The gap distance actually says only about the machining accuracy. With gap widths between 200 .µm and 300 µm using orders of magnitude for roughing. Ideally, to set the gap control so that a balance between erosion and feed produced. The different gap widths can be set only when it comes to deviations of this equilibrium. So could with larger gap widths (300 µm) increased idlers occur and no longer rising productivity. With smaller gap widths (200 µm) occur more frequently unwanted discharges that cause higher abrasion, but no spark discharges (poorer surface quality).

A proper analysis you can do only if they know the pulse parameters and their changes in gap length change measured.

A further influences the gap and flush the process sensitivity of the process control..

In principle, you can establish a relationship between machining rate and volumetric relative wear (ratio between eroded volume from electrode-tool and removed volume from workpiece) based on pulse time (a main working parameter - despite the fact there are other working parameters influencing the machining rate and relative wear).

When pulse time grows, machining rate increases (mainly due to discharge energy increasing) and volumetric relative wear decreases. The last one is also the result of carbon deposition on active surface of the tool (when working with hydrocarbons dielectric liquid) but mainly due to increase of removed volume from the workpiece.

Of course there is the relative tool wear as a ratio of the removal compositions of the workpiece and tool. These removals are dependent on the independent process parameters, ie, at a static pulse generator can do this, the pulse duration and the operating current to be, as it can only be the work flow at a relaxation generator.

A decisive characteristic is then the set ignition voltage, which is constant in the static pulse generator and the ignition ideally takes place after a small ignition delay. When relaxation generator, the ignition voltage is changed according to an exponential function, ie the specified idle voltage is rarely achieved, the optimum is at 72% of maximum voltage and the pulse energy does not change linearly. Consequently, the effects of process control are different.

For example, the carbon layer formation must be eliminated by appropriate choice of parameters and modified flushing.

The effect of the surface roughness of the electrode was investigated on the surface quality of the part, the wear and the productivity is just a side project. Taguchi DOE method was used, and the standard EDM programs were used, which are suggested by the EDM machine controller. During the process the parameters are not constant, the controller changed them (I don't want to develop the EDM parameters, I find the best practice).

I understand your objective already, that is, but not fundamental relationships of EDM not to know.

Your approach leading to a useful result for the system technology used, ideally to an optimal setting. General conclusions are hardly feasible, because of the ignorance of important process parameters. However, some knowledge of the machine and the process parameters must be enabled to apply Taguchi DOE. The multiple literature particularly from India and China, which Taguchi DOE are to apply to about 80% false or inaccurate because the present process energy sources and the process controls used restrictions that were not taken into account. The predetermined process parameter ranges are rarely ideal.

Actually, the optimization should be preceded by a preliminary study that defines a "meaningful" parameter field, or confirms the predetermined range.

Best achievements in optimizing

Hans-Peter Schulze