The hereby presented research, funded by the restricted grant LIDER, NCBiR, deals, in part, with the identification of the full implementation potential of the proposed optical measurement techniques in determination of surface flatness parameters, and their comparative assessment. The test methods included the photogrammetric measurement technique (TRITOP, GOM) and the structural light scanning approach (scanner ATOS, GOM), while the CMM measurement (DEA Global Image Clima) was the reference method. The accordingly designed and assembled experimental test stand comprised 2 steel plates. The test surfaces of the plates were appropriately ground; subsequently, the entire test stand was blackened to ascertain efficient optical scanning. Furthermore, the plates were connected by means of 8 screws, thus introducing considerable distortion. A measurement area of 140 × 240 mm was defined on the plate test surface, as determined by CMM, denoting 15 measurement paths of 240 mm in length, distributed every 10 mm, and characterized by measurement point densities of 1, 5, and 20 pt/mm. The reference CMM measurements were conducted on 3 consecutive days at different times (22 measurements in total) to exclude any possible surface modifications. Subsequently, optical scanning was applied and the measurement points lying at the cross-sections of the CMM measurement paths were isolated from the obtained polygon mesh. To further apply the photogrammetric method, the test surface was labeled with markers distributed every 10 mm and coinciding with the CMM measurement paths.
Comparative analysis of the flatness parameter for the selected CMM measurement and the measurement values obtained by means of the tested optical methods included:
- the entire measurement area,
- the sections comprising 80, 60, 50, 45, 40, 30, 20, 15, and 10 % of the entire measurement area, decreasing centrically,
- the measurement sub-areas of 30 × 50 mm allotted in the corners and in the center of the test plate.
The photogrammetric error of the tested parameter was established at 1.26–19.82 %, depending on the size of the measurement area. The corresponding error value, as determined by the structural light scanning technique, amounted to 0.03–4.31 %.