318
J. Eur. Opt. Society-Rapid Publ. 21, 30( 2025)
Fig
. 9. Frequency analysis of the rotational angle of the work piece.
Fig
. 10. Comparison qualified calculated material removal and real material removal.
workpiece is shown with its location-dependent relative average velocity in m / s. Using the Preston equation and assuming that the Preston coefficient and polishing pressure remain constant, the two images should be quantitatively comparable. It is not possible to determine the amount of material removed. Changes in the polishing tool are also not taken into account. The Pearson-correlation between thosetwoimagesis31.54 %.
6 Discussion
In the removal images it is clear that the result is concave in both cases. Nevertheless, there are deviations in the material removal: the main reason for this is probably the different area load. Unfortunately, the pressure on the polishing lever cannot be considered constant; there were pressure fluctuations during the tests. The work piece holder was also smaller than the work piece, so the work piece holder weight distorted the result. The weight of the work piece and the holder act as a constant area load over the radius( dead load). The force applied by the ball pin / calotte, on the other hand, acts as a point load, which depends on the diameter. This point can also be seen as an outlook to the future: taking area load into account. Overall, one should consider whether a six-axis robot and a rotating disk would not be a better production setup. With such a robot setup, you would use a rotary table for the polishing / lapping tool and a robot to guide the work piece over the tool. A robot offers more freedom when it comes to path planning. Generally, any NC motion system is suitable for this.
For a first attempt, the result is quite promising. Presumably, individual incorrectly taken images are statistically averaged out and therefore have little influence on the material removal. The Pearson correlation between