Journal of the Faculty of Englneerlng, Shinshu University, Ne. 67, 1990

tstixl･IJit*XefKSfEes ng67e1

Forming Limit and Its Improvement in Thin-

Walled Tube-End Flaring Using Orbital

Rotary Forming with Cylindrical Tools

Kimiyoshi KITAZAWA* and Jiro SEINO** (Received October 30, 1989)

The possibility of the tube-end fiaring using orbltal rotary forming with a

cylindrical tool has been experimentally investigated from the viewpoint of form-

ing limit by using thin-walled copper tubes. The forming limit is determined by

occurrence of rupture and various kinds of bucklings. The diameter of a cylin-

drlcal tool has a major influence on the forming limit in orbital rotary flaring. The

flaring 11mit of tube in orbital rotary forming with a cyHndrical tool is greater than

that with a conical tooL Furthermore, to prevent buckling a pass schedule of a

cylindrical tool is proposed and experimeRtally examined. The buckling mode is

prevented under the proposed preventing condition so that the forming limit is

dramatically improved. Therefore, the proposed orbital rotary fiaring with a cylin-

drical tool is likely to be a hopefully practical working process from the viewpQint

of forming limlt.

I. INTRODUCTION

The conventional press forming is a process in which a sheet metal is

formed to fit on a die surface, Due to this fitting necessity, the conven-

tional press forming is not so flexible for production of small lots. In consid-

eration of this problem of manufacture fiexibility, the,concept of orbital

rotary forming was first introduced in l986 by Kitazawa et al. i) and shortly

afterwards was applied to tube-end forming and sheet metal forming by

Kitazawa et al. 2)A"6). Orbital rotary forming is a process in which a tube end

or sheet metal is formed to fit on a tool-envelope-surface made by a relative

motion of a tool around the sheee metal or tube, In orbital rotary fiaring and

nosing of a tube end, the tube end fits well on the tool-envelope-surface. 2)

Therefore, orbltal rotary forming seems to be a hopeful process for fiexible

manufacture from the viewpoint of die-less or too!-less operation. It has also

* Asslstant, Department of

Shinshu University.** Professor, Department of lng, Shinshu Universlty.

Mechanical Systems Englneering, Faculty of Engineering,

Mechanical Systems Engineering, Faculty ef Engineer-

2 K. KITAzAwA and J. SEINobeen clarified that the forming limit of a tube end in erbital rotary forming

is greater than that in press forming.3) Recently, a computer-controlled

orbital rotary forming machine has been developed which has a mechanism

capable of forming continuously an appropriate tool-envelope-surface corre-

sponding to a product shape during a stroke of forming6). Also, an orbitai

rotary flanging of tube ends has been carried out by using this machine,

and it is found that the metal-fiow behavior is dependent on the pass sched-

ule of the tool during the forming stroke. 6) There exist such a large num-

ber of degrees of freedom of this pass schedule that the method for optimiz-

ing the pass schedule of the tool has not yet been found,

The orbital rotary forming process and the so-called "potter's wheel"

process are alike in the relative motion of tool or finger around a thin-walled

material such as a cup. It may be possible that the sight of the motion of

fingers in the potter's wheei process, when observed carefully, wil! lead to

a new idea for optimizing the pass schedttle of a tool in the orbital rotary

forming process. One of the more interesting problems in the application

of the motion of fingers to orbitai rotary forming is the to devise a method

for preventing the buckling in orbital rotary flaring of thin-walled tubes.

In the present work, thin-walled copper tubes are subjected to an orbital

rotary flaring using a cylindrical tool, designed to imkate the finger, ln an

attempt to study on the fiaring limit in pursuit of its improvement method.

2. EXPERIMENTALPROCEDURE

To clarify the forming limit, orbital rotary flarillg tests were carried

out on a modified lathe. As shewn in Fig.1, the proposed orbital rotary

forming is a process in which the tool-envelope-surface made by the rela-

tive motion of a cylindriacl tool around the tube moves in the axial dlrection

while the tube rotates, and thus, the tube-end is flared to fiow along the

tool-envelope-surface. The tube held by a chuck rotates ae a constant speed

of 315 rpm, and the cylindrical tool and its holder mounted on a tool rest

which is swiveled to a required angle r equal to the half apex angle of the

tool-envelope-surface, traveis in the axial direction at a constant feed of f

mm/rev. The materials used are two kiRds of seamless copper tubes (outer

diameters 35 mm; lengths 70 mm; wall thicknesses l,O and 1.5 mm). The

length of fiaring zone is equal to the outer diameter of the tube so that

Euler buckling should not occur. The mechcanical properties of the tube

are given in Table 1. Pasty molybdenum disulphide (MoS2) and machine oii

are used as lubrlcant and are ceated on both the tool and the tube. The tube

is fiared with the cylidrical tool until a failure such as buckling or rupturing

Forming Llmit and Its Imprevement in Thin-Walled Tube-End Flarlng 3

¢d

L/2

L/2

s£

rr --II

--p-

Ii

-1}t

,

,

v/

'

tro

¢do

Fig.

( ldler rotated )

/ / CyliRdrical/ ip D

Tu.bular specidieR

:::;::.:i'i:{:::: Chuck

flaring

N

with

L=N=do=

7S

315

35

mm

rpm

mm

a cylindrical

tool

1 Orbital rotary tool.

Table 1 Mechanical properties of the materials used.

Material

(JIS)6uterDia.do/mm 'ro/mm

ThicknessIWorkhardeniRgexponeRtn P}asticcoeff2cientF/MPa

Hardcopper(C1220-H)

e.6 o.e2 450'l3sl e.s e.o2 460

Softcopper(cm2e-e)

35 O.6 9.46 6eo

takes place. Then, the magnitude of the limiting flaring ratio 2c7) defined by

Eq. 1 at buckllng or rupturing can be obtained by measurlng the final outer

diameter, d, of the tube end and comparing k with the initial outer diameter,

do, of the tube end:

2c=(d-do>/d6- (1)Effects on Iimiting flaring ratio 2c were investigated of the diameter, D, of

the cylindrical tool, the half apex angle, r, of the tool-envelope-surface,

the feed displacement, L of the tool per revolution, the lubrication, the

wall thlckness, To, of the tube, and the work-hardening exponent, n, of the

tube material over the range shown in Table 2, To confirm the advantage

of using a cylindrical tool over using a con!cal tool, 2e for a cylindrical tool

have been compared wlth that for a conical tooL Furthermore, a comparison

4 K. KITAzAwA and J. SEiNO

Table 2 Experimental conditions.

Factors Conditions

Halfapexangletool-envelope-surfaee

o£

7/o15,20,3e,45

'6e,75

Feeddisplacementperrevoltitlon

oftoolf/MM･rev-t

o.el2s,o.os, o .1,O.7

Lubrication PastymolybdenumMachineoil

disulph!de,

Tubewallthickness To/mm O.6,O.8 (do=35 pata)

Diaraeterofcylindricaltool D/mm

15,2e,25,30

has also been made between Rc for the proposed rotary forming and that

for the conventional press forming.

3. EXPERIMENTALRESULTS

3. 1 Bnfavorable deformatieit modes

For the orbital rotary fiaring with a cylindrical tool, forming limits were

determined by the occurrence of severai unfavorable deformation modes such

as various kinds of bucklings and rupturings, as shown iR Fig. 2. The photo

in Fig.3 shows the defects formed on thin-walled hard copper tubes when

fiared to rupture or buckle with a cyliRdrical tooL

When r{iS200, necks occurred and propagated in the direction parallei

to the axis of tube as shown in Fig.3 (a), and finally, a rupturing occurred

along the necked regions. It should be stressed that this necking mode is

a characteristic mode in orbital rotary fiaring. In press flaring, it is well

(a)

Flg.

, ,

'f 11

Necking/ (b) Buckiing (c) Bulging (d) Buckling

Rupture at tool inle't near chuck2 Unfavorable deformation modes for tube-endrotary fiaring with a cylindrical tooL

(e) Week bulging in f!ared zone

fiaring by orbital

Forrning

Fig.

known that

meridlonal

Lirnit and Its Improvement iR Thin-Walled Tube-End Flaring

7=15e 7=20e 7=30e A.c=g.26 Ac=G.344 A.c=O.383 Buckling Bulging and Buckling with fiecking rupture at tool inlet

7=45e 7=6eO , 7=75e A.cmO.362 Ze=O.319 Ac=O.273 Buckling Blickling Buckling at tool inlet at too} iRlet at toel inlet

3 Unfavorable deformatioR modes in the process of orbitalrotary flarlng with a cylindrlcal tooL (Hard copper tube:

Outer diameter 35 mm; Wall thickness O. 6mm; D = 15mm;f=O.Imm/rev. Lubricant: Pasty rnolybdenum disulphide)

-!- tt -- -l.

tt :t

,

(a) Necking mode

Flg.4 Necking deve!oped press fiaring.

wall thickness disulphide)

necks started

direction as shown

ee,,ew=i'i"{''i"'afew/Li,k gee eqeegetw

a=300 a=45e Ac=e,325 Zc=O.351 (b) Necking and ruptupe

and rupture orlginated at the tube end and

at about 450 to the merldional direction iR (Hard copper tube: Outer diameter 35mm;

O.6mm. Lubricant: Pasty molybdenum

at the tube end aRd develQped at about

iR Fig. 4(a).

450 to

5

the

K. KITAzAwA and J. SEINo

On the other hand, in the range r}l:300, no necks occurred, and the form-

ing limit was determined by the onset of buckling at the tool iniet as shown

in Fig.2 (b). When r=30e, it is observed that the weak bulging mode shown

in Fig.2 (e) takes place after the above-mentioned buckling mode has taken

place. Furthermore, it is also found that the occurrence is restrained of

curling or waviness observed in the press flaring and orbital rotary flariRg

with a conical tool.

3.2 Forming limit

The relationship between the half apex angle of tool-envelope-surface r

and limiting flaring ratio 2c is shown in Fig.5 for two diameters of cylin-

drical tool D: 15 and 30 rnm. It is shown that 2c increases as r decreases,

O.6

o ct

o tsN o.4 tlt

.:

q O.2 ue ..H. 'G kH e.o o

Fig. 5 Influence between the ing falrlng thickness O.6 disulphide.

reaching a maximum

ases. As has been

occurrence ef neck

The shape of the 2c

case of D=15 mm,

Figure 6 shows the

2c. It is shown that

' -'-

/

OM :D :15 mm @tw:I)=3q mm

-. -ege --s.ee

"""-'Nhl op × NN

3o 6e gg Half apex angle of lt/e tool-envelope-surface

of the diameter of cylindrical tool D on the relationships

half apex angle of toel-envelpe-surface r and the limit- ratio 2c. (Hard copper tube: Outer diameter 35mm; wall

mm; f=O.1 mm/rev. Lubricant: Pasty rnolybdenum Otw: Buckling mode at tool inlet. []@: Other modes)

value at rr300, and then decreases as r further decre-

dlscussed above, in the range r)3eO, it is noted that the

and rupture observed in the range rl:i{200 is restrained.

curve in the case of D=:30 mm is similar to that in the

although 2c for D=:l5 mm !s grater than that for 30 mm.

dependence of tool diameter D on limiting fiarlng ratio

2c increases as D decreases and approaches a maximum

Forming Limit and Its Improvement in Thin･Walled Tube-End Flaring 7

O.5

x

.-O-jN

£ e.4

es:.

va

to O.3Tfi

".J

.E

A

--.e-NeNts

otw

r

:

ox

7=3oe7=45e

8x

Nx

7

×,.

× ±

gI

Fig.

/

e2 10 2e 30 4Q Diameter of cykndrical tool D/mm

6 Relationships between the diameter of cylindricalfiarlng ratlo 2c. (Hard copper

ness O.6 mm; f=O.1 mm/rev. Lubricant: Pastyphide. Otw: Buckling mode at tooKnlet)

ip D

tool and limitingtube: Outer diameter 35 mm; wall thick-

molybdenum disul-

value at about D= 20 mm. In the case of annealed copper tubes, the buckling

mode (Figs. 2 (c) and (d)) occurs in an early stage of fiaring. The effect of

tube wall thickness To on limiting flaring ratio Rc is shown in Fig. Z 2c in-

creases as To incerases. The effect of the feed displacement of tool per rev-

olution, f, on 2c is shown in Fig. 8. From thls figure, it is seen that 2c in-

creases as f increases.

4. DISCUSSION

From these experimental, results it is found that higher limiting fiaring

ratios are obtainable with cylindrical tools having smaller diameters, smaller

worltrhardening exponents of tube materials, smaller half apex angles of tool-

envelope-surfaces, Iarger wall thicknesses of tubes, and larger feed dis-

placements of tool per revolution as shown in Table 3. Considering production

cost, it is desirable that limiting flaring ratie increases as the feed displace-

ment of tool per revolution increases. It is recommended that the diame-

ter of the cylindrical tool is ha}f the outer diameter of the tube, considering

both the rigidity of cylindrical tool and the experimental results shown in

Fig. Z As shown in Fig. 8, it ls obvious that the limiting flaring ratio in

orbital rotary flaring increases as the feed displacement of tool per revolu-

tion increases, although the llmiting flaring ratio for the proposed rotary

forrning in Fig. 3 and those for the conventional press forming in Fig. 4

8

Table

K. KITAzAwA and J.

3 Method for improving

SEINO

flaring limit.

D/mpt f/mmrev To/mm 7/e n

X(--e.sdo) / / X(305) x/: High value, X :Low value

oKe--pptkua=.-fues

r--l

ig

bo=

'"'t

P.Hua

-"

e.6

O.4

O.2

o. o

g

"r'-gexastsz .-

l tw ee N N

N

OM : To= O.6 himtw pa : [I"o :O.8 mm

Fig.

o 3o 6o gg ' }lalf apex angle ef 7/e tool-envelope-surface '

7 Influence of tube wall thickness To on limiting flaring ratio 2c.(Hard copper tube: Outer diameter 35 mm; wall thicl<ness O.6 mm;

D=15 mm; f= O.1 mm/rev. Lubricant: Pasty molybdenum disulphide.O@: Buckling mode at tooHnlet. []tw: Other modes)

o K

o,r-

"ptkbo=J-suas

"--{

-tocr.etg-)lri

N.H"

O.5

O.4

e.3

O.2

-o

e/e7 -- 3oo

/x or×7=6oe

"t

Fig.

o.ol g.1 1.o Feed dispiacemeRt of the tool f/tum･rev'f

per revolution

8 Influence of the feed displacement of tool per reyolution f on !imiting

flaring ratio 2c. (Hard copper tube: Outer diameter 35 rnm; wall thick-

ness e,6 mm; D =15 mm. Lubricant: pasty molybdenum disulphide.O@: BuckHng mode at tool inlet)

Formlng Limit and Its Improvement in Thifi-Walled Tube-End Flaring 9

,.O, e.s O:Cylindrical teol : a=oo o tw:a=15e .H X @:ar3ee k O.4 to ¢ --.4 .r-l ts .･ ･-･a}-li･G g.:. o.3 N@k!iteta,.oge.N.ce-... '

IE S' O.2 o lg 2o 3e 4o (a) Eqtiivalent diameter Equivalent diameter of tool Deq/mm of tool Deq (b) 7=4se

Fig. 9 Infiuence of the equivalent diameter of tool Deq on limiting flaring ratio Rc. (Hard copper tube; Outer diameter 35 mm; Wall thickness O.6 mm. Lubricant: Pasty molybdenum disulphide)

are approximately the same. Thus, it is suggested that the forming limit

of orbital rotary forming is higher than that of press formlng.

In Re£ 3, orbital rotary flaring and nosing with a conical tool are de-scribed and forming limlts are reported except for the range r<450. From the

results of Ref. 3, the effects of D, r, f, To, and n on 2c are similar to the

above experlmental results. One of the main differences between the orbital

rotary formlng with a cylindrical tool and that with a conical tooi is in the

tool shape. To compare forming limits of both cylindrical and conical tools,

we have introduced the concept of "equivalent diameter of tool Deq" defined

in Flg. 9 (a). Figure 9 (b) compares fiaring limits of cylindrical and conical

tools by transforming results for conical tools into those for cylindrical tools

using Deq. It is found that the iimiting fiaring ratio for a cylindrical tool

is greater than that for a conical tooL Furthermore, the Iimiting fiaring

ratio increases with decreasing equlvalent diameter of tool for both cylin-

drical and conical tools. It is suggested that the equivalent diameter of tool

Deq has a major effect on the forming limit in orbital rotary forming.

The fitting behavior of tube to tool-envelope-surface is controlied by

contact area ratio defined as the ratio of the contact area to the total de-

formed area of the tube. 2) As has been discussed above, 2c increases with in-

creasing f and with decreasing Deq. On the other hand, the contact area ratio

decreases as both f and Deq decrease. Thus, it is suggested that the contact

area ratio does not seem to be a factor affecting 2c: the onset of buckling

at the too! lnlet ls indepeRdent of the contact area ratio. This burkling which

ssa

.tr.7 c

r;

;

:r?.1

×'.

:Z Deq× +-t

.-

10 K. KiTAzAwA and J. SEINOeccurred at the tool inlet is a local buckling of the mode, and propagated

in the circumferential direction of tube. 3) It was observed by Kitazawa et

al. 3) that it ls diMcult to predict the occurrence of the above 'buckling mode

by means of buckling load, and that the onses of buckling is affected b'y

local meal fiow. Thus, it is suggested that the effect of Deq on 2c is connected

closely with this local metai-flow phenomenon.

As has been mentioned above, from the viewpoint of how to improve

the formlng limit in orbital rotary flaring of thin-walled tubes, the method

for preventing buckling is the most important problem. In a previous paper, 3)

Kitazawa et al. have proposed the continuous lnclination method of the pass

schedule of conical tool in tube end flanging but we point out a difficulty in

applying this method to pi"eventing buckling that occurs at the tool iniet.

The motion of fingers in the potter's wheel process suggests a new idea,

the mothod for preventing buckling shown in Fig, 10, for the pass schedule

of a cy!indrical tool in the orbital rotary forming process. Then, we intro-

duced the modified inclination method of the pass schedule of a cyiindrical

tool shown in Fig. 10 (b). Th!s method is a process in which the tube at the

tool inlet is repeatedly flared to prevent buckling by using a cylindrical tool

( Idlep rotated )

Cylindrical tool

Z' , i' , 2[l'

l ,

Tubular

}

'

speclmeB

tC/ii3 '

(a) IRclination method

iZ ! ii])/1[ii

-t , ,

(b) Proposed method ( Modified inclination method )

method for the preventing of the buckling

'

Fig. Ie Proposed at tool inlet.

Forming Limit and Its Imprevement in Thin-Walled Tube-End Fiaring

tts,

wa

me

tttttt l

t"t

11

Single process IRciiRatioB method Proposed method (Ac=e.319) (Zc=O,455) (Ac=e.692)

Flg. 11 Effect of the proposed method on forming 11rnit. (r=60e. Hard copper tube: Outer diameter 35 mrn; Wal! thickness O.6 mm; D==15 mm; f=O.l rnm!rev. Lubricant: Pasty molybdenum disulphide)

with a small angle r uRder applicatioR of the above inclination method. Figure

11 shows that the buckling mode is prevented under the proposed prevent-

ing condition so that the forming Iimit is dramatically improved.

Therefore, the proposed orbital rotary flaring with a cylindrical tooi is

likely to be a hopefully practical working process from. the viewpoint of

formlng limit.

5. CONCLUSION

The possibility of the tube end flaring using orbital rotary £orming with

a cylindrical tool has been experimentally investigated from the viewpoint

of forming limit by using thin-walled copper tubes, The resuks are summa-

rized as follows:

1) Forming limits were determined by the occurrence of several unfavor-

ab!e deformation modes such as various kiRds of bucklings and rupturings.

In the range r:;E{200, a rupture mode occurred and propagated in the direc-

tion parallel to the axis of tube. On the other hand, ln the range rk300,

the forming limit was determined by the onset of buckling at the tool inlet.

2) It was found that higher limiting fiaring ratios are obtainab!e with

cylindrical tools having smaller diameters, smaller work-hardening exponents

of tube materials, smaller half apex angles of tool-envelope-surfaces, larger

wall thicknesses of £ubes, and larger feed displacements of tool per revoltt-

tion. The forming limit of orbital rotary forming is higher than that of

press forming.

3) It was found that the limiting fiaring ratio for a cylindrical tool is great-

er than that for a conical tooL The equivalent diameter of tool Deq has

12 K. KITAzAwA and J. SEINoa major infiuence on the forming limit in orbital rotary forming. It is recom-

mended that the diameter of cylindrical tool should be half the outer dia-

meter of tube.

4) A modified inclination method in which the tube at the tooi inlet is

repeatedly fiared by using a cylindricai tool with a small angle r during tube-

end fiaring, is introduced to prevent buckling. Then, lt was found that the

bucklng mode is prevented under the proposed preveRting condition so that

the forming limit is dramatically improved.

'

Ael{nowledgment

The authors would like to thank Dr. Matsuo Miyagawa, Dr. Hisashi

Nishimura and Dr. Ken-ichi Manabe for their advice. The authors are also

indebted to Mr. Akira Fujita and Mr. Nobukazu Tsutsumi for producing the

tooling.

REFERENCES

(1) K. Kitazawa, M. Kobayashi and H. Maruno: Proc. 1986 Japan Spring Conf. Tech-

nology of Plastlcity, 343-346 (1986).

(2) K. Kitazawa, M. Kobayashi and H. Maruno: Journal of the Japan Society for Tech-

nology of Plasticlty, Vol.29, No.330, 767-774 (1988).

(3) K. Kitazawa, M. Kobayashl and H. Maruno: Journal of the Japan Society for Tech-

nology of Plasticity, Vol.30, No.344, 1259-1266 (1989).

(4) K. Kitazawa and M. Kobayashi: Proc. 37th Japan Joint Con£ Technology of Pla- sticity, 349-352 (1986).

(5) K. Kitazawa, J. Seino and K. Fujita: Proc. 40th japan Joint Conf. Technology

of Plasticity, 37-4e (1989).

(6) K. Kitazawa, M. Iida and M. Kobayashl: Proc. 1989 Japan Spring Conf. Tech-

nology of Plasticity, 343-346 (1989).

(7) K. Manabe and H. Nishimura: Journal of the Japan Society for TechnQlogy of

Plasticity, Vol.24, No.266, 276-282 (1983).