STUDY ON THE HIGH SPEED MACHINING BY EDM USING A

MOVING COIL HEAD TYPE FEED CONTROL

T. MASUZAWA, K. TANAKA and M. FUJINO

Institute of Industrial Science, University of Tokyo

S U M M A R Y
The paper is concerned with an experimental study on a high speed
method of micro-machining (drilling) by EDM in which a moving coil (MC)
mechanism was used for the feed control of the electrode. The drilling
speed by this method is improved in comparison with either the electro-
hydraulic system or the ultrasonic vibration to the working fluid. As the
machining time is quite short (for example, 8secs time is required for 0.20
hole in a 0.5mm thick aluminum plate with the discharge pulses in very fine
finishing range), a swift responsing electrode feed control is desirable.
It is possible to measure metal removal efficiency in micro-machining range
and there is an upper limiting speed for different materials. Micro-drill-
ing in non-ferrous metal especially brass is commercially viable by the
process.

INTRODUCTION pulse width). Thus an optimal design may

In recent years the micro-machining demand a largest allowable value of 'w'
has been becoming more and more important and a shortest pulse width that may be
in engineering production. Among various selected by experience.
micro-machining techniques available at
present, EDM is one that can not be neg- The above mentioned condition can be
lected. But unfortunately, it is not achieved only if all the applied voltage
popularly used in practice because of two pulses produce discharge without fail.
main reasons such as (a) the lack of rec- However, all voltage pulses are not fol-
ognition of the method itself and (b) it lowed by discharge. The ratio of the
requires more machining time than the number of discharge actually occurring to
conventional method, for example, mechan- the number of the total applied voltage
ical drilling. However, the inferiority pulses is defined as the effective pulse
in machining speed is not essentially a ratio designated by 'n'. This effective
factor effecting its application, because pulse ratio 'n' is greatly influenced by
the amount of material to be removed is the base voltage of the servo. But the
very small. On the other hand, the rate most important factor is the combination
of material removal may be rather far of the working area and the pulse dimen-
behind the theoretical limit, when a ma- sions. A typical relationship among the
chining speed of several times as much as base voltage, pulse dimension and 'n' for
that of mechanical drilling is aimed. a fixed working area is shown in Fig. 1.
This paper deals with a new method It shows that n~l can be realised only in
to realize micro-drilling using EDM in a limited range of the pulse dimension.
which a machining speed comparable to On the other hand, for smaller working
mechanical drilling is achieved. Besides area such as micro-machining a similar
that, as the limit of machining speed is relation is observed as shown in Fig.2,
greatly influenced by the removal effi- but the range of n~l does not exist at
ciency and the efficiency is supposed to all. This figure does not include the
go down in micro-machining range, data case of the voltage pulses with very small
regarding the limit of removal rate by duty factor. Thus it can be said that a
causing discharge in micro-energy range is source with high pulse repetition does
also determined. not promise a high speed machining in
FACTORS EFFECTING THE MACHINING SPEED ~ r/-~r/-r·r~-r·~.r/~-­
Effective Pulse Ratio RI
The rate of material removal in EDM .j.J
....
is expressed as w x n, where 'w' is the 0
average amount of material removal by a >
single discharge and 'n' is the number of Q)
Ul
short
discharge in unit time. The removal rate Rl Q) ------------------

can thus be increased by increasing either ,.Q .3.____ _ __

00

'w' or 'n'. The 'w' can be varied by fine + fine +

changing the electrical conditions, but machining range machining range
an increase in 'w' will be accompanied by
an increase in the surface roughness. On Fig.l States of the Fig.2 States of the
the other hand, the maximum value of 'n' effective pulse effective pulse
is approximately limited to 1/T (T- the ratio ratio in micro-EDM

543

B. J. Davies (ed.), Proceedings of the Nineteenth International Machine Tool Design and Research Conference
© Department of Mechanical Engineering, University of Manchester Institute of Science and Technology 1979
544 HIGH SPEED MICRO-EDM WITH MC-HEAD

finish- and micro-machining when only the control system (see Appendix I) whose
base voltage is controlled. response is much faster is used in this
experiment.
Short Circuit Through Debris
Another factor influencing 'n' is the
behavior of debris and carbon produced in Three types of MC heads (designated
the working fluid. It can be confirmed as No.1, No.2 and No.3) are constructed
by the experiment conducted earlier and and used in the experiment. Full range
described briefly as follows /I7: loud speakers are used as the drivers of
When a 0.50 electrode an~a flat these heads. The mass of the electrodes
workpiece .are arranged with small gap and selected is too small to cause considera-
a pulse source is connected the discharge ble inertia force. The head No.1 has a
begins, first sparsely but gradually in- guide and a spindle made of acrylic resin
creasing in number. Finally the value of as shown in Fig. 3. The mass of its
'n' approaches nearly to unity, when short spindle is about 3.5g. The head No.2
circuit occurs. The time from the begin- shown in Fig. 4 has a guide of steel and a
ning of discharge to the short circuit spindle of aluminum. The mass of the
varies with the gap distance and is usual- spindle is about 27g and can be changed to
ly within lOms. Thus it is obvious that BOg when a copper ring is attached. The
the presence of debris and carbon increas- head No.3 shown in Fig. 5 does not have
es 'n' but a high value of 'n' does not a spindle but a small electrode is direct-
continue for longer period. The short ly connected to the driver and is guided
circuit seems to be basically due to the through a hole drilled in a we alloy guide
filling up of the gap with debris or car- block. The mass of the holder in this
bon, while the gap distance was kept case is about 0.5g and it posseses the
fixed. It may then be said that more highest self resonant frequency among all
debris cause larger 'n' while too much the three types explained above.
results in undesirable short circuits The block diagrams of the driving
that may be the main reason of speed re- circuits are shown in Figs. 6 and 7. In
duction in micro-EDM. both cases, a flat response exists in
As it appears from Fig. 2, it is audible frequency range. The servo input
difficult to realize a high 'n' in micro- is the difference between the average gap
EDM but actually it is possible in the voltage and the base voltage as usually
long run, because a higher 'n' is often adopted in EDM machines.
observed instantly. In a feed condition
of this type, the number of short cir-
cuits is rather large and most of these
occur within lOms in the above mentioned
volt.
camp.
H
-
LPF H
'----·----1 DC amp.
U
ll MC
head
experiment. Regarding micro-EDM, the ,uPC55AX2 L....-----J
following may be expected: pick-up 2SD144
(l) The increase in 'n' approaching 2SB242A
to l is possible.
(2) It is desirable to prevent the Fig.G Driving circuit for the MC-heads
change from high 'n' to the short circuit No.1 and No.3
immediately.
MC
MOVING COIL (MC) HEAD AND ITS head
CHARACTERISTICS
As the reason of frequent short 2SC960x2 2SD188
circuits in a conventional electrode feed 2SA607 2SA627
system is observed to be its poor response
feature, a new moving coil head type feed Fig.7 Driving circuit for the MC-head
No.2

L-----•-- - holder

Fig.3 MC head No.1 Fig.4 MC head No.2 Fig.S MC head No.3