Bosch-Shaft Tactile/Optical Comparison

<p style="text-align: center;"><img src="/ueditor/php/upload/image/20260911/1789138040116246.png" title="1789138040116246.png" alt="1.png"/></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">The physical measuring principle of fringe projection is fundamentally different from the tactile measuring principle. The physical limitations and special topics of optical measurement methods were discussed in detail in VDI Report No. 2393, 2021 (Ref. 5). The very good comparability of the tactile measurement results from section 2 was classified as sufficiently accurate for the system validation. The measurement and evaluation strategy of Zeiss Prismo verity was thus defined as the basis for the evaluations with Zeiss Atos Q. The special feature of the&nbsp;Gear Pro&nbsp;involute gear metrology software from Zeiss is that the same measurement program can be used for the evaluation of non-contact measurement data in STL format as for tactile measurement. This offers the advantage that, despite the different physical measurement principles, no further uncertainty contribution regarding the measurement strategy must be considered. In principle, however, as shown in Ref. 5, various restrictions must be observed for a high-quality measurement. The following relevant points are to be mentioned as examples: spraying-in of the surface to minimize reflections; ensuring accessibility in areas that are difficult for optical systems, such as the tooth root or the transition area to the protuberance; and sufficiently high point density to minimize the influence of polygonization (merge individual scans into a final mesh).</span></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">The measurement scope for evaluating the Bosch-Shaft with twelve teeth has remained identical to that used for the tactile validation measurements when evaluating the optically determined data with the Zeiss Atos Q. However, there is a significant difference in the recording of measuring points. For the subsequent mathematical evaluation regarding gear characteristics, the complete recording of the workpiece is necessary. Depending on the number of teeth, generating the data set takes significantly longer than tactile measurement with a GMI or CMM. An overview of the measurement scope and the measurement times is shown in Table 2.</span></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">The differences in the measurement results are again displayed using the actual-actual comparison function in the same way as in the section “Bosch-Shaft Tactile System Validation.” The classic characteristic profile is shown as an example in Figure 8 for the measurement data of Klingelnberg P40 (green), Zeiss Prismo verity (blue) and Zeiss Atos Q (yellow).</span></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">The system-related lower measuring point density of the fringe projection generates, especially in profile direction, rounding in edge areas due to polygonization and thus leads to falsification of the measurement result. At the Bosch-Shaft, there are only about 100 points in the profile direction after polygonization. Figure 9 visualizes the profile comparison results from Figure 8, enlarged for fringe projection systems’ “critical” edge areas after polygonization and gear evaluation, transition area involute zone into profile tip relief, and transition area into root zone/protuberance.</span></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">For all determined characteristics of profile and helix, the respective differences are Δ &lt; 6 μm. For the determined characteristics of pitch, runout, and spanwidth, the respective differences are Δ &lt; 3 μm. The differences determined are thus within the known orders of magnitude of tactile/optical comparison measurements of gears (Ref. 5). The differences for the waviness characteristic order spectrum are shown as an example for the profile measurement data in Figure 10. In Figure 10b, you can see the noise of the order spectrum of the measurement data determined by fringe projection (yellow) in the low-frequency range and the lack of detection of orders in the high-frequency range in an enlarged and colored variant. This makes it difficult to reliably detect frequencies. The numerical differences in the orders found are in the range Δ &gt; 50 nm.</span></p><p><span style="font-size: 14px; font-family: arial, helvetica, sans-serif;">The comparability of the waviness characteristic helix angle of waviness (βw) for the largest order found is shown in Figure 11. The differences between the two tactile measurement results and the optical measurement result are in the range of Δ &lt; 1.5°. In Figure 11, there is a visual comparison of the optical system given, but it is presented in a “damped” way. The reason for the visual differences can be explained by the noisy result of the order spectrum; the physical system-related restrictions cannot be dismissed out of hand at this point. The differences determined by the tactile system validation and the tactile/optical comparison are summarized in Table 3.</span></p><p><br/></p>