Electrical discharge machining (EDM) can produce cut-surface
quality that rivals or even surpasses surfaces finished by conventional
methods. In fact, one of the major benefits of the latest generation of
EDM equipment–both wirecut and diesinking–is that they often eliminate
the need for secondary finishing operations. This is especially
important with contours and shapes that are difficult to polish.
Although EDM is still relatively slow, particularly when a
high-quality cut surface is required, the newest equipment is capable of
much higher cutting speeds than was possible only a few years ago, with
no sacrifice in surface quality. Furthermore, advances such as computer
control, sophisticated software, and automation allow untended EDM
operations, releasing operators for other duties or automatically
machining workpieces overnight or during a weekend.
No machined surface is perfectly smooth. It doesn’t matter
how the surface was machined. An enlarged workpiece cross-section will
show peaks and valleys. These microscopic undulations determine surface
quality and whether a component will function as intended or even mate
with another component.
Before talking about EDM cut surfaces though, we must reach
agreement on what we mean by “high surface quality.” Since
there are various ways of specifying quality, and since the values can
be expressed in either English or metric, we must be careful in
evaluating and comparing surface-quality measurements. For example,
expressing surface roughness as so many microns is meaningless unless
the measurement method also is stated.
One way of expressing surface quality is the root-mean-square (rms)
method, which consists of squaring measurements taken over all the peaks
and valleys, adding the numbers, and taking the square root of the sum.
This sounds cumbersome and it is, but the result indicates surface
roughness and can be compared with similarly derived numbers to contrast
the quality of one surface against another. Although this method is
widely used, it overemphasizes maximum deviations that may only rarely
occur across the surface.
Another method gaining acceptance measures the arithmetic mean of
all peaks and valleys. This method measures all deviations and derives
a mean between the most prominent peaks and deepest valleys. For a given
surface, this measurement will be slightly larger than the rms value.
A third view of surface quality is maximum roughness, which
measures the distance between the highest peak and the deepest valley.
Obviously maximum roughness, R.sub.max in metric or H.sub.max in
English, will be considerably higher than the corresponding rms or
arithmetic mean measurement. The accompanying chart compares the
relative value of various finish measurements.
To understand how EDM affects cut-surface quality and integrity, we
must first look at how the process works. Metal is removed by erosion
caused by a controlled electrical spark. There is no direct contact
between electrode and workpiece. The rate at which metal is removed
depends on the electrical conductivity of the workpiece.
One terminal of the power supply is connected to the workpiece, the
other to the electrode. The workpiece and electrode are separated by a
dielectric fluid–usually deionized water for wire-cutting and a special
hydrocarbon for diesinking–which acts as an electrical insulator until
the spark occurs. The dielectric also cools the work area after the
spark ends, and flushes away metal particles before the next spark
DC voltage is applied between the electrode and the workpiece. At
first, no current flows because the two pieces are insulated by the
dielectric fluid, however, an electrical field does build up across the
gap. As the electrode approaches the part and the gap narrows, a point
is reached where the voltage ionizes the dielectric fluid and a spark
jumps the gap.
When the spark first occurs, a large amount of energy is released,
vaporizing material from the work surface. As current continues
flowing, intense heat melts additional material. When the voltage drops
to zero, current stops flowing, and the spark is quenched. At this
point material removal ceases.
As soon as the spark is quenched, the vapor bubble begins to
collapse. The dielectric cools the area, solidifying the material that
was melted. Some of this material is carried away by movement of the
dielectric fluid, leaving a small crater at the point where the
discharge occurred. Some of the melted material is redeposited into the
crater. This recast layer is an important factor in EDM surface
To some extent, this oversimplifies the sequence of events. The
accompanying drawings provide more detail. There are two important
points concerning EDM that should be emphasized here. First, it
isn’t simply the current flow that removes material–it is
switching the current on and off that actually vaporizes and melts the
Second, the most efficient part of the sequence is the initial
discharge when material is vaporized. The longer current flows, the
more heat builds up and the more material is melted rather than
vaporized. Also, some of the melted material always will be recast into
the cavity. There is presently no EDM technology to entirely eliminate
the recast layer. Newer power supply designs, however, are effective in
minimizing the recast layer as well as any heat-affected material
directly below it.
Pulses and quality
Metal removal, as already stated, depends on starting and stopping
the spark. The amount of metal removed is a function of the energy
released during the spark, which, in turn, is determined by frontal-gap
voltage (i.e., voltage between the workpiece and electrode), by the
electric current that flows, and by the time (pulse width) current flows
before the spark ends.
Older EDM equipment used a bank of capacitors in the power supply
to store energy for the spark. This design is known as a
“capacitant-discharge” power supply. The capacitors gather
and store electrical energy until the equipment senses proper
frontal-gap voltage, which creates the spark and releases this energy.
After the capacitors “dump” their charge, the spark is
extinguished and the capacitors begin a recharge cycle in preparation
for the next spark.
The problem is that the spark is initiated solely on the basis of
frontal-gap voltage sufficient to ionize the dielectric. For one spark
the capacitors may have become fully charged and will release a maximum
amount of energy thereby removing a significant amount of material. For
another spark the frontal-gap voltage may be reached before the
capacitors achieve full charge. This spark will remove less material.
Consequently, metal removal across a workpiece surface may be quite
uneven–there will be large craters at some points, small craters at
The pulse-type power supply was developed to eliminate this
problem. To control the spark more accurately, engineers designed a
power supply where each spark releases the same amount of energy. In
addition, each new spark is delayed until all the energy in the previous
spark has been used. The energy bursts may now occur at a more random
rate, but each time the spark creates an electrical current, the same
amount of energy is released and the same amount of material is removed.
The pulse-generator power supply provides more precise control of
the spark and enables more efficient use of each spark. This means
every eroded cavity will have essentially the same diameter and depth.
And, although there always will be some material recast into each
cavity, the depth of this recast layer is greatly reduced.
The latest generation of pulse-generator power supplies provides
highly refined cut-surface quality. The power supply furnishes a large
number of low current, extremely short pulses. Thus each spark is used
nearly 100 percent. Because of the low current and short spark
duration, material is removed primarily by vaporization. The recast
layer and underlying heat-affected zone is now about one-tenth of that
generated by a capacitant-discharge power supply. Surface finish quality
is such that secondary finishing operations are usually unnecessary. If
this recast layer must be removed from the workpiece in subsequent
steps, the amount of material will be less than with the irregular cut
produced by a capacitant-discharge power-supply system.
Surface quality is especially important when diesinking EDM is used
to make injection molds, pressure casting dies, and other components
whose cut surfaces must range from smooth-matte to highly-polished.
Modern pulse-control power supplies benefit both diesinking and wirecut
EDM. With the pulse-generator used in diesinking EDM equipment, a
highly-polished surface with an arithmetic mean roughness of 7 to 10
microinches is possible. Such EDM sink-polishing provides a consistent,
refined surface which is particularly desirable on delicate materials
and highly detailed components that could be damaged or destroyed by
traditional polishing methods.
Machining sintered materials
There is no problem with EDM cutting of conventional, homogeneous
metals such as tool and high-speed steel, or even titanium. Nonferrous
sintered materials, on the other hand, present an entirely different set
of problems. A major consideration with EDM cutting of sintered
materials is not so much a concern for surface quality as for surface
These materials include any of the various forms of carbide (i.e.,
carbides of silicon, tungsten, titanium, tantalum, and chromium). Also,
cubic boron nitride (CBN), because it has a hardness approaching diamond
and is less costly than diamond, is being used in some critical
applications. And then there is the “miracle material of the
’80s,” polycrystalline diamond.
Many people feel that EDM and sintered materials are incompatible;
that carbide and the other materials cannot be successfully machined by
this method. Others feel that the materials can be EDMed, but that the
necessary allowances, precautions, and limitations create so much
uncertainty that the chances for success are slim and unpredictable.
Unfortunately, there is some truth in both viewpoints, although certain
“wizards” seem to know a few secrets that they haven’t
shared with the rest of us.
In the case of proprietary materials, such as polycrystalline
diamond, material composition and special techniques for successfully
EDMing it have been withheld as trade secrets. Only recently has some
meager information surfaced.
The problem with electrical discharge machining any of these
sintered products is the structure and composition of the material
itself. Granules (or crystals) of the primary material are held
together by a matrix (or binder). The binder is generally much more
electrically conductive than the granules. Thus, during EDM cutting,
electrical current from the spark flows through the binder and around
the granules, eroding the binder and cutting the material.
The binder, however, does much more than merely hold the granules
in place. The sintering process captures the granules within the binder
under great tension. When the EDM spark flows through the binder, it
reacts electrochemically to the current, releasing some of that
tensional force. Following a high-energy EDM pulse, some of the
granules are totally free and will simply fall away. Others are only
partly held in place and may flake away under moderate pressure.
In studying the effects of EDM cutting on carbide materials,
researchers discovered that the shape and duration of the electrical
pulse had a marked influence on the surface integrity of EDM-cut
carbides. With older power supplies, varying amounts of energy
available in the EDM spark had inconsistent effects on the cobalt binder
material. The pulse-type spark generators that release an identical
amount of energy in each spark limit the electrochemical destruction of
the binder, resulting in improved surface integrity.
For more information on EDM equipment, circle E19.