Surface roughness evaluation and morphology reconstruction of electrical discharge machining by frequency spectral analysis

In this paper, it is proposed a method that using mathematical equation to describe the surface morphology of electrical discharge machining (EDM). In particular, response surfaces and the effect of current and pulse width are established by measuring diameter and depth of the pits. A matrix of surface heights is collected by scanning EDM surface with appropriate magnification. Combined with the fractal theory, the suitable fitting frequency range is obtained by analyzing power spectral density of one of the row profiles. Wavelet decomposition is used to separate the profile into signal layers in different frequency ranges. Through picking five suitable signal layers, the characteristic frequencies and corresponding amplitudes of each layer are obtained based on the sizes of the single-pulse pit. These five groups of characteristic parameters are used to reconstruct row profile and surface topography. The mathematical equation of the EDM surface is obtained, and it is verified effective.