A Preliminary Investigation on Surface Roughness Assessement of Complex Additive Manufactured Parts Scanned by X-ray Computed Tomography

Additive manufacturing technologies have many advantages over conventional subtractive counterparts. X-CT is currently the unique way to measure both the internal and external geometries of complex AM products, which is not possible using either the tactile or optical measurement methods. However one of the barriers that restrict X-CT usage from being useful for surface texture assessment is that X-CT generated data structure is not compatiable with the standard procedures of roughness characterization which requires uniform sampled data and also takes the prerequite that measured surface is basically planar. It is proposed to use the Gaussian filter based on the linear diffusion equation to achieve the ability for a distortion free filtration on non-Euclidean surfaces as well as the compability with X-CT generated triangular mesh measurement data. The widely used roughness parameters, such as Sa and Sq, are also extended for the same purpose.