The different effects of asymmetric rolling
The different effects of asymmetric rolling and surface friction on formation of shear texture in aluminium alloy AA5754
H.Jin*and D.J.Lloyd
Conventional rolling and asymmetric rolling(ASR)processes were applied to aluminium alloy AA5754with different roll metal frictional conditions.The rolling textures were determined by X-ray diffraction technique and the formation of shear texture was studied.It has been demonstrated that the ASR and friction have different effects on the generation of shear texture.The ASR forces the deformation texture to rotate about the transverse direction from the fcc plane strain compression texture,while a high friction generates a so called ideal fcc shear texture consisting of the{001}n110m and{111}n uvw m components.The effect of ASR penetrates throughout the sheet thickness,but that of friction exists only from the sheet surface to one-quarter thickness.
Keywords:Al–Mg alloy,Asymmetric rolling,Friction,Shear texture
Introduction
In recent years aluminium alloys have been widely used in automobile body applications,but the formability of aluminium alloy sheet,especially its deep drawability,is inferior to that of conventional steel sheet.Normally aluminium alloy sheet is supplied in the fully annealed condition,with a characteristic cube{001}n100m crystal-lographic texture component,which has a low r value, no more than1.1The r value,or the Lankford para-meter,represents the width to thickness reduction ratio in a uniaxial tensile test.On the contrary,the characte-ristic texture in a fully annealed low carbon steel sheet is the{111}n uvw m(c-?bre),with r values approaching 2?7.2,3Therefore,from the viewpoint of sheet form-ability enhancement,it is of great interest to generate the c-?bre in aluminium alloy sheet.
In conventional rolling of aluminium alloys,a so called ideal shear texture,consisting of the c-?bre and rotated cube H{001}n110m,can be generated from the sheet surface to one-quarter thickness by the shear strain due to roll sheet friction and rolling gap geometry.4,5A high friction and/or a large rolling gap geometry draught over5generates the ideal shear texture in the sheet surface,whereas a small rolling gap geometry draught, less than0?5,results in the ideal shear texture in the sheet quarter thickness layer.The rolling gap geometry draught is the ratio between roll sheet contact length L c and sheet entry thickness t0.
Asymmetric rolling(ASR),in which the circumferential velocities of the top and bottom rolls are different, generates an additional shear strain component through-out the sheet thickness.In recent years,intensive work has been carried out to apply ASR to aluminium alloys for texture modi?cation and grain re?nement.6–18The differ-ent circumferential velocities can be created by either different top and bottom roll diameters,or different top and bottom roll rotational speeds.Regarding the texture after ASR,both the ideal shear texture7,10,13,16,18and the fcc plane strain compression texture rotated about the sheet transverse direction(TD)have been reported.7,12,15,16 The fcc plane strain compression texture,or the b-?bre, consists of Bs{011}n211m,S{123}n634m and Cu {112}n111m components.While there are many observa-tions of the ideal shear texture after ASR,Jin and Lloyd15 found that in an Al–Mg alloy,unidirectional ASR makes the deformation texture deviate from the b-?bre by a10u rotation about TD,whereas reverse ASR with equivalent forward and backward strains will cancel the rotation and lead to the normal b-?bre.Moreover,they also observed that in Al–Si–Mg alloy AA6111the ASR process does not create the ideal shear texture without suf?ciently high friction,even when the velocity ratio is as high as4.12
To clarify the different effects of friction and ASR on the formation of shear texture,the present work on an Al–Mg alloy AA5754examined sheet processed by conventional rolling and ASR under three different lubrication condi-tions.The deformation textures after cold rolling were determined by the X-ray diffraction pole?gure technique. The relation between the deformation texture and roll metal friction during rolling process will be discussed. Experimental
Starting material and rolling process
The starting material was a commercially supplied DC cast2?5mm gauge O temper(fully recrystallised condition)AA5754sheet.The AA5754has a nominal composition of Al–3Mg–0?3Mn–0?2Fe(wt-%)and is used extensively in automotive inner panel applications.
Novelis Global Technology Centre,PO Box8400,Kingston,Ont.K7L5L9, Canada
*Corresponding author,email haiou.jin@https://www.wendangku.net/doc/1b8492971.html,
754?2010Institute of Materials,Minerals and Mining
Published by Maney on behalf of the Institute
Received21November2008;accepted22November2008
DOI10.1179/174328409X405634Materials Science and Technology2010VOL26NO6
A speci?c rolling mill,where the top and bottom rolls have the same diameter of161?5mm but each is driven by its own motor,was used in the present work.It is essentially a conventional rolling mill when the rolls rotate at the same rate,but performs ASR when the rolls rotate at different speeds.A total of nine sheets were rolled at room temperature to1mm?nal gauge(60% thickness reduction)with two different velocity ratios(1 and1?5)and three different frictional conditions(low, medium and high frictions).The sheet designation and the details of the rolling operation are listed in Table1. All the rolling process was unidirectional,and the rotation rate of the bottom roll was kept constant at 25rev min21,while that of the top roll was varied to achieve the desired velocity ratio.Nevertheless,during asymmetric rolling the velocity ratio was lower than the expected value,by around10–20%,due to the fact that the fast rotating roll experienced higher resistant force than the slow one.The sheet temperature after each rolling pass was measured by a thermal couple.The temperature in high friction rolling was slightly higher than the others,and the highest value recorded was65u C. Rolling gap geometry and roll sheet friction
To minimise the effect of rolling gap geometry,the rolling was carried out in two passes,?rst from2?5to 1?6mm,then to1mm.The rolling gap geometry draught is calculated by
L c t0~
R0
t0
cos{11{
D t
2R
(1)
where D t is the reduction in thickness,and R9is the ?attened roll radius given by Hitchcock’s equation19
R0 R ~1z
16(1{n2)F=W
eT
p E D t
(2)
where R is the roll radius,n is the Poisson’s ratio of the rolled material,aluminium(1:3),E is the Young’s modulus of the roll material,steel(261011N m22),F is the separating force and W is the sheet width.The separation force F is roughly the yield stress s YS of the aluminium sheet,which is102?8MPa for the starting material as measured by standard tensile testing in the rolling direction,multiplied by the contact area
F~s YS L c W(3) Combining equations(2)and(3)gives
R0~R1z 16(1{u2)(F=W)
p E D t
~R1z 16(1{u2)s YS L c
p E D t
~R1z 16(1{u2)s YS R0cos{11{(D t=2R0)
?
p E D t
(4)The?attened roll radius R9can be obtained by solving equation(4)numerically.
The low and medium frictional conditions were generated using a lubricant of Norpar15z9%Epal 1416z2%methyl laurate solution and a lubricant of Norpar15respectively,and the high frictional condition was ful?lled by dry rolling.The lubricants were sprayed thoroughly on both the rolls and the sheet top and bottom sides,whereas the dry rolling condition was made by cleaning the rolls with hexane.Generally,in commercial rolling mills,dry rolling gives the upper limit of friction coef?cient,the use of Norpar15z9%Epal 1416z2%methyl laurate solution gives the lower limit. The friction coef?cients under the designed conditions were estimated by measuring the forward slip S f.The forward slip is de?ned in terms of exit velocity of the strip and the surface speed of the roll.It is usually monitored by marking the roll surface with two parallel lines a known distance L0apart,and measuring the distance L between the impressions of those lines on the rolled strip
S f~
L{L0
(5) The average friction coef?cient m can be calculated by Ford’s equation20–23
m~
D t
2(R0D t){4(S f R0t0)
(6)
In the present work,a mark was made in the top roll surface and several very long starting material strips were rolled at room temperature down to1?9mm in one pass,under the appropriate frictional conditions. Rolling with other thickness reductions was tried,but all led to the sheets tending to coil around the rolls.As the mark was imprinted in the strips twice,L is the distance between the marks and L0is the roll circum-ference(507?5mm).Using the measured L and calcu-lated R9from equation(4),the friction coef?cient was estimated by equation(6).
Sample preparation and analyses
The crystallographic textures were measured using the X-ray diffraction pole?gure technique in the sheet top surface,one-quarter thickness from the top surface, centre,three-quarter thickness from the top surface,and bottom surface,in a Rigaku X-ray machine with a radiation source of RU-200B rotating anode and Cr target.The specimens were mechanically polished down to3m m diamond,followed by electropolishing in a solution comprising2%butylcellosolve,8%HClO4,30% alcohol and60%water with current density of 1?5A cm22for30s at210u C.The orientation distribu-tion functions were calculated using van Houtte’s MTM-FHM software.24No symmetry was forced dur-ing the calculation,so that low symmetric components
Table1Sheet designation and details of rolling conditions
Designation Velocity ratio Lubricant Friction coefficient
H11None010
H21.5None0.10
M11Norpar150.07
M21.5Norpar150.07
L11Norpar15z9%Epal1416z2%methyl laurate0.04
L21.5Norpar15z9%Epal1416z2%methyl laurate0.04
Jin and Lloyd Different effects of ASR and surface friction in AA5754
Materials Science and Technology2010VOL26NO6755
were divided into two or four variants,e.g.Bs 1,Bs 2,Cu 1,Cu 2,S 1,S 2,S 3and S 4,and an 11u Gaussian spread was used when calculating the texture component volume fraction.
Results
The rolling gap geometry draughts were 2?8for the reduction from 2?5to 1?9mm,3?4for the reduction from 2?5to 1?6mm,and 4?4for the reduction from 1?6to 1mm.The friction coef?cient is calculated to be 0?1for dry rolling,0?07with Norpar 15,and 0?04with Norpar 15z 9%Epal 1416z 2%methyl laurate solution.Some related important parameters are listed in Table 2.The mean grain size in the starting material was of 22m m in length and 14m m in thickness near the sheet
surface,and 27m m in length and 18m m in thickness in the centre,as measured in the longitudinal section.The through thickness (111)pole ?gures are shown in Fig.1.The texture of the starting material consists of weak cube and b -?bre texture components,which is typical in a fully annealed Al–Mg alloy sheet where recrystallisa-tion is due to particle stimulated nucleation.Table 3lists the volume fractions of pertinent texture components,including the cube and b -?bre,as well as the rotate cube H {001}n 110m and E {111}n 110m ,which are the major components of the ideal fcc shear texture.
Figures 2–4show the through-thickness (111)pole ?gures of the cold rolled sheets,with the volume fractions of pertinent texture components listed in Tables 4–9.Under the low frictional condition,the conventionally rolled sheet has the typical b -?bre texture
Table 2Some important parameters related to calculation of friction coef?cients
Dry rolling
Norpar 15Norpar 15z 9%Epal 1416z 2%methyl laurate D t 0.58mm 0.63mm 0.63mm L 521.1mm 513.5mm 508.1mm S f 0.0270.010.001R 983.0mm 83.0mm 82.9mm m
0.10
0.07
0.04
1(111)X-ray pole ?gures of 2?5mm thick AA5754starting material:contour levels are 1?0,1?5,2?0,2?5,3?0,3?5,4?0,5?0
and 7?0
Table 3Volume fractions of cube,b -?bre,H and E texture components in starting material
Surface
1/4thickness Centre 3/4thickness Backside surface Cube 3.263.784.744.963.76Bs 4.824.483.964.224.69S 8.358.778.028.549.06Cu 4.154.013.273.673.96H 0.671.020.580.700.72E
0.94
0.95
1.14
1.01
1.17
a)
b)
a conventionally rolled;
b asymmetrically rolled
2(111)X-ray pole ?gures of as rolled low frictionally rolled sheets:contour levels are 1?0,1?5,2?0,2?5,3?0,3?5,4?0,5?0
and 7?0
Jin and Lloyd Different effects of ASR and surface friction in AA5754
756Materials Science and Technology 2010
VOL
26
NO
6
Jin and Lloyd Different effects of ASR and surface friction in AA5754 a)
b)
a conventionally rolled;
b asymmetrically rolled
3(111)X-ray pole?gures of as rolled medium frictionally rolled sheets:contour levels are1?0,1?5,2?0,2?5,3?0,3?5,4?0, 5?0and7?0
a)
b)
a conventionally rolled;
b asymmetrically rolled
4(111)X-ray pole?gures of as rolled high frictionally rolled sheets:contour levels are1?0,1?5,2?0,2?5,3?0,3?5,4?0,5?0 and7?0
Table4Volume fractions of cube,b-?bre,H and E texture components in conventionally rolled sheet under low frictional condition
Surface1/4thickness Centre3/4thickness Backside surface
Cube187193196220212
Bs6.115.996.606.006.13
S15.1814.8415.0114.1614.37
Cu8.037.957.477.527.34
H0.290.330.260.350.33
E0.820.750.760.900.91
Table5Volume fractions of cube,rotated b-?bre,H and E texture components in asymmetrically rolled sheet under low frictional condition
Surface1/4thickness Centre3/4thickness Backside surface
Cube1.471.791.451.971.86
Bs6.275.356.596.938.37
S12.4311.7915.2615.3517.23
Cu5.975.847.997.838.06
H0.920.920.330.520.31
E1.301.190.830.880.75
Materials Science and Technology2010VOL26NO6757
and nearly no through thickness texture variation (Fig.2a and Table 4).The asymmetrically rolled one has a similar texture (Fig.2b ),but the b -?bre has been rotated about the TD by about 4–10u (Fig.5).There is limited through thickness texture variation (Table 5),but the rotation angle appears higher in the sheet top half thickness (Fig.5).
Under the medium frictional condition,the conven-tionally rolled sheet has the b -?bre texture,but the texture in the sheet surface is much weaker than that in the centre (Fig.3a and Table 6).Moreover,in the sheet surface,the b -?bre tends to rotate about the TD by y 5u (Fig.6).The asymmetrically rolled sheet has a very different texture (Fig.3b ).The sheet surface has an ideal shear texture,the sheet quarter thickness has a weak b -?bre rotated about the TD by 8u ,and the texture in the sheet centre consists of both the ideal shear texture and the rotated b -?bre,but very weak.
After dry rolling where the friction coef?cient is the highest,both the conventionally rolled and ASR processed sheets have an identical ideal shear texture
Table 6Volume fractions of cube,b -?bre,H and E texture components in conventionally rolled sheet under medium
frictional condition
Surface
1/4thickness Centre 3/4thickness Backside surface Cube 1.832.612.072.161.78Bs 5.026.076.436.434.85S 11.3614.3115.5115.2010.51Cu 5.37.377.627.624.93H 1.130.390.300.341.46E
1.42
0.67
0.81
0.80
1.44
Table 7Volume fractions of cube,rotated b -?bre,H and E texture components in asymmetrically rolled sheet under
medium frictional condition
Surface
1/4thickness Centre 3/4thickness Backside surface Cube 099132157156087Bs 1.763.681.513.901.55S 4.269.465.949.973.97Cu 1.994.972.835.141.92H 4.001.703.051.514.12E
3.78
1.69
3.13
1.71
3.65
Table 8Volume fractions of cube,b -?bre,H and E texture components in conventionally rolled sheet under high
frictional condition
Surface
1/4thickness Centre 3/4thickness Backside surface Cube 0.561.941.810.610.41Bs 1.522.475.551.181.69S 3.997.0212.643.344.25Cu 1.753.535.981.371.78H 4.202.390.364.374.68E
4.25
2.48
0.83
3.96
4.29
Table 9Volume fractions of cube,rotated b -?bre,H and E texture components in asymmetrically rolled sheet under high
frictional condition
Surface
1/4thickness Centre 3/4thickness Backside surface Cube 0.661.431.682.560.61Bs 1.501.343.991.101.52S 3.743.289.863.923.89Cu 1.551.425.132.001.71H 3.844.111.373.524.62E
4.31
4.12
1.60
3.44
4.13
5Rotation angle as function of sheet thickness in asym-metrically rolled sheet under low frictional condition
Jin and Lloyd Different effects of ASR and surface friction in AA5754
758Materials Science and Technology 2010
VOL
26
NO
6
in the surface(Fig.4).Through to the quarter thickness, the ideal shear texture persists in the asymmetrically rolled sheet,but it tends to evolve to a very weak rotated b-?bre in the conventionally rolled sheet.In the sheet centre,the conventionally rolled sheet has the b-?bre, the same as that in the centre of the low and medium frictionally rolled sheets.The asymmetrically rolled sheet has a weak b-?bre in the centre as well,rotated about the TD by8u.
Discussion
In the present work,there are four factors that may affect the development of rolling texture:the starting texture,the rolling gap geometry,the friction and the symmetry of rolling.Based on Fig.1,it is reasonable to assume that the starting material has no through thick-ness texture variation.Since the rolling gap geometry draught was controlled(Table2),it is considered to have a negligible effect.Therefore,only the friction and symmetry of rolling need to be considered.Although the measurement of friction coef?cients gives only an estimation of the prevailing level of friction,the experiments demonstrate the importance of friction in the development of deformation in the sheet.
Under the lowest frictional condition,the plasticity is essentially uniform through the thickness of the sheet, and the normal b-?bre texture is developed through the sheet thickness.Under asymmetric rolling,the b-?bre is rotated by up to10u around TD,but a signi?cant shear texture component is not present.However,the texture evolution is not uniform through the sheet,the rotation angle in the top half is higher than that in the bottom half(Fig.5).The rotation angle is de?ned as the angle rotated about the TD with respect to the Cu{112}n111m orientation.This may be attributed to the fact that the sheet top surface,which contacts the faster rotating roll, experienced higher friction.Non-uniform shearing in asymmetrically rolled sheet is not uncommon.6
Under the highest frictional condition,high shear texture components are developed in the surface regions under both symmetric and asymmetric rolling,but the texture is non-uniform through the sheet thickness.Under symmetric rolling,the texture reverts to the conventional b-?bre at the sheet centre.This is the expected response,where under the extreme condition of sticking friction,the?ow stress of the sheet at the surface is equal to the shear strength of the alloy in plane strain and produces high shear strain at the surface,and hence a shear texture.These conditions are relaxed with increasing depths towards the centre thickness of the sheet.Asymmetric rolling increases the depth to which the shear deformation,and hence the shear texture, persists.When the volume fractions of the b-?bre and the shear texture components are compared(Figs.7and 8),it is clear that the effect of the asymmetrical rolling under high friction is to reduce the b-?bre and raise the shear texture,in the region from the one-quarter thickness to sheet centre.It is also obvious that the effect of friction is much stronger than that of asymmetric rolling.
The medium frictional condition re?ects a transitional case between the two extremes.It is unclear why the
6Rotation angle as function of sheet thickness in conventionally rolled sheet under medium frictional condition 7Volume fractions of b-?bre and sum of H and E compo-nents as function of sheet thickness in conventionally
rolled sheet under high frictional condition
8Volume fractions of b-?bre and sum of H and E compo-nents as function of sheet thickness in asymmetrically
rolled sheet under high frictional condition
Jin and Lloyd Different effects of ASR and surface friction in AA5754 Materials Science and Technology2010VOL26NO6
759
b -?bre drops in the centre of the asymmetrically rolled sheet (Fig.9).The presence of an ideal shear texture at the surface of the asymmetrically rolled sheet demon-strates the ability of differential rolling speeds to introduce shear components.However,since there is an absence of shear texture under very low friction in asymmetrically rolled sheet,it is apparent that frictional condition is important in generating shear textures,even in asymmetrically rolled sheet under the present experi-mental conditions.
Conclusions
The development of rolling texture is strongly affected by friction and the symmetry of rolling in AA5754.Although both asymmetric rolling and high friction generate shear textures,their effects are different.Unidirectional asymmetric rolling produces a rolling texture similar to the b -?bre,but rotated about the TD,and this effect applies throughout the sheet thickness.Without a suf?ciently high friction between the rolls and metal,asymmetric rolling is unable to create an ideal shear texture.An increasing roll metal friction may weaken the rolling texture and force it to evolve to the ideal shear texture,but this effect is restrained within the range from the sheet surface to one-quarter thickness,
while the sheet centre retains the b -?bre regardless.When a high friction is combined with asymmetric rolling,the ideal shear texture is obtained from the sheet surface to one-quarter thickness,while the sheet centre has a weakened b -?bre rotated about the TD.
The present experiments demonstrate that investiga-tions of asymmetric rolling need to control the rolling friction,and the through thickness texture variation must be monitored and accounted for in any modelling of asymmetric rolling effects.
Acknowledgement
The authors are grateful to Novelis Inc.for permission to publish the present work.
References
1.W.F.Hosford and R.M.Caddell:‘Metal forming:mechanics and metallurgy’,274;1993,Englewood Cliffs,NJ,Prentice Hall.
2.W.B.Hutchinson:Int.Met.Rev.,1984,29,25–42.
3.R.K.Ray,J.J.Jonas and R.E.Hook:Int.Mater.Rev.,1994,39,129–172.
4.W.Truszkowski,J.Krol and B.Major:Metall.Trans.A ,1980,11A ,749–758.
5.W.Truszkowski,J.Krol and B.Major:Metall.Trans.A ,1982,13A ,665–669.
6.Q.Cui and K.Ohori:Mater.Sci.Technol.,2000,16,1095–1101.
7.K.H.Kim and D.Y.Lee:Acta Mater.,2001,49,2583–2595.
8.H.Jin and D.J.Lloyd:Metall.Mater.Trans.A ,2004,35A ,997–1006.
9.H.Jin and D.J.Lloyd:Scr.Mater.,2004,50,1319–1323.
10.M.G.de Lomana,A.R.Imaz,J.T.Ariztondo,C.Zubillaga,
C.Garcia-Rosales and J.G.Sevillano:Bol.Soc.Esp.Ceram.Vidrio ,2004,43,145–148.
11.H.Jin and D.J.Lloyd:Mater.Sci.Forum ,2004,467–470,381–386.12.H.Jin and D.J.Lloyd:Mater.Sci.Eng.A ,2005,A399,358–367.13.S.Kang,B.Min,H.Kim,D.S.Wilkinson and J.Kang:Metall.
Mater.Trans.A ,2005,36A ,3141–3149.
14.H.Jin and D.J.Lloyd:Mater.Sci.Forum ,2006,519–521,161–168.15.H.Jin and D.J.Lloyd:Mater.Sci.Eng.A ,2007,A465,267–273.16.J.-K.Lee and D.N.Lee:Int.J.Mech.Sci.,2008,50,869–887.17.J.Sidor,A.Miroux,R.Oetrov and L.Kestens:Acta Mater.,2008,
56,2495–2507.
18.J.Sidor,A.Miroux,R.Oetrov and L.Kestens:Philos.Mag.,2008,
88,3779–3792.
19.J.Hitchcock:‘Roll neck bearings’,ASME Research Publication,
Appendix 1;1935,New York,ASME.
20.H.Ford,F.Ellis and D.R.Bland:J.Iron Steel Inst.,1951,5,57.21.K.Kleinjohan:Z.Metallkd ,1965,10,685.22.I.C.Benedyk:Light Met.Age ,1971,2,6.23.S.F.Magyary:Aluminium ,1974,4,291.
24.P.van Houtte:‘The MTM-FHM software system’,Version 2;
2000,Leuven,Katholieke Universiteit.
9Volume fractions of b -?bre and sum of H and E compo-nents as function of sheet thickness in asymmetrically rolled sheet under medium frictional condition
Jin and Lloyd Different effects of ASR and surface friction in AA5754
760Materials Science and Technology 2010
VOL
26
NO
6