Int. J. Mach. Tools Manufact. Vol. 32, No. 1/2, pp. 183-193. 1992. 0890-6955/9255.00 + .

00
Printed in Great Britain Pergamon Press plc

Electrical Discharge Texturing

D.K. Aspin~all*, M.L.H. Wise*, K.J. Stout*, T.H.A. Goh*, F.L. Zhao** and M.F. EI-Menshawy***
*University of Bimfingham, UK; **Dalian Institute of Technology, PR China; ***Spark Tec Ltd., Birmingham, UK.

SYNOPSIS

The paper outlines a low cost electrical discharge machining arrangement for the texturing of steel rolls used in the cold rolling of
sheet steel and aluminium. The particular electrical discharge texturing (EDT) technique used affords a high degree of process
control compared with more conventional methods and, therefore, optimisation of roll texture is possible through the selection of
the appropriate operating variables.
Sample EDT operating dala are presented covering the effects of peak current and on-time relative to roll surface roughness (Ra)
and peak count (Pc). Scaaning electron micrographs are included of textured roll and sheet steel surfaces together with three
dimensional topographical representations and a simulation of the effects of roll wear and topography transfer. Roll surface
integrity data are also given which detail the structure and hardness of the white layer produced as a consequence of the EDT
process.

1 INTRODUCTION

Steel sheet for use in applications where formability and hardness of the shot. A further limitation to the use of SB is
appearance are important, e.g. automobile bodies, is usually the hardness of the roll since it is desirable to use rolls of high
produced with a controll~:l surface finish, or "texture", such hardness in order to reduce the amount of wear incurred
that it has a uniform matt ;appearance. The texture is applied during rolling. This militates against the use of SB because
to the sheet by the use ¢,f textured rolls during both cold the shot must be harder than the roils to cause effective plastic
rolling and temper rolling, Fig. 1. This is done for a number deformation of the roll surface and there is a limit to the
of reasons: after cold roiling it helps to prevent the rolled hardness that can be readily developed in the shot. There are
surfaces bonding together during batch annealing of the cold also environmental problems of dust and noise when using
rolled coils, which would otherwise result in the formation of SB and the roll journals need to be protected from abrasion.
"sticker-wrench marks"; it also provides a reservoir for
lubricating oil during forming, which affects both formability In LT 3,4,5 the texture is generated by locally melting the
and die wear and masks small imperfections in the sheet roll surface with a beam from, for example, a CO2 laser. In
surface; finally, it provides good keying for the decorative the CRM system a 3 kW unfocussed beam is chopped into
and protective paint finish t,2. In this respect it is important 40 kHz pulses by means of a rapidly rotating perforated disc
that the texture be random; i.e. without a particular lay, then focussed to a diameter of 0. I mm giving a power density
otherwise body panels of nominally the same colour would of 1011 W mm -2 which can be further assisted by use of an
differ according to their orientation and reflectivity. oxygen jet. The roll is rotated and the beam tracked axially at
a feed equal to the beam diameter to produce a helical pattern
I of craters, the rotational speed controlling the surface
roughness by altering the space between the craters. Whereas
LT is not influenced by the hardness of the roll, the regularity
of the texture produced by helical scanning can produce a
directionality in the rolled sheet with the problems described
[ above 6. The system developed by CRM can texture a roll
500 mm diameter by 1 m long roll within 2 hours and 700 t
of steel sheet can be rolled before retexturing is required.

EDT 7,8,9 is a version of electrical discharge machining in

cold roll coil anneal temper roll coil which the roll, as one electrode, is separated from a tool
10/20% <1% electrode by a dielectric medium, typically a hydrocarbon oil,
deformation deformation such as paraffin. The formation of minute surface craters on
per pass the roll results from the discharge of energy between the two
electrodes operating via a gap control servo-mechanism to a
pulse generator. As with LT, roll hardness is not a limiting
Fig. 1 Schematic representation of cold rol!ing. feature, but EDT offers the possibility of independently
controlling the height and number of surface peaks by
Traditionally, texture has been applied to the rolls by shot appropriate selection of the operating variables such as
blasting (SB), but this has been partially superseded by current and pulse duration 10. It is, therefore, possible to
electrical discharge texturing (EDT) and laser texturing (LT). produce a textured surface with a much higher density of
In SB hard shot is impacted on to the surface of the roll at peaks for a given roughness value. Electrical discharge ma-
high velocity to cause impact damage and roughening of the chining is a thermal process involving localised temperatures
surface. The problem with SB is that it is difficult to control estimated to be as high as 20,000 ° C 11. This means that the
the texture accurately due to the large number of variables microstructure of the surface layer of the rolls can be altered
inherent in the process, such as the size, shape, velocity and during EDT to give more desirable wear characteristics t/,

183
 184 D.K. ASPINWALt.et al.

however, as with all electrical discharge machining, the 2 EXPERIMENTAL WORK

process variables must be controlled to minimise surface
damage. Three sets of work rolls, 38 mm diameter by 196 mm long,
were manufactured from AISI M2 high-speed steel, fully
Several EDT tool electrode arrangements are possible: the hardened to 63 HRC and ground to final size. Texturing
simplest employs either a segmented strip or roll electrode was undertaken using a single DC servo quill arrangement on
spanning the length of the roll; alternatively, several simple an EMAG variable speed lathe with a Spark Tec generator as
finger electrodes each operated by a servo-mechanism, or a described previously 1°. The electrode materials used were
matrix of electrodes shaped to the curvature of the roll and Graphite Technologies' high density graphite (EC17) and
connected to a common servo-mechanism can be used. In copper graphite (EC15); the dielectric, BP 180, was
this case, each finger electrode is insulated from its neighbour circulated through a Filtermist Superfine dielectric filtration
and connected to a multi-channel pulse generator 13,14. unit capable of filtering to 1 /am. Surface roughness was
Fig. 2 shows two multi-finger servo heads mounted on the measured on site using a portable Mitutoyo Surftest 201
cross slide of a CNC lathe. Texturing is achieved by (evaluation length 4 ram, peak count level No. 2 setting -
sparking while rotating the work roll and simultaneously 0.5 I-tm) and, subsequently, by a Laser Form Talysurf; 3-D
scanning the electrode and servo assembly in the axial topography was determined on a modified Rank Taylor
direction. There are several manufacturers of EDT equipment Hobson Talysurf 6. Surface roughness measurements were
in Europe: e.g. Spark Tec's range varies from £150,000, for made on both the rolls and strip along and parallel to the
a 12 channel retro-fit system capable of processing rolls up to rolling direction. All data reported are the average of 5
measurements.

A specimen roll was textured in bands using the

parameters shown in Table 1 to determine the effects of
sparking conditions on both surface roughness and integrity.
Samples of the bands textured under conditions A to D were
cut from the roll using EDM wire erosion. Samples were
then mounted for metallographic investigation in conventional
bakelite moulds using either radial sections, in which the
surface layer could be viewed normally, or tangential sections
to magnify the surface layer and adjacent heat affected zone.
Specimens were polished to a ll.tm diamond finish and etched
in 2% nital. Microhardness measurements were made across
the samples using a load of 50 g. In order to ensure that
measurements made near the edge of the EDT surface were
representative of the hardness of the microstructure, and not
just an artefact of their proximity to an edge, microhardness
measurements were made across a section of uniform
microstructure from which it was found that there was no
change in hardness for measurements made within about
5 I.tm of the edge.
Rolling was undertaken on a Stanat Mann 4-high mill
equipped with 38 mm diameter work rolls driven by 130 mm
diameter back-up rolls; a light oil lubricant was used.

The workpiece material used in the initial trials was

1.2 mm thick, BS 1449 46/40 type steel sheet. Although
this was harder the ~100 HV10 BS 1449 CR/S1 material
typically used for car body work, it provided a useful
benchmark regarding the degree of texture transfer in
comparison with the softer sheet materials. Samples were cut
Fig. 2 Spark Tec EDT servo-heads mounted on into 100 mm square coupons, degreased and lightly pickled
CNC lathe. in concentrated hydrochloric acid to remove any superficial
650 mm in diameter, to £1M for a complete 72 channel corrosion. As this material was already textured, the texture
system employing a CNC lathe. was removed by first rolling with plain ground work rolls
using a light reduction of <2%; the samples were rolled four
The aims of the work were: to evaluate the surface times rotating 90 ° between passes then annealed in a vacuum
integrity of AISI M2 high-speed steel rolls on which fine and furnace at 650 °C for 2 hours to give a hardness of
coarse textures were produced by EDT and to examine the 150 HV10. The change in surface finish as a result of this
degree of transfer of these textures to steel sheet by different pretreatment is shown in Table 2 together with data from
amounts of cold reduction in a four-high rolling mill. untextured, aluminium alloy sheet (63 HVI0) used to

Table 1 Experimental Parameters

specimen current on time off time electrode polarity gain Ra* Pc*
A Its ~ts ~tm /cm
A 1 20 20 EC17 -ve 4 1.50 133
B 3 40 20 EC17 -ve 4 3.00 100
C 3 40 20 EC15 +ve 4 3.21 92
D 8 100 20 EC17 -ve 4 9.71 48
roll 1 1 10 20 EC17 -ve 4 1.24 166
roll 2 3 20 20 EC17 -ve 4 2.68 100
roll 3 2 20 20 EC17 -ve 4 1.91 118
roll 4 8 40 20 EC17 -ve 4 5.23 74
*Mitutoyo data
 Electrical Discharge Texturing 185

Table 2 Showing the effect of pre-treatment

on the surface finish of the steel sheet
and the initial condition of the
aluminium alloy sheet

Along rolling direction Ra#~.m

plain rolls 0.25
as received sheet 1.13
plain rolled sheet 0.24
as received AI sheet 0.19
Normal to rolling direction Ra#~m
plain roils 0.23
as received sheet 1.27
plain rolled sheet 0.28
0.38 Fig. 3 Cross section through roll 'A'; Ra 1.5 gm.
as received A1 sheel
#Form Talysurf data

compare the texture transfer efficiency with that of the

150 HVI0 steel sheet. The Ra values for the aluminium alloy
sheet should be treated with caution due to possible surface
damage caused by stylus contact.

In each case, three rotational speeds, 15, 20 and 45 rpm,

were used with a strain per pass of 6% for the aluminium
sheet and 6% and 10% far the steel specimens to determine
the influence of strain on transfer efficiency. Surface finish
measurements of the roll s and sheets were made before and
after rolling.

In order to investigate the transfer of the texture from roll

to sheet 4 mm and 2 mm squares were scribed onto the roll
surface. After rolling, the scribed areas on the rolls and Fig. 4 Cross section through roll 'B'; Ra 3,0 gm.
corresponding regions on the sheet were mapped using the
3-D topography system described above. The raw data from
the rolls were processed in order to remove the roll curvature
and so reveal the actual texture. The corresponding scribed
areas on the roll and sheet were located and the map of the
steel sheet texture was tken inverted to investigate the match
between the peaks and valleys of the two surfaces and hence
the degree of texture transfer.

3 RESULTS

Surface Integrity
Figs. 3-4 show cross '.;ections and Figs. 5-6 the surface
finish of rolls textured under conditions A and B to give Ra
values of 1.5 gm and 3.0 gin, respectively. In both cases,
the immediate sub-surface microstructure consisted of a
"white layer" extending for about 8 gm in from the surface
which was relatively featureless suggesting that it had been at
a sufficiently high temperature to melt the surface and Fig. 5 Surface of roll 'A'; Ra 1.5 gm.
dissolve the M6C and/VlC matrix carbides. Below this layer
can be seen austenite grain boundaries and undissolved
primary carbides which decrease in size towards the surface
(as a result of being taken into solution). Part of this light
etching region had retransformed to austenite, since its grain
size was smaller than that of the matrix, and in the rest the
fine secondary carbides, precipitated during the original
tempering, had redissolved leaving the prior austenite grain
boundaries clearly delineated. After this came a sharp
transition to the dark, tempered martensite matrix typical of a
fully hardened high-speed steel.

Figs. 7-8 show surface layers from specimens C and D

(Table 1) magnified as a result of the tangential section. The
irregular shape of the surface and the absence of eutectic
carbides suggest extremely rapid cooling from the melt.
Cracking can be seen in one of the largest white regions as a
result of thermal contraction strains.

The microhardness af the as-produced surface layers was Fig. 6 Surface of roll 'B'; Ra 3.0 I.tm.
measured before and after quenching in liquid nitrogen to
encourage any austenite present to transform to martensite.
186 D.K. ASPINWALLet al.

Roll 'C'
lOOO

9oo-
~ .
> 8oo-

m 700
c
2
600 B ~ ~ ~ e r EDT
[] Hardness after 77 K treatment
Fig. 7 Tangential section through roll 'C'; 5OO , i • i • i • i • i • i • i , i • i •

Ra 3.2 l.tm. 20 40 60 60 100 120 140 160 180 200

Distance from surface p.m
Fig. 9c Hardness profile for roll 'C'.

Roll 'D'
1000

900
o
BOO
x

700-
c
Fig. 8 Tangential section through roll 'D'; 2
:~ 600"
Ra 9.7 am.
[] Hardness after 77 K treatment
Roll 'A'
500 ' i • i , i • i • i • i , i , i • i •
1000 20 40 60 80 100 120 140 160 180 200
Distance from surface tim
-~ 900
0
u~
Fig. 9d Hardness profile for roll 'D'.
800
>
"r Fig. 9 shows that the surface layers were sorter than me
matrix, but increased in hardness following treatment at
700
Q 77 K, suggesting that they comprised a mixture of martensite
C
and retained austenite.
2
:E 600- [] Hardness after EDT
[] Hardness after 77 Ktreatment
5OO = i • i • i • i • i • i • i = i • i •
Texture Transfer
20 40 60 80 100 120 140 160 180 200 Tables 3-5 summarise the surface finish data from rolls 1-4
Distance from surface pm and the corresponding data from the aluminium and steel
sheets rolled by them. Although rolling was undertaken at
Fig. 9a Hardness profile for roll 'A'. three rotational speeds there was no significant difference
between the surface finish data obtained so these have been
averaged for each rolling reduction• There was very little
Roll 'B' change in the surface finish values of the rolls before and
I000 after use because of the small amount of material rolled.

The transfer of the roll texture to the aluminium sheet,

9O0 Table 3, seems to have been very effective with roll I,
O although the peak count was lower than that of the roll; with
800- roll 2 the reverse situation occurred; however, these data
should be treated with caution due to the problems of using a
stylus measuring method on soft materials, as described
700 ' above. Table 5 shows that at 6% reduction, the transfer of
texture to the steel sheet was less effective than for the softer
E aluminium, but improved on increasing the reduction to 10%.
.',r' 600 - ~ e r EDT
[] Hardness after 77 K treatment Fig. 10 shows the surface finish of steel sheet
500 • i = i • i • i • i • i • i • i , i • (Ra 1.61 lam) produced by a 10% deformation with roll 3
0 20 40 60 80 100 120 140 160 160 200 (Ra 1.96 gtm) as seen in the SEM while Figs. 11-18 show a
Distance from surface ~m sequence of sample information obtained from 3-D
topographical representations.
Fig. 9b Hardness profile for roll 'B'.
 Electrical Discharge Texturing 187

Table 3 Roll texture data

a) Before roiling

ong rolling Ra # Rsk # Rku # HSC # Pc*

rection peaks/length /cm
roll 1 1.11 -0.25 2.40 79
roll 2 2,59 -0.27 2.45 56
roll 3 1,95 -0.20 3.10 59
roll 4 5.11 -0.10 2.60 26
~rmal to Ra # Rsk # Rku # HSC#peaks/I P c *
iling direction ength /cm
roll 1 1.26 -0.30 2.75 83 166
roll 2 2.71 -0.25 2.30 46 100
roll 3 2.00 -0.10 2.90 56 118
roll 4 5.32 -0.15 3.00 29 74
#Form Talysurf data *Mitutoyo data
b) After rolling

long rolling Ra n Rsk # Rku n HSC # Pc*

irection peaks/length /cm
roll 1 1.18 -0.35 2.70 80
roll 2 2.62 -0.20 2.60 51
roll 3 1.80 -0.40 2.40 52
roll 4 4.98 -0.20 2.60 26
ormal to Ra # Rsk n Rku # HSC n Pc*
filing direction peaks/length /cm
roll 1 1.19 -0.20 2.70 80 166
roll 2 2.69 -0.25 2.45 45 103
roll 3 1.97 -0.50 2.40 52 116
roll 4 5.23 -0.10 2.55 29 71
#Form Talysurf data *Mitutoyo data

Table 4 Aluminium alloy sheet texture data

Along rolling Ra # Rsk # Rku # HSC # Pc*

direction peaks/length /cm
roll 1 (before) 1.11 -0.25 2.40 79
A1 sheet 6% 1.12 0.48 3.30 67 132
roll 1 (after) 1.18 -0.35 2.70 80
Normal to Ra # Rsk # Rku # HSC # Pc*
rolling peaks/length /cm
direction
roll 1 (before) 1.26 -0.30 2.75 83 166
A1 sheet 6% 1.17 0.44 3.27 67 136
roll 1 (after) 1.19 -0.20 2.70 80 166
Along rolling Ra n Rsk # Rku # HSC n Pc*
direction peaks/length /era
roll 2 (before) 2.59 -0.27 2.45 56
A1 sheet 6% 2.23 0.10 2.57 49
roll 2 (after) 2.62 -0,20 2.60 52
Normal to Ra n Rsk n Rku n HSC n Pc*
rolling peaks/length /cm
direction
roll 2 (before) 2.71 -0.25 2.30 46 100
A1 sheet 6% 2.27 0.18 2.63 46 108
roll 2 (after) 2.69 -0.25 2.45 45 103
~ForrnTalysurfdata *Mitu~yo data
 188 D.K. ASPINWALLet al.

Table 5 Steel sheet texture data

Along rolling Ra # Rsk n Rku n HSC # Pc*

direction peaks/length /cm
roll 3 (before) 1.96 -0.20 3.10 59
steel sheet 6% 1.47 0.05 2.70 52 110
steel sheet 10% 1.61 0.33 2.27 53 103
roll 3 (after) 1.88 -0.40 2.40 52
Normal to Ra # Rsk n Rku n HSC n Pc*
rolling direction peaks/length /cm
roll3 (before) 2.00 -0.10 2.90 56 118
steel sheet 6% 1.51 0.10 2.63 55 112
steel sheet 10% 1.63 0.37 2.87 54 105
roll 3 (after) 1.97 -0.50 2.40 53 116
Along rolling Ra n Rsk # Rku # HSC # Pc*
direction peaks/length /cm
roll 4 (before) 5.11 -0.10 2.60 26
steel sheet 6% 3.98 0.05 2.60 30 65
steel sheet 10% 4.53 0.05 2.73 26 60
roll 4 (after) 4.98 -0.20 2.60 26
Normal to Ra n Rsk # Rku # HSC n Pc*
rolling direction peaks/length /cm
roll 4 (before) 5.32 -0.15 3.00 29 74
steel sheet 6% 4.16 0.12 2.43 27 59
steel sheet 10% 4.53 0.02 2.73 26 60
roll4 (after) 5.23 -0.10 2.55 29 71
#Form Talysurf data *Mitutoyo data

4 DISCUSSION AND CONCLUSIONS

Conventional white layers 17 tend to occur when steels are

subjected to very high localised compressive or shear stresses
such as on the surface of a hammer or under the zone of
maximum compressive stress in single point turning 18. They
are extremely hard and generally consist of ferrite and carbide
mixtures which are so intimately mixed that they are
featureless, even in the transmission electron microscope.
The microstructures in Figs. 3, 4, 7 and 8, however, have
been produced by melting followed by rapid cooling. Under
these conditions it is possible that the resultant microstructure
could be soft due to the incomplete transformation of
austenite to martensite whereas in a conventional white layer
the temperature must be low enough to prevent recovery or
transformation. The absence of carbides suggests that the
Fig. 10 Surface of steel sheet after 10% "white layer" was supersaturated with carbon and the fact that
deformation; Ra 1.61 gin. M2 high-speed steel usually requires multiple tempering, to
remove retained austenite following conventional
austenitisation, implies that this "white layer" could also be
Fig. 11 shows the topography of an area on roll 3 defined retained austenite. Similar evidence was presented by
by a scribed 2 mm square. Fig. 12 shows the same area with Werquin 12 who suggested that the retained austenite would
the roll curvature filtered out and Fig. 13 the corresponding be beneficial in that it would transform to martensite during
area parameters. Fig. 14 shows the topography of the working and so harden the surface of the roll while having a
matching area on the steel sheet and Fig. 15 its area higher intrinsic toughness than martensite. The fact that
parameters. Fig. 16 is a magnified view of one corner of quenching in liquid nitrogen increased the hardness of the
Fig. 12 (roll) which can be matched against the white layers, as shown in Fig. 9, tends to confirm that they
corresponding imprinted area of the steel sheet in Fig. 17. To were a mixture of retained austenite and martensite rather than
make comparison easier, Fig. 18 (sheet) is an inversion of the hard martensitic layer reported to occur when steels are
Fig. 17 and should be compared with Fig. 16 (roll). electrical discharge machined in conventional die sinking
operations19. EDT and conventional EDM arrangements are
By inverting the profile of the topographical plot of the different in that the workpiece and electrode are constantly
roll surface, a theoretical plot of the topography of the sheet, moving relative to each other in the former (rotation of roll
assuming a 100% transfer, can be obtained. This technique and scanning of electrodes) which may cause the surface to
can be developed further to evaluate the effects of roll wear experience a different thermal cycle from that in die sinking.
by truncating the tops of the peaks and inverting the resultant
roll profile to give a simulation of the expected sheet surface:
this has been shown previously to give a good approximation
of the wear process ISA6. Figs. 19-20 show the roll surface
after 25% truncation and its inverted texture corresponding to
that of the sheet (assuming 100% transfer), respectively.
 Electrical Discharge Texturing 189

:i!iii

Fig. 11 Surface topography of roll 3 showing 2 mm square.

Fig. 12 As Fig. 11 with roll curvature removed.

17.93
A r e a Parameters

Loggod A r e a = 1 6 0 x 160 p o i n t B
Spacing = 20.000 um
Increase in A r e a = 0.203 m m ^ 2
Est V o l u m e 0.235 mm^3
Volumo 0 . 1 4 6 ram^3
Ra 2. 772 u m
Rq 3 . 4 7 2 um
Rt 23.242 um
Rp 8 . 8 1 6 um
Rv 14.426 um
Rsk -0,580
Rku 3.192
Mean 1 4 . 4 2 6 um
0.000 5,810 11,621 17.431 23.242
Height (um)

Fig. 13 Area parameters for roll 3.

190 D. K. ASPINWALL et al.

Fig. 14 Surface topography of steel sheet after 10%

deformation with roll 3 showing 2 mm square

Fig. 15 Area parameters for steel sheet after 10%

deformation with roll 3.

Fig. 16 Magnified view of one corner of the square on roll 3

shown in Fig. 11.
 Electrical Discharge Texturing 191

Fig. 17 Magnified view of the corresponding comer of the

square on the steel sheet shown in Fig. 14.

Fig. 18 Inverted view of Fig. 17 for comparison with

Fig. 16.

Fig. 19 Roll surface after 25% truncation compared to

Fig. 16
 192 D.K. ASPINWALLet al.

Fig. 20 Inverted view of Fig. 19 to simulate sheet steel

surface after rolling with 25% truncated roll.

Although the steel sheet was supplied with a texture, the 7 REFERENCES
preliminary treatment showed that this could be removed
successfully to give an almost perfect replica of the ground (1) T.G. Nilan, B.M. Perfetti and B.I. Scialabba;
rolls. The texture transfer (Ra) using the softer aluminium Relationship of Sheet Surface-Roughness Texture to
alloy sheet appeared to be about 95%, however with the Painted Sheet Appearance; Proc. 19th Conf. on
harder steel sheet the transfer depended on the amount of Mechanical Working and Steel Processing, Pittsburgh,
reduction, but not significantly on the surface roughness of USA, June 1977; 148-157.
the roll; typical transfer ratios being about 80-85% for 10%
reduction and -75% for a 6% reduction. (2) A. Bragard, V. Leroy, H. Bonnarens, J. Richelmi,
C. Marique and C. Emond; Present State of the CRM
Significant differences occurred between the absolute Work Related to Surface Analysis of Cold Rolled Steel
values of the 2-D surface roughness parameters generated by Sheets; Proc. 10th Biennial. Cong. Int. Deep Drawing
a single scan technique and those from the 3-D parameters Res. Group, 1978, 253-278.
obtained from the raster technique using the modified
Talysurf 6. This arises from the different sampling and (3) A. Bragard, J. Crahay and W. de Smul; Texturing the
filtering techniques used. For example, in the 2-D method Roughness of Work Rolls of Rolling Mills by Means of
the cut-off length is normally 0.8 mm so that frequencies of Laser Pulses: the Process of Tomorrow? Proc.
longer wavelength are removed, but in the 3-D method used Advances in Cold Rolling Technology, Institute of
here, the cut-off length is effectively determined by the length Metals, London, 1985, 94-101.
of the scan, which in the examples shown was over 3 mm.
In practice, the only surface data modification which has been (4) K.Z. Wang, and Y.X. Yang; Study on the Surface
performed is least squares levelling. Furthermore, in the 2-D Topography of Rolls by Means of Laser Pulses; Adv.
method, the deviations are measured about a mean line Technology of Plasticity, 1990, 2, 749-754.
whereas in the 3-D method they are measured about a mean
plane. (5) C. Marique, J. Crahay and A. Bragard; Surface
Roughness and User Properties of Cold Rolled Steel
The 3-D technique shows that with the use of a suitable Sheets; Proc. Developments in the Drawing of Metals,
location feature, corresponding areas on roll and sheet can be Metals Society, London, May 1983, 242-248.
compared and the ease with which the map of one, or the
other, can be inverted makes it possible to match directly and (6) B. Snaith, S.D. Probert and R. Pearce; Characterisation
so enable the degree of texture transfer to be assessed. This of Laser Textured Cold Rolled Steel Sheets; Wear,
should prove useful in the analysis of methods of producing 1986, 109, 87-97.
textured surfaces and the modelling of wear characteristics as
discussed previously. (7) M.S. Ahmed and J.A.G. Knight; Roll Texturing by
EDT; Proc 30th Conf. on Mechanical Working and
5 ACKNOWLEDGEMENTS Steel Processing, Pittsburgh, USA, June 1988; 405-
413.
The authors would like to thank Dr. P.J. Sullivan, School of
Manufacturing and Mechanical Engineering, University of (8) M.J. Godwin; Application of New Type of EDT for
Birmingham for providing the 3-D topographical plots, Surface Preparation of Cold Mill Work Rolls; Proc.
Professor M.H. Loretto, Director of the Interdisciplinary Advances in Cold Rolling Technology, Institute of
Research Centre in Materials for High Performance Metals, London, 1985, 102-107.
Applications and Professor J.F. Knott, School of Metallurgy
and Materials, University of Birmingham for the provision of (9) H. Kuwamoto, M.Kawano, Y. Koga, Y. Kachi and
laboratory facilities. M. Hiasa; Development of a New Roll Texturing
System; ibid, 108-114.
 Electrical Discharge Texturing 193

(10) D.K. Aspinwall, F.L. Zhao and M.F. E1-Menshawy;

EDT of Steel Rolls; Surface Topography, 1989, 2, 123-
141.
(11) J.A. McGeough; Advanced Methods of Machining;
Chapman and Hail, London, 1988.
(12) J.C. Werquin; Influence of the Texturing Method on the
Metallurgical Structure of Cold Mill Work Rolls;
Chavanne-Ketin Technical Report, M15, Cedex 33,
92070 Paris-La Defense, 1987.
(13) M.F. E1-Menshawy (for Spark Tec Ltd.); Electro
Discharge Machining Apparatus; United States Patent
No. 4950860, 21st Aug., 1990.
(14) M.F. E1-Menshawy (for Spark Tec Ltd.); Electro
Discharge Machining Apparatus; European Patent No.
0283205.B 1, 21K,9/88, Bulletin 88/38
(15) T.G. King, W. Watson and K.J. Stout; Modelling the
Micro-geometry of Lubricated Wear; 4th Leeds-Lyon
Symposium, Institut National des Sciences Appliqu6es
de Lyon, Sept., 1977.
(16) K.J. Stout and C.J. Bovington; The Role of Surface
Texture and its Characterisationin Monitoring Cylinder
Bore Manufacture and Wear Performance Prediction;
Leeds-Lyon Syrrposium on Tribology, Lyon, Sept.,
1987
(17) D. M. Turley; The Nature of White Etching Layers;
Materials Science and Engineering, 1975, 19, 79-86.
(18) R. Milovic, E.F. Smart and M.L.H. Wise; Comparison
between Use of TiN Coated and Uncoated High-Speed
Steel and Fine Grain Cemented Carbide Tools in
Cutting Resulpharised 0.08% C Free Cutting Steel;
Materials Science:and Techn., 1986, 2, 59-68.
(19) Machining Da:a Handbook; Metcut Research
Associates; 3rd ed., 1980, 2, 18-104.