Innovation leads EDM into the Industry 4.0 era

Electrical discharge machining (EDM) is an erosive manufacturing process for conductive materials. The process utilizes the discharges (sparks) between an electrode or wire (the erosion tool) and an electrically conductive workpiece. Each spark removes a part of a workpiece, a thermal-electrical process. The process is system-critical for everything from the automotive industry to information technology; cars or even smartphones could not be produced economically today without EDM. The process is used where other mechanical machining processes reach their limits, for example in the production of very deep and narrow holes or slots in extremely hard materials or with very complex surfaces.

Elaborate procedure
Spark erosion can be used to machine workpieces down to the smallest dimensions. "Each spark can be used to remove a specific point on the workpiece," explains Marco Boccadoro, Dipl. Ing. ETH. He is Head of EDM Research and Innovation at GF Machining Solutions in Losone (TI). This Swiss machine builder, a division of the mechanical engineering group +GF+, plays in the top league when it comes to EDM. And Marco Boccadoro confirms why EDM technology is not easy to master. After all, the challenge with EDM - as with many other industrial processes - is to ensure consistent and reproducible quality. EDM is still a relatively slow and costly process. In the case of die-sinking EDM, each electrode must first be manufactured to match the workpiece. Due to the high energy density of the discharges and the small gap between the electrodes, the EDM process is very complex to control. About 20 parameters need to be controlled in real time, and this is beyond the capabilities of a human operator. "For this reason, our machines contain an expert system, a kind of large database that provides optimized settings for most applications. But for several machining tasks, especially in production applications with repetitive tasks, there is a lot of room for improvement. This is where artificial intelligence can help by providing a learning capability for the machine," Marco Boccadoro continues. "Our machines are actually 80 percent pure electronics and computers, which makes them predestined for Industry 4.0."

Measurement and correction in the running process
Another important component of Industry 4.0 is the use of sensors, especially machine vision. Specifically, this involves the contactless measurement of holes drilled using EDM or of wire-eroded contours. Marco Boccadoro: "We have developed a system that inspects the contour of a hole by means of a high-precision camera, the so-called Integrated Vision Unit IVU. So it's a matter of optical measurement and logging of the measured values. What's new is that the information from the camera is fed back to the CNC and the machine can immediately derive suggestions for adjusting the process or making corrections." In principle, this is Industry 4.0 in its purest form: an optical device that is digitally fed back; in other words, a system that not only measures, but also acts on the basis of the measurement results, as it were, "in one fell swoop". The consequence of this: lower quantities of rejects, fewer process interruptions and shorter start-up times. In particular, this allows even more economical production of complex molds, such as dies and punches for the plastics and toolmaking industries or precision instruments for medical technology.

Inspect functional surfaces
Another area of application for this measuring system is the inspection of roughnesses and functional surfaces generated from the EDM process. A functional surface has properties such as self-cleaning effects caused by certain molecular structures. Such structures and, above all, defects in them can hardly be measured, except by using extremely expensive roughness measuring devices. As part of a CTI-funded research project, Marco Boccadoro's team is working on a so-called "surface interpreter" in collaboration with the Italian-Swiss University of Applied Sciences SUPSI and the Institute for Artificial Intelligence IDSIA in Lugano. The aim is to be able to measure surface quality and functionality as well as roughness while the process is running, and to detect and correct defects without having to remove a workpiece from the machine. The developers are drawing on experience from cancer research, among other things, where similar systems and the use of artificial intelligence can already be used to identify forms of skin cancer - i.e. a type of defect in the skin that is also a functional surface. This shows that thanks to interdisciplinary collaboration between a wide range of sciences, innovations are being created that will help Industry 4.0 projects achieve a breakthrough.