Because of their extreme hardness, ceramic materials and their composites are very popular for mechanical applications that are susceptible to wear and tear due to heavy loads. The other side of the coin is that conventional techniques do not usually suffice to work these materials. Some are even impossible to grind, because oxides deposit on the grinding wheel. For electrically conductive ceramics Electrical Discharge Machining (EDM) or spark erosion offers a solution. But in order to achieve optimum quality, process and material knowledge are indispensable. After all, there is a strong link between material properties and spark behaviour.
Ceramic materials and their composites are known for their extreme hardness (1400-4000 kg/mm2), high rigidity (200-600 GPa), high melting points (1600-3600 °C) and high chemical resistance to corrosion. In addition, certain ceramics (SiC, ZrB2, TiB2, Al2O3, ...) offer the advantage of low specific densities.
Top material in a wide range of application areas
Due to their high hardness, ceramics are very interesting for mechanical applications subjected to heavy loads, which are therefore susceptible to wear: tools, cutting slats, ball bearings, brake discs and other moving parts that make contact.
Technical ceramics can also be used in electrodes for metal processing or leading edges for aerospace appliances. Thanks to its high dimensional stability and low coefficient of expansion, SiC can be used as a construction and mirror material in high-quality aerospace instruments.
Figure 1: Relation between hardness and abrasion resistance
Conventional processing technologies fall short
The downside of ceramic materials: due to their high hardness, they cannot easily be worked with conventional techniques. For example, only tools in diamond or cBN are able to cut this material. This results in a limited geometric freedom of the designer.
Some types of ceramics are even impossible to grind, such as borides (Ultra High Temperature Ceramics, UHTCs) which are currently gaining in popularity. Oxides of these materials adhere to the grinding wheel and thus greatly complicate the machining process. Other processing technologies, such as spark erosion or Electrical Discharge Machining (EDM), must therefore be considered for this class of materials.
The solution: Electrical Discharge Machining (EDM) or spark erosion
The most popular technical ceramics - Al2O3SrO2 and Si3N4 - are non-conductive or low conductive (B4C, SiC). But by adding a conductive second phase, such as WC, TiB2 or TiN, for example, they do become electrically conductive (> 100 S/cm). They also become stronger, harder and tougher, resulting in finer microstructures. This is due to the inhibitory property of a second phase on the grain growth during sintering.
Figure 2: Conductive B4C-TiB2 composite
Materials that are electrically conductive are then eligible for machining with spark erosion or EDM. This process allows ceramics (and their composites) to be machined quickly and efficiently, no matter how high their hardness. After all, this process imposes a thermal load only on the workpiece. How does it work? A series of controlled electrical discharges between the workpiece (the ceramic) and the tool (wire, stamp or spark milling electrode) removes the material. During the process, both the workpiece and the tool are immersed in a dielectric (demi water or oil).
Figure 3: Principle of spark erosion
Different material removal mechanisms in play
Contrary to popular belief and current practice in the spark machining of metals, other material removal mechanisms also play a role in addition to evaporation and melting. For example, a so-called 'thermal shock' can occur, in which granules are dislodged by the thermal stress between two phases during the process. Increasing the strength and toughness of the material can therefore have a negative impact on the spark rate. What's more, it can even cause the dominant removal mechanism to shift to melting/evaporation. In general, materials for which the removal mechanism is thermal shock (e.g. TiCN, TiB2, TiN-ZrO2) have a higher surface roughness. A chemical reaction is also one of the options. Relatively high temperatures are reached during the machining process. Si3N4 does not melt, but decomposes from 1800°C into nitrogen and silicon (oxide). Also when machining tungsten carbides (WC, WC-ZrO2) the removal is mainly due to oxidation. The particle size plays an important role in the possible spark rates and roughness. Usually these surfaces are also very rough. The surface roughness of TiB2-B4C composites (Figure 5), materials which melt and evaporate under normal conditions is a good example. But a test sample with a certain amount of carbon on the grain boundary (TiB2 BT 60/40) always has a much higher roughness in the finishing phases. Why is this? The other grades have a strong grain boundary, and melting/evaporation is the primary method of material removal. The presence of carbon, which is used as a sinter additive, weakens the grain boundary and 'thermal shock' becomes the primary method of material removal. This affects the achievable surface roughness.
Figure 4: A spark erosion surface of Si3N4-TiN where the foamy structure is typical of a chemical reaction
Figure 5: Roughness after rough sparking (E501) and the various finishing phases (E502 to E506). TiB2 BT 60/40 is a material with a weaker grain boundary
Figure 6: Left normal B4C-TiB2 spark surface after finishing and on the right the material with a weaker grain boundary due to carbon addition. Grains are clearly visible which implies that 'thermal shock' takes place.
In short, in order to achieve optimum quality, process and material knowledge are indispensable. After all, there is a strong link between material properties and spark behaviour. Figure 7 shows an overview of some ceramic materials, their electrical conductivity (i.e. their suitability for sparking) and their hardness (often the parameter taken into account for use in wear-sensitive applications).
Figure 7: Overview of ultra-hard ceramic composites and their electrical conductivity