Tagungsunterlagen Dorp Eldorado

technische universität
dortmund
1st International Tube and Profile

Bending Conference
4th DORP 2011

Dortmunder Kolloquium zum
Rohr- und Profilbiegen
Dortmund, Germany
24th–25th November 2011
Institute of
Forming Technology and
Lightweight Construction
4th DORP 2011
Sponsored by:
IMT
International Metal Tube
2
Dortmund, Germany
4th DORP 2011 24th–25th November 2011
Index
Preamble 5
Keynote 7
Tube Bending of New Generation by MOS Bending Machine 9
Session 1: Nonconventional Processes 19
Machine Concept for Incremental Tube Forming 21
Tube Bending in Power Plants 25
Session 2: Standardization in Bending Technology 47
VDI 3430: A Guideline for Rotary Draw Bending of Profiles 49
Standard for tube and profile bending technology 51
Session 3: Profile Manufacturing 55
An Analysis of Bend Allowance Prediction in Folding 57
Curvature Optimized 3D-Profiles to Improve the Line

Speed of Flexible Roll Formed Profiles 61
Session 4: Profile Bending 83
Process Limit Extension of the TSS-Bending Process 85
Measurement System for Three Roll Bending of Beams 89
3
Dortmund, Germany
Session 5: Measurement and Simulation Techniques 99
Numerical Simulation of Profile Bending Using the TSS

Bending Method 101
Measurement of Bent Parts for the Correction of

Bending Machines 105
Variation of Contractile Strain Ratio of TA18 Titanium
Alloy Tubes with Plastic Deformation 121
Session 6: Tube Bending 125
Improving the Curvature of Bent Profiles by Adjusting the

Process Kinematics of the Three-Roll-Pushbending Process 127
Influence Factors of Three-Roll-Push-Bending 131
4
Dortmund, Germany
Preamble
In recent years, lightweight construction is more and more being realized by
structures made of profiles with complex cross sections and high strength materials. These
structures offer the advantage of high safety and good stiffness at the same time. This is
an important aspect for transport systems, also providing for lowered energy consumption
by the weight reduction. To realize esthetical and aerodynamic structures, there is a strong
demand in industry for 3D-bend contours made of these profiles. The integration of
functions and the high level of the ergonomic usage demands innovative solutions for the
production technology.
The Institute of Forming Technology and Lightweight Construction (IUL) and its
department of bending technology allocates a large range of solutions for these special
properties by process developments, process optimization, and basic research in the field
of bending and forming sheets and tubes. This aim is reached by the combination of basic
research and innovative ideas.
One example is the TSS-Bending (torque – superposed – spatial) process for 3D
freeform bending of profiles with arbitrary cross sections. The department has developed
a machine prototype and a special process planning tool based on analytical insights. By
this work the technology has reached a high industrial standard. Another example is the
innovative process “Incremental Tube Forming” (ITF), a combination of spinning and
bending. For this process an industrial machine prototype has been developed and a basic
research project with the aim of carrying out a fundamental analysis of ITF will start in the
near future. The semi-finished product sheet metal and the production of sheet metal-based
profiles is also a key aspect of the department. Particularly roll forming and free bending
of high strength materials and tailored blanks is under investigation. An important aim is
the investigation and prevention of crack failure to allow wider process windows for the
production of thin-walled lightweight profiles. For example, the process “free bending
with incremental stress superposition” tries to extend these limits by the superposition of
hydrostatic pressure in the bending area of the sheet metal bending process.
Not only the IUL has strong intentions to develop innovations in the special field of
bending, the whole sector and many researchers face up to the future challenges. So there
is a call for a forum for exchanging experiences between users, machine manufacturers,
and university researchers.
In response to the great feedback in the past years, DORP
2011 is a two-day as well as international conference for the
first time. In this way, there will be more room for interesting
lectures, stimulating discussions, and an intense exchange of
experiences and views between industry and science on an
international level.
The following lectures and papers are the result of the
DORP 2011 and show the excellent work of different research
institutes and innovative companies. However, it cannot reflect
the great interaction and fruitful discussion between our
guests that the DORP conferences, organized by the Technical
University of Dortmund and the IUL, are known for.
Prof. Dr.-Ing. A. Erman Tekkaya
5
Dortmund, Germany
6
Keynote
Dortmund, Germany
Tube Bending of New Generation by MOS

Bending Machine
Makoto Murata
1-5-1 Chofu Gaoka, Chofu-shi, Tokyo, 182-8585, Japan
University of Electro-Communications
murata@mce.uec.ac.jp
Abstract: The bending can be used to bend tubes and suitable only for long production
runs in limited sizes, not for short production runs in various sizes. Therefore the author
has invented a new flexible penetration bending method and named it the MOS bending
method. The bending is controlled by computer and is available for various bending radii
with high precision. The bender only needs to change the easy bending program when it
changes the bending radii. It does not need to replace dies. Bending radii and angles can
be easily adjusted by putting a tube through the die. The new CNC bender changes the
image of the conventional tube bender. The above and other benefits of MOS bending are
reviewed here, together with its mechanism and characteristics of tubebending.
Keywords: metal-forming, tube forming, tube bending, CNC, MOS bending
1. Introduction
There are many kinds of tube bending methods such as press bending, stretch
bending, draw bending and other various bending methods[1]- [5]. These bending
methods have been developed and are used depending on operation efficiency and other
requirements. Because bent circular tube products are employed to reduce production cost
and weight in manufacturing many kinds of products such as fluid arrangements, furniture,
transport apparatus, and mechanical parts.
A flexible bending[6] was invented by Dr. Murata and developed by Nissin
Precision Machines Co.,Ltd, And it challenges the conventional image of a tube
bender and is named MOS bending. The MOS bending is controlled by computer and is
available for various bending radii with high precision. With this unit there is no need for
replacement of dies when bending radii are changed only by a compute program change.
Bending radii and angles are readily adjusted as the tube is put through a die. These
and other benefits of MOS bending are reviewed here, together with its mechanism and
characteristics. Some examples of the products of MOS bending are noted in this study.
9
Dortmund, Germany
2. Mechanism of MOS bending by CNC ( Computer Numerical

2.1. MOS Bending Mechanism
Control )
Figure 1 shows the principal parts of the bending machine. The mechanism is a very
simple2.1.
structure,
MOS with an operating sequence as follows: The relative distance between
Bending
2.1. MOSMechanism
Bending Mechanism
theFigure
center 1ofshows
the guide cylinder③
the principal andofthe
parts center of the bendingThedie② is offsetisu.a A
Figure 1 shows the principal partsthe bending
of the machine.
bending machine. mechanism
The mechanism is very
a very
simple
tubular structure, with
workpiece① an operating
penetrates the diesequence
from the as follows:
guide Theby
cylinder relative
axial distance
force PL, between
then
simple structure, with an operating sequence as follows: The relative distance between the
the
bent ascenter
shown.
center of For
of the the guide
this
guide cylinder③
unit,
cylinder and and
the distance the center
(approach
the center of bending
V)
of the the bending
between the die②
die exit ofisthe
is offset offset
u.guide u. A
A tubular
cylinder
tubularand the
workpiece① center of the bending
penetrates the die
die is
from notthechanged.
guide In the
cylinder neighborhood
workpiece penetrates the die from the guide cylinder by axial force PL, then bentby axial force of
PL, the
thenas
bending
bent die,shown.
as the tube is this
pushed bythe the bending (approach
die as indicated by the arrow.
shown. For thisForunit, theunit,
distance distance
(approach V) between V) the
between
exit ofthe
the exit
guide ofcylinder
the guide and
cylinder
the center andofthe
thecenter
bending of die
the isbending die is In
not changed. notthechanged. In the of
neighborhood neighborhood
the bending of die,thethe
bending die, theby
tube is pushed tube
theisbending
pushed dieby the bending die
as indicated by as
theindicated
Bending
arrow. by moment
the arrow.
M=Pu×V+PL×u
Bending moment
M=Pu×V+PL×u
①Tubular work piece
②Bending die
①Tubular work piece
③Guide cylinder
②Bending die
⑪Spherical bearing
③Guide cylinder
Figure 1; Principal parts of MOS bending.
⑪Spherical bearing
Figure1;1;Principal
Figure Principalparts
partsofofMOS
MOSbending.
bending.
The extrusion load Pu is dependent on the magnitude of the offset u. A bending moment
M(=Pu×V+PL×u) operates to the bent tube. The position of the die is continuously
moved The
by extrusion
AC load Pu is dependent servomotors.
microcomputer-controlled on the magnitude ofbending
Thethe the offset u. A
is bending moment
The extrusion load Pu is dependent on the magnitude of offset u.die rotatedmoment
A bending on.
M(=Pu×V+PL×u)
Angle operates to the bent tube. The position of the die is continuously moved
M(=Pdirections
u×V+PL×u) as operates
shown Fig.1
to the and
bentthe twoThe
tube. angles are of
position controlled
the die is by continuously
two AC
by AC microcomputer-controlled servomotors. The bending die is rotated on. Angle
servomotors
moved by. AC microcomputer-controlled servomotors. The bending die is rotated on.
directions as shown Fig.1 and the two angles are controlled by two AC servomotors .
Angle directions as shown Fig.1 and the two angles are controlled by two AC
servomotors .
Photo 1: New CNC bender

10
Dortmund, Germany
2.2 CNC Tube Bending Machine

The AC microcomputer-controlled
2.2 CNC Tube Bendingservomotors Machine govern the position of the bending die
as shown
The AC in Figure 2. and Photo 1. Bending
microcomputer-controlled proceeds asgovern
servomotors follows: the position of the
A tubular
bending dieworkpiece①
as shown isininserted
Figure into
2. the
andguidePhotocylinder③.
1. Bending When AC servomoter
proceeds ⑥ is
as follows:
Aactivated, the tube is is
tubular workpiece pushed toward
inserted into Ztheaxis by cylinder.
guide plate⑩, andWhen it is fixed to chain⑨.The
AC servomoter is
tube thereby
activated, penetrates
the tube is pusheddie②. TheZdie
toward axisis by
shifted continuously
plate, toward
and it is fixed X and YThe
to chain. axes by
tube
thereby penetrates ④⑤.
AC servomotors die. The die is shifted continuously toward X and Y axes by AC
servomotors
Once a tube. is bent to a certain radius, the die is shifted to the position of offset u, which
depends
Once on the radius
a tube is bentR.toThe relationship
a certain radius,between
the die isoffsets andtoradii
shifted is shownofinoffset
the position Figureu,4.
Whendepends
which a tube is onbent to a certain
the radius R. Theangle, the tubebetween
relationship length isoffsets
measured by rotary
and radii encoder
is shown in
and the
Figure 4. tube
Whencontinuously
a tube is bentinserted into theangle,
to a certain shifted
thedie. Bending
tube length radii can be changed
is measured by rotaryat
the option
encoder andofthe thetube
operator for three-dimensional
continuously inserted into bending.
the shifted die. Bending radii can be
When the
changed topoption
at the of theoftube
the is sensed for
operator by three-dimensional
approach switch, bending
bending.commences.
When the topFigure
of the3
illustrates
tube is sensed thebycontrol system
approach of thebending
switch, CNC bender controlled
commences. by personal
Figure computer.
3 illustrates the control
system of the CNC bender controlled by personal computer.
2.3 MOS CNC Bending System
The bending2.3radius MOSRCNC of the tube isSystem
Bending determined by the magnitude of offset u. As
demonstrated in Figure 4, the relationship
The bending radius R of the tube is determined between offsets
by theand radii is a parameter
magnitude of offset u.ofAs
the
type of material. This relationship can be confirmed by bending experiment. Flexural
demonstrated in Figure 4, the relationship between offsets and radii is a parameter of
rigidity, Young’s modulus, and other material properties of the tube will vary with
the type of material. This relationship can be confirmed by bending experiment. Flexural
changes in material and dimensions of the workpiece. Accordingly, the bending is
rigidity, Young’s modulus, and other material properties of the tube will vary with changes
processed. Radius R varies even though the offset u does not change.
inBut
material
this unitandcan dimensions of the workpiece
bend the tubular workpiece.into Accordingly, the bending
a certain bending radius isbyprocessed.
adopting a
Radius R varies
suitable even though
relationship between the uoffset
and Ru does not change.
for that tube, even Butthough
this unitthecan bend the
tubular tubularor
material
workpiece into
dimensions change. a certain bending radius by adopting a suitable relationship between u and
R for that tube, even though the tubular material or dimensions change.
①Tubular workpice ②Bending die

③Guide cylinder
④AC servomotor for control of X axis
⑤AC servomotor for control of Y axis
⑥Motor for Z axis
⑦AC servomotor for control of α angle
⑧AC servomotor for control of β angle
⑨Chain for forward of tube
⑩Plate for warding of tube
⑪Base of Mandrel
Figure 2; Main
Figure components
2; Main of CNC
components bender.
of CNC bender.
11
Dortmund, Germany
Medais
Medais Approach Switch
Medais
Approach Switch Approach Switch
CNCCNC CNC
System
System Encoder
Encoder System Encoder
Counter
Counter Rotary encoder
Rotary encoder Counter
Motor for for
Motor TubeTube Rotary encoder
forwarding
forwarding
Precluded
Precluded Precluded
Mitsubishi
Mitsubishi Mitsubishi
Electronic
Electronic Electronic
Co.,Ltd. ACAC Co.,Ltd.
servo- ACAC
servomotor for for AC servo-
servomotor A
Co.,Ltd. servo-
Amp Controlling of Xofaxis
Controlling X axis
Amp C
Amp
Inter
Inter Inter
ACAC
servo-Amp ACAC
servomotor for for AC servo-Amp
servomotor A
-face
-face servo-Amp -face
Controlling of Yofaxis
Controlling Y axis C
ACAC
servo-Amp
servo-Amp ACAC servomotor
servomotor for for AC servo-Amp A
Controlling
Controlling of α of α C
angleangle an
ACAC servo-Amp
servo-Amp ACAC servomotor
servomotor for for AC servo-Amp A
Controlling
Controlling of β of β C
angleangle an
Figure 3;
Figure Control system of the bender. Figure 3; Control system of the bender.
Figure 3; 3; Control
Controlsystemsystemofofthe thebender.
bender.
TheTherelationship between the penetration lengths
relationship between the penetration lengths of tubes The ofrelationship
tubes Wt Wtand
between
the the
and bending
thebending
penetration
angle θ lengths
angle θ of tubes Wt an
whenwhenthe
Thebending
the radiiradii
relationship
bending change
between is shown
change the in Figure
penetration
is shown when 5. the
5. bending
lengths
in Figure of tubesradiiWt andchange is shownangle
the bending in Figure
θ 5.
Here,
when θ isthe
Here, θalmost
isbending
almostproportionate
radii change
proportionate to is
the penetration
toshown
the Here,
in Figure
penetration lengths
θ5.isHere,
almost
lengths of tubes proportionate
Wt.Wt.
θ istubes
of almost But theto bending
the
proportionate
But the penetration
bendingto the lengths of tube
radii of tubes
penetration
radii of tubesbent by this
lengths
bent method
of tubes
by this Wt.are,
method inthe
Butare, part, radii
not not
inbending
part, equal
of tubes
radii
equalto
ofthe
bent
tubes
to setby
the bending
bent
set this
bymethod
radii.
this
bending are,
These
method
radii. inare,
These part,
in not equal to the s
sections
sections
part, areequal
not known
are to as
known the“incomplete
as
set“incomplete R parts”
bending radii. R These
parts”because
sections
because
sections the
arearetube
known
the is inserted
tube
known as “incomplete
isasinsertedintointo
“incomplete theRthe
die
Rparts”
die because the tube i
parts”
while the the
while
because die istube
thedie moved
is moved toward the the
toward
is inserted into setting
die position.
thesetting while while
theAs
position. theAs
die toisdie
software
toissoftware
moved moved for toward
toward thethe
for CNC
the thebender,
CNC setting
setting bender, position. As to softwar
position.
the the
shape
shapeof the
of bentbent
the tubetubeis decided
is decidedby straight
by theparts
straight shapeIn,ofbending
parts In,the bentradii
bending tube Rn
radiiis Rn
and
decidedbending
and by straight parts In, bendin
bending
As to software for the CNC bender, the shape of the bent tube is decided by straight parts
angles θn, θn,
angles as shown
asradii in Figure
shown 6. 6. angles θn, as shown in Figure 6.
In, bending Rn inandFigure
bending angles θn, as shown in Figure 6.
Figure 4; Relationship between offset Figure 5; Relationship between

Figure 4; Relationship
Figure bendingbetween
4; Relationship
and R.offset
between
radius offset Figure
Figure
5; Relationship
Figure 4; Relationship
5; Relationship
penetration between
length between
tubepenetration
ofbetween offset
bending Figure 5; Relations
penetration
and
andand
bending radius
bending R. R.
radius and
length bending
of tube angle.
and radius
bendingR.angle. length of tub
length of tube and bending angle.
12
Dortmund, Germany
Figure
Figure 6;6; Control
Control system
system of the
of the bender.
bender.
TheThecoordinate
coordinate axes
axesforforthe
theinput
inputare
aredrawn
drawn with solid lines,
lines, and
andthethebending
bendingdirection
direction
forfor n-th
n-th bending
bending is is illustrated.
illustrated. TheThestraight
straightpart
partofofthe
thetube
tubePn-1-Pn
Pn-1-Pnisisoverlapped
overlappeduponupon
thethe
XX axis
axis ofofthethe input
input coordinateaxis
coordinate axisbybyrevolution
revolutionabout
aboutpoint
pointPn-1.
Pn-1.The
Then-th
n-thbending
bending
direction
direction cancan
bebe gotten
gotten byby(n-1)-th
(n-1)-thbending
bendingdirection.
direction.Rn, Rn,θn,
θn,In,In,andθn
andθnare
arerepresented
represented
by the functions given after the next equation. Therefore tubes
by the functions given after the next equation. Therefore tubes can be bent to the requiredcan be bent to the
required
form formof
by means bycontrolling
means of controlling
uxn, uyn, anduxn,Wnuyn,- and Wn - corresponding
corresponding to optionaltoRn,
optional Rn,
θn, In,
θn,
andφn. In, andφn.
Rn = f (uxn, uyn)
Rn = f (uxn, uyn)
θn = f (uxn, uyn, Wn)
θn = f (uxn, uyn, Wn)
I n = f (Wn)
I n = f (Wn)
φn φn==ff(u xn, uyn)
(uxn, uyn)
Where,
Where,
uxn, uyn= X axial and Y axial components of offset for the n-th bent part
uxn, uyn= X axial and Y axial components of offset for the n-th bent part
Wn=forwarded tube length toward Z axis for the n-th bent part.
Wn=forwarded tube length toward Z axis for the n-th bent part. But if the material and
But if the material and dimensions of the tube are changed, the bending radius and the
dimensions of the
bending angle may tubechange
are changed,
even ifthe bending
uxn, uyn, andradius andfigures
other the bending
do notangle mayFunctions
change. change
even
on Rnif uxn, uyn,vary
andθn andwith
otherthe
figures do not
material and change.
dimensions Functions on RnFigure
of the tube. andθn7vary with chart
is a flow the
material and dimensions of the tube. Figure 7 is a flow chart for inputting the
for inputting the bending data. Cross-points of two straight parts, bending radii, and the bending data.
Cross-points
first bending of two straight
direction areparts,
inputbending radii,
first. Then and the first
appropriate bending
offsets and direction
penetrationare lengths
input
first.
areThen appropriate
calculated offsets and penetration lengths are calculated by computer. The
by computer.
equations just above,
The equations obtained
just above, from experiments,
obtained are entered
from experiments, into theinto
are entered computer.
the computer. We
canWe cantubes
bend bendtotubes to complex
complex shapes
shapes as shown as in
shown in Photo2,
Photo2, Photo3 andPhoto3 and Photo4. An
Photo4.
outline of the of
An outline bending machine
the bending movement
machine for carrying
movement out the
for carrying outbending is shown
the bending in thein
is shown
flow
thechart
flowofchart
Figure 7. The bending
of Figure data, obtained
7. The bending from thatfrom
data, obtained program, are fed into
that program, are fed into
13
Dortmund, Germany
thethe computer.
computer. TheThe motormotorforfor tube tube formation
formation is activated,
is activated, andand thethetubetube
setset in the
in the bending
bending
machine
machine is is pushed
pushed toward
toward Z Z axis.axis.When When thethe tubetubeis is inserted
inserted intointo thethedie,die,thethe
microcomputer
microcomputer transmits
transmits pulsespulses to the
to the ACAC servomotors
servomotors to control
to control X Xandand Y axes.
Y axes. TheThe
the
number
the computer.
number of pulses
computer. The
of pulses
The motor
depends
depends
motor for
on tube
for tube
on
offset formation
offset magnitude.
formation is activated,
magnitude.
is activated,
TheThe and
AC
ACand the tube
tube set
servomotors
servomotors
the in the
inare
setare then
the bending
then activated
activated
bending
andmachine
and
thethe
machine is
dieisdie pushed
is shifted.
ispushed
shifted. toward
During
During
toward Z axis. Z axis.
the
theWhenbending When
bending
the tube the
recess,
recess, tube
thethe
is inserted is inserted
penetration
penetration
into into the
thelength
the die,length die,
of ofthethe
microcomputer thetube
tube is is
microcomputer transmits pulses to the AC servomotors to control X and Y axes. The
read
read by by
thethe
transmits rotary
rotary
pulses encoder
toencoder
the ACAnd And thethe
servomotors diedie is
istofixedfixed
control at and Y axes.
at the
X the stipulated
stipulated Theoffset
offset position
position
number while
of while
pulses thethe
number of pulses depends on offset magnitude. The AC servomotors are then activated
necessary
necessary
depends on length
length
offset of tube
ofmagnitude.
tube is The
inserted.
is inserted. AC servomotors are then activated and the die is shifted.
and the die is shifted. During the bending recess, the penetration length of the tube is
After
After
During
read bythe
the therequired
required
the
rotarytube
bending tube
recess,
encoder length
length Andhas has
the penetration
the diebeen
been pushed
ispushed
length
fixed theout
at out
of the fromfrom
tube isthe
stipulated the
die,
read die,
by
offset
theAC
AC servomotors
while the areare
servomotors
rotary
position
encoder
again
And
again the activated
die
activated is And And
fixed at
the the
the
die
necessary length of tube is inserted. die
is is
stipulated shifted
shifted offset
to to
the the next
position
next offset
while
offset position
the
positionnecessary
for for
the the
length ensuing
ensuing of tube bending
is
bending
inserted.
operation.
operation.
After theThese These
required procedures
procedures
tube length arearehas repeated
repeated in regular
in regular
been pushed out order.order.
from After
the After
die, bending
bending
AC is is
servomotors completed,
completed,
are
the
theagainAfter
motor
motor is the required
And thetube
is automatically
automatically
activated die length
isturned
turned hasoff.
off.
shifted been
to the pushed
next offsetout from the for
position die,the
ACensuing
servomotors are
bending
again activated
operation. TheseAnd the die isare
procedures shifted to the
repeated in next offset
regular position
order. After for the ensuing
bending bending
is completed,
operation. These procedures are repeated
the motor is automatically turned off. in regular order. After bending is completed, the
motor is automatically turned off.
Photo
Photo :2 :2
Photo :2
Photo 2
Photo
Photo:3 Photo :4:4 :4
Photo
Photo 3 :3
:3 Photo
Photo Photo 4
14
Dortmund, Germany
Figure 7; Flow chart of control for MOS bending.

3. CNC Bending Program
The3.bending
CNC Bending Program
system and software are input into the computer of the MOS bending
3. CNC
machine, Bending
enabling usersProgram
to produce complex bent tube products easily, different from a
The bending
conventional CNCsystem
bender.and software
Each input are inputshape
bending into the computerby
is produced of means
the MOS bending
of an
The bending
individual
machine, systemusers
program.
enabling and software
Figure givesare
to 8produce theinput intobent
six-process
complex thebending
computer
tube of theeasily,
program
products MOS
to bending
produce the from
different
amachine,
piece enabling
shown
conventional in theusers
CNC to produce
Photo.5. TheEach
bender. complex
scheme bent tubeshape
is as bending
input follows: products
is easily,
produceddifferent from a of an
by means
conventional CNC bender. Each input bending shape is produced by means of an
individual program. Figure 8 gives the six-process bending program to produce the piece
BENDING
individual PROGRAM
program. Figure 8 gives the six-process GO1bending
= startprogram to produce the
the process
shown in the Photo.5. The scheme is as follows:
piece shown in the Photo.5. The scheme is as follows: L200 = 200mm straight length
G01 L200 R40 T90 P90 F25 E25
R40 = 40mn bending radius
BENDING
R130 T90 PROGRAM
P90 GO1
R40 T90 P180 T90 = = start the process angle
90-deg.Bending
L200
P90 = bending
= 200mm straight
direction length
G01
R40 L200
T90 P180R40 T90 P90 F25 E25
R40
F,E == moving
40mn bending
speeds radius
of the die
R130
R130 T90
T90 P90P90
R40 T90
R40 T90 P180
P90 T90 = of
MO2=end 90-deg.Bending
the bending operationangle
M02 T90 P180
R40 P90 = bending direction
R130 F,E = moving speeds of the die
Figure T90 P90 program for production of tube
8;: Typical products by means of MOS banding.
R40 T90 P90 MO2=end of the bending operation
M02
Figure 8;: Typical program for production of tube products by means of MOS banding.
Figure 8; Typical program for production of tube products by means of MOS banding.
15
Dortmund, Germany
Photo 5 Photo
Photo :5 :5
4. CAD
4. and
CADCNC
andBending
CNC Bending
4. CAD and CNC Bending
It is difficult for us for
It is difficult to make
us to the
make program of required
the program bent shapes
of required of the of
bent shapes tubes
the without a
tubes without a
It is difficult
long time for
andusexcellent
to make skill.
the program you of required the bentshape
shapes of
thethe tubes without
longpractice
time practice and excellentWhen skill. Whendesignyou design the ofshape bent
of thetubes
bent by tubes by
a longthe
CAD, timeCAD practice and excellent skill. When to you designbent the shape of the by abent tubes by
CAD, the data
CADcan data becan
easily changed
be easily changed the to
CNC the CNCprogrambent program computer.
by a computer.
CAD, 9theshows
Figure CAD the dataexample
can be easily changed
of virtual to theofCNC
pictures the ofbent tubes
bent program by theaon computer. Figure
Figure 9 shows the example of virtual pictures the bent on tubes display of CNC
the display of CNC
9 shows
machine the example
control panel of virtual
at three pictures
directions of the bent
coordinates.tubes on the
Therefore, display
machine control panel at three directions coordinates. Therefore, everyone can easily everyoneof CNC can machine
easily
control
check thepanel
check bent attube’s
the three
bent directions oncoordinates.
shapesshapes
tube’s theon display Therefore,
of CNC
the display everyone
of bending
CNC can easily
machine.
bending checkPhoto.6
Photo.6
machine. the bentis the
is the
tube’spicture
picture shapes
of bentof on
tube the display of
corresponding CNC to bending
Fig.9.
bent tube corresponding to Fig.9. machine. Photo.6 is the picture of bent tube
Ifcorresponding
the If
chairs of tocomplex
the chairs Fig.9. and three
of complex and dimensional
three dimensionalshapesshapesare made are bymade thebyconventional
the conventional
CNC If the chairs
bending
CNC of machine,
machine,
bending complex and
the bending three dimensional
process
the bending shapes
is veryisdifficult
process are
andmade
very difficult manyandby the of
kinds
many conventional
complex
kinds of complex
CNCare
dies bending
dies are machine,
needed. Many the
needed. days
Manybending
are
days process
required is very
for
are required difficult
preparing
for and
preparingandarranging
many kinds
and arrangingthe ofdiescomplex
theanddies and
bending
dies are process.
bending But ifdays
needed.process.
Many the are
But new
if bending
required
the formachine
new bending preparing isand
machine used is toused
arrangingmake tothecomplex
dies and
make shaped
bending
complex shaped
chairs andBut
process.
chairs exteriors,
if the
and the CAD
new
exteriors, the data
bendingCAD isdata
inputis to
machine is the
usedcomputer
input to to
themake and can
complex
computer andbeshaped
easily
can produced
chairsproduced
be easily and
after a after
shortthe
exteriors, abending trial.is trial.
CADbending
short data input to the computer and can be easily produced after a short
bending trial.
Photo Photo
6: Sample of bentoftube
6: Sample bent tube
Photo 6: Sample of bent tube
16
Dortmund, Germany
Figure 9; Virtural pictures of bent tube

Figure 9; Virtural pictures of bent tube
5. Benefits of tube bending by MOS method

5. Benefits of tube bending by MOS method
There are many benefits of tube bending by MOS shown as follows.
There are many benefits of tube bending by MOS shown as follows.
(1) Bending radius can be easily set.
(1) Bending radius can be easily set.
Unlike the conventional bending by pressing a tube, rolls or bending dies are not
Unlike the conventional bending by pressing a tube, rolls or bending dies are not used
used any longer with this model so that desired radius can be specified freely.
any longer with this model so that desired radius can be specified freely.
(2)(2) Deformation
Deformation ratio
ratio of the
of the tube
tube cross-section
cross-section is minimized.
is minimized.
Unlike the
Unlike the conventional
conventional type,
type, the
the bending
bending section
section is
is securely
securely guided
guided by
by the
the die
die so
so that
that accurate
accurate roundnessroundness can be obtained
can be obtained on thesection.
on the cross cross section.
(3)(3) Bending
Bending radius
radius cancan
be be continuously
continuously changed.
changed.
Radius and
Radius and directions
directions (X/Y)
(X/Y) can
can be
be easily
easily controlled
controlled byby computer.
computer.
(4)(4) Versatile
Versatile bending
bending forms
forms areare available.
available.
Bending at an angle
Bending angle ofofmore
morethanthan180degree,
180degree,large-radius
large-radiusbending, andand
bending, spiral-form
spiral-form
bending
bending areare available
available now.now.
(5)(5) Sophisticated
Sophisticated bending
bending formform is obtained.
is obtained.
As there is no
As no difference
differenceininlevels
levelsininthethebending
bendingsection, a smooth
section, andand
a smooth even finish
even finish
cancan
be be
achieved.
achieved.
(6)(6)
Operability
Operabilityof various functions
of various functionsis improved.
is improved.
17
Dortmund, Germany
Inputting the coordinates of bending section is also available besides ordinary

inputting. Quiet environment is ensured because of motor-driven operation at high speed.
6. Conclusions
The many advantages offered by CNC bender based on the MOS method of computer
control of the relative die’s positions of die and tubes are being realized by users in Japan,
Europe, and the United States.
The many excellent MOS-generated products include seat, exterior pieces such
as monuments and buffer stops, and tubing to use in refrigerators, bicycles, and other
complex designs.
By means of programs of utmost simplicity, this new generation of benders makes
it available to meet needs of the most demanding and complex tube making operation.
REFERENCES
[1] Takahashi, A.; Japan Society for Techno. of Plasticity(JSTP),106th Symposium-

1986.
[2] Ochiai, I. ; JSTP Journal 23-255,1982,p.290.
[3] Tagaki, R.; JSTP Journal 23-255,1982,p.282.
[4] Yamatake, H.; JSTP Journal 10-103,1969,p.618.
[5] Takahashi, K.; Journal of Japan Light Metals, 5-56,2006,p.283.
[6] Murata, M. ; Conference Proceeding, International Tube Association,1998,p.329.
18
Session 1: Nonconventional Processes
Dortmund, Germany
Machine Concept for Incremental Tube Forming

Christoph Becker*, Matthias Hermes, A. Erman Tekkaya
Institute of Forming Technology and Lightweight Construction,
Baroper Straße 301, 44227 Dortmund, Germany
*Christoph.Becker@iul.tu-dortmund.de
Abstract: Incremental Tube Forming (ITF) is a recently developed process to manufacture

three dimensional bent tubes with variable diameter. The process principle as well as the
currently used experimental setup is described. Furthermore experimental results which
emphasize the advantages compared to conventional bending processes are mentioned.
These results will lead to the development of a new machine for Incremental Tube Forming.
Keywords: bending, tube forming, incremental, spinning
1. INTRODUCTION
In many industrial sectors the demand of three-dimensionally bent tubes increases

because of lightweight design as well as space and cost saving (Figure 1). This can be
seen in the automobile and aerospace industry as well as in civil engineering, where three-
dimensionally bent structures are oftentimes required [Chatti, 2005].
Figure 1: Lightweight structures [Chatti, 2005]
21
Dortmund, Germany
The
The aim
aim of
of recently
recently developed
developed processes
processes is
is to
to satisfy
satisfy especially
especially the
the lightweight design
lightweight design
for structures but alsobut
for structures to keep thekeep
also to stiffness constantconstant
the stiffness or to increase it. This can
or to increase be realized
it. This can be
by usingrealized
tubes withby varying crosswith
using tubes sections (e.g. cross
varying Tailored Tubes).
sections ForTailored
(e.g. example,Tubes).
these tubes
For
can currently be produced
example, with
these tubes canhydroforming [Koc, 2009].
currently be produced withTo manufacture[Koc,
hydroforming Tailored Tubes
2009]. To
manufacture
the Incremental TubeTailored
FormingTubes
(ITF) the
wasIncremental Tube Forming
developed [Hermes, 2009].(ITF) was developed
[Hermes, 2009].
2. PROCESS PRINCIPLE
2. PROCESS PRINCIPLE
With ITF, bent tubes with varying cross sections can be manufactured due to the
With ITF,
combination of abent tubesand
bending witha spinning
varying cross sections
process. During can
thebeprocess
manufactured
a pusherdue to the
transports
the tubecombination of a to
through a sleeve bending and a spinning
the spinning tool wherebyprocess. During
the tool the process
is consisting outaofpusher
three
transports the tube through a sleeve to the spinning tool whereby the tool is
spinning rolls. The spinning tool is rotating around the tube and is reducing the diameter
consisting out of three spinning rolls. The spinning tool is rotating around the tube
of the tube
and due to the infeed
is reducing of the of
the diameter spinning
the tuberolls.
due To varyinfeed
to the the diameter of the tube,
of the spinning rolls. the
To
infeed of thethe
vary spinning
diameterrolls
of can
the be changed
tube, accordingly.
the infeed To produce
of the spinning rollsthe
canbent tubes, a
be changed
bendingaccordingly.
process is superposed
To produce to the
the spinning process.
bent tubes, To realize
a bending processthreeis dimensionally
superposed to bent the
structures, the tube
spinning can be To
process. rotated.
realizeDue to the
three free movability
dimensionally bentofstructures,
the spinning the rolls, as well
tube can be
as the bending
rotated. head,
Due tofree formed
the free tubes regarding
movability tube diameters
of the spinning as as
rolls, as well welltheasbending
the bending
head,
line canfree formed tubes regarding tube diameters as well as the bending line can be
be manufactured.
manufactured.
Figure 2: Process principle of ITF process
The process
The process is
is currently
currently realized
realized on
on aa rebuilt
rebuilt lathe
lathe to
to generate
generate first
first experimental
experimental
knowledge
knowledge about
about the newthe new process
process. These [Becker, 2011].
experiments These
show experiments
a great advantageshow a great
of the ITF
process:advantage
Due to theofstress
the ITF process: Due
superposition to the stress
of spinning superposition
and bending of spinning
operation a drop in and
the
bending operation a drop in the bending forces can be measured. This is shown in
bending forces can be measured compared to conventional bending processes. These low
Figure 3.
bending forces can be used for an improved and smarter machine layout.
22
Dortmund, Germany
3. NEW MACHINE CONCEPT
The current aim is to realize a more realistic machine system with the intent of
establishing an industrial standard of the new ITF process. For this, a new machine was
developed and designed. It will have eight NC-controlled axes for the different tools of the
process. It will be possible to process tubes up to a diameter of 90 mm. The adjustment of
the spinning system will be possible during rotation and an additional mandrel will extend
the process limits significantly. In Figure 3 the design of the machine is shown [Hermes,
2011].
Figure 3: Machine concept and CAD-plot of new machine [Hermes, 2011]

Figure 4: Machine concept and CAD-plot of new machine
4. CONCLUSION AND OUTLOOK

4. CONCLUSION AND OUTLOOK
Theexperimental
The mentioned experimental
results showresults show theofpotentials
the potentials of the ITF
the ITF process. Thisprocess.
can be This can be
especially
especially
seenseen in the
in the lowlow bending
bending forces
forces which
which are are resulting
resulting fromfrom
the the stress
stress superposition
superposition of
of the spinning and bending
the spinning operation.
and bending Based
operation. on these
Based results
on these a new
results a machine
new machinewith with
eight
numerically
eight controlled
numericallyaxes has been
controlled developed.
axes has been This machine
developed. will
This enlargewill
machine theenlarge
application
the
area of application
the processarea
and will exploit
of the the and
process maximum potential
will exploit the of ITF. Within
maximum the next
potential steps
of ITF.
further Within the next
experimental as steps further
well as experimental
numerical as well will
investigations as numerical investigations
be conducted. willto
In addition
that an be conducted.
analytical In addition
approach to that
will be an analytical approach will be developed.
developed.
REFERENCES
[Becker, 2011] Becker, C., Hermes, Tekkaya, A.E.: New flexible Tube Forming Processes
-Tool Independent Forming of Tubes to Complex Workpieces with Workpiece-
Invariant Tools; Pipe and Tube World Conference, Düsseldorf, 2011.
[Chatti, 2005] Chatti, S.: Production of Profiles for Lightweight Structures, habilitation
thesis, Books on Demand GmbH, Norderstedt, 2005.
[Hermes, 2009] Hermes, M., Kurze, S., Tekkaya, A.E.: Verfahren und Vorrichtung zur
Umformung eines Stangenmaterials (process and device for the forming of tube
material), German Patent Application, DE102007046870A1, 2009.
[Koc, 2009] Koc, M., Altan, T.: An overall review of the tube hydroforming (THF)
technology, Journal of Materials Processing Technology, Volume 108, Issue 3,
p. 384-393, 2009. 23
Dortmund, Germany
REFERENCES
[Chatti, 2005] Chatti, S.: Production of Profiles for Lightweight Structures,

habilitation thesis, Books on Demand GmbH, Norderstedt, 2005.
[Hermes, 2009] Hermes, M., Kurze, S., Tekkaya, A.E.: Verfahren und
Vorrichtung zur Umformung eines Stangenmaterials (process and device for the
forming of tube material), German Patent Application, DE102007046870A1,
2009.
[Hermes, 2011] Hermes, M., Staupendahl, D., Becker, C., Tekkaya, A.E.:
Innovative Machine Concepts for 3D Bending of Tubes and Profiles, Key
Engineering Materials, Volume 473, p. 37-42, 2011.
[Koc, 2009] Koc, M., Altan, T.: An overall review of the tube hydroforming
(THF) technology, Journal of Materials Processing Technology, Volume 108,
Issue 3, p. 384-393, 2009.
24
Dortmund, Germany
Schwarze-Robitec GmbH
Jürgen Korte
Andreas Siems
Mandrelless-Bending
Residence
Schwarze - Robitec GmbH
Olpener Strasse 460-474

51109 Köln / Germany
Tel: +49 221 890 080

Fax: +49 221 890 0829
E-Mail: sales@sch-r.com
Web: www.sch-r.com
25
Dortmund, Germany
History
1903 Foundation of Gebrüder Schwarze
since 1930 Manufacture of tube bending machines (Peter Wirtz)
1953 Acquisition of Fa. Peter Wirtz, since then Schwarze-Wirtz
1962 First booster bender in the world developed together with
VGB and Babcock and delivered to Sweden
1965 Supply of first NC controlled automatic tube bender
1996 Merger with Robitec, since then Schwarze-Robitec
Portrait
Customer orientated, qualified and motivated staff

are the basis for technological advantage and
quality.
Decades of experience in the manufacture and

maintenance of tube & pipe cold bending machines.
Continuous new development and usage of most
advanced machinery is a matter of course.
Turnover 2010: Đ 20 million
Number employees: approx. 120
Surface area: approx. 43.000 m2
Workshop area: approx. 12.000 m2
26
Dortmund, Germany
Our experience for your future
All our machines are designed in such a

way, that they are ready to bend the newest
materials used in the latest developments
of (nuclear) power plants.
Booster Bending Machines
• In boiler industry people bent the products with the compression bending
system with a minimum bending radius of R ~ 2 – 2.5 D.
• In 1962 Schwarze-Robitec developed on request of VGB and Babcock the

first booster bending machine.
• Booster benders are suitable for cold bending of tubes without a mandrel
and bending radii less than R = 2D, for tube serpentines and single bends in
high strength and heat resistance steels in the same way as with thick-
walled high-pressure pipes.
27
Dortmund, Germany
Advantages of the
BOOSTER BENDING SYSTEM:
Manufacture of Tight Bending Radii:
Booster bending system facilities the manufacture of very tight bending radii
in thick-walled tubes (OD / t ~ 10) without mandrel down to R = D within the
tolerances according the former VGB Norms and the latest EN 12952-5
Norm established in 2000.
This enables the manufacture of boilers in smaller dimensions with higher

capacity.
Types of machines
Max. capacity at Max. section Min. tube OD Min. / max. CLR
Model 700 N/mm² modulus
mm cm³ mm mm
SR 60 DB 63,5 x 5 12,5 25 20 / 300
SR 80 DB 88,9 x 8 37,8 25 30 / 400
SR 100 DB 114,3 x 8,6 70,2 30 50 / 450
SR 165 DB 168,3 x 12,5 222 51 70 / 500
CNC 60 DB 63,5 x 5 12,5 25 20 / 200
CNC 80 DB 88,9 x 8 37,8 25 30 / 250
CNC 100/80 DB 88,9 x 11 46,9 30 30 / 350
CNC 100 DB 114,3 x 8,6 70,2 30 50 / 350
CNC 165 DB 168,3 x 12,5 222 51 70 / 425
Auch als TWIN lieferbar
28
Dortmund, Germany
Types of machines
Model Max. capacity at Max. section Min. tube OD Min. / max. CLR
700 N/mm² modulus
mm cm³ mm mm
SR 60 DB 63,5 x 5 12,5 25 20 / 300
SR 80 DB 88,9 x 8 37,8 25 30 / 400
SR 100 DB 114,3 x 8,6 70,2 30 50 / 450
SR 165 DB 168,3 x 12,5 222 51 70 / 500
CNC 60 DB 63,5 x 5 12,5 25 20 / 200
CNC 80 DB 88,9 x 8 37,8 25 30 / 250
CNC 100/80 DB 88,9 x 11 46,9 30 30 / 350
CNC 100 DB 114,3 x 8,6 70,2 30 50 / 350
CNC 165 DB 168,3 x 12,5 222 51 70 / 425
CNC 60 DB
29
Dortmund, Germany
CNC 100 - 80 DB with Flip-Over-Table
Optional:
For thin walled tubes:
-Mandrel retraction device
For burner openings pipes:
-Sawing device
-Bend-in-Bend device
-1D / 3D device
30
Dortmund, Germany
CNC 80 DB
with saw
1 D / 3 D execution
31
Dortmund, Germany
Special tools
1D/3D
1D/3D
For service and reparation shops

or small production units
SR 100 DB
32
Dortmund, Germany
CNC 165 DB
Tools
33
Dortmund, Germany
Tools
Bend quality and Bending Tolerances:
When bending without mandrel the booster bending system also guarantees
to meet the tolerances for wall thinning and ovality according to the former
V. G. B. prescriptions and the latest EN 12952-5 European Standard.
The following formula are used since 2000:
eext = eact (2 rb/dn + 0.5) / (2 rb/dn + 1), where
eext is the required minimum thickness at the extrados, in mm;

eact is the nominal thickness of the supplied tube minus the supplier’s
maximum negative thickness tolerance, in mm;
rb is the radius of the bend measured to the centre-line of the tube in mm;
dn is the nominal outside diameter of the tube, in mm.
34
Dortmund, Germany
Bending Tolerances:
The departure from circularity of the tube bends shall be calculated from the
equation:
u = 2 x (Do – do) / (Do + do) x 100%, where
u is the departure from circularity, in %;
Do is the maximum outside diameter measured at the tube bend apex, in

mm;
do is the minimum outside diameter measured at the same cross-section as

Do, in mm.
The permitted departure from circularity shall be within the limits of:
20 / (rb / dn) with a maximum of 10% for 1  (rb / dn )  2.
35
Dortmund, Germany
R=D
Solutions for
large serpentines
Serpentines up to approx. 10 - 12m

CNC 100 DB with Flip-Over-Table
CNC 100 DB + SR 80 DB with Flip-Over-Table
Serpentines above 10 m
CNC 80 DB-TW with Transport Carriage
CNC 100 DB-TW with Transport Roll device
CNC 100 DB-TW-2R
CNC 100 DB-TW-IH or OB
CNC 100 DB-TW-IH + OB
36
Dortmund, Germany
Flip-Over-Table
Magazine
with
Loading device
37
Dortmund, Germany
Video
Magazine with automatic loading
CNC 80 DB + SR 80 DB + Flip-Over-Table
38
Dortmund, Germany
CNC 80 DB + SR 80 DB + Flip-Over-Table
SR 80 DB on an elevator
CNC 60 DB-T
in production
39
Dortmund, Germany
Options
40
Dortmund, Germany
CNC 100 DB-TW-IH+OB

same bending direction + Induction Heating +
Offset Bending + Flip-Over-Table
U-Form Induction Heater
41
Dortmund, Germany
Offset Bending
Schwarze-Robitec
Membrane Tube Wall Bending Machines
• Our FL-machines are especially designed for bending of panels with

tubes up to max. Ø 114.3 mm
• Machine sizes for panels with a max. width of 1000, 2000, 3000 or
even 3600 mm
42
Dortmund, Germany
Membrane wall-production
diagonal
and
straight
bending
43
Dortmund, Germany
FL 2000
FL 3000
FL 3000
Maximum
bending
capacity:
114.3 mm
Maximum wall
width:
3000 mm
44
Dortmund, Germany
Control FL machines
End of
presentation
45
Dortmund, Germany
46
Session 2: Standardization in Bending Technology
Dortmund, Germany
VDI 3430: A Guideline for Rotary Draw Bending of

Profiles
Engel, Bernd; Hinkel, Markus; Kersten, Sebastian; Kopfer, Peter;
Kuhnhen,Christopher; Mathes, Christian
Chair of Forming Technology,
University of Siegen, Paul-Bonatz-Str. 9-11, 57076 Siegen
uts-sekretariat@uni-siegen.de
The guideline VDI 3430 deals with definitions and descriptions for the rotary draw
bending of profiles. It provides recommendations for the design of bent components and
shows the process boundaries as well as attributes of the rotary draw bending process.
Furthermore designations of tools and machine axes are defined. Separate chapters are
dedicated to the bending tools, failure and typical influences on components during the
bending process as well as the dimensioning and measuring of bent profiles. An overview
of the contents of the guideline is shown in Table 1.
Table 1; Content of guideline VDI 3430

Chapter Content Inhalt
1 Preliminary note Vorbemerkung
2 Process description Verfahrensbeschreibung
3 Definitions and designations Begriffe und Benennungen
4 Bending tools Biegewerkzeuge
5 Semi-finished profile Biegehalbzeug
6 Bent part and process boundaries Biegeteil und Verfahrensgrenzen
7 Dimensioning of bent parts Bemaßung von Biegeteilen
8 Measurement Messen und Prüfen
9 Optimized design of bent Konstruktionshinweise zum
components „Biegegerechten Konstruieren“
The guideline was developed by a group of experts attending the “Expertenworkshop”,

which is a part of the annual symposium „Biegen in Siegen“; organized by the Chair
of Forming-Technology of the University of Siegen. The participants were machine
manufactures, tool manufactures, profile fabricators, profile bending appliers and
acadamics. Thus, the knowledge of theory and practice is combined in the group “VDI-
GPL-Fachausschusses 113 – Biegetechnik”.
Another aim of this guideline is the establishment of a shared vocabulary. The
communication between machine manufacturers and users as well as customers needs to be
simplified. In the future, misunderstandings due to different designations can be avoided.
An increase of the efficiency in production and communication is the consequence.
49
Dortmund, Germany
Therefore, the time period between design and manufacturing of products will be
reduced.
With the VDI Guidelines the VDI (Verein Deutscher Ingenieure) fulfils its primary
function: the transfer of technical knowledge as a service to engineers and students.
Everyone may suggest a topic for a VDI guideline. If a committee is established after
examination of the appropriate VDI division, it consists of honorary experts coming
from all areas of research and teaching, planning and development, industry, technical
surveillance and authorities. The first result of this professional exchange of experience is
a draft, which is submitted to a public scrutiny. The Guideline VDI 3430 is published with
the status “green paper“. After examination of the comments received, the final version
called “white paper” will be published. This guarantees, besides being the state of the
art, neutrality with regard to individual commercial interests as well as relevance and
practicability. The “white paper” is usually published as a German/English version. [VDI,
2011]
For further information please visit: www.biegeninsiegen.de.
REFERENCES
[VDI, 2011] Verein Deutscher Ingenieure e.V.: „Wissenswertes zum VDI -

interesting facts about the VDI“
CO-AUTHORS
Becker, C. IUL TU-Dortmund Eckold, H.-H. Eckold GmbH & Co.

KG
Burkhardt, B. Firma B Burkhardt Gerlach, Dr. C. Tracto-Technik GmbH
& Co. KG
Grote, W. Carl Froh GmbH Haldenwang, U. MEWAG
Maschinenfabrik AG
Hermes, M. IUL TU-Dortmund Karwowski, M. Schlüter- tools GmbH
& Co. KG
Kibben, M. Thyssen Krupp Steel AG Kurz, J. Felss Burger GmbH
Riedel, M. Schuler Hydroforming Rosenberger, G. Rosenberger AG

GmbH & Co.KG
Schröder, M ESI GmbH Selter, O. UTS Universität
Siegen
Stahl, Dr. D. Gestamp Automoción Plettke, Dr. R. LFT Universität
S.L. Erlangen-Nürnberg
Reisinger, M. EDAG GmbH & Co. Tekkaya, Prof. IUL TU-Dortmund
KGaA Dr.-Ing. A. E.
50
Dortmund, Germany
Standard for tube and profile bending technology

Christian Gerlach; Matthias Hermes; Christopher Kuhnen; Christian Mathes;
Michael Rohrmann; Bernd Engel; A. Erman Tekkaya
TRACTO-TECHNIK GmbH & Co. KG, Werk 1 – Zentrale, Reiherstraße 2, 57368
Lennestadt, christian.gerlach@tracto-technik.de
Abstract: Bending of tubes and profiles is gaining more and more importance in all in-
dustrial branches. The manufacturing of a bending profile requires a broad experimental
knowledge. Universität Siegen, Technische Universität Dortmund and Tracto-Technik
GmbH & Co. KG are working on a cross-industrial standard, which can be used equally
by suppliers of semifinished products, users, and bending experts and facilitates the design
of the bending part.
Keywords: Standardisation, tube bending,
1. Starting situation
Bending of profiles and tubes as construction elements is gaining more and more
importance in the fields of automobiles, rail vehicles, chemical plants, materials handling,
and the furniture industry, especially regarding lightweight construction. Here, it is
necessary to manufacture precisely bent components, and, due to new fields of application,
ever more complex shapes are called for. The selection of the bending process and the
corresponding parameters for the manufacturing of the tools and the machine setting
depends on numerous factors: the customers´ demands, the fields of application of the
product, given tolerances, the materials as well as the economic frame conditions. As a
result of the high complexity, the problem can often only be solved by trial-and-error
method i.e. by cost and time intensive test production for each individual bending task.
This leads to waste of time and materials, which is disadvantageous to the application
of innovative bending techniques. The producers of bending components are often not
familiar with the techniques, machines and tolerances. This leads to misunderstandings
and, thus, needless iterative loops during the production enquiry until the appropriate
design data regarding the choice of the semifinished product, the bending contour,
and the bending process is achieved. Here, a cross-industrial standardization can be
helpful, providing a general minimum standard for tube and profile bending parts i.e. a
standardization of profile bending parts.
2. Project aim
For the first time, in this joint project, an industrial standard for profile and tube
bending is developed. This serves as the specific draft which will be attached to a
standardization claim at the German Standardization Institute (DIN). The claim will be
submitted to the standards committee „Technical Fundamentals“with the aim to achieve a
European and/or international standardization (ZEN/ISO).
51
Dortmund, Germany
In future, with the help of this new

cross-industrial standard, it shall be
possible to identify the optimal bending
process for profile and tube bending
parts. The bending process shall be
selected considering specified demands
like fields of application, materials,
and economic frame conditions. In the
project, these comprehensive data will
be entered into a selection matrix (see
Fig. 1), which will help the user make
a decision by means of a convenient
software tool. In addition to the basic
Fig. 1: Decision matrix for selecting a profile
technical parameters for the design of
and tube bending technique.
the individual process, a validation of
the feasibility shall be provided.
Concerning the profiles, the matrix will be developed by experts from Technische
Universität Dortmund, Institute of Forming Technology and Lightweight Construction
(IUL) and Universität Siegen, UTS, Chair for Forming Technology. The machine and plant
manufacturer Tracto-Technik GmbH & Co. KG will support the project as an expert (in
the field of roll and rotational bending) regarding economic and market relevant aspects.
3. Implementation
The project will be implemented in several steps. First, a study will examine the
standards for semifinished products and bent parts, which are partly very industry-
specific, and which shall be incorporated into the new standard. For example, in the field
of chemical plant building, elbows are standardized but, quite often, a corresponding
bending process is not or only insufficiently specified. The second step of the project will
be an examination of the capabilities of the different bending processes. For this purpose,
comprehensive market studies will be carried out, which, in case of doubts, will be backed
by feasibility studies and simulations and which will deliver first results regarding the
design criteria for the bending process. After this, the results will be entered into a data
bank, which will serve as the basis for the decision matrix and will be incorporated
into the standard. At the end of the project, a decision tool will be available which can
facilitate the co-operation between designers, builders, and manufacturers and, hence, help
save resources in the development of bent parts from tubes and profiles. The project is
scientifically based and is financially supported by the Federal Ministry of Economics and
Technology (Bundesministeriums für Wirtschaft und Technologie) on the grounds of a
decision by the German Bundestag (Deutscher Bundestag).
52
Session 3: Profile Manufacturing
Dortmund, Germany
An
AnAnalysis
analysis of Bend Allowanceprediction
bend allowance Predictionin
infolding
Folding
Dipl.-Ing.Philip
Dipl.-Ing. PhilipBeiter*,
Beiter*,M.
M.Sc.
Sc.Alexander
AlexanderHassis,
Hassis,
Prof.Dr.-Ing.
Prof. Dr.-Ing.Dipl.-Wirtsch.-Ing.
Dipl.-Wirtsch.-Ing. Peter
PeterGroche
Groche
Institutefor
Institute forProduction
ProductionEngineering
EngineeringandandForming
FormingMachines,
Machines,
Petersenstrasse30,
Petersenstrasse 30,64287
64287Darmstadt,
Darmstadt,Germany
Germany
**beiter@ptu.tu-darmstadt.de
beiter@ptu.tu-darmstadt.de
Abstract: The design process for bent parts includes more than the pure determination
ofAbstract: The design
the final cross process
section. It alsoforcovers
bent parts includes more
the calculation thannecessary
of the the pure determination
precut sheet
of respect
with the finaltocross section.
the bend It also In
allowance. covers the calculation
literature, of the necessary
several calculation precut
methods are sheet
known,
withofrespect
each to the on
them based bend allowance.
different In literature,
assumptions several calculation
or experimental methods
work. As are known,
a consequence,
each of them
considerable based on different
differences in resultsassumptions or experimental
may be obtained. Modern work.
CAD As a consequence,
software usually
considerable
includes differences
miscellaneous in results
methods may to
in order be support
obtained.
theModern
design CAD software
process, but it usually
is the
includes task
engineer’s miscellaneous
to choose methods in orderapproach.
the appropriate to support thepaper
This design process,two
analyzes butcommon
it is the
engineer’s
methods taskcalculation
for the to choose of thethe
appropriate approach.
bend allowance This paper their
by comparing analyzes twowith
results common
data
methods for the calculation
obtained from folding experiments. of the bend allowance by comparing their results with data
obtained from folding experiments.
Keywords: folding, unfolded length, bend allowance, stainless steel
Keywords: folding, unfolded length, bend allowance, stainless steel
1. INTRODUCTION
1. INTRODUCTION
WhenWhen designing
designing bent
bent parts,
parts, thethe determination
determination of the
of the final
final cross
cross section
section is only
is only the the
first
first step
step ofof
thethedesign
designprocess.
process.The Thesecond
secondstepstep consists
consists of of the
the calculation
calculationof ofthe
thenecessary
necessary
precut
precutsheet.
sheet.TheThe unfolded length
unfolded L0 isLusually
length less than the length of the geometric middle
0 is usually less than the length of the geometric
plane,
middlesince the shift
plane, sinceofthe
theshift
neutral fibre,
of the in terms
neutral of bend
fibre, allowance,
in terms of bendhas to be considered.
allowance, has to be
Inconsidered.
literature, several bend allowance
In literature, several bendcalculation
allowancemethods are described.
calculation methods Eachareofdescribed.
them is
based
Eachonofdifferent
them is based assumptions or experimental
on different assumptions work.
or experimental work.
InInGermany,
Germany,DIN DIN6935
6935[DIN [DIN6935,
6935,2010]
2010]isiswidely
widelyused.
used.ItItassumes
assumesthat thatthe
thebent
bent
part
partmay
may bebeseparated
separated into
intoa finite
a finitenumber
numberofofstraight
straightsections
sections(a(aandandb)b)andandcircular
circulararcs
arcs
(d)(d)
asasshown
shown ininFigure
Figure1 1a).a).The
Thecircular
circulararcarcsections
sectionsareareassumed
assumedtotohave haveconstant
constantradii.
radii.
InInFigure
Figure1 a),
1 a),ri rdescribes
i describes thetheinner
innerradius,
radius,rmrmisisthetheradius
radiusofofthe
thesheet’s
sheet’smidmidplane
planeandand
rurdenotes
u denotes thetheradius
radiusofofthetheunstretched
unstretchedfibre,
fibre,thethebending
bendingangleangleis isα αand
ands0s0stands
standsforforthe
the
sheetthickness.
sheet thickness. According
According to toDIN
DIN6935,
6935,thetheunfolded
unfoldedlength
lengthL0L0may maybebecalculated
calculatedbyby
summing
summing upup thethe lengths
lengths a and
a and b ofb the
of the straight
straight sections
sections andandthe the length
length of the
of the circular
circular arc
arc d,
d, represents
that that represents the unstretched
the unstretched fibrefibre
within within the blank
the blank (see equation
(see equation (1)). (1)).
ଶగ௥ೠ ఈ ଶగሺ௥೔ ା଴ǡହ௞௦బ ሻఈ
‫ܮ‬଴஽ூே଺ଽଷହ ൌ ܽ ൅ ܾ ൅ ݀ with ݀ ൌ ൌ (1)
ଷ଺଴ι ଷ଺଴ι
ToTodefine
definethe
the bend
bend allowance
allowance kk ininequation
equation(1),
(1), DIN
DIN 6935
6935 gives
gives the following
the following
equation, limiting
equation, k between
limiting k betweenzero and
zero one:
and one:
57
Dortmund, Germany
 r r 
0, 65
k  k0 , 65  0, 50, lg
5 lg i  i  (2)(2)
 s0 s0 
Thus, value k solely depends on the inner bending radius and the sheet’s thickness.
Thus,
Thus, value
value k solely
k solely depends
depends ononthethe inner
inner bending
bending radius
radius andand thethe sheet’s
sheet’s thickness.
thickness.
Neither material nor bending angle, or the type of bending process (i. e. press brake
Neither
Neither material
material nornor bending
bending angle,
angle, or orthethe type
type of of bending
bending process
process (i. (i.
e. e. press
press brake
brake
bending, folding, …), are taken into consideration.
bending,
bending, folding,
folding, …),…),areare taken
taken intointo consideration.
consideration.
In 1953, Mäkelt published on bend allowance prediction derived from experiments
In In 1953,
1953, Mäkelt
Mäkelt published
published onon bend
bend allowance
allowance prediction
prediction derived
derived from
from experiments
experiments
[Mäkelt,
[Mäkelt, 1953].
[Mäkelt, 1953]. InInaccordance
1953]. In accordance
accordance totoDIN
to
DIN 6935,
DIN6935, Mäkelt
6935, also
Mäkelt
Mäkelt suggests
also
also thethe
suggests
suggests decomposition
the of the
decomposition
decomposition of of
cross
thethesection
cross
cross into
section straight
section into
into sections
straight
straight and circular
sections
sections andand arcs. To
circular
circular take account
arcs.
arcs. ToTotake of
take the bend
account
account allowance,
of of
thethe bend
bend
he draws
allowance, a diagram
allowance, he he
draws thata returns
draws diagram thethat
a diagram value
that y, representing
returns
returns thethe value
value the
y, y,shift of the unstretched
representing
representing thethe offibre
shift
shift of
thethe
dunstretched
(compare
unstretched Figure
fibre
fibre 1 a), as
d (compare
d (compare a function
Figure
Figure of ri/s0.
1 a),
1 a), Hence,
as as the
a function
a function unfolded
of of
ri/sr0i/s length
0. Hence,
. Hence, according
thethe to
unfolded
unfolded
Mäkelt
length
length may be calculated
according
according to to
Mäkelt with
Mäkelt mayequation
may be be (3) andwith
calculated
calculated Figure
with 1 b). (3)(3)
equation
equation andand Figure
Figure 1 b).
1 b).
ଶగఈଶగఈ
‫ܮ‬଴ெ¡௞௘௟௧
‫ܮ‬଴ெ¡௞௘௟௧ ൌൌ ܽ ܾ൅൅
ܽ൅ ܾ ൅ ሺ‫ݎ‬௜ሺ‫ݎ‬ ൅ ‫ݏݕ‬
൅ ‫ݏݕ‬
ଷ଺଴ι ௜ ଴ሻ ଴ሻ (3)(3)
ଷ଺଴ι
Figure
Figure
Figure 1: 1:
1: a) a) Geometry
a) Geometry
Geometry andand
and indices
indices
indices of aof
of a bent
a bent
bent part,
part,
part, b) Determination
b) Determination
b) Determination of bend
of bend
of bend allowance
allowance
allowance
by by Mäkelt
Mäkelt [Mäkelt,
[Mäkelt,
by Mäkelt [Mäkelt, 1953] 1953]
1953]
Since Mäkelt defines factor y against the whole sheet thickness, the following
Since
Since Mäkelt
Mäkelt defines
defines factor y against thethe whole sheet thickness,
thethe following
relationship exists between yfactor y against
and factor k known whole
from DINsheet
6935.thickness, following
relationship
relationship exists
exists between
between y and
y and factor
factor k known
k known from
from DINDIN 6935.
6935.
k = 2y (4)
The performance of both approaches in folding is investigated within this paper.
k =k 2y
= 2y (4)(4)
The
The performance
performance of of both
both approaches
approaches in in folding
folding is investigated
is investigated within
within thisthis paper.
paper.
2. EXPERIMENTS AND MEASUREMENTS
Two different
2. 2.EXPERIMENTSmaterials
EXPERIMENTS were
AND
AND used for the experiments, mild deep drawing grade
MEASUREMENTS
MEASUREMENTS
DD11 (Rp0.2 = 170 – 340 MPa, Rm = 440 MPa) and the stainless steel X5CRNi18-10
(Rp0.2
Two
Two =different
240 – materials
different 260 MPa,were
Rm
materials =used
were750
usedMPa).
forfor
thethe experiments,
experiments, mild
mild deep
deep drawing
drawing grade
grade DD11
DD11
(Rp0.2
(Rp0.2 = 170
= 170 – 340
– 340 MPa,
MPa, RmR=m 440
= 440 MPa)
MPa) andand
thethe stainless
stainless steel
steel X5CRNi18-10
X5CRNi18-10 (Rp0.2
(Rp0.2 = =
58
Dortmund, Germany
240 – 260 MPa, Rm m = 750 MPa). All experiments were conducted by folding initially
flat specimen measuring
All experiments 200 mm inbylength
were conducted foldingand 40 mm
initially flatin width tomeasuring
specimen a bending 200angle
mm
in length
under and
load of40 mm
90°. in sheet
The widththicknesses
to a bendingand
angle underradii
bending loadused
of 90°. The sheetr/s-values
guaranteed thicknesses
of
and bending
either 1.5 orradii usedeach
4 for guaranteed r/s-values
experiment. Tableof Ieither
shows1.5the
or 4materials,
for each experiment.
thicknessesTable
and
Ibending
shows theradii used forthicknesses
materials, the experiments.
and bending radii used for the experiments.
Material
Material DD11
DD11 X5CrNi18-10
X5CrNi18-10
Thickness
Thickness 22 mm
mm 55 mm
mm 88 mm
mm 22 mm
mm 55 mm
mm
Radius
Radius 33 mm
mm 88 mm
mm 7.5 mm
7.5 mm 20 mm
20 mm 12 mm
12 mm 32 mm
32 mm 33 mm
mm 88 mm
mm 7.5 mm
7.5 mm 20 mm
20 mm
r/s-Value
r/s-Value 1.5
1.5 44 1.5
1.5 44 1.5
1.5 44 1.5
1.5 44 1.5
1.5 44
Table I: Overview of experiments
Table I: Overview of experiments
Preceding the bending operation, the unfolded length L0 of each specimen was
Preceding the bending operation, the unfolded length L00 of each specimen was
measured. The folding operation was conducted on a servodriven press, Synchropress
measured. The folding operation was conducted on a servodriven press, Synchropress
SWP 2500. A scematic representation of the tool is shown in figure 2 a). After the
SWP 2500.operation,
forming A scematiceach
representation
specimen’sof the tool is shown
geometry was in figure 2 a).
analyzed withAfter
a the forming
coordinate
operation, each specimen’s geometry was analyzed with a coordinate measuring machine
measuring machine as shown in figure 2 b).
as shown in figure 2 b).
2: a)
Figure 2: a) Schematic
Schematic of
of folding
folding tool,
tool, b)
b) Measurement of
of bent
bent specimen
specimen with
with
coordinate
coordinate measuring machine, c)
measuring machine, c) Measurement
Measurement ofof micro
micro hardness
hardness in
in bending
bending zone
zone
The coordinate
The coordinate measuring
measuring machine
machine was
was used
used to
to analyze
analyze the
the bent
bentpart’s
part’s geometry.
geometry. A
A
comparison of the measured cross section with the previously determined unfolded
comparison of the measured cross section with the previously determined unfolded length
length
L0 L00 the
reveals reveals the bend
inherent inherent bend Furthermore,
allowance. allowance. Furthermore,
micro hardnessmicro hardness
measurements
measurements along the sheet thickness expose the minima of the work hardening
along the sheet thickness expose the minima of the work hardening
59
Dortmund, Germany
within the bending zone (see figure 2 c). The location of the minimum hardness may
within
also be the bending
regarded zone
as an (see figure
indicator 2 c).
of the The location
unstreched of shift.
fibre’s the minimum hardness may also
be regarded as an indicator of the unstreched fibre’s shift.
3. COMPARISON OF RESULTS AND DISCUSSION

3. COMPARISON OF RESULTS AND DISCUSSION
Figure 3 compares the k-values proposed by DIN 6935 and Mäkelt with the data
obtained
Figurefrom folding experiments
3 compares the k-valuesand subsequent
proposed by DIN coordinate
6935 and(red dots)with
Mäkelt andthe micro
data
hardness
obtained from(yellow dots)experiments
folding measurements. First of all,
and subsequent good correlations
coordinate (red dots) and of micro
both
measurement
hardness (yellow techniques were observed
dots) measurements. Firstand analyzing
of all, the specimen
good correlations bymeasurement
of both coordinate
measuring
techniquesmachine and micro
were observed and hardness
analyzingleads to very similar
the specimen results. The
by coordinate reasonmachine
measuring for the
overall slightly
and micro increased
hardness leadsri/s
to0-values, compared
very similar to The
results. the initially
reason forintended ratios slightly
the overall of 1.5
and 4, may
increased be found compared
ri/s0-values, in the springback behavior
to the initially of the
intended bentofparts,
ratios which
1.5 and 4, may results in
be found
elevated ri values. behavior of the bent parts, which results in elevated ri values.
in the springback
Finally, it becomes
Finally, evident
it becomes that all
evident thatofall
theofexperimental k-values
the experimental correspond
k-values betterbetter
correspond to theto
k-value prediction
the k-value according
prediction according to Mäkelt.
to Mäkelt. TheThestrong
strongdecrease
decreaseinink-value
k-valuefor
for rri/s0-ratios
i/s0-ratios
below 2.5 as calculated by DIN 6935 cannot be confirmed.
below 2.5 as calculated by DIN 6935 cannot be confirmed. According to the presented
According to the presented work, it is preferable to determine the bend allowance in
work, it is preferable to determine the bend allowance in folding according to Mäkelt
folding according to Mäkelt instead of applying DIN 6935.
instead of applying DIN 6935.
Figure
Figure3:
3:Comparison
Comparisonofofk-values
k-valuesfrom
fromexperiments,
experiments,DIN
DIN6935
6935and
andMäkelt
Mäkelt
REFERENCES
REFERENCES
[DIN 6935, 2010] Kaltbiegen von Flacherzeugnissen aus Stahl, Issue 6935:2010-01,
[DIN 6935, 2010] Kaltbiegen von Flacherzeugnissen aus Stahl, Issue 6935:2010-
Beuth Verlag, 2010
01, Beuth Verlag, 2010
[Mäkelt, 1953] Mäkelt, H.; „Rationelles Schneiden und Biegen“; In: Schweizerisch
[Mäkelt, 1953] Mäkelt, H.; „Rationelles Schneiden und Biegen“; In: Schweizerisch
Technische Zeitschrift STZ, Vol. 50, Nr. 43, pp 675 – 689, 1953
Technische Zeitschrift STZ, Vol. 50, Nr. 43, pp 675 – 689, 1953
60
Dortmund, Germany
Curvature optimized 3D-Profiles to improve the Line

Speed of Flexible Roll Formed Profiles
Albert Sedlmaier
data M Sheet Metal Solutions, 83626 Valley - Germany
24/25 Nov. 2011 DORP – IUL 1

© 2011 data M SMS datam.de
Overview:
What is 3D roll forming?

History of flexible roll forming for 3D-profiles
Results of some research & development projects
data M‘s new concept to reach industrial accuracy
Load optimized designs
Curvature optimized contours of 3D Profiles
Subsequent processes like sweeping, welding
data M‘s offer to the industry
24/25 Nov. 2011 DORP – IUL 2

61
Dortmund, Germany
What is 3D Roll Forming?
Roll Forming of Profiles with discontinuous Cross Sections

( 3D Roll Forming, Flexible Roll Forming, … )
24/25 Nov. 2011 DORP – IUL 3

3D (flexible) Profile Families
… in depth/
height
… variable
in width
24/25 Nov. 2011 DORP – IUL 4

62
Dortmund, Germany
3D (flexible) Profile Families
… in depth/
height
… variable
in width
24/25 Nov. 2011 DORP – IUL 5

1st Generation: PtU Darmstadt (1999/ 2000)

1st Try-out-Line
COPRA® AMC
(Adaptive motion
control) from data M
Feasibility proven …
But: poor accuracy
24/25 Nov. 2011 DORP – IUL 6

63
Dortmund, Germany
1st Generation: Integrated European Project Proform:
Flexible roll forming line for automotive parts (European R&D Project PROFORM:
2007-2010 with 23 project partners including Daimler and FIAT)
COPRA® AMC (Adaptive Motion Control) delivered by data M in 2009
24/25 Nov. 2011 DORP – IUL 7

Comparison Simulation-Experiment
Warping and deviation effects on the profiles of more than 5 mm

But: feasibility - commercial, reproducability of forming has got proven
24/25 Nov. 2011 DORP – IUL 8

64
Dortmund, Germany
2nd Generation: data M 2010
24/25 Nov. 2011 DORP – IUL 9

24/25 Nov. 2011 DORP – IUL 10

65
Dortmund, Germany
24/25 Nov. 2011 DORP – IUL 11

2nd Generation: Experimental Results
Still very little warping to observe (as predicted by our FE Analaysis)

But much better than the 1st generation with an even wider spread 3D-profile.
24/25 Nov. 2011 DORP – IUL 12

66
Dortmund, Germany
Reasons for the Warping Effects:
Yellow/Grey: Tension -
danger of ruptures at
strains over ~ 20 %
Blue: Compression -
danger of wrinkles at
strains over ~ 10 %
Tension and compression stresses in the transition areas …
24/25 Nov. 2011 DORP – IUL 13

Simulation Results of Warping
… lead to warping effects - yellow and blue regions on the bottom
24/25 Nov. 2011 DORP – IUL 14

67
Dortmund, Germany
Analysis and Animation of 3D-Roll-Forming
24/25 Nov. 2011 DORP – IUL 15

Geometrical Relations
Profiles variable in width
Geometric necessary elongation and compression in the transition zones

allows rapid analytic calculation of adapted blank outline for cutting:
Strip edges stretched: thickness and height decrease
Strip edges compressed: thickness and height increase
Isotropic: height:thickness = 1:1 Anisotropic: height:thickness ~ R-value
24/25 Nov. 2011 DORP – IUL 16

68
Dortmund, Germany
Analytical Elongation on the Strip Edges

50%
45% Forming Limits of different Materials: H100
40% H 80
Low Carbon Steels
35% H 60
strip edge strains in %
30% H 40
High Strength Steels
25%
Ultra High Strength Steels
20%
Aluminum
15%
10%
5%
 Design should meet the material and process demands!
0%
100 200 300 400 500 600 700 800 900 1000
Desired transition radius of a flexible roll formed profile
24/25 Nov. 2011 DORP – IUL 17

The 3rd Generation 3D Roll Forming Line
Achieve an industrial acceptable accuracy

Achieve high rolling speed!
Use of improved bipod-concepts and new self riding blank
holder concepts
Production of 10 or 20 industrial prototypes of desired
profile geometries to proof the calculated feasibility
 Showing the industrial acceptable parts !
Proof of this new production process !
24/25 Nov. 2011 DORP – IUL 18

69
Dortmund, Germany
3rd Generation eD Roll Forming Line

Self-Riding Blank
Holders
Optimized new
Bipod-Concepts
New concepts for bipod-stands and self riding blank holder
24/25 Nov. 2011 DORP – IUL 19

Bipod - Parallel Kinematics
data M‘s patented bipod-system for more flexibility and higher accuracy in
flexible roll forming: presented at Tube & Wire Düsseldorf/ Germany 2010
24/25 Nov. 2011 DORP – IUL 20

70
Dortmund, Germany
Self Riding Blank Holders
Patented self riding blank holder to better guide the sheet between the
stands and avoid warping and scratching of the surface (principle sketch)
24/25 Nov. 2011 DORP – IUL 21

Comparison of Warping Effects
Investigation of the influence of the self riding blank holder:

Simulation shows only a fracture of warping than before - with the
use of the self riding blank holder at the critical points
24/25 Nov. 2011 DORP – IUL 22

71
Dortmund, Germany
Adaptive Control: COPRA® AMC
data M‘s control software COPRA® AMC (Adaptive Motion Control)

allows the precise control of up to 256 flexible axes from one PC
24/25 Nov. 2011 DORP – IUL 23

data M‘s Integrated Aproach:

Construction /
Simulation
Product
Mechanics
Control
24/25 Nov. 2011 DORP – IUL 24

72
Dortmund, Germany
Form Light Weight Design with 3D-Steel-Profiles

vs. Carbon Fibre
What about 3D-Steel-
Profiles instead of
CFK?
More than 3000 kg

weight reduction?
? But, at considerable
lower prices than
with CFRP?
Source: www.lightweightdesign.de
24/25 Nov. 2011 DORP – IUL 25

Load Adapted or Load Optimized Design
Equal Load
1. Constant cross sections without adaption: weight 100%
System
2. Load „adapted“ cross sections: weight reduction of 10%

Moment
Line
3. Load optimized cross sections: weight reduction of 20%
Variation 1: reducing weight by reducing beam ends: 10 – 20 %
24/25 Nov. 2011 DORP – IUL 26

73
Dortmund, Germany
Reduction-“Potencial“ of 3D-Profiles on Truck Members
H 2xH? 3xH?
Variation 2: Higher load by higher height H in the middle of the member:

 With double height: I = B * H³ / 12  2³ = 8-fold load capacity!!!
 Or with triple height:  3³ = 27-fold load capacity???
Weight savings through 3D-profiles are potenciated in the 3rd Power!!!
 either: 700% more load capacity with only 50% more weight
 or: 100% more load capacity with over 60% less weight !!!
 MUCH BETTER THAN CFRP AT CONSIDERABLE LOWER COSTS !!!
24/25 Nov. 2011 DORP – IUL 27

Toy Train derails at transitions from straight track to a curve
24/25 Nov. 2011 DORP – IUL 28

74
Dortmund, Germany
Velocity of the ‘Wild Mouse’
The ‚Wild Mouse“ is a roller coaster on Munich „Octoberfest“

without smooth transitions:  Speed is limited to 50 km/h
24/25 Nov. 2011 DORP – IUL 29

Roller Coaster Impression on Video
„Wild Mouse“-Video:
http://www.muenchs-wildemaus.de/rundfahrt.htm
24/25 Nov. 2011 DORP – IUL 30

75
Dortmund, Germany
Roll Forming Speed of 3D-RF

1
0 Forming Limits of different Materials: H100
0 H 80
strip edge strains in %
Low Carbon Steels

0 H 60
0 H 40
High Strength Steels
0
Ultra High Strength Steels
0
Aluminum
0
0
 Like on the toy train and the „Wild
0
Mouse“ – without easement curves the
0
 Smaller
200 300 forming
100
400 speed
500 of flexible
600 700 roll
800 forming
900 1000
transition radii is limited - unnecessarily!!!
limit the roll Desired transition radius of a flexible roll formed profile
forming speed !!!
24/25 Nov. 2011 DORP – IUL 31

Transition Curves at Train Tracks
Main difference at the „Wiener Bogen“:
Klothoide and „Wiener Bogen“ as transition tracks on real trains
Without such transition tracks a train would derail like the toy train!
24/25 Nov. 2011 DORP – IUL 32

76
Dortmund, Germany
Formula Rossa in Abu Dhabi: max. speed: 240 km/h

 Optimization with curves of 5. power needed!!!
24/25 Nov. 2011 DORP – IUL 33
Formula 1 video of turkish course in Istanbul:

http://www.sport.de/cms/formel-1/videos.html
Transmitting this to flexible roll forming means …
24/25 Nov. 2011 DORP – IUL 34

77
Dortmund, Germany
Continuously Changing Curvature

2x radius 600 mm
Sinus Curve
Continuously changing curv.
10% flattened
20% flattened
Testing of different mathematically calculated forming paths with

different curvatures: What tolerances are really needed?
Which forming speed can be achieved?
24/25 Nov. 2011 DORP – IUL 35
Why Forming Speed Counts …
Roll Forming Speed: 10 m / min 20 m / min 30 m / min

30ft / min 60 ft / min 90 ft / min
1m-parts / hour: 600 1.200 1.800
3m-parts / hour: 200 400 600
1m-parts / shift: 4.800 9.600 14.400
1m-parts / week: 75.000 150.000 225.000
3m-parts / week: 25.000 50.000 75.000
1m-parts / month: 300.000 600.000 900.000
1m-parts / year: 3.600.000 7.200.000 10.800.000
3m-parts / year: 1.200.000 2.400.000 3.600.000
12m parts and longer… (36ft parts and longer…)
1.000.000
24/25 Nov. 2011 DORP – IUL 36

78
Dortmund, Germany
Prduct families on one line: Seven at one Blow - or 11…?
By clever arrangement of profiles-ends

it is possible to produce complete
parts families simultaneuously
24/25 Nov. 2011 DORP – IUL 37

Subsequent Processes to Roll Forming:
Simulation of 3D-profiling to achieve a

high accuracy of coil edges
Simulation of welding and strip
thinning in order to calculate material
behavior for subsequent processes
like hydroforming…
Simulation of 3D-profiling to achieve a high accuracy of coil edges

Simulation of welding and strip thinning in order to calculate material
behavior for subsequent processes like hydroforming…
24/25 Nov. 2011 DORP – IUL 38
79
Dortmund, Germany
Subsequent Processes to Roll Forming:

3D-profiled Comparison of the
optimized sheet deformations
preform at 50 and 90 % filling
of the IHU-process
Effective
Final IHU-Geometry strain
© EFB/PtU Darmstadt
…Optimized preforms achieve higher deformation degrees on IHU
24/25 Nov. 2011 DORP – IUL 39

Our Offer to Industry: Design

Simulation
Product
Mechanics
Control
Feasibility studies for this new technology

Design and production of necessary tool sets: COPRA® RF
Development of adaptive motion control: COPRA® AMC
Making of prototype parts on our own prototype mill
Detailed planning of the overall flexible roll forming line
Promoting this new technology – world wide
24/25 Nov. 2011 DORP – IUL 40

80
Dortmund, Germany
Many Thanks for

Your Attention!
data M Sheet Metal Solutions GmbH I Am Marschallfeld 17 I D-83626

Valley Germany I Tel.: +49(0)8024/640-0 I Fax: +49(0)8024/640-300
e-mail: datam@datam.de I www.datam.de I www.roll-design.com
81
Dortmund, Germany
82
Session 4: Profile Bending
Dortmund, Germany
Process Limit Extension of the TSS Profile Bending

Process
Daniel Staupendahl*, Matthias Hermes, A. Erman Tekkaya
Institute of Forming Technology and Lightweight Construction (IUL), Technische
Universität Dortmund, Baroper Str. 301, 44227 Dortmund, Germany,
*Daniel.Staupendahl@iul.tu-dortmund.de
Abstract: Complex lightweight design in the automotive and utility vehicle industry
necessitates the use of tubes and profiles made of innovative high-strength steel grades.
The Torque Superposed Spatial (TSS) profile bending process is a roll based process that
has the potential of continuously bending load-optimized and high-strength tubes and
profiles with three-dimensional bending contours. The challenge in bending highstrength
steels versus standard steel grades is the great reduction of process limits. Experimental
results illustrating the advantages as well as the process limits of the bending process are
presented. Finally, process extensions to increase the process limits are discussed.
Keywords: Profile; Tube; Bending; Torque Superposed Spatial (TSS); High-strength
steel, Process limits
1. INTRODUCTION
In the field of tube and profile bending a steady increase in the demand of process
automation and flexibility can be observed. This can be traced back to continuously
decreasing batch sizes and the resulting growing model ranges, which oftentimes
make high tooling costs unprofitable. At the same time, product complexity increases,
meeting the current demand of unique design and light-weight construction. Kinematic
and thus flexible processes for complex bending operations are for instance the
Hexabend process [Neugebauer et al., 2002], the Nissin process [Murata et al., 2007],
and the TKS-MEWAG process [Flehmig et al., 2006]. The general requirement of high
accuracy and high process stability, however, presently is only met by the use of rotary
draw bending, stretch bending [Geiger, Sprenger, 1998], or by bending and calibrating
by hydroforming [Kleiner et al., 2006]. Since these processes and process chains are
based on the principle of tool-bound shaping, these processes have the disadvantage of
having a low flexibility, expensive tools, and high manufacturing costs when bending
long profiles and large cross-sections. Resulting from the absence of an appropriate
flexible procedure to bend three dimensional profiles with arbitrary cross-sections,
lengths, and contours at relatively low costs the Torque Superposed Spatial (TSS) profile
bending process was developed at the Institute of Forming Technology and Lightweight
Construction of TU Dortmund University [Hermes, Kleiner, 2008]. Chatti et al. [Chatti
et al., 2010] have shown the applicability of the TSS bending process in forming low
strength steel material using S235JR steel profiles with a tensile strength of 370 MPa.
85
Dortmund, Germany
University [Hermes, Kleiner, 2008]. Chatti et al. [Chatti et al., 2010] have shown the
applicability of the TSS bending process in forming low strength steel material using
S235JR steel profiles with a tensile strength of 370 MPa.
2. OBJECTIVES
2. OBJECTIVES
To investigate the influence of bending high strength steel profiles on the current TSS
Tobending
investigate the influence
process, of bending
process limit high strength
investigations were steel profilesThe
performed. on tested
the current TSSwas
material
bending process,
40x40x2.5 mmprocess limit investigations
square profile and 40x40x2.5 were
mmperformed. The tested
L shaped profiles madematerial was
of MW1000L
40x40x2.5 mm square
air hardening profile
steel from and 40x40x2.5
Salzgitter mm L-shaped
Mannesmann Precisionprofiles made ofstrength
with a tensile MW1000Lof 1100
airMPa.
hardening steel fromcould
Since Salzgitter Salzgitter Mannesmann
only supply Precision
closed profiles, with a tensile
the L-shaped strength
profiles of out
were cut
1100 MPa. Since Salzgitter could only supply closed profiles, the L-shaped profiles
of the 40x40x2.5mm square profile.
were cut out of the 40x40x2.5mm square profile.
3. EXPERIMENTAL SETUP
3. EXPERIMENTAL SETUP
The mainTheparts
mainofparts of thebending
the TSS TSS bending procedure
procedure are a feeding
are a feeding mechanism,
mechanism, which which
can becan
rotated (α1-axis) around its feeding axis (c-axis), and a deflecting, or bending unit,unit,
be rotated (α1 axis) around its feeding axis (c-axis), and a deflecting, or bending
which
which consist
consist of of a bending
a bending matrix,
matrix, movable
movable along
along a linear
a linear bending
bending axisaxis (x-axis).
(x-axis). TheThe
bending matrix has two compensation axes (α2- and τ-axis) to achieve a tangential runrun
bending matrix has two compensation axes (α2- and τ-axis) to achieve a tangential to to
thethe profile.
profile. (figure
(figure 1) 1)
Figure 1: TSS-bending process and realized machine

Figure 1: TSS-bending process and realized machine
4. 4.PROCESS
PROCESSLIMIT
LIMITINVESTIGATIONS
INVESTIGATIONSAND
AND EXTENSIONS
EXTENSIONS
To investigate the geometric
To investigate limits limits
the geometric of bending the square
of bending profile,
the square a contour
profile, withwith
a contour
-1 -1
curvatures starting
curvatures at at
starting 0.0005 mm
0.0005 mm-1andandcontinuously increasing
continuously to to
increasing 0.0024
0.0024mmmm-1waswas
chosen. Figure 2 shows the target curvature in comparison to the actual inner and outer
chosen. Figure 2 shows the target curvature in comparison to the actual inner and outer
curvature of the profile produced by the bending test. The actual curvature after bending
curvature of the profile produced by the bending test. The actual curvature after bending
86
Dortmund, Germany
was measured using the 3D scanning system GOM Atos. Trails with bending curvatures
was
was measured
measured using
using thethe
3D3D scanning
scanning system
system GOM
GOM Atos.
Atos. Trails
Trails with
with bending
bending curvatures
curvatures
above 0.0024 mm-1 directly
-1 lead to failure.
above
above 0.0024
0.0024 mmmm
-1 directly
directly lead
lead toto failure.
failure.
Figure
Figure 2: Target Figure
ramp 2:2: Target
curvature
Target(κL,ramp
ramp curvature
blue) (κ(Lκ, Lblue)
and actual
curvature , blue)
curvature
andand actual
(κU, inner curvature –
actual
curvature (κU, green,
inner curvature
outer – green,
curvature – outer
yellow) curvature
curvature (κ , inner curvature – green, outer curvature – yellow)– yellow)
U
Additionally,thetheL-profiles
Additionally, L-profileswere werebentbentinindifferent
differentbending
bendingplanes.
planes.Target
Targetcurvatures
curvaturesof
Additionally, the L-profiles were
-1 -1bent in different bending planes.-1Target curvatures of
of 0.0022
0.0022 mm-1mm -1 0.0007 mm-1
and0.0007 mm -1 were investigated,
investigated,0.00220.0022
0.0022mm mm-1
mm -1 being the
beingthe maximum
themaximum
maximum
0.0022 mm and mm were investigated, being
producible
producible curvature.
curvature.The L-profile
The self twisted
L-profile self when
twisted notwhenbeing bent
not over
being its bent
symmetrical
overitsits
producible curvature. The L-profile self twisted when not being bent over
axis, which could
symmetrical notwhich
axis, be prevented
could with the currentwith
process setup. The self-twisting of the
symmetrical axis, which could notnot bebe prevented
prevented with thethe current
current process
process setup.
setup. The
The self-
self-
profile at a of
twisting curvature
the of κL
profile at = curvature
a 0.0022 mm-1 of κ Lis=shown
0.0022 inmm -1
figure
twisting of the profile at a curvature of κL = 0.0022 mm is shown in figure 3 (left).
-1 is3shown
(left). in figure 3 (left).
Following
Following thesethese experiments,
experiments, the TSS bending machine was extended by two
Following these experiments, thethe TSS
TSS bending
bending machine
machine waswas extended
extended byby twotwo additional
additional
additional
servo servo
axes thataxes thatthe
allow allow the numeric
numeric control control
of the ofα2the
- andα2-τ-axis
and τ-axis
of theofbending
the bending
head.
servo axes that allow the numeric control of the α2- and τ-axis of the bending head.
head.
Using
Using Using a systematic
a systematic
a systematic engineering
engineering
engineering design
design design
approach
approach approach
different
different different
drive
drive drive were
concepts
concepts concepts
were were
evaluated
evaluated
andfinally
evaluated
and finally
andone oneultimate
finally ultimate
one ultimatedesign
design waschosen
design
was chosen
was chosen(figure 3).3).3).
(figure
(figure ThisThis
This
newnew designoffers
newdesign
design offersthe
offers the
the
possibility
possibility
possibility todynamically
dynamically
totodynamically twist twist
twistthethe theprofile
profileprofile
during during
bending
during bending totocompensate
to compensate
bending compensate thetheself-
the self-twisting self-
of twisting ofof unsymmetrical
unsymmetrical
twisting unsymmetrical
profiles. A compensated profiles. A
profiles. A compensated
bending compensated
curvature of bending
κL = 0.0022
bending curvature
curvaturemm-1 of isof
κ κ=L =0.0022
-1
shown 0.0022
in figure
mm mm is is shown
3-1(right).
shown inin figure
figure 3 (right).
3 (right).
L
Figure
Figure 3:3: Extension
Extension of of machine
machine toto optimize
optimize process
process limits
limits
Figure 3: Extension of machine to optimize process limits
87
Dortmund, Germany
5. CONCLUSION
Process limits of bending symmetrical square profile and unsymmetrical L-profile

using the TSS bending process with controlled α1 , x-, and c-axis were investigated. The
investigations lead to the extension of the process by further controlled α2 and τ-axes.
With this extension it was possible to compensate the self-twisting of the unsymmetrical
L-profile. In further investigations, the analytic process model will be extended by the
additional controlled axes. To investigate the possibility of an increase of the maximum
curvature producable in bending operations, furthermore, different bending levers will be
analyzed.
6. ACKNOWLEDGEMENTS
The research leading to these results has received funding from the Research
Program of the Research Fund for Coal and Steel [RFSR-CT-2009-00017] and by the
DFG (German Research Foundation).
REFERENCES
[Chatti et al., 2010] Chatti, S., Hermes, M., Tekkaya, A.E., Kleiner, M.: The new
TSS bending process: 3D bending of profiles with arbitrary cross-sections, In:
CIRP Annals - Manufacturing Technology, 315-318 (2010), 59/1
[Flehmig et al., 2006] Flehmig, T., Kibben, M., Kühni, U., Ziswiler, J.: Device
for the Free Forming and Bending of Longitudinal Profiles, Particularly Pipes,
and a Combined Device for Free Forming and Bending as well as Draw Bending
Longitudinal Profiles, Particularly Pipes, Int. patent with application no. PCT
EP2006/00252 (2006)
[Geiger, Sprenger, 1998] Geiger, M., Sprenger, A.: Controlled Bending of
Aluminum Extrusions, In: Annals of the ClRP 197-202 (1998), 47/1
[Hermes, Kleiner, 2008] Hermes, M., Kleiner, M.: Vorrichtung zum Profilbiegen
(dev. for profile bending), German Patent Application, DE102007013902A1 (2008)
[Kleiner et al., 2006] Kleiner, M., Chatti, S., Ewers, R., Hermes, M., Homberg, W.,
Shankar, R.: Process Chain for the Improvement of Hydroforming Processes Using
Tailored Semi-Finished Tubes, In: Annals of the WGP, Production Engineering,
57-62 (2006), 13/1
[Murata et al., 2007] Murata, M., Kuboti, T., Takahashi, K.: Characteristics of
Tube Bending by MOS Bending Machine, In : Proc. of the 2nd Int. Conf. on New
Forming Technology, 135 144 (2007)
[Neugebauer et al., 2002] Neugebauer, R., Drossel, W.-G., Lorenz, U., Luetz,
N.: Hexabend - A New Concept for 3D-free-form Bending of Tubes and Profiles
to Preform Hydroforming Parts and Endform Space-frame-components, In :
Advanced Technology of Plasticity, 1465–1470 (2002), 2
88
Dortmund, Germany
Overview
• Geometry and Rolls Positioning;
• Rolls diameter
• Profile forms and tolerances
• Profile deformation
• Measurement system for the radius
• Limits of the Measurement system
Measurement System for three 2

roll bending of beams
89
Dortmund, Germany
The Universal Rolls

The guide rolls

90
Dortmund, Germany
Rolling beams/channels on edge
Retaining
Roll
Push-Pull
Unit
Principles of Profile Rolling 5
The Universal Profiles bending machine

91
Dortmund, Germany
The Universal Profiles bending machine
Principles of Profile Rolling 7
Variable Geometry Profile Bender

92
Dortmund, Germany
The Beams Bender

Necessary measurements for automatic

bending
 Measurement of rolls diameter necessary to
make a proper measurement of the radi
 Positions of the rolls defining the radius of
the bended part
 The Positioning of the rolls can be measured
by hydraulics up to 0,1 mm, or even smaller
if necessary. Linear transducers are used.

93
Dortmund, Germany
The Shapes
• Angles
• Tees
• Bars
• Beams
• Pipes
• Channels
•…

Optical measurement system - TubeInspect
TubeInspect system components
OK
not OK
The TubeInspect measuring

software is designed for
application in production and
features an easy operability.
94
Dortmund, Germany
The Bending Loads

L= Load
Wf = Resistance of Section
σ = Yield Strength
L
4xWf x σ
Load =
D

The Pressure on the profile
L 1) In order to reduce the waste of material

(flat end) the rolls distance has to be
minimum;
D 2) Minimum rolls distance means higher load
necessary for bending the same profile;
Wf x σ 3) Higher load means higher compression
Load = stresses on the profile cross-section.
4xD

95
Dortmund, Germany
Typical cross section deformation
Thickening of bars
Ovalisation of pipes
Web deformation of beams and channels
Torsion of asymmetrical sections

Measurement of the radii manual
2
4P 2 + C
R=
8P
R is the Radius
C is the length of chord
P is the depth

96
Dortmund, Germany
Measurement of the radii automatic
2
4P 2 + C
R=
8P
R is the Radius
C is the distance of the Lasers (fix)
P is the depth, measured by Laser
depending on rolls diameters and
height of profile
Laser 1 Laser 2

Possible automatic cycle with Laser

Needed parameters
• desired outer diameter of the part
• outer diameter of contact rolls
• Inner height of profile
• percentage of tolerance where to stop
Bending action
• Bending rolls are positioned at the first bending point
• rotation and bending is done
• Measurement of the diameter by the laser
• calculation of new position of the bending rolls, to reach halfway
the desired diameter
• These steps are repeated until the diameter is reached within the tolerances

97
Dortmund, Germany
Achievable results and limits

The results achieved in real customer work are
- +/-3 % tolerances on the diameter in the measurement
- Automatic bending safe until 90% of the desired diameter
- no manual measuring necessary
- no complicated calculation and measuring necessary
Limits of this system are

- The tolerances in profile dimensions
- The tolerances in material properties
- The deformation of the profile
- only suitable for big radii

98
Session 5: Measurement and Simulation Techniques
Dortmund, Germany
Numerical Simulation of Profile Bending Using the

TSS Bending Method
M. Hudovernika, D. Staupendahlb, M. Hermesb, K. Kuzmanc,
A. E. Tekkayab, J. M. Slabea
a
Tecos, Slovenian Tool and Die Development Centre, Kidričeva 25. SI-3000,
Celje, Slovenia, matej.hudovernik@tecos.si
b
Institute of Forming Technology and Lightweight Construction (IUL),
Technische Universität Dortmund, Baroper str. 301, Dortmund, Germany,
Daniel.Staupendahl@iul.tu-dortmund.de
c
University of Ljubljana, Faculty for mechanical engineering, Aškerčeva 6,
SI-1000, Ljubljana, Slovenia
Abstract: The high demand for the realization of cost effective light weight construction,
reduction of energy, and material consumption employ needs for constant development of
innovative incremental, adaptive, and robust solutions capable of accurate 3D bending of
tubes and profiles made from high strength steels. A kinematic bending method known as
Torque Superposed Spatial – TSS Bending, capable of flexible, continuous, and adaptive
2D and 3D bending of profiles with symmetrical and asymmetrical cross-sections has
been developed at the IUL – TU Dortmund. To understand the forming mechanisms
of the process numerical investigations of the TSS bending method are introduced. An
elastic-plastic numerical model using solid elements and high-strength steel material
characteristics is presented and compared to experimental data.
Keywords: 3D Profile Bending, 2D profile bending, TSS - Torque Superposed Spatial,
Numerical simulations, FEM
1. INTRODUCTION
The decrease of material, weight, and energy consumption play an important role
in reaching new market demands and environmental standards. The reduction of overall
weight generates an additional need for the development of high strength materials
and economical solutions for robust, flexible, and cost effective forming technologies.
Three-dimensionally bent tubes and profiles made of high strength materials show great
potential in new design of spatial structures. The TSS bending method, developed at
the IUL – TU Dortmund [1, 2], in combination with the use of high-strength materials
shows great industrial potential in achieving complex continuous 3D shaped contours
of profiles with symmetrical and asymmetrical cross-sections in a vast amount of
applications [3]. In the following, a numerical simulation model for 2D profile bending
using the TSS method is presented. The model was generated and analyzed with
101
Dortmund, Germany
Deform 3D.Results
Deform 3D. Results of numerical
of numerical analysis
analysis were analyzed
were analyzed and with
and compared compared with
experimental
experimental data provided by the IUL, TU Dortmund.
data provided by the IUL, TU Dortmund.
2. THE TSS PROFILE BENDING METHOD

2. THE TSS PROFILE BENDING METHOD
The TSS bending method offers a new approach to solving technological tasks and
The TSSnew
to withstand bending method restrictions
challenges, offers a newand approach
processtolimits.
solving Thetechnological
system showntasks and
in
Figure 1 operates in several parallel phases, which include the transportationinofFigure
to withstand new challenges, restrictions and process limits. The system shown the
1 operates
profile alongin its
several parallel axis,
longitudinal phases,thewhich include the
plasticization transportation
of material duringofbending,
the profile
andalong
the
application of bending forces to realize the 2D and 3D geometry of the finished product.of
its longitudinal axis, the plasticization of material during bending, and the application
bending
The forces tofeed
continuous realize the 2D
of the and is
profile 3Dprovided
geometrybyof thethe roll
finished
basedproduct.
feedingThe continuous
mechanism,
feed oftransports
which the profile theisprofile
provided
alongby the
the croll based
axis. feeding mechanism,
2D bending of the profilewhich transports
is realized by
the profile along
pre-defined the c axis. movement
NC programmed 2D bendingofof thethe profilehead
bending is realized
along the by xpre-defined
axis. The 3D NC
programmed
contour is thenmovement
producedof bytherotating
bendingthe head along bearing
torsion the x axis.
(α1),The 3D contour
mounted around is then
the
producedmechanism,
feeding by rotating thereby
the torsion bearing (α1),
superposing mounted
a torque, andaround
changingthe feeding mechanism,
the bending plane
thereby to
relative superposing
the profile.a The
torque,
twoand changing theaxes
compensation bending
(α2 andplane relative
τ-axis) to the
of the profile.
bending The
head
self-align during the
two compensation axesprocess
(α2 andaccording
τ-axis) oftothethebending
movement head of the profile
self-align to the
during achieve a
process
tangential
according run of the
to the bendingofhead
movement the relative
profile toto achieve
the profile. [1-3] run of the bending head
a tangential
Figure 1; Working scheme for TSS profile Figure 2; Comparison of the bending head position in
bending process [1]. the actual TSS bending process (a), and the position of
the single roll for the 2D bending simulation (b).
relative to the profile. [1-3]

3. NUMERICAL ANALYSIS
All tools, the feeding rolls and the bending head, were defined as rigid elements
within3.an
NUMERICAL
elastic-plasticANALYSIS
numerical model – Figure 3a. The simulation for 2D bending of
square 40 x 40 x 2.5 mm profile, with an output of variable continuous bending radii
Allfrom
ranging tools,Rmin
the=feeding
400 mm,rolls
to Rand the
max = bending
2000 mm ishead,
shownwere defined
in Figure 3b.as rigid elements
within an elastic-plastic numerical model – Figure 3a. The simulation for 2D bending
of square 40 x 40 x 2.5 mm profile, with an output of variable continuous bending radii
ranging from Rmin = 400 mm, to Rmax = 2000 mm is shown in Figure 3b.
102
Dortmund, Germany
a b
Figure 3; Input geometry model (a) – target contour (b).
The
Thematerial
materialproperties
propertiesassigned
assignedtotothethemodel
modelareare
taken
takenfrom
fromthethe
airair
hardening
hardening steel
steel
1
MW1000L
MW1000L 1
. The . The yield
yield strength
strength measures
measures 900900
MPa,MPa, while
while thethe tensile
tensile strength
strength measures
measures
1100
1100 MPaMPa [4]. [4].
TheThe
Von Von
Mises Mises
yieldyield function
function and isotropic
and isotropic hardening
hardening were assigned
were assigned to the to
the Lagrangian
Lagrangian incremental
incremental type of simulation
type of simulation model. Allmodel.
loads All
suchloads
as thesuch as the
torque, torque,
bending
bending
force, force, and
and clamping clamping
force force were
were replaced withreplaced withInmovements.
movements. In thethe
the TSS process, TSS process,
bending
theconstantly
head bending head constantly
adapts adapts
its position its position
with regard towith
theregard to the localbycurvature
local curvature by free
free rotation
rotation
around aroundaxis.
its vertical its The
vertical axis.head
bending Theisbending
thereforehead
alwaysis positioned
therefore always
tangentialpositioned
to the
tangential
local to the local
profile curvature profile
(Figure 2a).curvature
For the 2D (Figure 2a). simulation,
bending For the 2D the bending
bendingsimulation,
head wasthe
bending
replaced by head was replaced
one single rigid roll,byandone singlemovements
bending rigid roll, were
and bending
applied in movements
both x andwerec
direction both xinand
appliedasinshown c direction
Figure 2b. as shown in Figure 2b.
4. EXPERIMENTAL
4. EXPERIMENTAL ANALYSIS
ANALYSIS
The experimental analysis for the 2D bending of 40x40x2,5 mm profile made from
The experimental analysis for the 2D bending of 40x40x2,5 mm profile made from
MW1000L steel was carried out at the IUL, TU Dortmund. Experiments were done with
MW1000L steel was carried out at the IUL, TU Dortmund. Experiments were done with
increasing single radii, as shown in Figure 3b, from Rmin = 400 mm, in 100 mm steps to
increasing singlemm.
Rmax = 2000 radii,Toasachieve
shown in Figure 3b, from
a continuous Rminwere
curvature = 400 mm, in 100tomm
interpolated steps with
a spline to
Rmax = 2000 mm. To achieve a continuous curvature
an effective radius trend of 416 mm to 1985 mm. [3] were interpolated to a spline with an
effective radius trend of 416 mm to 1985 mm. [3]
5. RESULTS
5. RESULTS
Optimal results for simulating 2D profile bending using the TSS bending method
Optimal
were results
gained for simulating
by using 2D profile
the following bending
settings: using with
1600 steps the TSS bending
a 0.1s method
time increment,
were gained
aprox. by using
75.000 mesh the following
elements. The settings: 1600
calculation timesteps
usingwiththesea settings
0.1s time increment,
amounted up to
aprox. 75.000 At
42 hours. meshthe elements. Thethe
point where calculation time reached
bending head using these settings
its final amounted
position on theup x to
axis,
42 maximum
hours. At the point
stress where
of 773 MPathe bending
(Figure 4) head
and a reached
maximum its strain
final position
of 0,181 on the calculated.
were x axis,
Approximately
maximum stress of 5%
773ofMPageometrical
(Figure 4)deviation was measured
and a maximum strain ofregarding
0,181 werecontinuality
calculated.and
accuracy of the
Approximately 5%profile curvature deviation
of geometrical per overallwas
length of the profile
measured - Figure
regarding 5.
continuality and
accuracy of the profile curvature per overall length of the profile - Figure 5.
1
Air hardening steel, named specifically by the manufacturer Salzgitter Mannesmann Precision GmbH.
1
Air hardening steel, named specifically by the manufacturer Salzgitter Mannesmann Precision GmbH.
103
Dortmund, Germany
MPa
Figure 4; Detailed state of stresses within critical Figure 5; Comparison of the contours gained in
area. the numerical and experimental investigations.
6. CONCLUSIONS
The
6. numerical
CONCLUSIONSsimulation of two-dimensional profile bending by the TSS profile
bending method was introduced. Results were obtained by a numerical simulation using
the Deform 3D software.
The numerical Material
simulation data for the airprofile
of two-dimensional hardening steel
bending by MW1000L was bending
the TSS profile used.
Formethod was introduced.
the geometrical Results
accuracy were obtained
between by a numerical
the numerical simulation results,
and experimental using the Deform 3D
a deviation
5%software. MaterialIn
was measured. data for the
further air hardeningthe
investigations, steel MW1000L
numerical was developed
model used. For the
willgeometrical
be used
accuracy between the numerical and experimental results, a deviation 5% was measured. In
as a basis for detailed numerical analysis of 3D profile bending.
further investigations, the numerical model developed will be used as a basis for detailed
numerical analysis of 3D profile bending.
ACKNOWLEDGEMENT
ACKNOWLEDGEMENT
The research was part financed by the European Union’s European Social Fund, from
the European Union’s
The research wasResearch Fund by
part financed forthe
Coal and Steel
European (RFCS)
Union’s under grant
European Socialagreement
Fund, fromn°
the
[RFSRCT-2009-00017], and Fund
European Union’s Research the German
for CoalResearch
and Steel Foundation (DFG).
(RFCS) under grant agreement n° [RFSR-
CT-2009-00017], and the German Research Foundation (DFG).
REFERENCES
REFERENCES
[1] S. Chatti, M. Hermes, A.E. Tekkaya, M. Kleiner, The new TSS bending process:
[1] 3D
S. Chatti,
bending M.ofHermes, A.E.
profiles withTekkaya, M. Kleiner,
arbitrary The new TSS
cross-sections, CIRP bending
Annals process: 3D bending of
- Manufacturing
profiles with arbitrary cross-sections, CIRP Annals - Manufacturing Technology Volume 59,
Technology Volume 59,
Issue 1, str. 315-318, 2010. Issue 1, str. 315-318, 2010.
[2] [2] Hermes
Hermes M, Kleiner
M, Kleiner M (2008) M (2008) zum
Vorrichtung Vorrichtung
Profilbiegenzum Profilbiegen
(device (deviceGerman
for profile bending), for
profile bending), DE102007013902A1,
Patent Application, German Patent Application,
registr. Date DE102007013902A1,
20.03.2007. registr. Date
D. Staupendahl, C. Becker, M. Hermes, A.E. Tekkaya, M. Kleiner, New methods for
[3] 20.03.2007.
manufacturing 3D-bent lightweight structures, In: Proceedings of the 3rd International Conference
[3] D. Staupendahl,
on Steel C. 2011
in Cars and Trucks, Becker, M. Hermes, A.E. Tekkaya, M. Kleiner, New
S. 120-129.
[4] methods
Salzgitterfor manufacturing
Mannesmann 3D-bent
Precision GmbH., lightweight
Precision Tubesstructures,
acc. To DINIn:EN
Proceedings of the
10305-1 – MW1000L
3rd International
- For Conference
highly stressesd components,on Steel in Cars
Werkstoffblatt 049and Trucks,
R, Jan. 2009. 2011 S. 120-129.
[4] Salzgitter Mannesmann Precision GmbH., Precision Tubes acc. To DIN EN
10305-1 – MW1000L - For highly stressesd components, Werkstoffblatt 049 R,
Jan. 2009.
104
Dortmund, Germany
Measurement of bent
parts for the correction of
bending machines AICON 3D Systems GmbH
Biberweg 30 C
D-38114 Braunschweig
Germany
Dr.-Ing. Christoph Dold http://www.aicon3d.com
Topics
• Company profile: AICON 3D Systems GmbH

• Optical measurement system TubeInspect
• Measurement of bent parts
• Inspection of roll-bended parts
• Inspection of free-form bended parts
• Contour inspection
• Process control with TubeInspect
• Summary
105
Dortmund, Germany
AICON 3D Systems GmbH – company profile
AICON Company Profile
AICON History
• Founded in 1990 in Braunschweig, Germany,

by Carl-Thomas Schneider and Werner Bösemann
• Spin-Off from Volkswagen R&D and
Braunschweig’s Technical University
• Located in Northern Germany near Hanover and Wolfsburg
• Easy to reach by train and air (40 minutes to Hanover Airport)
106
Dortmund, Germany
AICON Company Profile
AICON Product Lines
Optical Gauge Portable CMMs
TubeInspect DPA MoveInspect
Application Line Vehicle Testing
TubeInspect S / TubeInspect HS
MoveInspect DPS WheelWatch ProCam

TubeInspect HD EngineWatch
107
Dortmund, Germany
History of tube measurement at AICON
1994:
First installation at an
automobile manufacturer
2011:
• More than 140 installations world
wide in all tube manufacturing
sections
• Continuous software development
TubeInspect is dedicated to ...
... programmable
optical gauging
... setting up and correcting

bending programs
... digitizing of master tubes

for Reverse Engineering
108
Dortmund, Germany
High resolution digital cameras

acquire the tube details.
The transillumination
assures equal
conditions of brightness.
109
Dortmund, Germany
Very precise reference points

guarantee reliable measuring
results also in a tough
production environment.
OK
not OK
The TubeInspect measuring

software is designed for
application in production and
features an easy operability.
110
Dortmund, Germany
How does it work?
From measuring images to tube geometry
Automatic extraction of the tube contour
Hochauflösend wird das

Rohr in seiner ganzen
Länge erfasst ...
The tube is captured by cameras from multiple views…
111
Dortmund, Germany
Automatic extraction of the tube contour
... and a 3D cylinder model is generated.
Technical data for different models
Maximum
Measuring Measurable
measurable
Model volume tube Accuracy
tube
[mm] diameters
lengths
2.500 x
TubeInspect 1.100 x 5 – 7m 3,2 – 200mm ± 0.1mm
700
1.100 x
TubeInspect S 1.100 x 1-2m 3,2 – 200mm ± 0.1mm
700
1.080 x
TubeInspect HS 980 x 1.5m 2 – 100mm ± 0.050mm
500
450 x
TubeInspect HD 400 x 0.6m 1 – 20mm ± 0.025mm
200
112
Dortmund, Germany
Measurement of bent parts
Measurement features
• Bending data PRB / LRA

• Tangent points / bending coordinates
• Sheath deviation
• End-to end distance, tube length


  
 


 

 




113
Dortmund, Germany
Tube with bending points (I)
• Measurement of a cylindrical model

• Principles of photogrammetry
• Complete 3D acquisition
• Detection of bending points 5
• Search and assignment 7

• Prediction of not detected bending
points, e.g. in case of flat bends
6
Tube with bending points (II)
• Calculation of bending elements 4
• Based on measured cylinder model 6
• Adjustment of bending model (PRB)

5
• Nominal-actual comparison
• Graphical view of nominal and actual
• Different views including sheath
deviation
• Deviation and correction values BS BE

BS BE
• Sheath deviation for BS/BE BS BE
• Based on bending model
114
Dortmund, Germany
Tube with bending points (III)
• Reports and output

• Display of different views
• Reports containing measured data
• Correction of bending machines

• Data transfer to bender
• Selected data only
• Damping of values possible
Freeform with bend data
• Freeform BD
• Bending data is available
• Correction of bender is possible
115
Dortmund, Germany
Freeform with bend data
• Characteristics of Freeform BD tubes

• Variable bending radii
• Push values are zero
• New evaluation method with TubeInspect
Measurement of a freeform geometry
• 3D cylinder model is measured

• Special processing algorithm to
calculate corrections
• Segments are detected
automatically
• Bending data of measured
geometry is derived
• Correction values for the
bending machine are calculated
116
Dortmund, Germany
Evaluation of a Freeform BD tube
• Results:
• Nominal-Actual view
• Sheath tolerance (contour)
• Bending data
Freeform (contour inspection)
• 3D cylinder model is
measured
• Contour inspection
between nominal and
actual data
• Any cylindrical geometry
is compared
• Visual evaluation only
• Orientation is possible
117
Dortmund, Germany
Range of applications
• Possibilities to evaluate a 3D measurement
Process Control with TubeInspect
118
Dortmund, Germany
Data transfer to bending machines
SERVER
Bender 1
Bender 2
Bender 3
Bender 4
Data transfer to bending machines
measuring result: measuring result:

tube bent with coordinates tube bent with correctional data
119
Dortmund, Germany
Summary
Measurement of bent parts for the correction of benders
• AICON provides optical camera based 3D measurement

systems
• TubeInspect allows the measurement of different tube types
• Measurement and evaluation of tubes with classical bending
points and freeform geometries
• Wide tube range
• Software provides additional functionalities
• Automatic process control is possible
• Correction of benders
• Automatic mode for remote control
120
Dortmund, Germany
Variation of Contractile Strain Ratio of TA18

Titanium Alloy Tubes with Plastic Deformation
T .Huang, M. Zhan*
State Key Laboratory of Solidification Processing, Northwestern Polytechnical
University, Xi‘an 710072, China
zhanmei@nwpu.edu.cn (M.Zhan)
Abstract: TA18 Titanium alloy tubes are materials with an obvious anisotropy. This
paper studied the variation laws of contractile strain ratio (CSR, an index for anisotropy)
of a TA18 high-strength titanium alloy tube by two computation methods. It is found
that variation laws and range of CSR at different locations are different. And the CSRs
obtained by two methods vary with strain similarly, but the CSR obtained by Method 2
is larger a little than that by Method 1 due to the thickness effect included in Method 2.
Keywords: anisotropy, CSR, tube bending, FEM, uniaxial tension
1. INTRODUCTION
As one kind of key lightweight components, titanium alloy tubes have been
increasingly used and have a promising future for finding a wider application in the
aerospace, aviation and related high-technology industries, due to their advantages of
high-pressure resistance, a high strength/weight ratio and long life, and so on. However,
titanium alloy tubes are materials with an obvious anisotropy. The anisotropy of tubes is
usually described by contractile strain ratio (CSR). A tube with CSR>1 means that the
deformation in width direction is easier than that in thickness direction. That is to say, it is
not easy for the material to thin or thicken.
The CSR relates to the preferred orientation of crystal, evolution of texture, strain
paths, plastic deformation degree, etc [Gilmour et al., 2004]. Xiangdong Wu et al.,
(2004) found that the CSR affected the plastic deformation behavior of anisotropic sheet
metal. And it also affected the springback of tubes after bending [JIANG Zhiqiang et al.,
2010]. Bending, bulging, flattening, and flaring are main plastic forming processes for
the application of titanium alloy tubes. During these plastic forming processes, titanium
alloy tubes will be deformed into various components under different stress states and
deformation degree. For example, in the NC bending process of titanium tubes, defects
of rupture of outer wall is due to over-thinning of outer wall under the action triaxial
tensile stress, and defects of wrinkling of inside wall is due to over-thickening of inside
121
Dortmund, Germany
wall
wall under the action
under the action of
oftriaxial
triaxialcompress
compressstressstress(shown
(shown in in Fig.1).
Fig.1). TheThe occurrence
occurrence of
of these
these
defectsdefects hasrelation
has the the relation
to thetovariation
the variation
of CSRof during
CSR during the plastic
the plastic formingforming process.
process. But the
But
CSRthe CSR isregarded
is always alwaysasregarded
a constant asindependent
a constant of independent of plasticprocess
plastic deformation deformation
in recent
researches
process on theresearches
in recent plastic forming
on the of titanium
plastic alloyoftubes,
forming which
titanium may
alloy result
tubes, in analysis
which may
error at
result in some degree.
analysis errorTherefore, it is essential
at some degree. to study
Therefore, variationto
it is essential laws of CSR
study of titanium
variation laws
alloy
of CSR tubes with plastic
of titanium alloydeformation
tubes with to develop
plastic the plastictoforming
deformation developofthe
titanium
plasticalloy
formingtubes
andtitanium
of extend applications
alloy tubes andof titanium alloy tubes. This
extend applications paper studied
of titanium the variation
alloy tubes. This ofpaperCSR
of a TA18 high-strength titanium alloy tube by experiment and FEM,
studied the variation of CSR of a TA18 high-strength titanium alloy tube by experiment and also analyzes
measurement
and FEM, and results.
also analyzes measurement results.
1200
1000
True Stress  (MPa)

800
Experimentation
600
FEM
400
200
0
0.00 0.02 0.04 0.06 0.08 0.10 0.12
True Strain  
Figure 1. Characteristic of tube bending Figure 2. Stress-strain curves of TA18 tubes
2. MATERIALS
2. MATERIALSAND
ANDMETHODS
METHODS
2.1. Material
2.1. Material
The tubes used are TA18 high-strength titanium alloy ones made by Western Titanium
The tubes used are TA18 high-strength titanium alloy ones made by Western Titanium
Company. The tubes are 12 mm in outside diameter and 0.9 mm in thickness. The true
Company. The tubes are 12 mm in outside diameter and 0.9 mm in thickness. The true
stress-strain curves of the tube were obtained by tensile testing, as shown in Fig. 2.
stress-strain curves of the tube were obtained by tensile testing, as shown in Fig. 2.
2.2. Methods
Tube2.2. Methods
specimen with a plug for each end were used to measure the CSR, as shown in
Tube
Fig.3.specimen
Uniaxialwith a plug
tensile for of
testing each end
tube were used
specimen wastocarried
measureoutthe
byCSR, as shown
a CMT5205 in Fig.
electronic
3. Uniaxialtesting
universal tensilemachine.
testing of tube
The specimen
tension wasiscarried
velocity out In
3mm/min. by experiment
a CMT5205a longitudinal
electronic
universal testing machine. The tension velocity is 3mm/min. In experiment a
extensometer and a transverse extensometer were used to record the variation of gauge
longitudinal extensometer and a transverse extensometer were used to record the
length and diameter of TA18 tube specimens during uniaxial tensile deformation. The
variation of gauge length and diameter of TA18 tube specimens during uniaxial tensile
transverse extensometer
deformation. The transversewasextensometer
fixed on the middle of tube
was fixed specimens.
on the middle of tube specimens.
To accurately reflect the variation laws of CSR of titanium tube in the plastic
To accurately reflect the variation laws of CSR of titanium tube in the plastic
deformation process, two methods are used in this study. Method 1: In accordance
deformation process, two methods are used in this study. Method 1: In accordance with
with aerospace standard of contractile strain ratio testing of titanium hydraulic
aerospace
tubing, thestandard
CSR can of contractile
be determinedstrainasratio testing of titanium
CSR=-Ec/(Ea+Ec) , in hydraulic
which, Ec,tubing,
Ea andtheEr
CSR can circumferential
are true be determined strain,
as CSR=-E /(Ea+Estrain
true caxial c) , inandwhich, Ec, Ea strain,
true radial and Erespectively.
r are true
circumferential strain, true
Method 2: According axialdefinition
to the strain andoftrue
theradial
CSR,strain,
the CSRrespectively. Method 2:as
can be determined
According to the definition of the CSR, the CSR can be determined as
122
Dortmund, Germany
CSR=ln[(D1+d1)/(D0+d0)]/ln(t1/t0), in which,
CSR=ln[(D1+d1)/(D0+d0)]/ln(t1/t0), in which, Do, Do,
do , todo,
andtoD1and
, d1,D1,
t1 ared1,outside
t1 arediameter,
outside
diameter, inside diameter
inside diameter and wall and wall thickness
thickness of tube
of tube before andbefore and after
after tensile, tensile, respectively.
respectively.
An
AnFEFEmodel
modelforforuniaxial
uniaxialtensile
tensile of
of aa TA18
TA18 tube
tube specimen
specimen was was established
established using
using
software
software Abaqus, as shown in Fig. 4. The gauge length of tube specimen isis 50mm.
Abaqus, as shown in Fig. 4. The gauge length of tube specimen 50mm. By By
simulating the deformation process of TA18 tube specimens in the uniaxial tension, the
simulating the deformation process of TA18 tube specimens in the uniaxial tension, the
variation of gauge length and diameter at different locations with time can be obtained.
variation of gauge length and diameter at different locations with time can be obtained.
Figure 3. Tube specimen and plug Figure 4. An FE model for

uniaxial tensile of a TA18 tube specimen
3. RESULTS AND DISCUSSIONS

3. RESULTS AND DISCUSSIONS
Fig. 5 shows variations of the CSR of end point of gauge length (Point A in Fig. 4)
Fig.the
and 5 shows variations
midst point of thelength
of gauge CSR of end point
(Point of gauge
B in Fig. lengththe
4) during (Point
FEMAsimulation
in Fig. 4) and
for
the midst
tensile point of gauge
deformation length
of a tube (Point B
specimen. in Fig.
From Fig.4)5,during
it can the FEMthat
be seen simulation for tensile
the variation laws
deformation
and range of CSR of a oftube specimen.
points Fromlocations
at different Fig. 5, itarecandifferent.
be seen that the variation
As shown in Fig. laws
5(a), and
the
range
CSR of of
theCSR
midstofpoint
points at different
(Point locations
B) increases are different.
from about 0.3 to 1.6Aswith shown in Fig. 5(a),
the increase the
of strain
CSR of
during thethe midstplastic
tensile point deformation,
(Point B) increases
howeverfrom the about
increase 0.3slope
to 1.6 withathe
varies lot increase
during the of
process. In the beginning
strain during the tensilethe increase
plastic is slow, in however
deformation, the midstthe stage the increase
increase slope isvaries
rapid,aandlot
atduring
final stage the increase
the process. In thebecomes
beginning sharp.
the As shownisin
increase Fig.5(b),
slow, in thethe CSRstage
midst at thetheend point
increase
(Point A) increases
is rapid, and at final at stage
first, then varies a becomes
the increase little and sharp.
decreases finally,in
As shown and CSR changes
Fig.5(b), the CSR in at
a
range of about
the end point 0.3 to 0.65.
(Point The different
A) increases variation
at first, laws of
then varies CSR and
a little at different
decreases locations
finally,may
and
beCSRresult from different
changes in a range plastic deformation
of about degrees.
0.3 to 0.65. The The more variation
different close to thelawsmidst of the
of CSR at
tube specimen the location is, the more thorough plastic deformation occurs. This leads to
different locations may be result from different plastic deformation degrees. The more
an obvious change of CSR in the midst.
close to the midst of the tube specimen the location is, the more thorough plastic
Fig. 6 shows variations of the CSR of the midst point of gauge length during the
deformation
experiment foroccurs.
tensileThis leads to an
deformation of obvious change ofAs
a tube specimen. CSR in the
seen frommidst.
Fig. 6, an increase
occursFig. 6 shows
in the beginning variations of thedecrease
and a slight CSR ofinthe themidst point of gauge
later deformation length
stage. Thisduring the
variation
experiment for tensile deformation of a tube specimen. As seen
trend is similar to that of the midst point by FEM. But the variation ranges of them are from Fig. 6, an increase
occurs inThe
different. the CSR
beginning
of theand
midsta slight
point decrease in the later
by FEM changes deformation
in the stage. This
range of 0.3-1.6 variation
and the CSR
bytrend is similar
experiment to that inofthe
changes therange
midstofpoint by FEM.
0.09-0.61. ThisBut the variation
difference rangesfrom
may result of them are
the slip
and displacement
different. The CSR change
of thebetween the sample
midst point by FEM and the extensometer.
changes in the range of 0.31-0.67 and the
CSRAsbyseen from Fig.changes
experiment 5 and Fig. 6, the
in the rangeCSR of values obtained
0.09-0.61. by two methods
This difference almost
may result equal
from the
toslip
each
andother in the former
displacement change plastic deformation
between the sample stage.
and But in the latter plastic deformation
the extensometer.
stage, Astheseen
CSRfrom obtained
Fig. 5 by
andMethod 2 becomes
Fig. 6, the CSR values larger than by
obtained thattwobymethods
Methodalmost1. Andequal
the
difference
to each otherbecomes
in the obvious as plastic
former plastic deformation
deformation stage.increasing.
But in the Thislatter may bedeformation
plastic due to the
thickness effect are included in Method 2.
stage, the CSR obtained by Method 2 becomes larger than that by Method 1. And the
difference becomes obvious as plastic deformation increasing. This may be due to the
thickness effect are included in Method 2.
123
Dortmund, Germany
1.6 0.8 0.8
1.4 0.7 0.7
1.2 0.6 0.6

Method 1
1.0 0.5 0.5
Method 2
Method 1
CSR
CSR
CSR
0.8 0.4 0.4
Method 1
0.6 0.3 Method 2 0.3
0.4 0.2
Method 2
0.2
0.2 0.1 0.1
0.0 0.0 0.0

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
True strain  True strain  True strain 

(a) Point B (b) Point A Figure 6. CSR -strain curve of
Figure 5. CSR -strain curves of TA18 tubes from FEM TA18 tubes from experiment
4.4.CONCLUSIONS
CONCLUSIONS
(1)The
(1) The variation
variation lawslaws
andand range
range of CSR
of CSR at different
at different locations
locations areare different.
different. TheThe
CSRCSRin
inthethemidst
midstincreases
increasesfaster
fasterand
andfaster
fasterwith
withthe
theincrease
increaseofof strain
strain during
during the
the tensile
tensile plastic
plastic
deformation. The CSR at the end increases at first, then varies a little and decreases finally,
deformation. The CSR at the end increases at first, then varies a little and decreases finally,
and the CSR changes in a range is less a lot than that in the midst. (2)The CSRs obtained
and the CSR changes in a range is less a lot than that in the midst.
by two methods vary with strain similarly, but the CSR obtained by Method 2 is larger a
(2) The
little than that CSRs obtained
by Method 1 duebyto two methods effect
the thickness vary with strainin similarly,
included Method 2.but the CSR
obtained by Method 2 is larger a little than that by Method 1 due to the thickness effect
included in Method 2.
ACKNOWLEDGEMENTS
ACKNOWLEDGEMENTS
The authors would like to thank the National Natural Science Foundation of China
(51175429), the Fund of New Century Excellent Talents in University (NCET-08-
The authors
0462) would like
, the Fundation to thank(JC201136),
of NWPU the National
theNatural
Fund ofScience Foundation
the State of China
Key Laboratory of
(51175429), the
Solidification Fund of in
Processing New Century
NWPU Excellentand
(KP200919) Talents in University
the Project (NCET-08-0462),
of 111(B08040).
the Fundation of NWPU (JC201136), the Fund of the State Key Laboratory of
Solidification Processing in NWPU (KP200919) and the Project of 111 (B08040).
REFERENCES
[Gilmour et al., 2004] K.R. Gilmour, A.G. Leacock, M.T.J. Ashbridge. The
REFERENCES
influence of plastic strain ratios on the numerical modelling of stretch forming.
[Gilmour
Journal
et al.,of ] K.R. Gilmour,
Materials
2004 ProcessingA.G. Leacock,152
Technology M.T.J. Ashbridge.
(2004) 116–125.The influence of
plastic strain Wu
[Xiangdong ratios on 2004]
et al., the numerical
Xiangdong modelling of stretch
Wu. Research on theforming. Journal of
plastic deformation
Materials
behaviorProcessing Technology
of anisotropic 152 (2004)
sheet metal under 116–125.
different loading paths[D]. Beihang
[Xiangdong Wu et al.,
University, 2004] Xiangdong Wu. Research on the plastic deformation behavior
2004.
of[JIANG
anisotropic sheet metal
Zhiqiang under2010]
et al., different loading
JIANG paths[D].Deformation
Zhiqiang. Beihang University，
laws of
2004.
mediumstrength titanium alloy thick-walled tubes in NC bending process [D].
[JIANG Zhiqiang et al.,
Northwestern 2010] JIANG
Polytechnical Zhiqiang.
University, Deformation laws of medium-strength
2010.
titanium alloy thick-walled tubes in NC bending process [D]. Northwestern
Polytechnical University, 2010.
124
Session 6: Tube Bending
Dortmund, Germany
Improving the Curvature of Bent Profiles by

Adjusting the Process Kinematics of the Three-Roll-
Pushbending Process
Engel, Bernd; Kopfer, Peter
Chair of Forming Technology,
University of Siegen, Paul-Bonatz-Str. 9-11, 57076 Siegen
peter.kopfer@uni-siegen.de
Abstract: Three-Roll-Pushbending is a conventional Freeform-Bending process to

manufacture bent profiles with variable radii and radii distributions along their longitudinal
axis. The shaping is done kinematically by varying the position of the forming-tool and
the axial feed of the profile. Measurements of the resulting curvature of the bent tube
have shown a characteristic periodical distribution over an extensive length of the part
until a constant radius is reached. Such inadvertent transition zones have to be avoided.
This paper presents a strategy for the Three-Roll-Pushbending to smooth and shorten the
transition zones by adjusting the process kinematics.
Keywords: Freeform-Bending, Three-Roll-Pushbending, process kinematics
1. INTRODUCTION
The task of forming technology is to manufacture parts of low weight providing a

maximum of functionality at the same time. Therefore formed profiles are increasingly
used these days [Gantner et al., 2007]. They feature a beneficial ratio of rigidity to weight
and therefore are commonly applied for lightweight construction. The manufacturing
processes require a high flexibility. They have to be both suitable for series production
as well as cost-efficient. Freeform-Bending processes meet these standards by bending
straight-lined profiles with variable radii and radii distributions along their longitudinal
axis.
2. PROCESS KINEMATICS OF THE THREE-ROLL-PUSHBENDING
The shaping of the Freeform-Bending-Processes is done kinematically. Although

there exist different methods of Freeform-Bending, the principle of the forming
procedure is basically the same. The profile is guided axially through stationary
supporting tools, either a cavity or rolls. The forming force is transmitted to the profile
by positioning a movable die (a cavity, a roll or a combination of rolls) relatively to the
127
Dortmund, Germany
stationary tools, while the profile is simultaneously fed forward in longitudinal direction
bystationary
the meanstools,
of a while theAt
pusher. profile is simultaneously
the University fed research
of Siegen forward in
is longitudinal
done on the direction
Three-
by the means of a pusher. At the University of Siegen research is done on the Three-
Roll-Pushbending (TRPB) process. TRPB is a Freeform-Bending process, which is - in
Roll-Pushbending (TRPB) process. TRPB is a Freeform-Bending process, which is - in
accordance with Plettke et al., 2010 - mainly used for bending geometries consisting of
accordance with Plettke et al., 2010 - mainly used for bending geometries consisting of
several plain bending curves on different bending planes. A schematic illustration is shown
several plain bending curves on different bending planes. A schematic illustration is
in shown
figure 1.
in figure 1.
Figure
Figure 1:1: Schematic
Schematic illustration
illustration ofof
thethe TRPB
TRPB process.
process.
The bending-roll (2), the supporting-rolls (4) and the pusher- and rotation-unit (5)
guideThe the
bending-roll
profile (1)(2), the supporting-rolls
in longitudinal direction.(4)The andforming-roll
the pusher-(3) andeffects
rotation-unit (5)
the bending
guide
moment the profile
by being (1)positioned
in longitudinal
towards direction.
the profile.The Using
forming-roll (3) effects
the pusher- the bending
and rotation-unit (5)
moment by being
the profile is fedpositioned
forward. towards
The TRPB the profile.
can be Using appliedtheonpusher- and rotation-unit
a conventional (5)
rotary-draw-
thebending
profile ismachine.
fed forward. TheThe TRPB can beisapplied
forming-roll mounted on aonconventional rotary-draw-bending
the bending-arm, whereas the
machine. The forming-roll
supporting-rolls and theis bending-roll
mounted on the bending-arm,
replace whereas
the pressure dietheandsupporting-rolls
the bend die,
andrespectively.
the bending-roll replace the pressure die and the bend die, respectively.
TheTheprocess
processkinematics,
kinematics,thethecoordinated
coordinatedmovements
movementsofofthe theforming-roll
forming-rolland andthethe
profile
profile during
during thethebending
bending process,
process,define
define thethecurvature
curvaturedistribution
distributionofofthethebent
bentprofile.
profile.
In In
simplified
simplified terms,
terms,two twokinematic
kinematic conditions
conditions cancanbebedistinguished.
distinguished.IsIsthetheposition
positionofofthethe
forming-roll
forming-roll changing
changing while
while thetheprofile
profileis isfed
fedforward
forwardconcurrently,
concurrently,a aradii
radiidistribution
distribution
is obtained.
is obtained. A constant
A constant radius
radius is being
is being bentbentif theif forming-roll
the forming-rollstaysstays stationary
stationary during during
the
the feed
axial axialoffeed of the In
the profile. profile.
additionIn toaddition to the positioning
the positioning of the forming-roll,
of the forming-roll, the deflection the
of deflection
the machine, of the
the machine,
springbacktheof springback
the profile and of thetheprofile and the
positioning positioning
accuracy accuracy
especially of
theespecially
forming-rollof the forming-roll
influence influence
the bending result.theThe bending
shorterresult. The shorter
the horizontal the horizontal
distance between
thedistance between
centre points ofthe
thecentre points ofand
forming-roll thetheforming-roll
bending roll andinthex-direction
bending roll dx,inthex-direction
bigger
x, the
is dthe biggerofisthe
deviation theresulting
deviationradiiof while
the resulting radii while
slightly changing theslightly changing
y-position. the y-
Therewith,
theposition.
inaccuracy Therewith, the inaccuracy
in positioning in positioning
the forming-roll the forming-roll
has an increasing negative hasimpact
an increasing
on the
negative
bending impact
result whenonthe thedistance
bendingdxresult when theFurthermore,
is decreasing. distance dx isthe decreasing.
shorter dx,Furthermore,
the bigger
is the
the shorter
needed dforming
x, the bigger is the needed forming force to effect the same bending
force to effect the same bending moment. Therefore a bigger
moment. Therefore a bigger distance dx is preferred.
distance dx is preferred.
Although constant radii should be bent, measurements by Engel and Kersten, 2010
Although constant radii should be bent, measurements by Engel and Kersten, 2010
of the resulting curvature of the bent tube (Fig. 2) have shown a characteristic periodical
of the resulting curvature of the bent tube (Fig. 2) have shown a characteristic periodical
distribution over an extensive length of the part until a constant radius is reached. This
distribution over an extensive length of the part until a constant radius is reached. This
128
Dortmund, Germany
transition zones
transition occur
zones with
occur every
with initial
every bending
initial step
bending when
step thethe
when straight profile
straight is loaded
profile is loaded
loaded
transition zones occur with every initial bending step when the straight profile is
withwith
the transmitted
the force
transmitted for
force the
for first
the time.
first time.
with the transmitted force for the first time.
Figure
Figure
Figure 2: Radii
2: Radii
2: Radii distribution
distribution
distribution along
along
along the
thethe longitudinal
longitudinal
longitudinal axis
axis
axis ofbent
of aof aa bent
bent profile.
profile.
profile.
From
From this
this point
point of
of view,
view, aa shorter
shorter distance
distance ddxx resulting
resulting in shorter
shorter transition zones is
From thisHowever,
favoured. point of view, a shorter
the forming-roll
forming-rolldistance
does dx
notresulting in in
necessarily shorter transition
have transition
to be
zones
zones
be fixed
fixed
is
in isone
one
favoured. However, the does not necessarily have to in
favoured. However,
position to
to bend the forming-roll
bend aa constant
constant radius. does not
radius. Regardingnecessarily
Regarding the the TRPB have to
TRPB process, be fixed
process, the in one position
the forming-roll
forming-roll maymay
position
to bend
as a constant radius. Regarding the TRPB process, the forming-roll may as well
as well
well move
move with
with an
an appropriate
appropriate speed
speed to
to the
the axial
axial feed
feed of
of the
the profile
profile on
on aa circular
circular path,
path,
move withisan
which appropriate
concentric speed
to the
the to the radius
bending axial feed
(Fig.of3).
the profile on a circular path, which is
3).
which is concentric to bending radius (Fig.
concentric to the bending radius (Fig. 3).
Figure 3: Comparison
Figure
Figure 3: of startup-strategies
3: Comparison
Comparison of A and
of startup-strategies
startup-strategies A B. B.
A and
and B.
The
TheThe movement
movement
movement of the
of the
of the forming-roll
forming-roll
forming-roll can
cancan be
be be achieved
achieved
achieved by
by by moving
moving
moving the
thethe bending-arm
bending-arm
bending-arm by
byby
the the
the value
value of βof
value β
β while
ofwhile advancing
advancing
while advancingthethe forming-roll
forming-roll
the in in
forming-roll direction
direction
in of of
direction thethe
of bending-arm
bending-arm
the bending-arm byby
thethe
by the
value
value
value of ∆y
of
of ΔyFR∆yFR (Fig.1).
(Fig.1).
(Fig.1).
FR TheThe
The on-line
on-line
on-line startup-strategy
startup-strategy
startup-strategy (Strategy
(Strategy
(Strategy B)
B)B) is
is is shown
shown
shown in
in in comparison
comparison
comparison
to the
to
to the the conventional
conventional
conventional linear
linear
linear startup-strategy
startup-strategy
startup-strategy A.
A. A. Several
Several
Several tests
tests
tests on
on on both
both
both strategies
strategies
strategies were
were
were run
runrun at
at at
129
Dortmund, Germany
the Chair of Forming Technology in Siegen (UTS) using defined tube specimen (diameter:
the Chair of Forming Technology in Siegen (UTS) using defined tube specimen
20 mm, wall thickness: 1 mm). The resulting radii distributions along the length of the
(diameter: 20 mm, wall thickness: 1 mm). The resulting radii distributions along the
tubes are illustrated in figure 4 by one example of each startup-strategy.
length of the tubes are illustrated in figure 4 by one example of each startup-strategy.
Figure 4: Resulting radii distributions for startup-strategies A and B.

Figure 4: Resulting radii distributions for startup-strategies A and B.
3. CONCLUSIONS
3. CONCLUSIONS
It can be stated, that the transition zone is shortened by more than one half using the
It can startup-strategy
on-line be stated, that the transition
B instead of zone is shortened linear
the conventional by more than one halfA.using
startup-strategy In the
the same
on-line startup-strategy
way the oscillationBamplitudes
instead of of
thethe
conventional linear
transition zone arestartup-strategy
reduced as well.A.Thus,
In thethe
same way the
contour oscillation
accuracy of amplitudes of the transition
the bent profile zone are reduced
can be improved as well. the
by adjusting Thus, the
process
contour accuracy
kinematics of themuch
without bent ado.
profile can be improved by adjusting the process kinematics
without much ado.
REFERENCES
REFERENCES
[Gantner et al., 2007] Gantner, P.; Harrison, D. K.; Silva, A. K. de; Bauer, H.; “The
Development
[Gantner et al., of a Simulation
2007] Gantner, P.;Model
Harrison,andD.the
K.;Determination
Silva, A. K. de;ofBauer,
the Die
H.; Control
“The
Data for the ofFree-Bending
Development a Simulation Technique.“;
Model and the In:Determination
Proceedings of the the Die
Institution
Control of
Mechanical
Data for the Engineers, Part Technique.“;
Free-Bending B: Journal of In:
Engineering Manufacture
Proceedings 221 (2), 2007,
of the Institution of
pp. 163–171.
Mechanical Engineers, Part B: Journal of Engineering Manufacture 221 (2), 2007,
[Plettke et al., 2010] Plettke, R.; Vatter, P. H.; Vipavc, D.; Cojutti, M.; Hagenah, H.
pp. 163–171.
(2010): et“Investigation
[Plettke on the
al., 2010] Plettke, R.;Process
Vatter, P.Parameters
H.; Vipavc,and
D.;Process
Cojutti, Window of Three-
M.; Hagenah, H.
Roll-Push-Bending.”;
(2010): “Investigation onIn: the
Proceedings of the 36thand
Process Parameters Matador
ProcessConference;
Window ofpp. 25-28;
Three-
14.-16. Jul. 2010, Manchester,
Roll-Push-Bending.”; England.
In: Proceedings of the 36th Matador Conference; pp. 25-28;
[Engel andJul.
14.-16. Kersten, 2010] Engel,England.
2010, Manchester, B.; Kersten, S.; “Sensitivitätsanalyse beim Freiform-
biegen and
[Engel von Rohrprofilen.”;
Kersten, 2010] In: 30. EFB-Kolloquium
Engel, B.; Kersten, S.; Blechverarbeitung;
“Sensitivitätsanalyse 2010,
beimBad
Boll, Germany.
Freiform-biegen von Rohrprofilen.”; In: 30. EFB-Kolloquium Blechverarbeitung;
2010, Bad Boll, Germany.
130
Dortmund, Germany
Influence factors of three-roll-push-bending

InﬂuenceChair
Factors ofVipavc,
R. Plettke, D. Three-Roll-Push-Bending
P. Vatter
of Manufacturing Technology,
Friedrich-Alexander-University Erlangen-Nuremberg,
R. Plettke, D. Vipavc, P. Vatter
Egerlandstr. 13, 91058 Erlangen, Germany
Chair of Manufacturing Technology, Friedrich-Alexander-University Erlangen-
d.vipavc@lft.uni-erlangen.de
Nuremberg, Egerlandstr. 13, 91058 Erlangen, Germany d.vipavc@lft.uni-erlangen.de
Abstract: Three-roll-push-bending is a flexible, kinematic bending process for tubes. The

bending tools
Abstract: can be positioned to is
Three-roll-push-bending achieve the bending
a flexible, geometry
kinematic bending by process
pushing forthrough
tubes.and
rotating
The the tube.
bending tools This
can beprocess has many
positioned influencing
to achieve the factors
bendingwhich can bebydivided
geometry pushing into
controland
through androtating
disturbance factors.
the tube. ThisThe paperhas
process willmany
present the known
influencing factors.
factors A simulation
which can be
based study
divided of the infl
into control anduence of friction
disturbance between
factors. Thebending tool present
paper will and tubethein three-dimensional
known factors.
Afree form bending
simulation will beofdiscussed
based study in detail.
the influence In order
of friction to investigate
between bendingthistoolinfl
anduence
tubea in
new
simulation modelfree
three-dimensional based on bending
form the finitewill
element method in
be discussed was built In
detail. uporder
and validated against
to investigate
experimental
this influence avalues.
new simulation model based on the finite element method was built up
Keywords:
and validated against experimental values.
Three-roll-push-bending, finite element method, influencing factors
Keywords: Three-roll-push-bending, finite element method, influencing factors
1. INTRODUCTION
1. INTRODUCTION
Bent metal tubes find widespread application in many sectors of industry. A new free
Bent
formmetal tubesprocesses
bending find widespread application in many
is three-roll-push-bending andsectors of industry.
has been researched A new free
by Gerlach
form bending
[Gerlach, processes
2010] is three-roll-push-bending
and Merklein [Merklein et al., and has While
2009]. been researched
this process by Gerlach
is flexible
[Gerlach,
and able2010] and Merklein
to produce non-constant 2D- et
[Merklein andal.,3D-curvatures,
2009]. While this process
it was shown is flexible and
that numerous
able to produce
influencing non-constant
factors 2D- predictability
lead to poor and 3D-curvatures,
of the itprocess.
was shown that numerous
A possible solution to
influencing
compensatefactors
this islead to poor predictability
the integration of the process.
of optical measurement A possible
devices solution
[Kim et al., 2009]toand
compensate integration of optical measurement devices [Kim et
automatic correction. This trial-and-error approach leads to a certain amount of and
this is the al., 2009] waste
automatic correction. This trial-and-error approach leads to a certain amount of waste
material and production downtime. To avoid this a numerical model of the three-roll-
material and production downtime. To avoid this a numerical model of the three-roll-
push bending process has been developed. The working principle of the three-roll-push-
push bending process has been developed. The working principle of the three-roll-push-
bending process, which bases on two holding rolls, one bending roll and one setting roll,
bending process, which bases on two holding rolls, one bending roll and one setting roll,
is isbased
basedononP-,P-,Y-,
y-,C-C-and
andA-axis,
A-axis,asasshown
shownininFigure
Figure1.1.
A-axis
Holding rolls
C-axis
Bending roll
Setting roll
P-axis
Y-axis
Figure 1; Tools and axis of the three-roll-push-bending machine.
Figure 1; Tools and axis of the three-roll-push-bending machine.
131
Dortmund, Germany
An arc as a bending geometry is defined depending on the position of the setting

roll given by P- and Y-axis and the feeding rate of the C-axis. Investigations have shown
that the bending radius as target value by bending an arc is affected by many influences.
For better comprehension of these factors Taguchi´s method has been applied. These
factors were classified after G. Taguchi [Kleppman, 2006] in two groups of control and
disturbance factors. The identification of the most significant influencing factors was done
after extensive experimental work had been finished and discussed by Plettke [Plettke
et al., 2010]. These factors have been published and discussed by Hagenah [Hagenah et
al., 2010]. The primary aim of Taguchi´s method is to minimize variations in the target
value caused by disturbance factors which are present in the process. The process is then
said to have become robust. This can be applied to three-roll-push-bending processes by
evaluating the effect of each single disturbance factor on the target value. For this sake the
numerical investigation of the process has been done.
Numerical investigation of the process has begun using simple finite element models,
where only the kinematic of the P-, Y- and C-axis has been modeled. This model suffered
from a deviation up to 20% between simulated and experimentally defined bending
radius R [Cojutti et al., 2009]. To improve simulation results the machine deflection as
disturbance factor has been simulated in the second numerical model [Hagenah et al.,
2010]. Respecting the machine deflection, the predictability of the process was improved
by 10% and this gives an overall precision of about 90% in prediction of the FE-model.
The remaining 10% are caused by further disturbance factors and therefore lead to
variations in the target value.
All investigations up to now were focused on the bending geometry arc, with the
bending radius R being the target value, which can be bent by setting the P-, Y- and C-
axis. Beside these axes the tube can also be rotated during the bending process by the
A-axis, (see Figure 1). To investigate the influence of the tube´s rotation by the A-axis, the
FE-model was enhanced and validated against experiments.
2. SIMULATION OF THE THREE-DIMENSIONAL FREE FORM BENDING
The three-roll-push-bending process allows us to bend different kinds of bending

geometries. These geometries cannot necessarily be described by a set of bending radii
in all cases. Especially when employing the A-axis, the need for the redefinition of the
general target value, other than the bending radius R, has appeared. Investigating three-
dimensional bent tubes needs the definition of the bending result not only in terms radius,
arc length and angle, but also in terms of torque. To understand what happens during
the bending process two simplified models have been set up. These models distinguish
two-dimensional free form bending (2DFB) from three-dimensional free form bending
(3DFB). The 2DFB model allows only the observation of the bending moments Mb
produced by the position of the setting roll, while in 3DFB an additional deformation
mode, i.e. the torque Mt of the tube around its axis of symmetry, produced by the A-axis,
can be defined, Figure 2.
132
Dortmund, Germany
Mb Compression Mb Compression
Mb
y
y Mb φ
x
x
Tension z Tension Mt
Figure2;
Figure 2;Left
Lefthand
handside
sidepicture
pictureshows
shows2DFB
2DFBand
andright
righthand
handside
side3DFB.
3DFB.
As
As result
result ofof the
the torque
torque aa helix
helix will
willbe bebent [Soleretetal.,
bent[Soler al.,2006].
2006].ThisThisbending
bendinggeometry
geometry
can
can be now used used totoinvestigate
investigatethe theinfluencing
influencingfactors factors during
during 3DFB.
3DFB. TheThe target
target value
value will
will
be thebe comparison
the comparison of theof position
the position of end
of the the endpartpart of the
of the helixhelix in reality
in reality andand model.
model. The
The difference
difference shallshall be minimized.
be minimized.
The experiments were
The experiments wereconducted
conductedonon a three-roll-push-bending
a three-roll-push-bending machine
machine Wafios Wafios
BMZ
BMZ
61. The 61.working
The working principleprinciple
has been hasexplained
been explained
in detailinindetail in [Hagenah
[Hagenah et al.,The
et al., 2010]. 2010].
tube
The
usedtube used
in this in this experiment
experiment was madewas made ofsteel
of carbon carbonSt 37 steel
withSt an
37outer
with an outer diameter
diameter of 20mm
of
and20mm
a walland a wall thickness
thickness of 1mm. A of helix
1mm.was A helix
chosen wasaschosen as target geometry.
target bending bending geometry.
Since this
Since this was a preliminary investigation a helix
was a preliminary investigation a helix with a randomly defined radius with a randomly defined radius and
of 100mm of
100mm
torsion of and80°torsion
for theof 80° for
feeding the feeding
length of 400mm length of 400mm
was chosen. was chosen.
To bend the selectedTo bend the
geometry,
selected geometry, the process parameters for P- and Y-axis were defined based on
the process parameters for P- and Y-axis were defined based on characteristic lines and set
characteristic lines and set to P=31.5° and Y=37.43°. In order to achieve the torsion, the
to P=31.5° and Y=37.43°. In order to achieve the torsion, the rotation of the A-axis was
rotation of the A-axis was programmed accordingly.
programmed accordingly.
For the numerical investigation of 3DFB the improved FE-model presented by
For the
[Hagenah numerical
et al., 2010] was investigation
used as a basis. of 3DFB the improved
This FE-model uses theFE-model
programpresented
Abaqus 6.8 by
[Hagenah et al., 2010] was used as a basis. This FE-model
and is solved by the Abaqus explicate solver. Since it was meant to simulate the 2DFB, uses the program Abaqus 6.8
and isadaptions
some solved byhad the to
Abaqus explicate
be carried solver.
out. The Since 2DFB
original it was FE-model
meant to simulate
was remodeledthe 2DFB, to
include also the lower half of the geometry, the symmetrical plane was removed. Sinceto
some adaptions had to be carried out. The original 2DFB FE-model was remodeled
includedemands
3DFB also the lowerrotation half
of ofthethetube,
geometry, the symmetrical
the kinematic of the tube planehaswas removed.
been updated Since
to
3DFB demands
simulate the A-axis.rotation of the parameters
All other tube, the kinematic of the tube
of the FE-model has beenthe
remained updated
same as to already
simulate
the A-axis.inAll
described other parameters
[Hagenah et al., 2010].of the FE-model remained the same as already described
in [Hagenah
To compare et al., the
2010].simulation results with the experiment, the bent helix was
To compare
measured with thetheoptical
simulation results with
measurement the experiment,
system ATOS I [Bergmann the bent helix
et al.,was measured
1997]. The
result
with thewasoptical
an STL-model.
measurement Thissystem
model ATOS
was compared
I [Bergmann to theetcomputational
al., 1997]. The model.
resultThewas
two models were
an STL-model. Thisregistered
model was in one coordinate
compared to the system. The partsmodel.
computational of the Thetubetwothatmodels
were
clamped and not in
were registered bentonewere fit to eachsystem.
coordinate other using
The partsa bestof fit the
algorithm
tube that[Schneider
were clampedet al.,
2008].
and not Doing
bent were this fita todifference
each otherofusing 29.7a bestmm fitbetween
algorithm computation-Abaqus
[Schneider et al., 2008]. and
measurement-STL
Doing this a difference can beofseen,29.7asmm shown in Figure
between 3 on the left handand
computation-Abaqus side. In order to
measurement-
reduce
STL canthebe deflection
seen, as shown the 3DFBin FigureFE-model
3 on the was leftfurther
hand side.improved
In order by tointroducing
reduce the
anisotropic
deflection the friction
3DFB conditions.
FE-model The was
rotation of theimproved
further tube causes bysliding-friction μg on the
introducing anisotropic
walls
frictionof conditions.
rolls besides Thethe roll-friction
rotation of the μtube
r. Tocausessimulate this frictionμg
sliding-friction phenomenon
on the wallstheof
anisotropic
rolls besides friction model for deformable
the roll-friction μr. To simulate surfaces this[Pabst
frictionet al., 2010] has the
phenomenon beenanisotropic
adopted.
The roll friction was defined as μ
friction model for deformable surfaces [Pabst et al., 2010] has been adopted. Theroll
r = 0.05, which is comparable to the case of a steel roll
rolling
frictionon wasa steel
definedplateas[Fischer
μr = 0.05, et al., 2005].
which For the sliding
is comparable friction
to the case of a value
a steelof roll
μg =rolling
0.25
was
on adefined
steel platewhich is similar
[Fischer et al.,to2005].
the friction
For thecase whenfriction
sliding a steela plate
valueslides
of μg on another
= 0.25 was
steel
definedplate [Fischer
which et al., to
is similar 2005]. On the right
the friction case handwhen side of figure
a steel 3 the improvement
plate slides on another steel of
plate [Fischer et al., 2005]. On the right hand side of figure 3 the improvement of
133
Dortmund, Germany
the results
the results in terms
in terms of maximal
of maximal deflection
deflection whichwhich was reduced
was reduced to 24.7mm
to 24.7mm onend
on the theof
end of
the Helix.
the Helix.
Max deflection Max deflection
29.7mm 24.7mm
STL Abaqus STL Abaqus
Figure 3; Comparison between experiment and computational model – left hand side
Figure 3; Comparison between experiment and computational model – left hand side
with isotropic friction model, right hand side with anisotropic friction model.
with isotropic friction model, right hand side with anisotropic friction model.
3. SUMMARY AND OUTLOOK

3. SUMMARY AND OUTLOOK
Considering friction this numerical based investigation of 3DFB shows noticeable
Considering
improvement friction this numerical
in computational precision based investigation
of the of 3DFB
FE-model. shows noticeable
Respectfully this
improvement
phenomenon in computational
cannot be neglected.precision
In futureof the
workFE-model. Respectfully
this preliminary this phenomenon
model will be
cannotbybe
improved neglected.the
determining Inprecise
future friction
work this preliminary
coefficients model will be improved by
experimentally.
determining the precise friction coefficients experimentally.
REFERENCES
REFERENCES
[Gerlach, 2010] Gerlach, C.: “Ein Beitrag zur Herstellung definierter Freiformbiegegeometrien bei Rohren und Profilen”.
Dissertation, Universität Siegen, Shaker Verlag Aachen 2010, ISBN 978-3-8322-9484-7
[Merklein et al., [Gerlach,
2009] Merklein, 2010] Gerlach,
M.; Hagenah, C.: M.:“Ein
H.; Cojutti, Beitrag
“Investigation zur Herstellung
on Three-Roll definierter
Bending of Plain Tubular
Freiformbiegegeometrien
Components”. bei Rohren
In: Key Engineering Materials 410-411(2009), und Profilen”. Dissertation, Universität
S. 325-334
[Kim et al., 2008] Kim, H.-S.; Naranbaatar, E.; Ahammad, S.A.; Bae, Y.-H.; Lee, B.-R.: “Real-time forming error inspection
Siegen,
system using computerShaker
vision forVerlag Aachen
small-sized tubes”. In:2010, ISBN
Proceedings 978-3-8322-9484-7
of ICROS-SICE International Joint Conference 2009,
S. 160.
[Merklein et al., 2009] Merklein, M.; Hagenah, H.; Cojutti, M.: “Investigation on
[Kleppman et al., 2006] Kleppman, W.; “Taschenbuch Versuchsplanung, Produkte und Prozesse optimieren”. 4. Überarbeitete
Three-Roll
Auflage, Carl Hanser Verlag,Bending
München 2006,of Plain Tubular Components”. In: Key Engineering Materials
ISBN 978-3-446-40617-9
[Plettke et al., 2010] Plettke, R.; Vatter, P. H.; Vipavc, D.; Cojutti M.; Hagenah. H.: “Investigation on the Process Parameters and
410-411(2009),
Process Window S. 325-334
of Three-Roll-Push-Bending”. In: Proceedings MATADOR 36th Conference, Manchester 2010, S. 25-28
[Hagenah et al.,[Kim et al., 2008]
2010] Hagenah, H.; Vipavc,Kim, H.-S.;R.;Naranbaatar,
D.; Plettke, E.; Ahammad,
Merklein, M.: “Numerical Model of Tube S.A.; Bae, Y.-H.;
Freeform Bending by Lee,
Three-Roll-Push-Bending”. In: Proceedings 2nd EngOpt International Conference on Engineering Optimization, Lisabon
2010 B.-R.: “Real-time forming error inspection system using computer vision for
small-sized
[Cojutti et al., 2009] Cojutti, M.;tubes”.
Vipavc, D.; In:Hagenah,
Proceedings of M.:
H.; Merklein, ICROS-SICE International
“An Innovative Approach JointDesign
for the Process Conference
for
Three-Roll Bending of Plain Tubular Components”. In: Proceedings of the International Congress on Efficient Rollforming,
2009,
Bilbao 2009, S. 160.
S. 133-139.
[Kleppman
[Soler et al., 2006] et al.,
Soler, J. M.; Rangel, R. 2006] Kleppman,
H.: “Geometrical W.; “Taschenbuch
characterization
Part C: J. Mechanical Engineering Science, IMechE 2006, S. 1831-1841
Versuchsplanung,
of the canted coil springs. Proc”. In: IMechE Vol.Produkte
220
[Bergmann et al., und1997]

Prozesse
Bergmann,optimieren”.
D.; Galanulis, K.; 4. Winter,
Überarbeitete
D.: “Advanced Auflage, Carl Hanser Verlag,
3D-Fringe-Projection-Systems”. GOM München
mbH,
Braunschweig 1997, S. 1-8.
2006, ISBN 978-3-446-40617-9
[Schneider et al., 2008] Schneider, M.; Friebe, H.; Galanulis, K.: “Validation and optimization of numerical simulation by optical
[Plettke
measurements of tools et
and al.,
parts”.2010] Plettke,
In: Proceedings R.; Vatter,
of International DeepP. H.; Vipavc,
Drawing D.; IDDRG,
Research Group Cojutti M.; Hagenah.
Olofström 2008,
S. 1-12.
H.:Pabst,
[Pabst et al., 2010] “Investigation
S.; Thomaszewski, on B.;
theStraßer,
Process Parameters
W.: “Anisotropic andforProcess
Friction DeformableWindow
Surfaces andofSolids”.
Three-Roll-
In:
ProceedingsPush-Bending”. In: ProceedingsSymposium
of the 2009 ACM SIGGRAPH/Eurographics MATADOR 36thAnimation,
on Computer Conference,
New YorkManchester
2010, ISBN: 978-2010,
1-60558-610-6
S. 25-28
[Fischer et al., 2005] Fischer, U., et.al.: “Tabellenbuch Metall”. 43. Auflage, Verlag Europa Lehrmittel, Haan-Gruiten 2005, ISBN:
3-8085-1723-9
134
Dortmund, Germany
[Hagenah et al., 2010] Hagenah, H.; Vipavc, D.; Plettke, R.; Merklein, M.:
“Numerical Model of Tube Freeform Bending by Three-Roll-Push-Bending”. In:
Proceedings 2nd EngOpt International Conference on Engineering Optimization,
Lisabon 2010
[Cojutti et al., 2009] Cojutti, M.; Vipavc, D.; Hagenah, H.; Merklein, M.: “An
Innovative Approach for the Process Design for Three-Roll Bending of Plain
Tubular Components”. In: Proceedings of the International Congress on Efficient
Rollforming, Bilbao 2009, S. 133-139.
[Soler et al., 2006] Soler, J. M.; Rangel, R. H.: “Geometrical characterization of the
canted coil springs. Proc”. In: IMechE Vol. 220 Part C: J. Mechanical Engineering
Science, IMechE 2006, S. 1831-1841
[Bergmann et al., 1997] Bergmann, D.; Galanulis, K.; Winter, D.: “Advanced
3D-Fringe-Projection-Systems”. GOM mbH, Braunschweig 1997, S. 1-8.
[Schneider et al., 2008] Schneider, M.; Friebe, H.; Galanulis, K.: “Validation
and optimization of numerical simulation by optical measurements of tools and
parts”. In: Proceedings of International Deep Drawing Research Group IDDRG,
Olofström 2008, S. 1-12.
[Pabst et al., 2010] Pabst, S.; Thomaszewski, B.; Straßer, W.: “Anisotropic
Friction for Deformable Surfaces and Solids”. In: Proceedings of the 2009 ACM
SIGGRAPH/Eurographics Symposium on Computer Animation, New York 2010,
ISBN: 978-1-60558-610-6
[Fischer et al., 2005] Fischer, U., et.al.: “Tabellenbuch Metall”. 43. Auflage, Verlag
Europa Lehrmittel, Haan-Gruiten 2005, ISBN: 3-8085-1723-9
135
Dortmund, Germany
136
Imprint:
Institute of
Forming Technology and
Lightweight Construction
IUL - Institute of Forming Technology

and Lightweight Construction
Technische Universität Dortmund
Baroper Strasse 301
44227 Dortmund
Germany
Telefon: +49 (0) 231 755 2660
Telefax: +49 (0) 231 755 2489
http://www.iul.eu
Professor Dr.-Ing. Matthias Kleiner

Professor Dr.-Ing. A. Erman Tekkaya
Editorial office:
Dipl.-Ing. Andres Weinrich
Organization team:
Dipl.-Ing. Andres Weinrich
Dipl.-Ing. Matthias Hermes
Dipl.-Ing. Christoph Becker
M.Sc. M.Eng. Pleul Christian
M. Sc. Mohammad Malekipour Gharbi
Dipl.-Ing. Daniel Staupendahl
Dipl.-Des. Patrick Cramer

Tagungsunterlagen Dorp Eldorado

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tagungsunterlagen Dorp Eldorado

Uploaded by

Copyright:

Available Formats

technische universität

1st International Tube and Profile

4th DORP 2011

Tube Bending of New Generation by MOS Bending Machine 9

Session 1: Nonconventional Processes 19

Machine Concept for Incremental Tube Forming 21

Tube Bending in Power Plants 25

Session 2: Standardization in Bending Technology 47

VDI 3430: A Guideline for Rotary Draw Bending of Profiles 49

Standard for tube and profile bending technology 51

Session 3: Profile Manufacturing 55

An Analysis of Bend Allowance Prediction in Folding 57

Curvature Optimized 3D-Profiles to Improve the Line

Session 4: Profile Bending 83

Process Limit Extension of the TSS-Bending Process 85

Measurement System for Three Roll Bending of Beams 89

Session 5: Measurement and Simulation Techniques 99

Numerical Simulation of Profile Bending Using the TSS

Measurement of Bent Parts for the Correction of

Variation of Contractile Strain Ratio of TA18 Titanium

Alloy Tubes with Plastic Deformation 121

Session 6: Tube Bending 125

Improving the Curvature of Bent Profiles by Adjusting the

Influence Factors of Three-Roll-Push-Bending 131

Prof. Dr.-Ing. A. Erman Tekkaya

Tube Bending of New Generation by MOS

2. Mechanism of MOS bending by CNC ( Computer Numerical

Photo 1: New CNC bender

Photo 1: New CNC bender

2.2 CNC Tube Bending Machine

①Tubular workpice ②Bending die

Figure 4; Relationship between offset Figure 5; Relationship between

Figure 7; Flow chart of control for MOS bending.

Figure 9; Virtural pictures of bent tube

5. Benefits of tube bending by MOS method

Inputting the coordinates of bending section is also available besides ordinary

[1] Takahashi, A.; Japan Society for Techno. of Plasticity(JSTP),106th Symposium-

Machine Concept for Incremental Tube Forming

Abstract: Incremental Tube Forming (ITF) is a recently developed process to manufacture

In many industrial sectors the demand of three-dimensionally bent tubes increases

Figure 1: Lightweight structures [Chatti, 2005]

Figure 2: Process principle of ITF process

3. NEW MACHINE CONCEPT

Figure 3: Machine concept and CAD-plot of new machine [Hermes, 2011]

4. CONCLUSION AND OUTLOOK

[Chatti, 2005] Chatti, S.: Production of Profiles for Lightweight Structures,

Schwarze - Robitec GmbH

Olpener Strasse 460-474

Tel: +49 221 890 080

1903 Foundation of Gebrüder Schwarze

since 1930 Manufacture of tube bending machines (Peter Wirtz)

1953 Acquisition of Fa. Peter Wirtz, since then Schwarze-Wirtz

1962 First booster bender in the world developed together with

VGB and Babcock and delivered to Sweden

1965 Supply of first NC controlled automatic tube bender

1996 Merger with Robitec, since then Schwarze-Robitec

Customer orientated, qualified and motivated staff

Decades of experience in the manufacture and

Turnover 2010: Đ 20 million

Number employees: approx. 120

Surface area: approx. 43.000 m2

Workshop area: approx. 12.000 m2

Our experience for your future

All our machines are designed in such a