Sheet Metal Profiles With Variable Height: Numerical Analyses On Flexible Roller Beading

Journal of
Manufacturing and
Materials Processing
Article
Sheet Metal Profiles with Variable Height: Numerical
Analyses on Flexible Roller Beading
Tianbo Wang * and Peter Groche
Institute for Production Engineering and Forming Machines, 64287 Darmstadt, Germany;
groche@ptu.tu-darmstadt.de
* Correspondence: wang@ptu.tu-darmstadt.de

Received: 18 December 2018; Accepted: 21 January 2019; Published: 1 February 2019
Abstract: In the spirit of flexible manufacturing, the novel forming process “flexible roller beading”
was developed, which allows the incremental production of height-variable sheet metal profiles.
After designing the process and realizing a test facility for flexible roller beading, the feasibility was
experimentally shown. The following step addresses the expansion of the process limits. With this
aim, the mechanical behavior of the sheet metal during the process was investigated by means of FEA.
Due to the variable cross-section development of the sheet metal profile, a multidimensional stress
distribution was identified. Based on the present state of stress and strain, conclusions about the
origin of appearing defect formations were drawn. Observed defects were sheet wrinkles as a result
of compressive stresses in the profile flange and material thinning in the profile legs and bottom due
to unintendedly exceeding tensile stresses. The influences of the forming strategy as well as tool- and
workpiece-side variations on the quality of the manufacturing result were investigated. From the
results of the analyses, measures to avoid component failure were derived. Given all the findings,
guidelines were concluded that are to be considered in designing the forming sequence. With the
insights into the occurring processes and the mastery of this novel forming process, important
contributions are made to its industrial suitability. The approach of lightweight and load-oriented
component design can be extended by realizing new families of sheet metal profiles. With respect to
Industry 4.0, on-demand manufacturing is increasingly required, which is why flexible roller beading
is of substantial relevance for the industrial sheet metal production.
Keywords: flexible manufacturing; load-adapted component design; sheet metal profiles; flexible
roller beading; flexible roll forming
1. Introduction
As a consequence of the so-called fourth industrial revolution or Industry 4.0 [1], the trend is
moving away from conventional mass production [2]. The industrial change is moving towards mass
customization, which implies increasing component variation and changeovers. Consequently, the
requirements for the manufacturing systems are higher flexibility and on-demand manufacturing [2].
This change is accompanied by new manufacturing challenges and the further development of
production technologies. The realization of several families of parts on the same system with shortened
changeover time constitutes one objective of flexible manufacturing systems [3]. Another purpose
of flexible processes is the economization of material expenditure by producing load-oriented
components [4]. According to Yang et al. [2], the flexibility level can be raised by “increasing the degree
of freedom by using non-dedicated tools and by varying forming path”. In incremental profile forming,
this means the implementation of flexible roll movements. The most widespread manufacturing
process for cold-rolled profiles is roll forming as a pure bending process [5]. The product family of
conventional roll forming is limited to sheet metal profiles with constant cross-sections along the
J. Manuf. Mater. Process. 2019, 3, 19; doi:10.3390/jmmp3010019 www.mdpi.com/journal/jmmp

J. Manuf. Mater. Process. 2019, 3, x FOR PEER REVIEW 2 of 15
J. Manuf. Mater. Process. 2019, 3, 19 2 of 15
bending process [5]. The product family of conventional roll forming is limited to sheet metal
profiles with constant cross-sections along the longitudinal direction. Flexibility in roll forming can
longitudinal
be achieved direction.
by using Flexibility
movable in rollwhich
rolls forming can bethe
enable achieved by using
production of movable rolls which
width-variable enable
transverse
the production of width-variable transverse sections [6,7]. Lindgren
sections [6,7]. Lindgren et al. [8] presented a process extension by manufacturing profiles with et al. [8] presented a process
extension
conical by manufacturing
width and depth development.profiles with Theconical
profilewidth
flanges and depth
were development.
formed alternatelyThe andprofile flanges
the gradient
were formed alternately and the gradient of the conically shaped height
of the conically shaped height contour was constant. In the work of Sedlmaier et al. [9], an approach contour was constant. In the
to produce height-variable profiles is presented. In this process, the sheet metal is clamped in a this
work of Sedlmaier et al. [9], an approach to produce height-variable profiles is presented. In die
process, the sheet metal is clamped in a die with prefabricated shape,
with prefabricated shape, where the profile bottom already takes the final height-variability. where the profile bottom already
takes the final height-variability.
Subsequently, roll forming tools Subsequently,
gradually form roll forming tools gradually
the profile flanges by form the profile
moving alongflanges
flexible by
moving along
trajectories. flexible
In this trajectories.
approach, In this
the profile approach,
height contour theisprofile
predefinedheight bycontour
the given is die-shape
predefinedand by the
the
given die-shape and the
profile length is limited as well. profile length is limited as well.
The developed
The developedincremental
incremental forming
forming process called
process “flexible
called roller beading”
“flexible expandsexpands
roller beading” the potentialthe
of lightweight design by manufacturing sheet metal profiles
potential of lightweight design by manufacturing sheet metal profiles with height-variablewith height-variable cross-sections [10].
The flexible manufacturing
cross-sections [10]. The flexible system enables the production
manufacturing system enables of novel theprofile familiesofbynovel
production enabling the
profile
customization of the profile height contours. In this case, the shaping
families by enabling the customization of the profile height contours. In this case, the shaping of the of the height contour of the
profile contour
height is not given
of theby profile
the toolisgeometry
not givenbut by the
the forming tool paths.
tool geometry but theThis allows short
forming changeover
tool paths. times
This allows
and the ability of on-demand production, which comply with the
short changeover times and the ability of on-demand production, which comply with the trend of trend of the modern manufacturing
industry.
the modern The manufacturing
resulting sheet metal components
industry. can be used
The resulting sheet as load-oriented
metal components carrier profiles
can be as well as
used as
semi-finished products with height-variable beads [10]. The beads in
load-oriented carrier profiles as well as semi-finished products with height-variable beads [10]. The the semi-finished sheet metals
can beinused
beads the as material buffers
semi-finished sheet in downstream
metals can be used forming processes.
as material Due in
buffers to downstream
the height-variability
forming
of the beads, varying amounts of material can be provided at critical
processes. Due to the height-variability of the beads, varying amounts of material can be provided at positions of the component.
The aimpositions
critical of using of these
thekinds of semi-finished
component. The aim of products is to expand
using these kinds of thesemi-finished
limits of forming processes
products is to
like hydroforming.
expand the limits of forming processes like hydroforming.
The manufacturing
The manufacturingofof profiles withwith
profiles inconstant cross-sectional
inconstant developments
cross-sectional involvesinvolves
developments the presence the
of a multidimensional state of stress during the process. In flexible
presence of a multidimensional state of stress during the process. In flexible roll forming, roll forming, longitudinal strains
are necessarystrains
longitudinal to obtain arethe intendedto
necessary profile
obtain geometry in the transition
the intended zone, where
profile geometry in the thetransition
profile changes
zone,
its cross-section [11]. The transition area is divided into a stretching zone,
where the profile changes its cross-section [11]. The transition area is divided into a stretching zone, where the profile shape is
concave and a compression zone with a convex contour. The control
where the profile shape is concave and a compression zone with a convex contour. The control over over the appearing compressive
longitudinal
the appearingstresses avoids longitudinal
compressive component defects stresses in avoids
the form of flange buckling
component defects in [12].
theCurrent
form ofresearch
flange
addresses the development of countermeasures to reduce the
buckling [12]. Current research addresses the development of countermeasures to reduce geometric deviations resulting from thethe
multidimensional
geometric deviations stress distribution
resulting from the [13].multidimensional
To optimize the novel stressprocess of flexible
distribution [13]. roller beading,
To optimize thea
fundamental understanding of the appearing stress and strain distribution
novel process of flexible roller beading, a fundamental understanding of the appearing stress and during the process is to
be built and the causes of occurring errors are to be investigated.
strain distribution during the process is to be built and the causes of occurring errors are to be During initial investigations [10],
the process limits
investigated. During were detected
initial in terms of[10],
investigations wrinkles, sheet limits
the process thinning,were and other geometric
detected in terms of instabilities.
wrinkles,
sheet thinning, and other geometric instabilities. So far, the producibility of height-variable order
So far, the producibility of height-variable profiles has been proven by laboratory tests. In profiles to
analyze the component behavior and to reduce the expenditure on
has been proven by laboratory tests. In order to analyze the component behavior and to reduce the parameter variations for process
optimizationon
expenditure a numerical
parameterstudy is carried
variations out. optimization a numerical study is carried out.
for process
2. Materials and Methods

2. Materials and Methods
The target component is a hat profile with a customizable profile depth course. The geometric
The target component is a hat profile with a customizable profile depth course. The geometric
characteristics and nomenclature are shown in Figure 1.
characteristics and nomenclature are shown in Figure 1.
z transverse section
y
bending angle
α flange
profile
z depth legs
x
y profile
(a) (b) bottom (c)
Figure 1. Characteristics and nomenclature of the target geometry.

Figure 1. Characteristics and nomenclature of the target geometry.
A numerical model of the novel process is created in order to characterize the present stress
A numerical
distribution model
during of the novel
the process. In thisprocess is created
way, possible in order
types to characterize
of component failure the present stress
are identified and
distribution during the process. In this way, possible types of component failure are identified and
the according defect emergences are determined. By identifying the effects of process-, tool- or
the according defect emergences are determined. By identifying the effects of process-, tool- or
workpiece-sided variations on the production result and the appearing stress distribution, process
workpiece-sided variations on the production result and the appearing stress distribution, process
design-related conclusions can be drawn.
design-related conclusions can be drawn.
2.1. Flexible Roller Beading
2.1. Flexible Roller Beading
For experimental investigations, a flexible roller beading test facility for the production of
For experimental investigations, a flexible roller beading test facility for the production of
height-variable hat profiles is developed [10]. The profile is formed incrementally by a number
height-variable hat profiles is developed [10]. The profile is formed incrementally by a number of
of roll passes. The tool components directly participating in the forming zone consist of the upper
roll passes. The tool components directly participating in the forming zone consist of the upper
forming roller, the lower forming roller, and blank holder rolls. The blank passes through the forming
forming roller, the lower forming roller, and blank holder rolls. The blank passes through the
stand with a constant feed velocity. During the process, the blank holder rolls keep the profile flanges
forming stand with a constant feed velocity. During the process, the blank holder rolls keep the
on the initial height. The distance between the blank holder rolls determines the profile width and can
profile flanges on the initial height. The distance between the blank holder rolls determines the
be adjusted before every roll pass. The motions of the upper and lower rolls shape the depth-varying
profile width and can be adjusted before every roll pass. The motions of the upper and lower rolls
profile bottom. To avoid material stretching, the needed material is to be pulled from the flanges, which
shape the depth-varying profile bottom. To avoid material stretching, the needed material is to be
leads to desired lateral material inlet (Figure 2e). Geometrically fitted metal sheets can be used to
pulled from the flanges, which leads to desired lateral material inlet (Figure 2e). Geometrically fitted
provide the necessary material in accordance with the profile bottom curve. For the realization of profile
metal sheets can be used to provide the necessary material in accordance with the profile bottom
height variability the manufacturing system needs to provide certain degrees of freedom. Both the
curve. For the realization of profile height variability the manufacturing system needs to provide
upper roller and the lower roller have a vertical movability which allows the forming of varying profile
certain degrees of freedom. Both the upper roller and the lower roller have a vertical movability
depths. Additionally, the lower roller has a degree of freedom in horizontal longitudinal direction in
which allows the forming of varying profile depths. Additionally, the lower roller has a degree of
order to adjust its position in the transition zones and guarantee the permanent perpendicular support
freedom in horizontal longitudinal direction in order to adjust its position in the transition zones and
of the blank by both forming rollers during the process. The described process principle of flexible
guarantee the permanent perpendicular support of the blank by both forming rollers during the
roller beading is shown in Figure 2.
process. The described process principle of flexible roller beading is shown in Figure 2.
Side view Front view (feed direction)

state 1 upper roller z upper roller
y
blank blank
holder holder
z rolls rolls
(a) x lower roller (d) lower roller

state 2
z
lateral
(b) material
x
inlet
state 3
profile flange
z profile bottom
Bead width
(c) x (e) = blank holder distance
Figure 2. Process principle and degrees of freedom of a flexible roller beading stand.
Figure 2. Process principle and degrees of freedom of a flexible roller beading stand.
The upper roller is mounted to the ram and the lower roller to the die cushion of a servo press.
The upperdegree
The horizontal roller is
ofmounted torealized
freedom is the rambyand the lower
a linear roller
module to the
where thedie cushion
lower rollerofisaguided
servo press.
along.
The
The horizontal degree
flexible roller of freedom
beading is realized
stand also containsbya adrive
linear module
unit, whichwhere the lowerfor
is responsible roller is guided
the sheet feed
along. The flexible roller beading stand also contains a drive unit, which is responsible
motion and a guidance unit. The CAD construction in Figure 3 visualizes the essential components of for the sheet
feed motionstand
the forming and ina mounted
guidance condition.
unit. The CAD construction in Figure 3 visualizes the essential
components of the forming stand in mounted condition.
b) drive unit a) forming unit c) guidance unit

b) drive unit a) forming unit c) guidance unit
press ram
press ram
upper
upper
roller blank
Sheet feed roller blank
Sheet feed holder
direction holder
direction rolls
rolls
lower
lower
roller
roller
linear module
linear module
z
z
x press bed
x press bed
die
die
cushion
cushion
Flexible roller beading

Figure 3. Flexible
Figure beading system.
Figure 3.
3. Flexible roller
roller beading system.
system.
Initial experiments show
Initial show the
the feasibility
feasibility ofof the process.
process. The
The manufacturedprofile profile depth
depthcontour
contour
is Initialinexperiments
shown experiments
Figure 4. Theshow
DC04thesheet
feasibility
metal of the
the process.
profile was The manufactured
formed manufactured
in ten roll profile
passes, in depththe
which contour
bead
is
is shown
shown in
in Figure
Figure 4.
4. The
The DC04
DC04 sheet
sheet metal
metal profile
profile was
was formed
formed in
in ten
ten roll
roll passes,
passes, in
in which
which the
the bead
bead
depthconstantly
depth constantlyincreased
increasedby by10%
10% ofthe the final depth during each roll pass. The nominal curve is
depth
the constantly
targeted increased
profile bottom by 10% of
contour of
and
final
thethe
final
red
depth
depth
curve
during
during
depicts
each
each
the
roll
roll pass.
pass. The
manufactured The nominal
nominal
bead
curve is
is the
curvewhich
bottom the
targeted
targeted profile
profilebybottom
bottom contour
contour and
and the red
red curve
the Atos curve depicts the
depicts to manufactured
theflexible
manufactured bead bottom
bottom which
beadprocess was
whichwere
was
was measured
measured by a a 3D-digitizer
3D-digitizer “GOM “GOM Atos III”.
III”. Similar
Similar to flexible roll
roll forming,
forming, process limits
limits were
measured
detected inby a 3D-digitizer
terms of wrinkles“GOM
in the Atos III”.
the flange which Similar to longitudinal
indicate flexible roll forming, process limits were
compression.
detected
detected in in terms
terms of
of wrinkles
wrinkles in
in the flange
flange which
which indicate
indicate longitudinal
longitudinal compression.
compression.
flange
flange
wrinkling
wrinkling
xx
(a) (b) zz
(a) (b)
Experimental results: (a)(a)manufactured

Figure 4. Experimental
Figure manufactured profile bottom
bottom curve; (b)
(b) picture of
of the
Figure 4.
4. Experimental results:
results: (a) manufactured profile
profile bottom curve;
curve; (b) picture
picture of the
the
manufactured
manufactured profile.
manufactured profile.
profile.
2.2. FE Model for Flexible Roller Beading
2.2.
2.2. FE
FE Model
Model forfor Flexible
Flexible Roller
Roller Beading
Beading
The numerical model of the flexible roller beading process is based on preceding researches on
The
The numerical
numerical model
modeldue of
of the
the flexible
flexible roller
roller beading
beading process is is based
based on preceding researches on
roll forming simulations to the resemblance of theprocess
two processes. on
Forpreceding
roll forming researches
processes, on
roll
roll forming
forming simulations
simulations due
due to
to the
the resemblance
resemblance of
of the
the two
two processes.
processes. For
For roll
roll forming
forming processes,
processes,
FE-simulations are a common and powerful tool for process evaluations. The three-dimensional
FE-simulations
FE-simulations are
are aa common
common and
andmetalpowerful
powerful tool
tool in for
forthe process
process evaluations. The three-dimensional
deformation behavior of the sheet is shown work ofevaluations.
Kuichi et al.The [14].three-dimensional
Rebelo et al. [15]
deformation
deformation behavior
behavior of the sheet
of the sheet metal
metal is shown in the work of Kuichi et al. [14]. Rebelo et
et al.
al. [15]
made a comparison between implicit andisexplicit
shownalgorithms
in the work of Kuichi
with et al.
the result [14].
that Rebelo
implicit calculations[15]
made
made a
a comparison
comparison between
between implicit
implicit and
and explicit
explicit algorithms
algorithms with
with the
the result
result that
that implicit
implicit
are less sensitive to interferences and show a better conformity with experimental tests. An
calculations
calculations are
are less
less sensitive to to interferences and show aa better conformity with experimental tests.
solver was also usedsensitive
by Senanayake interferences and
et al. [16] andshowDaniel better
et al.conformity
[17], who with experimental
also found tests.
that penalty
An
An implicit
implicit solver
solver was
was also
also used
used by
by Senanayake
Senanayake et
et al.
al. [16]
[16] and
and Daniel
Daniel et
et al.
al. [17],
[17], who
who also
also found
found that
that
contact with an exponential behavior is more suitable than a linear pressure–overclosure relationship.
penalty
penalty contact with
contact these
with anan exponential
exponential behavior is more suitable than a linear pressure–overclosure
Following findings about behavior
roll formingis more suitable than
simulations, the FE a linear
modelpressure–overclosure
of the flexible roller
relationship.
relationship.
beading process is created in MSC (Marc/Mentat 2012) using an implicit solver and penalty contact.
Following
Following these findings about
about roll roll forming simulations, the FE
FE model of
of the flexible roller
The process wastheseassumedfindings
to be quasi-static. forming
To reduce simulations,
the simulation the time, model
only one the
halfflexible roller
of the system,
beading process
beadingincludes is
process the created
is created in MSC
in MSC (Marc/Mentat 2012) using an implicit solver and penalty contact.
which blank and the (Marc/Mentat
tool rollers, was 2012) using anunder
simulated implicitthesolver and penalty
assumption contact.
of symmetry.
The process
process was
waswas assumed
assumed to
totobe
bethe quasi-static.
quasi-static. To
To reduce
reduce the simulation
the rollers,
simulation time,
time, only
only one half
half of
one holder the
ofrolls
the
The tool system reduced upper and lower forming as well as the blank
system,
system, which
which includes
includes the
the blank
blank and
and the
the tool
tool rollers,
rollers, was
was simulated
simulated under
under the
the assumption
assumption of
of
and the feed and guidance rolls. Each roll pass was represented by a separate forming stand with
symmetry. The tool system was reduced to the upper and lower forming
symmetry. The tool system was reduced to the upper and lower forming rollers, as well as the blank rollers, as well as the blank
holder
holder rolls
rolls and
and thethe feed
feed andand guidance
guidance rolls.rolls. Each
Each rollroll pass
pass was
was represented
represented by by aa separate
separate forming
forming
J. Manuf.
J. Manuf. Mater.
Mater. Process.
Process. 2019,
2019, 3,
3, 19
x FOR PEER REVIEW 55 of 15
of 15
stand with its corresponding roller motions. The forming stands were connected in series with a
stand
its
stand with
corresponding
distance its corresponding
rollermm.
of 1300 motions.
Theroller
Themotions.
motion curvesThe
forming stands
were forming stands
were connected
given were
to the forming connected
in series
rollswith in seriesdistance
a stand
via position with a
control.
stand
of distance
1300 mm.
According to Theof 1300
Grochemotionmm. The
et al.curves motion
[18], thewere curves
given to theof
consideration were given
forming
the stand to the
rolls forming rolls
via position
stiffness generated via
control. position control.
According
improved to
results
According
Groche
regarding to
et al. Groche
the[18], et al. [18],
the consideration
applied the
contact forces. consideration
of the stand
Still, of
the stiffnessthe stand
definitiongenerated stiffness generated
improved
of the forming rolls improved
results results
regardingrigid
as analytical the
regarding
applied the applied
contact
bodies appeared to becontact
forces. Still, theforces.
sufficient definition
for theStill,pursued
the
of thedefinition
forming of the
rolls
investigations. as forming
Due to rolls
analytical rigid
highly asbodies
analytical rigid
appeared
time-consuming
bodies
to appeared
be sufficient
computational fortothe
times, be
thesufficient
pursued
sheet metal forwas the modelled
investigations. pursued investigations.
Due with
to highly
a 4-node Dueelement
time-consuming
shell to highly time-consuming
computational
(type 75), which times,
was
computational
the sheet metal
considered times,
as a was the sheet
thickmodelled
shell element metal
with aand was
4-node modelled
assumed with
shell strain
element a 4-node shell
(type 75), which
formulation. element
Bilinearwas (type 75),
considered as
interpolation which
was was
a thick
used
considered
shell element
for the as aand
coordinates,thick shell
assumed element and assumed
strain formulation.
displacements, and rotations strain
Bilinear
here formulation.
interpolation Bilinear
[19]. The 4-node was used interpolation
shell for was used
the coordinates,
element combines
for the coordinates,
displacements,
efficient computation displacements,
and rotations
effort here
with [19]. and
accurateTherotations
4-node inhere
results shell [19].
element
roll formingThe 4-node
combines shell
efficient
simulations [15,element
16]. Thecombines
computation effort
selected
efficient
with
reduced computation
accurate effort
resultsmethod
integration with
in roll forming accurate
was used results
simulations
to overcome in roll
[15,16]. forming
volumetric simulations
The selected
locking.reduced [15, 16].
With aintegration
step time The of
selected
method
20 for
reduced
was rollintegration
eachused pass and a method
to overcome relaxation was
volumetric used
steplocking.
timeto of
overcome
With a step
4 between volumetric
time of 20locking.
two forming eachWith
forstands, roll a step
pass
adaptive and time of 20 for
a relaxation
time stepping
each
step roll
with time pass
of 4 and
an initial a relaxation
between
fraction oftwo step
0.01forming
and time of 4fraction
stands,
a minimal between
adaptive two
of time forming
10 -6 was used.stands,
stepping withsheet
The anadaptive
initial
and mesh timeproperties
fractionstepping
of 0.01
with
and an initial
a minimal
are shown fraction
fraction
in Figure of 0.01 and
− 6 a minimal
of 10elastic–plastic
5a. The fraction
was used. behaviors
The sheetofandof 10 -6 was used. The sheet and mesh properties
the mesh
sheet properties
metal materialsare shown(DC04 in Figure
and Al2024)5a.
are
The shown in Figure
elastic–plastic 5a. The
behaviors elastic–plastic
of the sheet behaviors
metal of
materials the sheet
(DC04
are modelled using the von Mises yield criterion and the flow curve obtained from tensile tests metal
and materials
Al2024) are (DC04
modelled and Al2024)
using the
are
von modelled
Mises
(Figure 5b).yieldusing the von
criterion and Mises
the flow yield
curvecriterion
obtained and fromthetensile
flow curve obtained
tests (Figure 5b).from tensile tests
(Figure 5b).
Solver Implicit 600
material flow curves
Solver Implicit 600
Element type Type 75 500 material flow curves
Element typelength Type 500
Sheet metal 120075
mm 400
Sheet metalelement
length edge length (x,y) 1200 mm σ 400
300
Horizontal 2.5 mm
σ 300 Al2024
Horizontal
Sheet thickness edge length (x,y)
element 2.5 mm
1.0 mm 200
Al2024
DC04
Sheet thickness 1.0 200
Elements in thickness direction 2 mm 100
DC04
Elements in thickness direction 253760 100
Total elements 0
0 0.1 0.2 0.3 0.4
0
(a) Total elements
Materials 53760
DC04, Al2024 (b) 0 0.1 ε
0.2 0.3 0.4
(a) Materials DC04, Al2024 (b) ε
Figure 5. (a)
Figure 5. (a) Mesh
Mesh properties and (b)
properties and (b) material
material flow
flow curves
curves of
of DC04
DC04 and
and Al2024.
Al2024.
Figure 5. (a) Mesh properties and (b) material flow curves of DC04 and Al2024.
Gehring [20]
Gehring [20]showed
showed thatthat friction
friction has no has no notable
notable influence influence on the
on the results results ofsimulations
of numerical numerical
Gehringof[20]
simulations showed processes.
roll forming that friction Theofhas no notable
application influence
of friction showed on no thesignificant
results ofdeviation
numerical to
of roll forming processes. The application friction showed no significant deviation to the frictionless
simulations
the frictionlessof roll forming results
numerical processes. The application
regarding process of friction
forces and showed
geometry. no Therefore
significant adeviation
frictionless to
numerical results regarding process forces and geometry. Therefore a frictionless numerical model
the frictionless
numerical model numerical
with was results regarding
non-rotating rolls wasprocess
chosen. forces and geometry.
Thenegligence
justification of theTherefore
negligence a frictionless
of friction
with non-rotating rolls chosen. The justification of the of friction has also been shown
numerical
hasa also model
beenofshown with non-rotating
in a variety rolls was
of numerical chosen. The justification of the negligencein theof friction
in variety numerical investigations on rollinvestigations
forming, e.g.,oninroll theforming,
works of e.g.,
Groche etworks
al. [18],of
has also been
Grocheetetal.al. shown
[18],Boman in
Traub et a variety
et al.
al. [22], of numerical
[21], Boman investigations
et al. [22], on roll forming, e.g., in the works of
Traub [21], and Goertan et al.and Goertan
[23]. et al. [23].
Additionally, Additionally,
kinematic reversalkinematic
is used
Groche
reversaletis al.used
[18],meaning
Traub et al. [21], Bomanto
that, et al. [22], and Goertan theisetwork
al. [23]. Additionally, is kinematic
meaning that, contrary to the realcontrary
process, thethe work real process,
piece position piece
fixed and position
motion fixedtoand
is applied the
reversal
motion isapplied
is used meaning to the that,
roll contrary
sets. The to theofreal
fixation the process,
blank is the work piece
implemented by position
fixing 18 is fixedofand
nodes the
roll sets. The fixation of the blank is implemented by fixing 18 nodes of the flange at the front edge
motion
flange is the
at applied to the ofroll sets. The infixation of thex-direction
blank is implemented 6b). by fixingdepicts
18 nodes of the
of the sheet infront edge
longitudinal the sheet
x-direction longitudinal
(Figure 6b). Figure 6 depicts(Figurea section Figure
of the6numerical a section
model
flange at the
of the numerical front edge of the sheet in longitudinal x-direction (Figure 6b). Figure 6 depicts a section
including the sheetmodel mesh including
and the first thetwosheet
rollmesh
passes and the first
(Figure two the
6a) and rolldeformed
passes (Figure
profile6a) and 6c).
(Figure the
of the numerical
deformed profile model
(Figureincluding
6c). the sheet mesh and the first two roll passes (Figure 6a) and the
deformed profile
Boundary (Figure 6c).
condition:
18 nodes condition:
Boundary fixed in x-direction
18 nodes fixed in x-direction
upper
roller
upper
roller
(b)
(b)
z
z x
lower y x z
roller
lower y x z
roller (c)
(c) yx
(a)
y
(a)
Figure
Figure 6. (a)Numerical
6. (a) Numericalmodel
modelwith
withtwo
tworoll
rollpasses,
passes,(b)(b) sheet
sheet fixing
fixing boundary
boundary conditions,
conditions, andand
(c)
Figure
(c) 6. (a)
deformed
deformed Numerical
profile.
profile. model with two roll passes, (b) sheet fixing boundary conditions, and (c)
deformed profile.
2.3. Principles of Forming Sequence Design
2.3. Principles of Forming Sequence Design
J.2.3.
Manuf.
J. Manuf. Mater.
Principles
Mater. Process.
of2019,
Process. 2019,
Forming 3,Sequence
3, x FOR x FOR
PEERPEER REVIEW
Design
REVIEW 6 of 615of 15
The geometrical
geometrical characteristics
characteristics of
of the
the targeted
targeted profile depth contour are
are shown in
in Figure 7.
The The
geometrical characteristics of the targeted profile
profile depth
depth contour
contour shown
are shown Figure
in Figure 7. 7.
z z
x Initial height z = 0 mm
The
J. Manuf. geometrical
Mater. Process. 2019,
x
characteristics ofInitial
theheight
3, x FOR PEER REVIEW
z = 0 mm
targeted profile depth contour are shown in Figure 7.
6 of 15
z Profile
Profile
The geometrical characteristics
bottom
bottom x
curve curve of theInitial
targeted
height z =profile
0 mm depth contour are shown in Figure 7.
z
Profile
x Initial height z = 0 mm
Geometric
Figure 7. bottom curve characteristics of the target profile bottom curve.
Figure
Figure 7. Geometric
7. Geometric characteristics
characteristics oftarget
of the the target profile
profile bottom
bottom curve.
curve.
Profile
hmax,a describes the maximum bottom curveprofile height of the bottom curve and thus contains the highest
h hmax,a
max,a describes
describes the the maximum
maximum profileprofile
height height of theof bottom
the bottom curve curve
and and thusthus
containscontains the highest
the between
highest
risk of material thinning Figureas7.aGeometric
result of insufficient
characteristics lateral
of thematerial
target profile inlet. With
bottom the
curve. distance the
risk risk
blank
of material
of material
holders thinning thinning
remaining
as
as aunaltereda result
result of insufficient
ofduring
insufficientall rolllateral
lateral
passes,
material
material inlet.inlet.
plastic stretching WithWith the distance
the distance
(Figure 8a) andbetween
between
the forming
the blank
the blank holders remaining unaltered during all roll passes, plastic stretching (Figure 8a) and the
hholders
of multiplemax,a describes
remaining
bending edges unaltered
the maximum
Figure (Figure
7. Geometric 8b)during
is
profile all roll
expected.
height
characteristics ofpasses,
Theofthe
the
plastic
adjustment
bottom
target profile
stretching
of
curve the blank
and
bottom
(Figure
thus holder
curve. contains 8a) sets
roll and the
the for each
highest
forming
forming of of multiple bending edges (Figure 8b) is expected. The adjustment of the blank holder roll
roll of multiple
riskpass makes sure
material bending
thinningthat edges
the
as amaterial(Figure
result 8b) is in
oflength
insufficient expected.
the lateral The
cross-section adjustment
material of inlet.
hmax,aof the the
remains
With blank holderbetween
consistent
distance rolland is
sets sets
for for each
each roll
hmax,a holders
equivalent
the blank
roll
pass
describes
to the material
pass
makes makes
the maximum
remaining sure
length
sure
that that
the
profile
needed
unaltered
the
material
inheight
during
material length
allofroll
the target the
length
in in the
the
bottomplastic
geometry
passes, curve
(Figure
cross-section
cross-section
and8c).thus of
Thiscontains
stretching hof
max,a
approach
(Figure
hmax,a remains
remains
thesupports
8a) highest
and the
consistent
consistent and and
is is equivalent
equivalent to to material
the the material length length neededneeded in thein target
the target geometry
geometry (Figure (Figure 8c). 8c).
This This
risk
the of
material
forming material thinning
inlet andbending
of multiple as a
reduces edges result
the risk of insufficient
of sheet
(Figure lateral
8b)stretching. material inlet. With
is expected. The adjustment of the blank holder roll the distance between
approach
approach supportssupports the material
the remaining
material inlet inlet and
andthat
reduces reduces the risk of sheet stretching.
the
sets blank
for eachholders roll pass makes unaltered
sure the the
during allrisk
materialrollofpasses,
sheet
length stretching.
plastic
in stretching (Figure
the cross-section of hmax,a 8a) remains
and the
forming
consistent of multiple
and
Impeded Impeded
lateral bending
is equivalent
lateral
materialmaterial edges
inlet inletto (Figure Multiple
the material 8b) is
length expected.
needed
Multiple
bendingbending
edges edgesThe
in the adjustment ofProfile
target geometry theProfile
blank width holder
(Figure
width adaption roll
8c). This
adaption
roll
roll pass 1 pass 1 upper roll
upperroll pass n pass n roll
roll pass 1 pass 1 roll
roll pass 2 pass 2
sets for each
approach supports roll roller
pass makes sure
theroller
material inlet that the material
and reduces the risklength
of sheet in the cross-section of hmax,a remains
stretching.
roll 1pass 2 1 3 425 3 4 roll
5 roll
consistent and is equivalent
Impeded lateral material inlet
to the material length needed in
Multiple bending edges
the target geometry
roll pass (Figure 8c). This
pass pass
Profile width adaption
approach roll supports
pass 1 theupper
material inlet
roll pass n and reduces roll pass 1 the risk of sheet stretching.
roll pass 2
blank holder
blank holder roller 1 1
rolls rolls Plastic Plastic elongation
elongation bending roll Profile
pass 1width2 adaption
3 4 25 2roll
Impeded lateral material inlet Multiplebending
bending bendingbending
edges
in the profile
in the profile legs legs edge (roll edge 3 3pass
roll pass 1 upper roll pass n roll pass 1 edge (roll edgeroll pass (roll
(roll 2 4 4
(a) roller (b) pass 1)pass 1)
pass 2)pass 2) (c) 5 5
(a) (b) (c)
blank holder roll pass 1 2 3 4 5 roll
1
rolls Plastic elongation bending bending pass
2
Figure
Figure 8. Blank
8. Blank holder
holder roll
rollinset set legs
adaption:
adaption: (a) risk
(a) risk of sheet stretching without adaption, (b) risk
of of
Figure 8. Blank holder the profile
roll set adaption: (a) ofrisksheet stretching
of sheet edgestretchingwithout
(roll edge adaption,
without
(roll (b) risk
adaption, (b) risk of 3
4
multiple bending without adaption, and (c) principle of pass 1)holder
blank pass 2)
roll set adaption.
multiple
(a) bending without adaption, and (c)
blank
multiple rolls holder
bending without (b)principle of blank holder roll set adaption. (c) 15
Plasticadaption,
elongation and (c) principle of blank bending
holder roll set adaption.
bending 2
in the profile legs 3
Figure 8. Blank holder roll set adaption: (a) risk of sheetedge (roll edge (roll
stretching without adaption, (b) risk of 4
By
By analyzing
By analyzing
analyzing
(a) the
the effect
the effect
effect of
of the the forming
of forming
the forming sequences
sequences on
on the
sequences the geometric
on geometric accuracy
accuracy
the geometric
(b) various
various
accuracy strategies
strategies
variouspass 1)
strategies pass 2)
(c) 5
multiple bending without adaption, and (c) principle of blank holder roll set adaption.
are considered:
are considered:
are considered:
Figure 8. Blank holder roll set adaption: (a) risk of sheet stretching without adaption, (b) risk of
• •Constant
Constant
By
profile
profile
analyzing depthdepth increase
increase
the effect
(Figure
(Figure
of the forming
9).
9). sequences
• multiple
Constant bending
profile without adaption,
depth increase principle of on
and (c) 9).
(Figure
theholder
blank geometric
roll setaccuracy
adaption.various strategies
are considered: ∆h = ∆h = const.
const.
• Constant bending angle increase (Figure 10).
• By analyzing
Constant thedepth
profile effect increase
of the forming 9).∆h ∆h
(Figure sequences on the geometric accuracy various strategies
• Constant material inlet increase (Figure 11). ∆h ∆h
are considered: ∆h = const. ∆h ∆h
• Constant profile depth increase (Figure 9).∆h ∆h ∆h
∆h = const. ∆h
Figure 9. Forming sequence ∆h
Figure 9. Forming sequence withwith constant
constant profile
profile
∆h depthdepth increase
increase in each
in each roll pass.
roll pass.
∆h
∆h
• •Constant
Constant bending
bending angleangle increase
increase (Figure
(Figure 10). 10). ∆h
Figure 9. Forming sequence with constant∆hprofile depth increase in each roll pass.
∆α = ∆α = const.
const.
• ConstantFigure 9. Forming
bending
Figure
sequence with ∆α
angle increase
9. Forming
constant
sequence(Figure ∆α profile
10).
with constant
depth increase in each roll pass.
profile depth increase in each roll pass.
∆α ∆α
∆α = const.
∆α ∆α
• ∆α ∆α 10).
Constant bending angle increase (Figure
∆α
∆α = const. ∆α
Figure
Figure 10. Forming
10. Forming sequence
sequence withwith
constant∆α bending
constant bending
angleangle increase
increase in each
in each roll pass.
roll pass.
∆α ∆α
∆α
• •Constant
Constant material
material inlet
inlet increase
increase (Figure
(Figure 11).
11).constant
∆α
Figure 10. Forming sequence with
Figure 10. Forming sequence with∆α constant bending
bending angle
angle increase
increase in
in each
each roll
roll pass.
pass.
∆uy = ∆u y = const.
const. ∆u
∆uy ∆u y ∆u ∆uy
y ∆u y y
• Constant material
Figure inlet increase
10. Forming (Figure
sequence 11). bending angle increase in each roll pass.
with constant
∆u = const.
• y
Constant material inlet increase (Figure 11). ∆uy ∆uy ∆uy
∆uy = const. ∆uy ∆uy ∆uy

∆α
∆α
∆α
Figure 10. Forming sequence with constant bending angle increase in each roll pass.
• Constant material inlet increase (Figure 11).
∆uy = const. ∆uy ∆uy ∆uy

Figure 11. Forming sequence with constant lateral material inlet increase in each roll pass.
Figure 11. Forming sequence with constant lateral material inlet increase in each roll pass.
Additionally, a distinction regarding the bending radii at each roll pass was made. In the first
Additionally,
approach, a distinction
the bending regarding
radii of the the bending
upper and radii atedges
lower bending each remain
roll passunchanged
was made.(Figure
In the first
12a).
approach,
The upperthe bending
bending radiiwas
radius of the upperby
defined and lower
the edgebending
radius ofedges remain
the lower unchanged
blank holder (Figure
rolls and12a).
the
The upper
lower bending
bending radius
radius was defined
was given by the
by the edge edgeofradius
radius of the
the upper lower blank
forming roller. holder rolls and the
lower bending radius was given by the edge radius of the upper forming roller.
(a) (b)
R1 R1>R2>R3
R1=R2=R3
R2
R3 R2 R1
R3
Figure
Figure 12.
12. (a)
(a) Constant
Constant bending
bending radii
radii in
in each
each roll
roll pass,
pass, and
and (b)
(b) constant
constant bending
bending arc
arc length
length in
in each
each
roll
roll pass.
pass.
Alternatively, the
Alternatively, thebending
bendingradii
radiican
canbebeadapted
adapted to to
thethe bending
bending sequence
sequence so that
so that the the bending
bending arc
length remains
arc length remainsconstant for for
constant each rollroll
each passpass(Figure
(Figure12b).
12b).InInreal
realexperiments
experimentsthisthis means
means that
that
custom-made
custom-made lower lower blank
blank holder
holder rolls
rolls and
and upper
upper forming
forming rollers
rollers need
need to to be
be manufactured
manufactured andand
re-tooled every roll
re-tooled before every rollpass.
pass.The
Theimplementation
implementationisisnot not practicable
practicable in in reality
reality andand conflicts
conflicts with
with the
the
basicbasic
ideaidea of flexible
of flexible manufacturing.
manufacturing. Even so, Even
for so,
the for the purpose
purpose of understanding
of process process understanding this
this approach
approach is investigated.
is investigated.
3. Results
3. Results and
and Discussion
Discussion
To determine
To determine the the process
process limits
limits and
and toto derive
derive optimization
optimization measures
measures onon the
the process
process or or tooling
tooling
side, aa fundamental
side, fundamental understanding
understanding of of the
the appearing
appearing stress
stress condition
condition during
during the
the forming
forming process
process
needs to be built. For this purpose, numerical investigations were carried out by
needs to be built. For this purpose, numerical investigations were carried out by means of the model means of the model
presented in
presented in Section
Section 2.2
2.2 and
and aa workpiece
workpiece consisting
consisting ofof DC04
DC04 material.
material. The
The results
results ofof the
the numerical
numerical
analyses are
analyses are presented
presented hereafter.
hereafter.
The stress
The stress distribution
distribution in in the
the lateral
lateral direction
direction along
along the
the transverse
transverse section
section during
during the the process
process
appears to
appears to be
be aa superposition
superposition of of tensile
tensile stresses
stresses and
and bending
bending stresses
stresses (Figure
(Figure 13).
13). In
In the
the area
area of
of the
the
bending edges, bending stresses are predominant, while tensile stresses prevail in
bending edges, bending stresses are predominant, while tensile stresses prevail in the profile bottom the profile bottom
and leg
and leg as
as well
well asas in
inflange
flangesections
sectionsclose
closetotothe
theupper
upper bending
bending edge. ByBy
edge. exceeding
exceedinga critical value,
a critical the
value,
tensile
the stresses
tensile leadlead
stresses to material stretching
to material in theinlateral
stretching direction,
the lateral whichwhich
direction, causescauses
sheet thinning and is
sheet thinning
therefore to be avoided.
and is therefore to be avoided.
In the transition zones, the length of the profile bottom is greater than in the original state. In order
to obtain the zintended height-varying profile geometry, longitudinal strains along the profile bottom
are necessary.
y
Accordingly,
x longitudinal stresses were intentionally introduced to the profile bottom
during the process (Figure 14). To reduce the longitudinal stresses in the profile bottom, sharp height
transitions are to be avoided.
z
The stress distribution in the lateral direction along the transverse section during the process
appears to be a superposition of tensile stresses and bending stresses (Figure 13). In the area of the
bending edges, bending stresses are predominant, while tensile stresses prevail in the profile bottom
and leg as well as in flange sections close to the upper bending edge. By exceeding a critical value,
J.the
Manuf. Mater.
tensile Process. 2019,
stresses lead3,to
19 material stretching in the lateral direction, which causes sheet thinning
8 of 15
and is therefore to be avoided.
z
y x
J. Manuf.
J. Manuf. Mater.
Mater. Process.
Process. 2019,
2019, 3,
3, xx FOR
FOR PEER
PEER REVIEW
REVIEW 88 of
of 15
15
In the
In the transition
transition zones,
zones, the
the length
length ofof the
the profile
profile bottom
bottom isis greater
greater than
than in
in the
the original
original state.
state. In
In
z
order to obtain the intended
order to obtain the intended height-varying profile geometry, longitudinal strains along the profile
y x height-varying profile geometry, longitudinal strains along the profile
bottom are
bottom are necessary.
necessary. Accordingly, longitudinal
Accordingly, longitudinal stresses
stresses were
were intentionally
intentionally introduced
introduced toto the
the profile
profile
bottom during the process (Figure 14). To reduce the longitudinal stresses in the
bottom during the process (Figure 14). To reduce the longitudinal stresses in the profile bottom, profile bottom,
heightFigure
sharp height 13. Lateral
transitions stresses
are to
to along a transverse section of the profile during the
be avoided.
avoided. the process.
process.
sharp transitions are be
process
process upper
upper
state:
state: roller
roller
bead bottom
bead bottom
zz
yy xx
Figure 14.
Figure Longitudinal stresses
14. Longitudinal stresses along
along the
the profile
profile bottom
bottom during
during the
the process.
process.
Figure 14. Longitudinal stresses along the profile bottom during the process.
To avoid sheet stretching along the transverse sections lateral material inlet is intended.
To avoid
To avoid sheet
sheet stretching
stretching along
along the
the transverse
transverse sections
sections lateral
lateral material
material inlet
inlet is
is intended.
intended.
Subsequently, longitudinal stresses in the flanges and particularly along the sidewise profile edges
Subsequently,
Subsequently, longitudinal
longitudinal stresses
stresses in the flanges and particularly along the sidewise profile edges
occur. The deeper the profile depth,inthe
themore
flanges and particularly
material along in
has to be pulled thefrom
sidewise profileflanges.
the profile edges
occur. The
occur. The deeper
deeper thethe profile
profile depth,
depth, the more
more material
material has
has to
to be
be pulled
pulled in from
from the
the profile
profile flanges.
flanges.
Consequently, the lateral displacementthe of the profile edge varies along the inlongitudinal direction. As a
Consequently,
Consequently, the lateral
the lateral displacement
displacement of the
of and profile
the profile edge
edge varies along
variesare
along the longitudinal
the longitudinal direction. As
result, profile edge contours with convex concave sections formed. Depending direction. As
on the either
aa result,
result, profile
convex
profile edge contours
or concaveedge
contours with
with convex
shape of the profile
convex andand concave sections
edge section,concave
sections are formed.
local longitudinal are
formed. Depending
displacements
Depending on on the
the
occur which lead
either
either convex
convex oror concave
concave shape of
shapethe the
ofprocessprofile
the profile edge
edge section, local longitudinal displacements
section, local longitudinal displacements occur occur
to longitudinal stresses during (Figure 15).
which lead
which lead to
to longitudinal
stresses during
during thethe process
process (Figure
(Figure 15).
15).
zz
yy xx
Figure 15.
Figure 15. Displacement
Displacement and
and longitudinal
stresses along
along the
the profile
profile edge during the process.
The
The resulting
resultingdisplacements
displacementsindicate that
indicate thethe
that material is not
material is drawn in purely
not drawn
drawn laterally.
in purely
purely In convex
laterally. In
The resulting displacements indicate that the material is not in laterally. In
zones,
convex the displacement
zones, the increases
displacement in the positive
increases in the longitudinal
positive direction
longitudinal while concave
direction while zones zones
concave cause
convex zones, the displacement increases in the positive longitudinal direction while concave zones
cause increasing
cause increasing displacements
displacements in
in the
the negative
negative direction.
direction. The
The locations
locations ofof the
the maximum
maximum positive
positive and
and
the maximum negative displacement border the compression zone asshown
the maximum negative displacement border the compression zone asshown in Figure 16. in Figure 16.
concave compression
concave compression zone
zone
xx
convex
convex convex
convex
yy
z
y x
Figure 15. Displacement and longitudinal stresses along the profile edge during the process.
The resulting displacements indicate that the material is not drawn in purely laterally. In
convex zones,
increasing the displacement
displacements increasesdirection.
in the negative in the positive longitudinal
The locations direction
of the maximumwhile concave
positive andzones
the
cause increasing displacements in the negative direction. The locations of the
maximum negative displacement border the compression zone asshown in Figure 16. maximum positive and
the maximum negative displacement border the compression zone asshown in Figure 16.
concave compression zone
x
y convex convex
stretching stretching
zone zone
Sheet edge
Maximum Maximum (non-deformed
positive negative state)
longitudinal longitudinal
displacement displacement
Figure 16. Displacement

Figure16. Displacementpaths
pathsof
ofthe
theconcave/convex shaped profile
concave/convex shaped profile edge.
edge.
In sum, the dominant stress state during the process reveals lateral stretching and the presence
In sum, thecompressive
of longitudinal dominant stress state during
and tensile stress the process
areas in the reveals lateral stretching
profile flange, and the presence
which, by exceeding critical
of longitudinal compressive and tensile stress areas in the profile flange, which, by exceeding
values, cause failure and geometric deviation. By selection of an appropriate forming strategy, the critical
risk
values, cause failure
is to be minimized. and geometric deviation. By selection of an appropriate forming strategy, the
risk is to be minimized.
3.1. Process- and Tool-Side Variations: Effects of the Forming Sequence
3.1. Process- and Tool-Side Variations: Effects of the Forming Sequence
In Section 2.3, the different forming strategies and the two types of bending radii modification
In Section The
are presented. 2.3, the different forming
combination strategies
of the forming and theand
strategies twothetypes of bending
bending radii modification
radii modification gives
are
the presented.
selection ofThe combination
the considered of the forming
forming sequences strategies and the
(Figure 17a). Figurebending radii the
17 depicts modification gives
resulting profile
the selection
bottom of when
curves the considered
applying forming sequences
the different forming(Figure 17a). Figure 17 depicts the resulting profile
sequences.
bottom curves when applying the different forming sequences.
(a) (b)
(c) (d)
Figure 17.
Figure Forming strategies
17. Forming strategies and
and the
the resulting
resulting profile
profile bottom
bottom contours.
contours.
The nominal curve shows the striven target geometry. The profile bottom curves produced by
the forming sequences FS2_angle, FS3_arc, and FS3_angle show a better accuracy. The impracticable
approach of adapting the bending radii shows no improvement. The profile accuracy appears to be
depending on the combination of the bending radii (upper bending radius r , lower bending radius
depending on the combination of the bending radii (upper bending radius r uu, lower bending radius
r ) and the bending angle increase ∆α of each roll pass. In Figure 18, the quotient of ∆α/(r + r ) at
roo) and the bending angle increase ∆α of each roll pass. In Figure 18, the quotient of ∆α/(r u+rou) at xo = 0
x = 0 for each roll pass of the considered forming sequences is shown.
for each roll pass of the considered forming sequences is shown.
1.2 1.2
1 1
0.8 0.8
∆α/(ru+ro) FS1_angle ∆α/(ru+ro) FS2_angle
[°/mm] 0.6 0.6
[°/mm]
0.4
FS1_arc 0.4
FS3_angle
0.2 FS2_arc 0.2 FS3_arc
0 0
depending on the combination of the bending radii (upper bending radius r u, lower bending radius
ro) and the bending angle increase ∆α of each roll pass. In Figure 18, the quotient of ∆α/(r u+ro) at x = 0
for each roll pass of the considered forming sequences is shown.
1.2 1.2
1 1
0.8 0.8
∆α/(ru+ro) FS1_angle ∆α/(ru+ro) FS2_angle
[°/mm] 0.6 0.6
[°/mm]
0.4
FS1_arc 0.4
FS3_angle
0.2 FS2_arc 0.2 FS3_arc
0 0
0 1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11
(a) roll pass
(b) roll pass
Figure 18.∆α/(r
Figure18. ∆α/(ruu ++ rroo))atateach
eachroll
rollpass
passof
ofthe
theforming
formingsequences.
sequences.
The forming sequences with larger deviations all have a strong increase of ∆α/(ru + ro ) in the last
The forming sequences with larger deviations all have a strong increase of ∆α/(ru + ro) in the last
the roll passes in common. Additionally, a large quotient in the first roll passes (FS3_angle) appears to
the roll passes in common. Additionally, a large quotient in the first roll passes (FS3_angle) appears
have no negative effect on the geometric accuracy. Derived from the results, it can be noted that large
to have no negative effect on the geometric accuracy. Derived from the results, it can be noted that
increases of ∆α/(ru + ro ) in the last roll passes are to be avoided. Compared to constant bending radii,
large increases
J. Manuf. of ∆α/(r
Mater. Process. + xroFOR
2019,u3, ) in PEER
the last roll passes are to be avoided. Compared to constant bending
REVIEW 10 of 15
the application of adaptive bending radii leads to an undesired reduction of ∆α/(ru + ro ) in the first
radii, the application of adaptive bending radii leads to an undesired reduction of ∆α/(r u + ro) in the
roll passes and an increase in the last roll passes. With regard to the results of FS2 in Figure 17c, where
first roll passes and an increase in the last roll passes. With regard to the results of FS2 in Figure 17c,
the largest deviation is observed, this effect is pronounced most significantly (Figure 19b).
where the largest deviation is observed, this effect is pronounced most significantly (Figure 19b).
1 1 1
0.8 0.8 0.8
0.6 0.6 0.6

∆α/(ru+ro) ∆α/(ru+ro) ∆α/(ru+ro)
[°/mm] FS1_angle [°/mm] FS2_angle [°/mm] FS3_angle
0.4 0.4 0.4
FS1_arc FS2_arc FS3_arc
0.2 0.2 0.2
0 0 0
0 1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11
(a) roll pass
(b) roll pass
(c) roll pass
Figure 19. Effect of adapting the bending radii to ensure constant bending arc length on ∆α/(ru + ro ).
Figure 19. Effect of adapting the bending radii to ensure constant bending arc length on ∆α/(r u + ro).
Provided that ru and ro remain constant, ∆α can be set comparatively large during the first roll
pass with Provided that rutendency
a declining and ro remain through constant, ∆α can be set comparatively large during the first roll
to the last.
pass with a declining tendency through to the last.
3.2. Effect on the Sheet Thickness
The largest deviations in the profile bottom curve appear in the section x = 0–350 mm. In the
following, the resulting
The largest deviations profile in theof the forming
profile bottom sequence
curve appearwith constant bending
in the section x =angle
0–350 increase
mm. Inand the
constant
following, bending radii (FS2_angle)
the resulting profile ofisthe examined.
forming The effect of
sequence the constant
with partly exceededbending and subjacent
angle increase profile
and
depths
constant onbending
the material radiithickness
(FS2_angle) is investigated
is examined. along
The the cross-sections
effect of the partly at xexceeded
= 0 mm, and 200 mm, and
subjacent
350 mm,depths
profile whichon is the
the location
materialof maximum
thickness profile depthalong
is investigated (Figurethe20).
cross-sections at x = 0 mm, 200 mm,
and At 350xmm,= 0 mm,which areas
is the with both marginal
location of maximum sheet thinning
profile depth and thickening
(Figure 20). are observed. The sheet
thinning is caused by the occurring tensile stresses along the cross-section (Figure 13). The thickening
is a result of the blank holder roll set adaption (Figure 8c). By changing the lateral distance between
x = 0 mm x = 200 mm x = 350 mm
the blank holder z roll sets, the profile width is adjusted z so that the material amount z at the position with
maximum y profile x depth (h max,a ) is kept constant
y for
x each roll pass. As a consequence,
y x the remaining
transverse sections contain
cross section at x = 0 mm surplus materialcross which
sectionisatpushed
x = 200 mm back into the flange
cross sectioninat xthe
= 350course
mm of
5
profile depth z [mm]
5 5
the following roll passes. This material return creates local compressive stresses which lead to the
-5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120
observed material thickening. Such areas of thinning and thickening appear at x = 200 mm as well.
-15 -15 -15
Even though the maximum thickness is larger, the total cross-section
cross-section
thickness increase is lower than at x = 0 mm.
cross-section
-25 -25 -25
Since the profile depth is greater at x = 200 mm, a lower amount of material needs to be returned to the
-35 -35 -35
flange, whereby lower y [mm]compressive stresses occur. The y [mm] cross-section at x = 350 mm contains y [mm] no surplus
material that needs to be pushed back which is why no sheet thickening is observed. Yet material
thickness [mm]
1.1 1.1 1.1

1.05 1.05 1.05
thinning1 occurs in the bending edges, in
1 the lower bending edge particularly
1 since the lower bending
0.95 0.95 0.95
radius
0.9 (6 mm) is smaller sheetthan the upper
thickness 0.9 radius (12 mm). sheetThis material
thickness 0.9 thinning indicates that the
sheet thickness
0.85 0.85 0.85
0.8
initial thickness initial thickness initial thickness
0.8 0.8
0 20 40 60 80 100 120 0 20 40 60 80 100 120 0 20 40 60 80 100 120
y [mm] y [mm] y [mm]
Figure 20. Sheet thicknesses in the cross-sections at x = 0, 200, 350 mm.

Figure 19. Effect of adapting the bending radii to ensure constant bending arc length on ∆α/(r u + ro).
Provided that ru and ro remain constant, ∆α can be set comparatively large during the first roll
pass with a declining tendency through to the last.
The largest deviations in the profile bottom curve appear in the section x = 0–350 mm. In the
lateral material inlet is impeded by the bending edges. As a result, the necessary material is obtained
following, the resulting profile of the forming sequence with constant bending angle increase and
from the areas with the highest plastification, which are the bending edges. Figure 21 shows the
constant bending radii (FS2_angle) is examined. The effect of the partly exceeded and subjacent
diminishing material thickness with progressing roll passes and the correlation with the increasing
profile depths on the material thickness is investigated along the cross-sections at x = 0 mm, 200 mm,
difference between the desired and the actual material inlet.
and 350 mm, which is the location of maximum profile depth (Figure 20).
x = 0 mm x = 200 mm x = 350 mm
z z z
y x y x y x
cross section at x = 0 mm cross section at x = 200 mm cross section at x = 350 mm
5
profile depth z [mm]
5 5
-5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120
-15 -15 -15
cross-section cross-section cross-section

-25 -25 -25
J.J. Manuf.
-35 Mater.
Manuf. Mater. Process.
Process. 2019,
2019, 3,
3, xx FOR
FOR PEER
PEER REVIEW
REVIEW
-35 -35 11
11 of
of 15
15
y [mm] y [mm] y [mm]
no
no sheet
sheet thickening
thickening is is observed.
observed. Yet Yet1.1material
material thinning
thinning occurs
occurs in in the
the1.1bending
bending edges,edges, in in the
the lower
lower
thickness [mm]
1.1
1.05 1.05 1.05
bending
bending 1 edge
edge particularly
particularly since since thethe lower
lower
1 bending
bending radiusradius (6 (6 mm)
mm) is is smaller
smaller
1 than
than the
the upper
upper radius
radius
0.95 0.95 0.95
sheet thickness sheet thickness sheetbending
thickness
(12
(12 mm).
mm).
0.9 This
This material
material thinning
thinning indicates
indicates
0.9 that
that thethe lateral
lateral material
material inlet 0.9 is
inlet is impeded
impeded by by the
the bending
0.85 0.85 0.85
initial thickness initial thickness initial thickness
edges.
edges. 0.8 As
As aa20result,
result, the
the necessary
necessary material
material
0.8 is
is obtained
obtained from from the
the areas
areas with
with
0.8 the
the highest
highest plastification,
plastification,
0 40 60 80 100 120 0 20 40 60 80 100 120 0 20 40 60 80 100 120
which
which are are the
the bendingy [mm]edges.
bending edges. FigureFigure 21 21 shows
shows the the ydiminishing
diminishing
[mm] material
material thickness
thickness with y [mm]progressing
with progressing
roll
roll passes
passes andand the
the correlation
correlation with with the the increasing
increasing difference
difference betweenbetween the the desired
desired and and thethe actual
actual
Figure 20. Sheet thicknesses in the cross-sections at x = 0, 200, 350 mm.
material
material inlet.
inlet. Figure 20. Sheet thicknesses in the cross-sections at x = 0, 200, 350 mm.
1.2
1.2 25
25
initial
initialthickness
thickness
[mm]
inlet [mm]
At x = 0 mm, areas 1.1

1.1 with both marginal sheet thinning and thickening 20
20
are observed. The sheet
[mm]
thickness [mm]
sheet
sheetthickness
thickness
material inlet
thinning
xx==350
350mm
mm
is caused 1by
1 the occurring tensile stresses along
roll pass22 the cross-section
rollpass
sheet
sheetthickness
thickness
15
15 (Figure 13). The
thickness
actual
actual
lateral material
thickening is a result of the blank holder roll set adaption

0.9
0.9 roll
rollpass
sheet (Figure 8c). By changing the
pass44
sheetthickness
thickness
10
10 lateral
material
materialinlet
inlet
zz 0.8
0.8 roll
rollpass
pass66
distance
yy xx between the blank holder roll sets, the profile width sheetis
sheet adjusted
thickness
thickness
55 so that the materialnominal nominal
amount
lateral
0.7
0.7 material
materialinlet
inlet
roll
rollpass
pass88 00
at the position with 0.6
0.6
maximum profile depth (hmax,a) is sheet kept
sheet constant
thickness
thickness 00 11 2for
2 33 44 each
55 66 77 8roll
8 99 10 pass.
10 11
11 As a
roll
rollpass
pass10
10
consequence, the remaining 00 20
20 40
40
transverse 60
60
sections80
80 100
100
contain 120
120
surplus material which roll
rollpass
passpushed back into
is
yy[mm]
[mm]
the flange in the course of the following roll passes. This material return creates local compressive
stresses which Figure
Figure 21.
21.
Figurelead Correlation
21. Correlation
Correlation between
between material
to the observed materialthinning
material thinning and
thickening.and insufficient
insufficient
Such areas lateralof material
lateral material inlet.
thinning inlet.
and thickening
appear
3.3. at x = 200Variations
Workpiece-Side mm as well. Even though the maximum thickness is larger, the total thickness
3.3.
3.3. Workpiece-Side
is lower Variations
Workpiece-Side
increase Variations
than at x = 0 mm. Since the profile depth is greater at x = 200 mm, a lower amount of
Besides
material the
needs influence of forming
to be returned to sequence
the design,
flange, wherebythe impact
lowerbyby varying workpiece
compressive stresses properties
occur. The is
Besides
Besides the
the influence
influence of
of forming
forming sequence
sequence design,
design, the
the impact
impact by varying
varying workpiece
workpiece properties
properties
significant. Workpiece properties can be divided into geometric and material properties.
iscross-section
is significant. at x = 350 mm
significant. Workpiece
Workpiece contains
properties
properties can
cannobebesurplus
divided
divided material
into that needs
into geometric
geometric andtomaterial
and be pushed
material back which is why
properties.
properties.
3.3.1. Geometric Properties: Workpiece Thickness
3.3.1.
3.3.1. Geometric
Geometric Properties:
Properties: Workpiece
Workpiece Thickness
Thickness
To examine the influence of the DC04 workpiece thickness, the same roll pass is performed on
ToTo examine
examine the
the influence
influence of
of the
the DC04
DC04 workpiece
workpiece thickness,
thickness, the
the same
same roll
roll pass
pass is
is performed
performed on
on aa
a sheet blank with 1.0 mm and 0.5 mm. The results demonstrate that, with decreasing workpiece
sheet
sheet blank
blank with
with 1.0
1.0 mm
mm and
and 0.5
0.5 mm.
mm. The
The results
results demonstrate
demonstrate that,
that, with
with decreasing
decreasing workpiece
workpiece
thickness, identical tool motions achieve larger profile depths (Figure 22).
thickness,
thickness, identical
identical tool
tool motions
motions achieve
achieve larger
larger profile
profile depths
depths (Figure
(Figure 22).
22).
0.5
0.5
profile
profile bottom
bottom curves
curves
z = profile depth [mm]
00
00 200
200 400
400 600
600 800
800 1000
1000 1200
1200 1400
1400
-0.5
-0.5
-1
-1
thickness
thickness 0.5
0.5 mm
mm
-1.5
-1.5
thickness
thickness 1.0
1.0 mm
mm
-2
-2
nominal
nominalcurve
curve
-2.5
-2.5
xx ==longitudinal
longitudinal direction
direction[mm]
[mm]
Figure
Figure 22.
Figure 22. Resulting
22. Resulting profile
Resulting profile bottom
profile bottom curves
bottom curves with
curves with workpiece
with workpiece thickness
thickness variation.
workpiece thickness variation.
variation.
When
When regarding
regarding the
the stress
stress and
and strain
strain distribution
distribution during
during the
the process,
process, itit is
is evident,
evident, that
that the
the
additionally
additionally gained
gained profile
profile depth
depth isis not
not aa result
result of
of advantaged
advantaged lateral
lateral material
material inlet
inlet but
but of
of material
material
stretching.
stretching. Figure
Figure 23
23 depicts
depicts the
the comparison
comparison of of the
the lateral
lateral stress
stress distributions
distributions along
along the
the transverse
transverse
section
section at
at the
the position
position of
of maximum
maximum depth depth (h max,a)) of
(hmax,a of the
the considered
considered blanks
blanks with
with different
different material
material
When regarding the stress and strain distribution during the process, it is evident, that
the additionally gained profile depth is not a result of advantaged lateral material inlet but of
material stretching. Figure 23 depicts the comparison of the lateral stress distributions along the
transverse section at the position of maximum depth (hmax,a ) of the considered blanks with different
material
J. Manuf.
600 thicknesses.
Mater. Process. 2019, 3, x FOR PEER REVIEW
workpiece thickness = 1 mm
0.005 12 of 15
[N/mm²][N/mm²]
lateraliny-lateral y-
lateralε22strain ε22
400 0.004
200
thickness 0.5 mm
600 0.005
0.003
0
workpiece thickness = 1 mm thickness 1.0 mm
[N/mm²][N/mm²] directiondirection
400 0.004
0 20 40 60 80 100 120 140 160 0.002
thickness 0.5 mm
lateraliny-lateral y- stress instress
lateral strain
-200
200
0.003
lower surface 0.001
-4000 thickness 1.0 mm
0 20 40 60 80 100 120 top surface
140 160 0.002
0
-600
-200 0 20 40 60 80 100 120 140 160
a. tensile y = lateral coordinate [mm] lower surface c. 0.001 y = laterial coordinate [mm]
-400
600 top surface 0
-600 workpiece thickness 0.5 mm 0 20 40 60 80 100 120 140 160
a. 400
tensile y = lateral coordinate [mm] c. y = laterial coordinate [mm]
200
600
0
400
workpiece thickness 0.5 mm z
directiondirection
0 20 40 60 80 100 120 140 160

tensile tensile
tensile tensile
-200
x
stress instress
200
tensile lower surface y
-4000
0 20 40 60 80 100 120 top surface
140 160
z
-600
b.
-200
y =tensile
lateral coordinate [mm] y x
lower surface
-400
top surface
-600
b. Figure 23. (a–b) yLateral
= lateral coordinate [mm] (c) strain along the cross-section at x = 350 mm (hmax,a) for sheet
stress and
metal thicknesses 1.0 mm and 0.5 mm.
Figure 23. (a–b)
Figure 23. (a–b) Lateral
Lateral stress
stress and
and(c)
(c)strain
strainalong
alongthe
thecross-section
cross-sectionatatxx==350
350mm
mm(h(h
max,a ) for sheet
max,a) for sheet
metal
metal thicknesses
thicknesses 1.0 mmmm and 0.5
0.5 mm.
By regarding the1.0stress and mm.
distributions (Figure 23a,b), a significant expansion of the areas with
dominant
By tensile stresses are observed in (Figure
the workpiece awith smaller thickness. Increasing lateral
By regarding
regarding the
the stress
stress distributions
distributions (Figure 23a,b),
23a,b), a significant
significant expansion
expansion ofof the
the areas
areas with
with
stretching
dominant of the material is the result of growing tensile stresses (Figure 23c).
dominant tensile
tensile stresses
stresses are
are observed
observed inin the
the workpiece
workpiece with
with smaller
smaller thickness.
thickness. Increasing
Increasing lateral
lateral
stretching
stretching of
of the
the material
material is
is the
the result
result of
of growing
growingtensile
tensilestresses
stresses(Figure
(Figure23c).
23c).
3.3.2. Material Properties
3.3.2. Material Properties
3.3.2.The influence
Material of material variation was investigated by performing the same forming sequence
Properties
on a The
DC04-steel
influenceand an Al2024
of material blank with
variation identical dimensions.
was investigated The resulting
by performing profile bottoms
the same forming sequenceand
on
The influence of material variation was investigated by performing the same forming sequence
aplastic strainand
DC04-steel along the cross-section
an Al2024 blank withat the position
identical of maximum
dimensions. profileprofile
The resulting heightbottoms
hmax,a isand
shown in
plastic
on a DC04-steel and an Al2024 blank with identical dimensions. The resulting profile bottoms and
Figurealong
strain 24. the cross-section at the position of maximum profile height hmax,a is shown in Figure 24.
plastic strain along the cross-section at the position of maximum profile height hmax,a is shown in
Figure 24.
0.14
0
profile bottom curves DC04, Al2024
0.12
0 200 400 600 800 1000 1200 1400
[mm] z [mm]
strain εyy
-2 0.1
0.14
-4
0
profile bottom curves DC04, Al2024 0.08
0.12 Al2024
-6 0 200 400 600 800 1000 1200 1400
profilezdepth
lateralεyy
-2 0.06
target geometry 0.1
DC04
-8 0.04
lateral strain
-4 0.08
Al2024 0.02 Al2024
-10
profile depth
-6 0.06
DC04
target geometry
-12
-8
0
0.04 DC04
0 50 100 150
-14 Al2024 -0.02
0.02
-10
longitudinal direction x [mm] lateral direction y [mm]
DC04 0
-12
0 50 100 150
-14 -0.02
Figure 24. (a)
Figure 24. (a) Profile
Profile bottom
bottom curves
directionand
curves
longitudinal and (b)
(b) lateral
x [mm] lateral strain
strain along
along the
the cross-section
cross-section at
at xx ==y [mm]
lateral direction 350
350 mm
mm of
of aa
DC04 and Al2024 sheet
DC04 and Al2024 sheet blank.blank.
Figure 24. (a) Profile bottom curves and (b) lateral strain along the cross-section at x = 350 mm of a
The
DC04profile
The bottom
and Al2024
profile sheetcurve
bottom blank.of
curve of thethe Al2024
Al2024 workpiece
workpiece showed
showed anan improved
improved geometricgeometric accuracy
accuracy
compared
compared toto the
the DC04
DC04 blank,
blank, while
while the
the lateral
lateral strain
strain increased
increased marginally.
marginally. The
The decisive
decisive explanation
explanation
for The profile bottom curve of the Al2024 workpiece showed an improved geometric accuracy
for the achievement of larger profile depths is the improved lateral material inlet during the first
the achievement of larger profile depths is the improved lateral material inlet during the first roll
roll
compared
passes. to the
Figure 25 DC04
depicts blank,
the whilematerial
lateral the lateral strain
inlet and increased
the marginally.
longitudinal stressThe
in decisive
the profileexplanation
edge at the
passes. Figure 25 depicts the lateral material inlet and the longitudinal stress in the profile edge at
for themoment
same achievement
of the of larger
first roll profile depths is the improved lateral material inlet during the first roll
pass.
the same moment of the first roll pass.
passes. Figure 25 depicts the lateral material inlet and the longitudinal stress in the profile edge at
the same moment of the first roll pass.
0.06 lateral material inlet
material inlet
0.05
0.04
[mm]
0.03 Al2024
0.02
0.01 DC04
0
-0.01 0 200 400 600 800 1000 1200 1400
(a) longitudinal direction x[mm]
20
longitudinal stress in the profile edge
stress [N/mm²]
longitudinal
0
0 200 400 600 800 1000 1200 1400 Al2024
-20
DC04
-40
-60
(b) longitudinal direction x[mm]
Figure 25. Material inlet (a) and longitudinal stress in the profile edge (b) of the DC04 and
Figure 25. Material inlet (a) and longitudinal stress in the profile edge (b) of the DC04 and Al2024
Al2024 blanks.
blanks.
The larger elastic region of the Al2024 material results in less plastic deformation and strain
The larger
hardening in theelastic region
bending of during
edges the Al2024 material
the initial rollresults
passes.inThis
lessbenefits
plastic the
deformation and strain
lateral material flow
hardening in the bending edges during the initial roll passes. This benefits the lateral material
leading to significantly increased material inlet (Figure 25a). The positive effect of the enhanced lateral flow
leading
materialtoinlet
significantly
outweighsincreased
the largermaterial inletrecovery
springback (Figure of25a).
theThe positive
Al2024 effect
material, of the
which enhanced
results in an
lateral material
improved inlet outweighs
size accuracy. the larger springback
As a consequence recovery
of increasing of theinlet,
material Al2024
the material,
accordinglywhich results
increasing
in an improved size accuracy. As a consequence of increasing material inlet, the accordingly
longitudinal stress in the profile edge (Figure 25b) enhances the risk of wrinkles. As a counter measure,
increasing longitudinal stress in the profile edge (Figure 25b) enhances the risk of wrinkles. As a
additional roll passes can be performed in total to reduce the material inlet for each roll pass.
counter measure, additional roll passes can be performed in total to reduce the material inlet for each
roll pass.
4. Conclusions
In the context of the presented research, the stress and strain distribution during the process
4. Conclusions
“flexible roller beading” has been identified and understood. The potential component error patterns
and In thecause
their context
have ofbeen
the presented
determined. research, the stress
The challenge and strain
of process designdistribution during the
is the unification process
of opposing
“flexible roller beading” has been identified and understood. The
guidelines. Longitudinal elongation of the profile bottom is necessary for the manufacturing of potential component error
the
patterns and their cause have been determined. The challenge of process design
target geometry and therefore indispensable. To reduce the longitudinal stresses in the profile, bottom is the unification of
opposing guidelines. zones
sharp height-transition Longitudinal
should beelongation
avoided in of the profile design.
the component bottomDuring is necessary
the process, forlateral
the
manufacturing of the target geometry and therefore indispensable. To
tensile stresses are to be minimized in order to avoid plastic elongation, and relating thereto, materialreduce the longitudinal
stresses
thinning. in To
theprevent
profile,the bottom
tensilesharp height-transition
stresses from exceeding zones should value,
the critical be avoided in thematerial
the lateral component inlet
design.
must beDuringenhanced.the process, lateral
A significant tensile stresses
dependence are the
between to be minimized
material choiceinand order
theto avoid plastic
material inlet is
elongation,
shown. Theand relating
material thereto,
inlet materialwith
is associated thinning. To preventofthe
the appearance tensile stresses
stretching from exceeding
and compression zones
the critical
in the profilevalue,
edgestheand
lateral material
flanges, whichinletcanmust be enhanced.
cause component Afailure
significant
in thedependence
form of sheet between the
wrinkles.
material choice and the material inlet is shown. The material inlet is associated
The compression zones are caused by non-pure lateral displacement of the material in the profile with the appearance
of stretching
flanges. andtocompression
In order avoid the formingzones of in convex–concave
the profile edges and flanges,
profile whichand
edge contours can reinforce
cause component
a uniform
failure
material ininlet,
the form
tailored of workpieces
sheet wrinkles. The compression
are conceivable. By using zones are caused
customized by non-pure
semi-finished lateral
blanks, the
displacement of the material in the profile flanges. In order to avoid the
necessary material for the height-variable profile is accordingly provided. By performing additional forming of convex–concave
profile edgeand
roll passes contours
thuslyand reinforce
reducing theadegree
uniform of material
deformationinlet,per
tailored workpieces
step, the compressive are conceivable.
stresses canBy be
using
furthercustomized
reduced. In semi-finished
the design of the blanks,
forming thesequence,
necessary material
large increasesfor ofthe
theheight-variable
bending angle, during profile the is
accordingly provided. By performing additional roll passes and thusly
last roll passes particularly, are to be avoided to improve the geometric accuracy of the profile bottom reducing the degree of
deformation
curve. With regardper step, the findings
to the compressive
obtained stresses can
in the be further
presented reduced.
work, process In optimization
the design ofmeasures
the forming can
sequence, large increases of the bending angle, during the last roll passes
be derived and applied to the experimental setup and experiment design. By means of the identified particularly, are to be
avoided to improve the geometric accuracy of the profile bottom curve.
influence of process-, tool-, and workpiece-side variations, corresponding process adaption can be With regard to the findings
obtained
actioned whenin the faced
presented work, process
with varying processoptimization
conditions. The measures can beobjective
overarching derivedofand theapplied
presented to and
the
experimental setup and experiment design. By means of the identified influence
future work is the determination of process limits and the reliable derivation of forming sequences to of process-, tool-,
and workpiece-side
obtain variations, corresponding
individually customized height-variableprocess profiles.adaption can be actioned
By accomplishing when faced
these objectives, with
a highly
varying process conditions.
flexible, industrial-suited sheetThe overarching
metal profile forming objective
process,of the
which presented and futuremanufacturing
allows on-demand work is the
determination of process limits
of light-weight profiles, can be introduced. and the reliable derivation of forming sequences to obtain
individually customized height-variable profiles. By accomplishing these objectives, a highly
Author Contributions:
flexible, industrial-suitedConceptualization,
sheet metal T.W.profile
and P.G.;forming
methodology, T.W.; software,
process, which allows T.W.; validation,
on-demand T.W.;
formal analysis, T.W.; investigation, T.W.; resources, P.G.; data curation, T.W.; writing—original draft preparation,
manufacturing of light-weight profiles, can be introduced.
T.W. and P.G.; writing—review and editing, P.G. and T.W., visualization, T.W., supervision, P.G., project
administration, P.G.; funding acquisition, P.G.
Funding: This research was funded by the German Research Foundation (DFG), grant number GR1818_53-1
“Production of Multi-Directional Widened Profiles”.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to
publish the results.
References
1. Schwab, K. The Fourth Industrial Revolution; Crown Business: New York, NY, USA, 2017.
2. Yang, D.Y.; Bambach, M.; Cao, J.; Duflou, J.R.; Groche, P.; Kuboki, T.; Sterzing, A.; Tekkaya, A.E.; Lee, C.W.
Flexibility in metal forming. In CIRP Annals—Manufacturing Technology; Elsevier Ltd.: Amsterdam,
The Netherlands, 2018; pp. 743–762.
3. Mehrabi, M.G.; Ulsoy, A.G.; Koren, Y.; Heytler, P. Trends and perspectives in flexible and reconfigurable
manufacturing systems. J. Intell. Manuf. 2002, 13, 135–146. [CrossRef]
4. Allwood, J.M. Steel and Aluminium in a Low Carbon Future. In Proceedings of the 10th International
Conference on Technology of Plasticity, ICTP 2011, Aachen, Germany, 25–30 September 2011; pp. 27–42.
5. Broer, G.; Martin-Bullmann, R. Kaltprofile; Verlag Stahleisen mbH: Düsseldorf, Germany, 1993.
6. Groche, P.; Von Breitenbach, G.; Jockel, M.; Zettler, A. New Tooling Concepts for Future Roll Forming
Applications. In Proceedings of the ICIT, 4th International Conference on Industrial Tools, Celje, Slovenia,
10–12 December 2003; pp. 121–126.
7. Ona, H.; Sho, R.; Nagamachi, T.; Hoshi, K. Development of Flexible Cold Roll Forming Machine Controlled
by PLC. In Steel Research International; WILEY-VCH: Weinheim, Germany, 2010; pp. 182–185.
8. Lindgren, M.; Ingmarsson, L.-O. 3D roll forming of hat-profile with variable depth and width. In Proceedings
of the 1st International Congress on RollForming, Bilbao, Spain, 14–15 October 2009; pp. 1–8.
9. Sedlmaier, A.; Dietl, T.; Harrasser, J. 3D Roll Forming in Automotive Industry. In Proceedings of the 5th
International Conference on Steels in Cars and Trucks, Amsterdam, The Netherlands, 18–22 June 2017.
10. Groche, P.; Storbeck, M.; Wang, T. Continuous forming of height-variable beads by flexible roller beading.
J. Adv. Manuf. Technol. 2019, under review.
11. Berner, S.; Storbeck, M.; Groche, P. A Study on Flexible Roll Formed Products Accuracy by Means of FEA and
Experimental Tests. In Proceedings of the 14th International ESAFORM Conference on Material Forming,
AIP Conference Proceedings, Belfast, Northern Ireland, 27–29 April 2011; pp. 345–350.
12. Groche, P.; Zettler, A.; Berner, S.; Schneider, G. Development and verification of a one-step-model for the
design of flexible roll formed parts. Int. J. Mater. Form. 2011, 4, 371–377. [CrossRef]
13. Park, J.C.; Yang, D.Y.; Cha, M.H.; Kim, D.G.; Nam, J.B. Investigations of a new incremental counter forming
in flexible roll forming to manufacture accurate profiles with variable cross-sections. Int. J. Mach. Tools Manuf.
2014, 86, 68–80. [CrossRef]
14. Kiuchi, M.; Koudabashi, T. Automated design system of optimal roll profiles for cold roll forming.
In Proceedings of the 3rd International Conference on Rotary Metalworking Processes, Kyoto, Japan,
8–10 September 1984; pp. 423–436.
15. Rebelo, N.; Nagtegaal, J.C.; Taylor, L.M.; Passman, R. Comparison of implicit and explicit finite element
methods in the simulation of metal forming processes. In Proceedings of the ABAQUS Users Conference,
Newport, Rhode Island, 1 January 1992.
16. Senanayake, R.S.; Cole, I.M.; Thiruvarudchelvan, S. The application of computational and experimental
techniques to metal deformation in cold roll forming. J. Mater. Process. Technol. 1994, 45, 155–160. [CrossRef]
17. Daniel, W.J.T.; Meehan, P.A. Implicit finite element study of non-steady effects in cold roll forming.
In Proceedings of the 5th Australasian Congress on Applied Mechanics (ACAM 2007), Brisbane, Australia,
10–12 December 2007.
18. Groche, P.; Mueller, C.; Traub, T. Experimental and Numerical Determination of Roll Forming Loads; Wiley:
Weinheim, Germany, 2013; pp. 112–122.
19. MARC. User Information—Volume A, Element Library—Volume B; MARC Analysis Research Corporation:
Palo Alto, CA, USA, 2014.
20. Gehring, A. Beurteilung der Eignung von Metallischem Band und Blech zum Walzprofilieren. Ph.D. Thesis,
Universitätsverlag Karlsruhe, Karlsruhe, Germany, 2006.
21. Traub, T.; Miks, C.; Groche, P. Force Measurements Supporting the Set-up Process in Roll Forming.
In Proceedings of the 1st Annual International Conference on Mechanical Engineering, ATINER’s Conference
Paper Series, MEC2017-2346, Athens, Greece, 17–20 July 2017.
22. Boman, R.; Papeleux, L.; Bui, Q.V.; Ponthot, J.P. Application of the Arbitrary Lagrangian Eulerian formulation
to the numerical simulation of cold roll forming process. J. Mater. Process. Technol. 2006, 177, 621–625.
[CrossRef]
23. Görtan, M.O.; Vucic, D.; Groche, P.; Livatyali, H. Roll forming of branched profiles. J. Mater. Process. Technol.
2009, 209, 5837–5844. [CrossRef]
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Sheet Metal Profiles With Variable Height: Numerical Analyses On Flexible Roller Beading

Uploaded by

Document Information

Original Title

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

Sheet Metal Profiles With Variable Height: Numerical Analyses On Flexible Roller Beading

Uploaded by

Copyright:

Available Formats

Journal of

J. Manuf. Mater. Process. 2019, 3, 19; doi:10.3390/jmmp3010019 www.mdpi.com/journal/jmmp

2. Materials and Methods

Figure 1. Characteristics and nomenclature of the target geometry.

Side view Front view (feed direction)

(a) x lower roller (d) lower roller

b) drive unit a) forming unit c) guidance unit

Flexible roller beading

Experimental results: (a)(a)manufactured

∆uy = const. ∆uy ∆uy ∆uy

∆uy = const. ∆uy ∆uy ∆uy

J. Manuf. Mater. Process. 2019, 3, x FOR PEER REVIEW 7 of 15

Figure 16. Displacement

0.8 0.8 0.8

0.6 0.6 0.6

1.1 1.1 1.1

Figure 20. Sheet thicknesses in the cross-sections at x = 0, 200, 350 mm.

-5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120 -5 0 20 40 60 80 100 120

-15 -15 -15

cross-section cross-section cross-section

At x = 0 mm, areas 1.1

thickening is a result of the blank holder roll set adaption

0 20 40 60 80 100 120 140 160

0.06 lateral material inlet

You might also like