (non-split) blank holder plate, the flexible blank holder is expected to deform elastically to compensate for the gap caused by the thin side of a multi-gage w e l d e d blank. The blank holder transmits the force from six nitrogen cylinders by deforming elastically to press on the thin side of the sheet in order to apply the BHP to the flange of the d e e p d r a w n cup. Thickness was the primary attribute of the ring shaped blank holders which were investigated.
3.2. Experimental Tooling and Data Collection The tooling used in these experiments is the same as shown in Fig. 1, except a flat blank holder plate is placed on top of the nitrogen cylinders instead of the load cell holding plate and the blank holder plate with inserted ring (Fig. 5). The selected blank holder plate thicknesses used in the experiments are based on the results of the FEM simulations [4]. The thicknesses are: 1.00 in., 1.25 in., and 1.35 in. Each blank holder is deformed under 12, 14, 16, and 18 tons of force. Both sides of the blank holder are subjected to the same BHF to prevent the blank holder plate from tipping. A circular welded blank with a 1 m m thickness difference is placed between the blank holder and die to create a gap of I mm. In order to simulate the deep d r a w i n g conditions without actually drawing the blank, a hole is cut in the middle of the blank to provide space for the punch (Fig. 5).
Welded Blank' I Thick Side, 1.8 mm
J
~
l DIE : I Welded Blank, Thin Side, .,- .................. 0.8 mrn
I
I Bl"nk"°'d°r] I ~
J
I I Nitrogen L_..J Cylinders
Fig. 5. Tooling for experiments with flexible blank holder The pressure distribution between the holder and the blank was evaluated with a pressure sensitive film placed between the welded blank and the blank holder. This paper (Pressurex®), supplied by Sensor Product Inc., East Hanover, NJ, had already been
used successfully for measuring BHP in forming cups from flat sheet metal blanks. The pressure paper becomes pink when pressure is applied that is within the designed range of the film. The paper used for these experiments has a low sensitivity rating with a pressure range of 350-1400 psi. A single point scanner supplied by Sensor Products is used to read the pressure from the film. The scanner is linked to a computer program which outputs the pressure based on the color density.. The paper consist of two sheets of a polyester film substrate (Fig. 6). The substrate of one sheet of the p a p e r is b o n d e d to a color d e v e l o p i n g layer. The other sheet has tiny microcapsules filled with ink which are b o n d e d to the substrate. When pressure is applied to the two sheets after they have been p u t together, the microcapsules break and release the ink onto the color developing layer. ~
1
Substrate(polyester film) Microcapsule layer Color developing layer Substrate (polyester film)
Fig. 6. Cross section of Pressurex®pressure film
3.3. Pressure on the Thick Side of the Welded Blank During the 18 ton BHF tests, a sheet of pressure sensitive film is placed between the 1.00 and 1.35 in, thick blank holder and the thick side of the welded blank (Fig. 5). Since there is no gap present between the blank holder and the sheet, it was expected that the pressure distribution w o u l d be fairly uniform across the entire surface of the blank holder. A p h o t o c o p y of an actual contact patches from the pressure paper (Fig. 7) shows that the pressure distribution was not at all uniform. The pressure on 1.00 in. thick blank holder surface is highly concentrated around the locations of the nitrogen cylinders and along the weld line where the blank holder is bent to touch the thin side of the welded blank. The pressure on the 1.35 in. thick blank holder is more uniformly distributed but still mostly concentrated around the nitrogen cylinders and the weld line. Fig. 8 illustrates this and compares the pressures measured on both blank holders. The pressure was measured along the half circle with diameter O 381 m m (15") corresponding to the location of the nitrogen cylinders (Fig. 7). Note the dramatic drop in pressure b e t w e e n the cylinders. The pressure paper recorded no contact between the
38
L. Shulkin et al. /Journal of Materials Processing Technology 59 (1996) 34-40
blank holder and sheet in these areas. The results of these tests clearly show that even the thickest blank holder does not behave as a rigid body.
Fig. 9 compares the magnitudes of contact areas measured on the blank holder surface at the thin side of the welded blank for all three thicknesses of the blank holder plates and all four BHF values. The contact area is defined as the area of the colored spot on the sheet of pressure paper. The contact areas are almost directly proportional to the applied BHF. 12
Blank holder thickness
9
[ ] 1.35 in.
[ ] 1.25 in.
6
[ ] 1.00 in.
~ 3
a)
L v
b)
18
Fig. 7. Photocopy of contact patches on the blank holder surface for different thicknesses of the blank holder plate: a) 1.00 in b) 1.35 in (thick side of welded blank)
Pressure (psi)
]Nitrogen Cylinders ]
I
1.00in. . . . . 1.35 in.
1400
OOO
'/
1200
800
16 14 12 Blank Holder Force (tons)
Fig. 9. Comparison of contact area magnitude for each blank holder and BHF at thin side of welded blank
Fig. 10 shows the shape and location of the contact area for each blank holder with 18 ton BHF. The illustration in Fig. 10 is intended to give a side by side comparison of each contact area at a glance. Fig. 11 shows the pressure along arc A - B (Fig. 10) located at the outer diameter of the blank holder ring. Fig. 12 shows the pressure along radial line O - A (Fig. 10).
60O 400 B ~
200 0
,
;;
, , , ,
,
, . . , ,
45
, , , , . , , ,
, . . , ,
90
, , ,
. . . . .
135
/Weld Line
,,
180
Angle (degree) Fig. 8. Pressure distribution on 1.00 in. and 1.35 in. blank holders at thick side of welded blank with 18 tons BHF [4]
1.00" Fig.
3.4. Pressure on the Thin Side of the Welded Blank
The literature on the pressure film stated that temperature, humidity, and contact time can change the reaction of the film under load [4]. These factors could not be controlled during the experiments. High points, low points, tool marks, and surface finish could also affect the contact pressure distribution. Each experiment was repeated three times and the results of the three trials were averaged to reduce the variations in the results.
10.
1.25"
1.35"
Side by side comparison of contact area locations for each blank holder at 18 tons BHF on thin side of welded blank
All contact areas are located near the outer edge of the blank holder ring [4]. There is no contact between the blank holder and the sheet at 6 in. from the blank holder center (Fig. 12). Six inches marks the edge of maximum (12 in.) diameter welded blank which was successfully used in previous experiments with this tooling [1]. This is clear evidence that the present single piece blank holder can not be used to draw a
L. Shulkin et al. /Journal of Materials Processing Technology 59 (1996) 34-40
cup from a welded blank, with a thickness difference of I mm.
39
elements. The simulations took an average of about 3 hours of CPU time each on an IBM Risc 6000 Workstation.
2000
1600
,~1200
800
Fig. 13. FEM model of elastic blank holder 40O
0 10
20
30
40
50
Angle O, degree
Fig. 11. Pressure magnitude along arc A - B (Fig. 10) at 18 tons BHF for three blank holder thicknesses 2000
I
iiii1.35in. t " ,25 1
1600-
K
;
t
................................
1200-
800-
The experimental contact areas are larger than those predicted by the simulations for every BHF except for the 12 ton experiment. As the BHF decreases, the.predicted contact areas become closer to the measured values. Only at 12 tons BHF, the simulation reflects a larger contact area than those of the experiments. The 1.00 in. thick blank holder simulations have the best correlation with the experimental data (Fig. 14). The pressure predicted by the simulations has nearly two times the value of the largest pressure of the experimental values. A comparison of the experiments and simulations with 1.00 in. blank holder at 18 tons BHF is shown in the Fig. 15. The pressure is m e a s u r e d in the radial direction from the center of the blank holder towards its edge.
400-
0
7
7.5
6.5
9
9.5
12 ~
[ ] Experiment
9
[]Simulation
0
Distance from Blank Holder Center (in)
Fig. 12. Pressure magnitude along line O - A (Fig. 10) at 18 tons BHF for three blank holder thicknesses
6
83 0
3.5. FEM simulations
Simulations have been performed using PAMSTAMP, an explicit 3D finite element package, and the results, in terms of the contact areas and pressures, are compared with the experiments. Eight n o d e brick elements with elasto-plastic material model are used for the blank holder. The welded blank is placed on top of the blank holder and acts as an upper boundary (Fig. 13). The 1.8 m m thick sheet which is connected (welded) to the 0.8 m m sheet created a gap of 1.0 mrn between the blank and blank holder. The nodes of the welded blank are fixed in all directions and defined as undeformable shell
18 16 14 12 Blank Holder Force (tons)
Fig. 14. Contact area comparison for 1.00 in. blank holder: experiment vs. simulation Elastic deflection and tilting of the die and ram are not considered in the simulations. This may explain w h y the contact areas are larger in the simulation than in the experiments at 12 tons of force. At high blank holder forces, the tooling and the press can elastically deflect and tilt resulting in larger contact areas. Twelve tons of BHF might not have been enough to deform the tooling and the press elastically during the experiments and therefore the contact area
L. Shulktn et al. / Journal of Materials Processing Technology 59 (1996) 34-40
40
was small. In the simulations, the upper boundary of the elastic blank holder is the rigid welded blank. Therefore, the contact area changes are due to the blank holder deflection only. As a result, the predicted pressure increases linearly with increasing BHF.
Pressure (psi)
[~_~_Experiment[ Simulation [
.......................
ooot
.
.
.
.
.
.
.
.
.
.
.
.
.
;/--;:1 .
.
.
.
.
.
1
7.5 8.0 8.5 9.0 0.5 10.0 Radial Distance From Blank Holder Center (in)
This study illustrates that it is very difficult if not impossible to predict the elastic deflections of the blank holder, not to even mention the deflections of the press and remaining parts of the tooling, within reasonable level of accuracy. Thus, a practical approach in controlling the elastic deflections of the blank holder in deep drawing will be to use multiple h y d r a u l i c or n i t r o g e n c y l i n d e r s that can be individually controlled during the process. This latter approach has been also selected by the industry in designing and m a n u f a c t u r i n g m o d e r n large transfer presses and cushions equipped with multiple point hydraulic BHF control cylinders [6]. References
[1]
Fig. 15. Contact pressure comparison for 1.00 in. blank holder at 18 tons BHF: experiment vs. simulation [4] [2] 4. C o n c l u s i o n s
Variation of the BHP on the flange surface can be achieved by elastically deforming the blank holder in certain areas. This results in improved control over the material flow and ultimately in improved deep drawn part quality. When a uniform thickness blank is used and the blank holder surface is in close contact with the sheet, a slight deflection of the blank holder results in a dramatic change of the pressure distribution on the blank holder-sheet interface, and a relatively thick steel blank holder design can be used. In order to predict the elastic deflections of a blank holder and their influence upon the drawing process, additional w o r k is necessary. Elastic deflections of the blank holder along with that of the press and tooling must be considered in order to m a k e reasonable predictions using theory-ofelasticity and FEM calculations.
[3]
[4]
[5]
[6]
D. Brouwers, L. Taupin, L. Shulkin, M. A. Ahmetoglu, G. L. Kinzel, and T. Altan, Blank Holder Force Control Using Nitrogen Cylinders in Deep Drawing of Round and Rectangular Parts from Welded Blanks, ERC/NSM-93-59, The Ohio State University, Columbus, Ohio (1993). E. Doege, Static and Dynamic Stiffness of Presses and Some Effects on the Accuracy of Workpiece, Annals of the CIRP, 29 (1980) 167. E. Doege, and G. Stock, Lock Tool - DeepDrawing Process and Examples of Operations with Elastic Blank-Holders, SAE Paper #950917 (1995). S. Jansen, L. Shulkin, M. Ahmetoglu, G. Kinzel, and T. Altan, Design of a Flexible Blank Holder for Deep Drawing of Tailor Welded Blanks, ERC/NSM-S-95-29, The Ohio State University, Columbus, Ohio (1995). T. Jimma, and F. Sekine, Effects of Rigidity of Die and Press on Blanking A c c u r a c y of Electronic Machine Parts, Annals of the CIRP, 44 (1992) 319. K. Siegert, Compensation of Tilting and Horizontal Displacement of Upper Die, Relative to the Lower Die, and Out-of-Center Forming Load by a Closed-Loop Control System. Annals of the CIRP, 43 (1994) 267.
Materials Processing Technology Journalof MaterialsProcessingTechnology59 (1996)34-40
ELSEVIER
Elastic deflections of the blank holder in deep drawing of sheet metal L e o n i d S h u l k i n a, S t e v e n W . J a n s e n b, M u s t a f a A. A h m e t o g l u a, G a r y L. K i n z e l c, T a y l a n A l t a n a'*
a ERCfor Net Shape Manufacturing, Ohio State University, Columbus, OH 43210, USA b Utilase, Inc., 20530 Hoover Road, Detroit, M148205, USA c Department of Mechanical Engineering, Ohio State University, Columbus, OH 43210, USA
Industrial Summary
In general, elastic deformations of deep drawing tools influence the process negatively; thus, they are undesirable. At the same time, an elastically deforming blank holder that adjusts itself to the changing process conditions can be instrumental in achieving uniform distribution of the blank holder pressure over the flange area. This will result in uniform material flow during deep drawing and ultimately in improved quality of the parts. A series of experiments and FEM simulations have been conducted in the ERC for Net Shape Manufacturing in order to investigate the influence of the elastic deflections of the blank holder on the pressure distribution between the blank holder and the sheet.
1. I n t r o d u c t i o n
In deep drawin~ a punch is used to force a sheet metal blank into the cavity of a die. Thus, the clearance between the die and the punch gives the blank the desired shape. The sheet metal is held in place b y means of a blank holder. The blank holder m a y either transmit a certain a m o u n t of predetermined force to the blank or it may be used to create a gap for sheet metal flow which is slightly larger than the thickness of the blank. There are many variables, such as blank holder pressure (BHP) or force (BHF), lubrication, die surface quality, blank shape, etc., which affect the quality of formed parts. All of these variables need to be carefully considered when designing a part and tooling for deep drawing. In most studies of deep drawing, the blank holder is assumed to be nearly "rigid" and deforms little. Thus, it does not adjust to the thickness changes that occur in the sheet metal during deformation. The blank holder touches the sheet metal where the thickness is a maximum and applies pressure only at these points. A uniform pressure distribution may be obtained b y first determining the locations of the increased thickness and then grinding (spotting) the
* Corresponding author. 0924-0136/96/$15.00© 1996ElsevierScience S.A.All rightsreservod Pl10924-0136 (96)02284-4
blank holder at these areas. BHF m a y be applied to the blank holder b y m e a n s of a die cushion, mechanical springs, or nitrogen or hydraulic cylinders. Many studies have shown that the combined rigidity of the dies and press has a great influence on the accuracy in precision blanking and deep drawing [2]. Elastic deflections of press and tooling are difficult to predict and control. Thus, they affect the quality of the parts and are undesirable. These deflections are caused by eccentric forming and cutting forces and result in u n e q u a l distances between the comers of the blank holder and the die. T. Jimma and F. Sekine analyzed deformations of the dies subjected to various loads and p r o p o s e d a method to select a press suitable for a given die set and a given off-center loading [5]. K. Siegert p r o p o s e d a closed loop control system using hydraulic cylinders and proportional valves for c o m p e n s a t i o n of the tilting and h o r i z o n t a l displacement of the upper die relative to the lower die [6]. An elastically deforming blank holder that adjusts itself to the sheet thickness can provide a uniform distribution of the BHP over the flange. This will result in uniform material flow during deep drawing and ultimately in improved quality of the parts. Also, spotting of the blank holder is not required in
L. Shulkinet al. /Journal of Materials Processing Technology 59 (1996) 34-40
this case. E. Doege and G. Stock had some success using a rubber elastic blank holder for forming flat sheets into rectangular pans. They observed several a d v a n t a g e s such as elimination of wrinkles, reduction of local BHP, and easier die try-outs and process adjustments [3]. The elastic deflections imposed intentionally on a blank holder can be roughly d i v i d e d into two categories: "small" and "large." "Small" elastic deflections are n e e d e d w h e n the blank holder is initially in full contact with the sheet, and a very small deformation of the blank holder is enough to cause significant changes in the BHP distribution. "Large" elastic deflections, on the other hand, are n e e d e d w h e n the blank holder first needs to be brought in contact with the sheet surface as in the case with the multi-gage w e l d e d blanks. The deflections must be large enough to change the blank holder surface geometry in order to compensate for the initial gap(s) due to the different thickness of the blank. This m a y require a large BHF and a good understanding of h o w the blank holder contacts the sheet surface and h o w the BHP is distributed. A series of experiments have been conducted at the E R C / N S M in order to investigate the influence of the elastic deflections of the blank holder on the p r e s s u r e distribution at the blank holder-sheet interface.
2. "Small"ElasticDeflections A p h o t o g r a p h of the tooling u s e d for the experiments is shown in Fig. 1 [1]. The experiments are conducted in a 160 ton single action hydraulic MINSTER/Tranemo DPA-160-10 press. The die is m o u n t e d to the u p p e r die shoe. The six nitrogen cylinders as well as a stationary punch are mounted on the lower die shoe. The press brings the die in contact with the sheet which allows the nitrogen cylinders to apply the BHF. A 2.5 inch thick load cell plate is located on top of the nitrogen cylinders. It is provided with 6 pockets for placing the load cells. A 2 inch thick blank holder plate with an inserted blank holder ring is placed on top of the load cells. The inserted ring is divided in two halves. Each half of the inserted ring can be adjusted in its height by shimming. Thus, the tooling can be used for forming axisymmetric n o n - w e l d e d blanks, w e l d e d blanks with same sheet thicknesses on both sides of the weld, or welded blanks of different thicknesses with the weld line in the middle. The main dimensions of the tooling are given in Fig. 2.
35
WeldLine Position / ~ ~.--Nitrogen Blank~ f ~ , ~ C y l i n d e r Holder / V
Plate ~
kd
\
Fig. 1. Phetograph of the experimental tooling and location of nitrogen cylinders
~ 158.2
SHEET
Fig. 2. Main dimensions of the experimental tooling (in mm) The six nitrogen cylinders are arranged into two individually controlled pressure groups consisting of three cylinders each as shown in Fig. 1. By setting a different nitrogen gas pressure for each pressure
L. Shulktn et al. / Journal of Materials Processing Technology 59 (1996) 34-40
36
group, separately controlled values of the BHF can be obtained on each side of the blank holder. This is possible because the blank holder plate deflects elastically under the force produced by the nitrogen cylinders. Laser-beam welded circular blanks from AKDQ steel are used to form the round cups. A 1.8 m m thick sheet is welded to a 0.8 m m thick sheet along the center line. Each half of the inserted blank holder ring is adjusted in its height to compensate for the 1 m m thickness difference between the thick and thin sides of the blank. Thus, the welded blank placed on top of the blank holder is in full contact with the blank holder ring inserts and the BHP is applied to both thick and thin sides of the blank. Fig. 3 s h o w s h o w d i f f e r e n t forces were transmitted from the six nitrogen cylinders through the 2.5 in. thick load cell holding plate to the six load cells. The force values read by the load cells are almost equal to the values of force set by nitrogen cylinders (2.2 tons on the thin side of a sheet metal blank and 0.5 tons on the thick side). It can be concluded that the blank holder acts not as a rigid plate but rather deforms elastically at the areas where the forces of the nitrogen cylinders are applied. This indicates that it m a y be possible to develop an "elastic" blank holder which can adapt its shape to the profile of a sheet metal blank making the deep drawing process more flexible.
kept at 1 ton while on the thin side it varied as given in Table 1. The lubricant was a heavy duty, lightly pigmented, synthetic d r a w i n g c o m p o u n d . The quality of the cups in these examples improves from a fractured cup with significantly distorted weld line to a somewhat non-symmetric cup and finally to a completely symmetric cup w i t h the weld line remaining centered. As seen in Table 1, the 13 inch blanks could not be drawn without failure. Even with very low BHF tearing could not be avoided on the thin side.
a)
b)
Tailor Welded Blank 0.8 / 1.8 mm, Exp. 58 Thin side - Load Cells 1,2,3 Thick side - Load Cells 4,5,6 20
.
.
.
.
.
.
.
.
.
.
Punch Force
c)
- - - 0 - - - Load Ceil#1
r:O ,-.
15
"-'On
Lead Cell#2
+
Load Cell#3
42_
...... i ........... ~ - - " ' " ' i - ~
-~--Load 0o,.
i
× L0adCe,.S
i
/ /
i
i
i
i
iI . , iiiiiiii
uO . ~
.1:2 " ~
f...
,, .................!........
...........
.
Fig. 4. Photographs of welded cups deformed under different BHF ratios. Process parameters are given in Table 1.
Table 1 Experimental conditions for the examples in Fig. 4
rn
-20
0
20
40
60
80
Punch Displacement [mm]
Example
Blank BHF, ton Diameter, in thick side thin side
Fig. 3. BHF variation: load cell readings a) fractured cup Fig. 4 shows the photographs of welded cups deformed u n d e r different BHF ratios between the thick (1.8 mm) and the thin (0.8 mm) side of the blank. The BHF on the thick side of the blank was
13
1
15
b) non-symmetric cup 11
1
5
c) symmetric cup
1
12
11
37
L Shulkin et al./Journal of Materials Processing Technology 59 (1996) 34-40
3. "Large" Elastic Deflections 3.1. Flexible blank holder experiments In deep drawing of tailor welded blanks with a single (non-split) blank holder plate, the flexible blank holder is expected to deform elastically to compensate for the gap caused by the thin side of a multi-gage w e l d e d blank. The blank holder transmits the force from six nitrogen cylinders by deforming elastically to press on the thin side of the sheet in order to apply the BHP to the flange of the d e e p d r a w n cup. Thickness was the primary attribute of the ring shaped blank holders which were investigated.
3.2. Experimental Tooling and Data Collection The tooling used in these experiments is the same as shown in Fig. 1, except a flat blank holder plate is placed on top of the nitrogen cylinders instead of the load cell holding plate and the blank holder plate with inserted ring (Fig. 5). The selected blank holder plate thicknesses used in the experiments are based on the results of the FEM simulations [4]. The thicknesses are: 1.00 in., 1.25 in., and 1.35 in. Each blank holder is deformed under 12, 14, 16, and 18 tons of force. Both sides of the blank holder are subjected to the same BHF to prevent the blank holder plate from tipping. A circular welded blank with a 1 m m thickness difference is placed between the blank holder and die to create a gap of I mm. In order to simulate the deep d r a w i n g conditions without actually drawing the blank, a hole is cut in the middle of the blank to provide space for the punch (Fig. 5).
Welded Blank' I Thick Side, 1.8 mm
J
~
l DIE : I Welded Blank, Thin Side, .,- .................. 0.8 mrn
I
I Bl"nk"°'d°r] I ~
J
I I Nitrogen L_..J Cylinders
Fig. 5. Tooling for experiments with flexible blank holder The pressure distribution between the holder and the blank was evaluated with a pressure sensitive film placed between the welded blank and the blank holder. This paper (Pressurex®), supplied by Sensor Product Inc., East Hanover, NJ, had already been
used successfully for measuring BHP in forming cups from flat sheet metal blanks. The pressure paper becomes pink when pressure is applied that is within the designed range of the film. The paper used for these experiments has a low sensitivity rating with a pressure range of 350-1400 psi. A single point scanner supplied by Sensor Products is used to read the pressure from the film. The scanner is linked to a computer program which outputs the pressure based on the color density.. The paper consist of two sheets of a polyester film substrate (Fig. 6). The substrate of one sheet of the p a p e r is b o n d e d to a color d e v e l o p i n g layer. The other sheet has tiny microcapsules filled with ink which are b o n d e d to the substrate. When pressure is applied to the two sheets after they have been p u t together, the microcapsules break and release the ink onto the color developing layer. ~
1
Substrate(polyester film) Microcapsule layer Color developing layer Substrate (polyester film)
Fig. 6. Cross section of Pressurex®pressure film
3.3. Pressure on the Thick Side of the Welded Blank During the 18 ton BHF tests, a sheet of pressure sensitive film is placed between the 1.00 and 1.35 in, thick blank holder and the thick side of the welded blank (Fig. 5). Since there is no gap present between the blank holder and the sheet, it was expected that the pressure distribution w o u l d be fairly uniform across the entire surface of the blank holder. A p h o t o c o p y of an actual contact patches from the pressure paper (Fig. 7) shows that the pressure distribution was not at all uniform. The pressure on 1.00 in. thick blank holder surface is highly concentrated around the locations of the nitrogen cylinders and along the weld line where the blank holder is bent to touch the thin side of the welded blank. The pressure on the 1.35 in. thick blank holder is more uniformly distributed but still mostly concentrated around the nitrogen cylinders and the weld line. Fig. 8 illustrates this and compares the pressures measured on both blank holders. The pressure was measured along the half circle with diameter O 381 m m (15") corresponding to the location of the nitrogen cylinders (Fig. 7). Note the dramatic drop in pressure b e t w e e n the cylinders. The pressure paper recorded no contact between the
38
L. Shulkin et al. /Journal of Materials Processing Technology 59 (1996) 34-40
blank holder and sheet in these areas. The results of these tests clearly show that even the thickest blank holder does not behave as a rigid body.
Fig. 9 compares the magnitudes of contact areas measured on the blank holder surface at the thin side of the welded blank for all three thicknesses of the blank holder plates and all four BHF values. The contact area is defined as the area of the colored spot on the sheet of pressure paper. The contact areas are almost directly proportional to the applied BHF. 12
Blank holder thickness
9
[ ] 1.35 in.
[ ] 1.25 in.
6
[ ] 1.00 in.
~ 3
a)
L v
b)
18
Fig. 7. Photocopy of contact patches on the blank holder surface for different thicknesses of the blank holder plate: a) 1.00 in b) 1.35 in (thick side of welded blank)
Pressure (psi)
]Nitrogen Cylinders ]
I
1.00in. . . . . 1.35 in.
1400
OOO
'/
1200
800
16 14 12 Blank Holder Force (tons)
Fig. 9. Comparison of contact area magnitude for each blank holder and BHF at thin side of welded blank
Fig. 10 shows the shape and location of the contact area for each blank holder with 18 ton BHF. The illustration in Fig. 10 is intended to give a side by side comparison of each contact area at a glance. Fig. 11 shows the pressure along arc A - B (Fig. 10) located at the outer diameter of the blank holder ring. Fig. 12 shows the pressure along radial line O - A (Fig. 10).
60O 400 B ~
200 0
,
;;
, , , ,
,
, . . , ,
45
, , , , . , , ,
, . . , ,
90
, , ,
. . . . .
135
/Weld Line
,,
180
Angle (degree) Fig. 8. Pressure distribution on 1.00 in. and 1.35 in. blank holders at thick side of welded blank with 18 tons BHF [4]
1.00" Fig.
3.4. Pressure on the Thin Side of the Welded Blank
The literature on the pressure film stated that temperature, humidity, and contact time can change the reaction of the film under load [4]. These factors could not be controlled during the experiments. High points, low points, tool marks, and surface finish could also affect the contact pressure distribution. Each experiment was repeated three times and the results of the three trials were averaged to reduce the variations in the results.
10.
1.25"
1.35"
Side by side comparison of contact area locations for each blank holder at 18 tons BHF on thin side of welded blank
All contact areas are located near the outer edge of the blank holder ring [4]. There is no contact between the blank holder and the sheet at 6 in. from the blank holder center (Fig. 12). Six inches marks the edge of maximum (12 in.) diameter welded blank which was successfully used in previous experiments with this tooling [1]. This is clear evidence that the present single piece blank holder can not be used to draw a
L. Shulkin et al. /Journal of Materials Processing Technology 59 (1996) 34-40
cup from a welded blank, with a thickness difference of I mm.
39
elements. The simulations took an average of about 3 hours of CPU time each on an IBM Risc 6000 Workstation.
2000
1600
,~1200
800
Fig. 13. FEM model of elastic blank holder 40O
0 10
20
30
40
50
Angle O, degree
Fig. 11. Pressure magnitude along arc A - B (Fig. 10) at 18 tons BHF for three blank holder thicknesses 2000
I
iiii1.35in. t " ,25 1
1600-
K
;
t
................................
1200-
800-
The experimental contact areas are larger than those predicted by the simulations for every BHF except for the 12 ton experiment. As the BHF decreases, the.predicted contact areas become closer to the measured values. Only at 12 tons BHF, the simulation reflects a larger contact area than those of the experiments. The 1.00 in. thick blank holder simulations have the best correlation with the experimental data (Fig. 14). The pressure predicted by the simulations has nearly two times the value of the largest pressure of the experimental values. A comparison of the experiments and simulations with 1.00 in. blank holder at 18 tons BHF is shown in the Fig. 15. The pressure is m e a s u r e d in the radial direction from the center of the blank holder towards its edge.
400-
0
7
7.5
6.5
9
9.5
12 ~
[ ] Experiment
9
[]Simulation
0
Distance from Blank Holder Center (in)
Fig. 12. Pressure magnitude along line O - A (Fig. 10) at 18 tons BHF for three blank holder thicknesses
6
83 0
3.5. FEM simulations
Simulations have been performed using PAMSTAMP, an explicit 3D finite element package, and the results, in terms of the contact areas and pressures, are compared with the experiments. Eight n o d e brick elements with elasto-plastic material model are used for the blank holder. The welded blank is placed on top of the blank holder and acts as an upper boundary (Fig. 13). The 1.8 m m thick sheet which is connected (welded) to the 0.8 m m sheet created a gap of 1.0 mrn between the blank and blank holder. The nodes of the welded blank are fixed in all directions and defined as undeformable shell
18 16 14 12 Blank Holder Force (tons)
Fig. 14. Contact area comparison for 1.00 in. blank holder: experiment vs. simulation Elastic deflection and tilting of the die and ram are not considered in the simulations. This may explain w h y the contact areas are larger in the simulation than in the experiments at 12 tons of force. At high blank holder forces, the tooling and the press can elastically deflect and tilt resulting in larger contact areas. Twelve tons of BHF might not have been enough to deform the tooling and the press elastically during the experiments and therefore the contact area
L. Shulktn et al. / Journal of Materials Processing Technology 59 (1996) 34-40
40
was small. In the simulations, the upper boundary of the elastic blank holder is the rigid welded blank. Therefore, the contact area changes are due to the blank holder deflection only. As a result, the predicted pressure increases linearly with increasing BHF.
Pressure (psi)
[~_~_Experiment[ Simulation [
.......................
ooot
.
.
.
.
.
.
.
.
.
.
.
.
.
;/--;:1 .
.
.
.
.
.
1
7.5 8.0 8.5 9.0 0.5 10.0 Radial Distance From Blank Holder Center (in)
This study illustrates that it is very difficult if not impossible to predict the elastic deflections of the blank holder, not to even mention the deflections of the press and remaining parts of the tooling, within reasonable level of accuracy. Thus, a practical approach in controlling the elastic deflections of the blank holder in deep drawing will be to use multiple h y d r a u l i c or n i t r o g e n c y l i n d e r s that can be individually controlled during the process. This latter approach has been also selected by the industry in designing and m a n u f a c t u r i n g m o d e r n large transfer presses and cushions equipped with multiple point hydraulic BHF control cylinders [6]. References
[1]
Fig. 15. Contact pressure comparison for 1.00 in. blank holder at 18 tons BHF: experiment vs. simulation [4] [2] 4. C o n c l u s i o n s
Variation of the BHP on the flange surface can be achieved by elastically deforming the blank holder in certain areas. This results in improved control over the material flow and ultimately in improved deep drawn part quality. When a uniform thickness blank is used and the blank holder surface is in close contact with the sheet, a slight deflection of the blank holder results in a dramatic change of the pressure distribution on the blank holder-sheet interface, and a relatively thick steel blank holder design can be used. In order to predict the elastic deflections of a blank holder and their influence upon the drawing process, additional w o r k is necessary. Elastic deflections of the blank holder along with that of the press and tooling must be considered in order to m a k e reasonable predictions using theory-ofelasticity and FEM calculations.
[3]
[4]
[5]
[6]
D. Brouwers, L. Taupin, L. Shulkin, M. A. Ahmetoglu, G. L. Kinzel, and T. Altan, Blank Holder Force Control Using Nitrogen Cylinders in Deep Drawing of Round and Rectangular Parts from Welded Blanks, ERC/NSM-93-59, The Ohio State University, Columbus, Ohio (1993). E. Doege, Static and Dynamic Stiffness of Presses and Some Effects on the Accuracy of Workpiece, Annals of the CIRP, 29 (1980) 167. E. Doege, and G. Stock, Lock Tool - DeepDrawing Process and Examples of Operations with Elastic Blank-Holders, SAE Paper #950917 (1995). S. Jansen, L. Shulkin, M. Ahmetoglu, G. Kinzel, and T. Altan, Design of a Flexible Blank Holder for Deep Drawing of Tailor Welded Blanks, ERC/NSM-S-95-29, The Ohio State University, Columbus, Ohio (1995). T. Jimma, and F. Sekine, Effects of Rigidity of Die and Press on Blanking A c c u r a c y of Electronic Machine Parts, Annals of the CIRP, 44 (1992) 319. K. Siegert, Compensation of Tilting and Horizontal Displacement of Upper Die, Relative to the Lower Die, and Out-of-Center Forming Load by a Closed-Loop Control System. Annals of the CIRP, 43 (1994) 267.