See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235655689

Analysis of Strip Residual Curvatures in Anti-Crossbow Cassette during

Tension Levelling

Article  in  Steel Research International · January 2008

CITATION READS

1 1,580

5 authors, including:

Mostafa Jamshidian Ali Beheshti

Isfahan University of Technology George Mason University
20 PUBLICATIONS   152 CITATIONS    26 PUBLICATIONS   236 CITATIONS   

SEE PROFILE SEE PROFILE

Susan Dolatabadi Mahmoud Salimi

Gonabad University of Medical Sciences Isfahan University of Technology
1 PUBLICATION   1 CITATION    82 PUBLICATIONS   799 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

High temperature tribological performance of Ni alloys under helium environment for very high temperature gas cooled reactors (VHTRs) View project

Stochastic analysis of interphase effect on mechanical property of polymer/clay nanocomposites View project

All content following this page was uploaded by Mostafa Jamshidian on 28 December 2018.

The user has requested enhancement of the downloaded file.

STEEL RESEARCH INTERNATIONAL Pages: 513-520 Published: September 2008

ANALYSIS OF STRIP RESIDUAL CURVATURES IN ANTI-

CROSSBOW CASSETTE IN TENSION LEVELING PROCESS
M. Jamshidian1, A. Beheshti1, A. Sadeghi Dolatabadi2, M. Olfat Nia1, M. Salimi1
1
Department of Mechanical Engineering, Isfahan University of Technology, P.O. Box 84154,
Isfahan, Iran
2
Mobarakeh Steel Company, Isfahan 84815-161, Iran

ABSTRACT performed to elongate the strip plastically in

Tension leveling is a process used in steel a combination of tensile and bending strain
industry in order to remove any shape so that all longitudinal fibres in the strip are
defects present in temper rolled strip. of approximately the same length. The
Although strip bad shapes come into sight tension levelling process is designed
as edge waves, centre buckles and quarter primarily to remove shape defects resulting
buckles can be corrected giving the strip a from unequal elongation over the width of
suitable amount of elongation in work rolls rolled strips such as wavy edges, side
cassette, such defects of so-called crossbow buckles and centre buckles. This
and coil set known as residual curvatures improvement occurs in work rolls cassettes
are inevitable because in the process the in a typical tension leveller as shown in
strip undergoes repeated bending. These Figure 1. The repeated bending and
undesired curvatures can be corrected in stretching in this step however, tends to
anti-crossbow and anti-coil set cassettes, induce subsequent curl or coil set and gutter
giving the strip an optimum amount of or crossbow which are kinds of global
curvature. In this paper a theoretical shape defects. Crossbow and coil set are
analysis based on incremental theory of usually corrected in the second stage of the
plasticity and considering mixed mode process when the strip passes through anti-
hardening for material behavior has been crossbow and anti-coil set cassettes
employed in order to study the effect of respectively.
primary process parameters namely work
curvature and tension on strip residual
curvatures coming out of the work rolls
cassette and then to analyze variation of
these curvatures with respect to second
bending curvature in anti-crossbow
cassette. Review of the results presents a Figure 1. A typical tension leveller machine
method for determining the optimal work
curvature in anti-crossbow cassette to Because of its crucial role in delivering
remove residual curvatures especially perfect flat strips to customers, tension
transverse residual curvature or crossbow. levelling has been studied by many
researchers. Misaka and Masui studied
1. INTRODUCTION analytically the mechanism of the process
Tension levelling is an indispensable for curling and guttering after repeated
process to improve the flatness of metallic bending and developed a method for
strip products in order to satisfy the calculating the curl and gutter [2]. The
stringent requirements of the costumers. amount of strip elongation, power loss and
Tension levelling is a metal forming tension that occurs in the process were
technique to produce perfectly flat metallic studied by Patula [3]. Kawaguchi studied
strips [1]. The tension levelling process is
numerically the actual strip curvature on a curvature in anti-crossbow cassette that can
roll and the strip deflection between rolls in remove residual curvatures especially
the process [4]. Hattori et al. developed a transverse residual curvature or crossbow.
mathematical model to predict the curl and
crossbow induced in a strip during the 2. THEORETICAL ANALYSIS
process[5]. Kajihara et al. developed an To simplify and simulate the complicated
analytical method of the work radius, behavior of strip in tension leveling
elongation, curling and guttering of strips process, the following assumptions are
[6]. Hira et al. introduced theoretical considered in the present analysis:
analysis of deformation behaviour under 1. The strip elongates under plane strain
repeated bending in strip processing lines conditions. Transverse strain is neglected
[7]. Hibino devised a practical equation for because the strip width is sufficiently large
estimating the levelling strain from compared to the strip thickness.
experimental results and revealed that the 2. The stress in thickness direction is
extreme point of the strain did not lie at the ignored with respect to large in plane stress
contact point between the sheet and the roll values, that is plane stress condition in
[8]. Yoshida and Urabe carried out a thickness direction dominates.
computer aided process design for tension 3. The shear stresses and shear strains in
levelling with finite element analysis longitudinal (rolling) and transverse
involved with a sophisticated constitutive directions are neglected.
model of cyclic plasticity [9]. Morris et al. 4. The applied tension on strip is assumed
performed a series of low cycle fatigue tests to be constant as strip passes through roll
under constant strain amplitude control gaps i.e. longitudinal tension on strip is
[10]. The results from their experiments constant during bending over free rotating
suggest that a mixed mode hardening rolls.
process is present in tension levelling, with 5. The tensile stress is applied uniformly
the application of a Bauschinger effect. Huh across the strip width.
et al. performed a simulation-based process 6. The material behaviour is considered to
design for the tension levelling of metallic be in mixed or combined hardening mode
strips based on the elastic-plastic finite i.e. both isotropic and kinematic hardening
element analysis [11]. As in tension modes are deemed to be present as material
levelling process the material undergoes deforms plastically.
repeated bending unbending deformation, it From the assumptions 2 and 3 and Mises
experiences cyclic tension compression yield criterion, the equivalent stress is
loadings under which the effect of given by:
Bauschinger effect dominates. Meanwhile
the effect of isotropic hardening in cyclic
loading affects the material behaviour as
well. In this study a theoretical analysis 3
based on incremental theory of plasticity 2
and considering mixed mode hardening for (1)
material behaviour has been employed in
order to study the effect of primary process Where: , , – subscripts indicating
parameters on strip residual curvatures and longitudinal or rolling, transverse and
then to analyze variation of these curvatures thickness direction of the strip, respectively,
with respect to second bending curvature in , , – Deviatoric stress components in
anti-crossbow cassette. Finite element , , directions, respectively, , , –
simulations have also been employed to Back stress components in , , directions,
verify the proposed solution. The Results respectively. The equivalent plastic strain
are used to determine the optimal work
increment is expressed by the
following expression: (7)

2 The associated flow rule using the above

(2)
3 yield function is:

Where: , , – Plastic strain

(8)
increments in , , directions,
respectively. It is assumed that the
relationship between stress and strain is
governed by the Hooke’s law for elastic Where: – Constant coefficient of
deformation. Then the following expression associated flow rule.   is defined as:
relates stress increments to strain
increments while the deformation occurs
elastically: ⁄
⁄ (9)
⁄
d (3)
In order to relate stress and strain
increments during plastic deformation, an
Where: , , – Total strain equation to evaluate the yield locus
increments in , , directions, movement in terms of plastic deformation is
respectively, , , – Stress essential. The simplest equation is Prager’s
linear kinematic hardening rule that for our
increments in , , directions,
case is defined as [12]:
respectively. The matrix is defined as:

1 1
(10)
1 (4)
1

Where: – Poisson’s ratio, – Elastic Where: , , – Back stress

modulus. In case of plastic deformation, the
increments in , , directions,
associated flow rule together with Mises
respectively, – kinematic hardening
yield criterion as the yield function is used
modulus. Coefficient is a scalar that
to relate stress and strain increments. The
indicates the kinematic hardening rate or
yield function is defined by the following
the rate at which yield locus moves during
equation:
the development of plastic deformation. If
3 the relation between and is known as:
0
2
(5) σ
(11)
Where: – Current yield stress or flow
stress. and are defined as: Then H, the tangent modulus or isotropic
hardening modulus will be defined as:

(6)
(12)
In the present study the isotropic hardening assumptions 1 and 2 respectively. Therefore
is approximated by the exponential law as if strain increment is known at a given
the following [13]: time, the incremental stresses and
and strain increment can be calculated
σ 1 (13) thorough some computational efforts. The
method used for identification of elastic and
Where: – Yield stress at zero plastic plastic boundary and method of calculating
strain, – Material parameter indicating reverse direction stress in this study is
the maximum change in the size of yield similar to what has been proposed by Hira
surface, – Material parameter that defines et al. [7]. Generally the procedure involves
the rate at which size of the yield surface strain increment decomposition into elastic
changes as plastic straining develops, – and plastic parts. When through thickness
Accumulated equivalent plastic strain. distributions of stresses and are
Differentiating equation (13) gives: calculated as above, the bending moments
in longitudinal and transverse directions,
(14) and are calculated by integration.
After successive bending and unbending the
Finally doing some algebra, the following strip is released or is unloaded. Unloading
equation will be obtained as the relationship means that the integrated longitudinal stress
between stress increments and total strain through thickness and the longitudinal
increments during plastic deformation: and transverse bending moments, and
, are reduced to zero. and at
3 0 are first calculated by varying the
in the middle of the strip thickness so
2
as to set to zero. Subsequently,
calculations are made to reduce the bending
3 moments to zero. When the curvature of the
2 strip is zero under loading of bending
moments and , the residual
σ longitudinal and transverse curvatures after
unloading, and are given by the
following equations, respectively:
(15)

Where I is the three dimensional identity

1 (18)
matrix and and are defined as:

⁄ 1 (19)
⁄ (16)
⁄ Where: t – strip thickness. is a constant
being defined by the following equation:
⁄
⁄ (17)
12 1 (20)
⁄

In Eq. (15) the transverse strain increment Since the present constitute equations for
d and the stress increment in thickness strip behaviour are in terms of longitudinal
strain increment , here an equation is
direction dσ are set to zero according to
needed to calculate through thickness
bendingg strain. Whhen a strip with
w thicknness 3. RESULT TS AND DISCUSSION N
is bent
b to a curvature with the In this sectionn the simulaation code has
h been
longituddinal strainn in middlee fibre , the used to sim mulate proocess of residual
longituddinal strainn at a diistance froom urvatures eliimination in a typicall tension
cu
middle fibre in thiickness direection is givven levveller as shhown in Figgure 1. Calcculations
by: aree made forr a steel ssheet with material
paarameters as a shown in table 1 and
  (221) thiickness of 1 mm.

Using above eqquation toogether with w Taable 1. material

m paarameters used
u in
incremeental theoryy of plasticity, a code hash sim
mulations (S
SI Units)
been deeveloped inn order to simulate sttrip E b C
behavioour in tensioon levellingg process. ForF 200e6
2 0.3 200e6 500 e6 500 10e9
the purppose of connciseness details
d are not
n
presenteed here; details
d can be found in 3.11. Effect of
o primary process
previouus works [2,,7]. In order to verify the pa arameters on
o residuall curvatures
model presented here, the bending and a Prior to analyyse the antii-crossbow cassette
unbendiing processs of steel sheets is also a in tension levelling proocess, the effect
e of
simulateed using thhe finite eleement packaage priimary proceess parametters i.e. tenssion and
ABAQU US [13]. Thhe steel waas assumedd to wo ork curvatuure on resiidual curvaatures in
be elastic-plastic with isotroopic-kinemaatic wo ork rolls cassette hhas been studied.
hardeninng behaviour. For the sake of Caalculation results
r for residual cuurvatures
brevity the simuulation dettails are not n i.ee. crossbow
w and coil sset are presented in
presenteed here; details
d can be found in Fig gure 3 for a strip that is bent and
d unbent
present authors’ previous works [114]. byy a curvaturre varying from 0 to 60 1/m
Figure 2 shows thee equivalentt plastic strrain unnder constannt tension oof 100 MPaa. On the
distribuution after straightenin
s g as a typiical oth her hand ini Figure 4 longitudiinal and
examplee of the resuults. All Strress and strrain traansverse ressidual curvatures are depicted
distribuution resultss from botth simulatioons forr a stripp that unndergoes bending- b
like Figgure 2 weere seen too be in cloose unnbending prrocess by a constant curvature
c
agreemeent, which demonstratees the validdity off 25 (1/m) under
u a tenssion varying
g from 0
and accuuracy of booth approachhes. MPa to 200 MPa.
M
Distance from middle fibre (m)

Residual curvatures ,1/m

1/m

Equiva
alent plastic strain
s Curvature
e , 1/m

Figure 2.
2 Comparisson of Equivvalent plasttic Figgure 3. Efffect of curvvature variaation on
strain prrofile from F.E.M. andd analytical ressidual curvvatures undeer constantt tension
solutionns off 100 MPa
 seccond bennding currvature off κ
0 tto 50 1/m . The fo following can be
Residual curvatures (1/m)

obbtained From m Figure 5:

(1)) When thee curvatures κ and κ are the
samme that is κ κ 5
50 1/m, thee sign of
ressidual curvvatures are reversed. In I other
wo ords residuual curvatuures occur by the
samme absoluute value but in opposite
dirrection. (2)) The signss of crossb bow and
cooil set are noot necessariily the samee i.e. for
a special
s amoount of κ , ccrossbow is positive
wh hile coil seet is negatiive. (3) Co oil set is
Te
ension (1//m)
mo ore sensitivve to secondd bending curvature
c
Figure 4. Effect of tension variation on thaan crossboow, as whhat was true t for
residuall curvaturess under connstant bendiing inttroduction of o residual curvatures in strip
unbendiing curvaturre of 25 1/m
m duue to the first bendiing in wo ork rolls
cassette. (4) Both
B crossbbow and coil set can
These figures show s that longitudinal bee removed by reversee bending with an
residuall curvature κ is moree sensitive to opptimum cuurvature off second bending
the appplied bendiing-unbendding curvatuure wh hich is different
d foor eliminaation of
with resspect to trannsverse residdual curvatuure croossbow or coil set. Inn this speccial case
κ . Inn other words w appllication off a croossbow κ is zero w when κ 13 1/m
curvaturre even sm mall, leads tot a relativvely wh hile coil sett disappears at κ 5 1
1/m. As
large chhange in κ and afterr a maximuum cooil set is morre sensitive to reverse bending,
b
value thhere is no perceptible change inn κ it reaches zerro in an am mount of κ that is
with inncreasing κ. But transvverse residdual sm
maller than what is neeeded for crossbow
c
curvaturre κ inccreases sm moothly with
w eliimination. When
W crosssbow is zeero there
wo ould be a negative
n coil set that sh
hould be
bendingg curvaturee κ , with a relativvely
remmoved by a similar proocedure in anti-coil
sudden change in a point aftter the suddden
sett cassette that is m mounted aftter anti-
increasee in κ .
croossbow caassette in a typical tension
levveller. The necessary
n aamount of curvature
c
3.2. Elimination
E n of crossboow by secoond
in both cassetttes can be applied to strip by
bendingg in anti-crrossbow casssette
djusting corrresponding roll intermeesh.
adj
The elim mination off crossbow introducedd to
the stripp due to bennding-unbennding in woork
rolls cassette is invvestigated in this section.
This unndesired currvature is reemoved by the
Residual curvatures (1/m)

applicattion of seecond bendding in annti-

crossbow cassettee. Second bending iss a
longituddinal curvatture appliedd to the stripp in
a direction oppossite to the last bendiing
direction in work rolls
r cassettte. The trickk is
that crossbow is rem moved by using
u the saame
factor thhat causes its appearaance. Figuree 5
shows thet calculaation resultss for residdual
c set when a
curvaturres i.e. crosssbow and coil
strip is bent
b and strraightened byb a curvatuure Second bending ccurvature (1/m)
of κ 50 1/m and thenn bent and a
straighteened in thee reverse direction
d byy a Fig
gure 5. Varriation of crossbow andd coil set
wiith respect to second beending curv
vature
3.3. Efffect of tenssion and sttrip thickness vaaries with sttrip thickneess variation
ns. From
on residdual curvattures elimination theese figures it is clear tthat as both
h tension
After quantitative
q e analysis of crossbow nd strip thicckness incrrease the necessary
an
eliminattion, here the effect of two othher mount of κ to eeliminate residual
am
importaant process parameterrs i.e. tensiion cu
urvatures decreases.
d This meaans that
and strip thicknesss on the opttimum amouunt ressidual curvvatures in thinner strip are
of κ haas been invvestigated, following the remmoved by greater am mounts off second
same prrocedure ussed for plottting Figuree 5. beending curvvature. Furthhermore co omparing
Figures 6 and 7 shhow the efffect of tensiion cu
urvature andd tension shhows that the t main
variations on residual curvatuures facctor in residual curvattures elimin nation is
eliminattion and Fiigures 8 annd 9 illustrrate beending curvvature becaause the effecte of
how thee optimum second
s bendding curvatuure tennsion is sligght.

Crossbow (1/m)
Coil set (1/m)

Second be
ending curvatture (1/m)) Seco
ond bending curvature (1/m)

Figure 6. Effect of o tension variations

v o
on Figure
F 7. efffect of tension variatiions on
coil sett eliminatioon crrossbow eliimination
Crossbow (1/m)
Coil set (1/m)

Second be
ending curvatture (1/m)) Seco
ond bending curvature (1/m)

Figure 8. effecct of striip thickness Figure

F 9. effect off strip thickness
variatioons on coil set elimination variations
v onn crossbow elimination
n

4. CON NCLUSION NS em
mployed too verify accuracy of the
In this paper
p a theooretical anaalysis follow
wed an
nalytical soolution. Thhen the sim
mulation
by its numerical
n immplementattion throughh a proogram was used to exxamine elimmination
simulatiion prograam was developedd to off crossbow in a typical tension leveller.
simulatee strip behaaviour in tennsion levelliing Thhe calculaation resullts showeed how
process. Finite element sim mulations were
w ressidual cuurvatures are inttroduced
 simultaneously to a strip due to longitudinal Japan Society for Technology of
bending and straightening and how they Plasticity, vol. 28, 1987, pp. 34-41.
vary with bending curvature variations in 6. Kajihara T., Furumoto H., Takemasa
work rolls cassette. Furthermore, T., Taniguchi N., “Study in levelling
elimination of residual curvatures by process”, Technical report in
second bending in anti-crossbow cassette Mitsubishi Heavy Industry, vol. 25,
was investigated. It was revealed that an 1988, pp. 315-322.
optimal amount of second bending exists at 7. Hira T., Abe S., Azuma, S., “Analysis
which crossbow is removed while coil set of sheet metal bending deformation
occurs in opposite direction; so the behaviour in processing lines and its
additional anti-coil set cassette subsequent effectiveness”, Kawasaki Steel
to anti-crossbow cassette is required to Engineering Report, vol. 19, 1988, pp.
eliminate this undesired curvature. Finally 54-62.
effect of tension and strip thickness on 8. Hibino F., “Fundamental nature of
crossbow elimination was studied and it tension levelling”, Journal of the Japan
was found that greater the tension and Society for Technology of Plasticity,
thicker the strip, the smaller second bending vol. 37, 1996, pp. 431-439.
curvature is required to eliminate residual 9. Yoshida F., Urabe M., “Computer
curvatures. The same procedure can be aided process for the tension levelling
followed to determine the optimum of metallic strip”, Journal of Materials
operating conditions in anti-coil set Processing Technology, 89, 1999, pp.
cassette. In general this simulation program 218-223.
can also be used to determine optimum 10. Morris J.W., Hardy S.J., Lees A.W.,
process parameters in process online Thomas J.T., ”Cyclic behaviour
control. concerning the response of material
subjected to tension levelling”,
5. ACKNOWLEDGMENTS International Journal of Fatigue, vol.
The support of Mobarakeh Steel Company 22, 1999, pp. 93-100.
is gratefully acknowledged. The authors 11. Huh H., Lee H., Park S., Kim G., Nam
also wish to thank Mr. Rasouli for his S., “The parametric process design of
contributions to this work. the tension levelling with an elasto-
plastic finite element method”, Journal
6. REFERENCES of Materials Processing Technology,
1. Roberts W.L., “Cold rolling of steel”, vol. 113, 2001, pp. 714-719.
Marcel Dekker, New York, 1978. 12. Khan A.S., Huang S., “continuum
2. Misaka Y., Masui T., “Shape correction theory of plasticity”, John Wiley &
of steel strip by tension leveller”, Trans Sons Inc., New York, 1995.
ISIJ, vol. 18, 1978, pp. 475-485. 13. Hibbit, Karlsson, Sorensen,
3. Patula E.J., “Tension-roller-levelling “ABAQUS Standard User’s Manual”,
process; elongation and power loss”, ABAQUS Inc, 2003.
Trans ASME, Journal of Engineering 14. Salimi M., Jamshidian M., Beheshti
Industry, vol. 101, 1979, pp. 269-277. A., Sadeghi Dolatabadi A., “Bending-
4. Kawaguchi, K., “A numerical study on unbending analysis of anisotropic sheet
the wrapping angle of strip in the under plane strain condition”,
tension leveller”, Journal of the Japan Esteghlal Journal of Engineering, vol.
Society for Technology of Plasticity, 26(2), 2008, pp. 187-196.
vol. 21, 1980, pp. 807-814.
5. Hattori S., Maeda K., Matsushita T.,
Murakami S., Hata J., “Strip curvature
in tension levelling”, Journal of the

View publication stats