Abstract

The characteristics of additively manufactured parts (topological optimisation, freedom of design …) present a challenge from a metrological point of view. In this respect, non-contact inspection is of great importance for the rapid verification of this type of parts. However, their particular characteristics (stair-effect, high roughness, surface finish) can contribute to measurement errors and deviations when using different optical inspection equipment. In relation to this, 3D scanning point-clouds filtering can improve the results by increasing precision during inspection. In particular, this paper evaluates the influence of applying filters to digitize additive manufactured metal parts, which are scanned using different optical inspection systems commonly used in industry (based on laser triangulation, conoscopic holography and structured light techniques). The main objective is to establish the most suitable inspection procedure depending on the inspection system used. According to the dimensional results achieved, it can be concluded that the filtering does not have significant influence, regardless of the used sensor. However, the geometrical results are strongly influenced by the point-cloud quality; consequently, it is recommended to apply filters for most of the evaluated optical systems. One of the main contributions of this study is the definition of the filtering process methodology prior to non-contact inspection, depending on both the used optical system and the part to be evaluated, in order to define the most suitable filtering parameter to achieve a precise optical inspection.

Keywords

Optical measuring systems, Non-contact inspection, Point cloud filtering, Metrological evaluation, 3D scanning, Metrological software filters, Geometrical and dimensional analysis, Metal additive manufacturing