J. Eur. Opt. Society-Rapid Publ. 2025, 21, 30 Ó The Author( s), published by EDP Sciences, 2025 https:// doi. org / 10.1051 / jeos / 2025026 Available online at: https:// jeos. edpsciences. org
Journal of the European Optical Society-Rapid Publications
RESEARCH ARTICLE
Knowledge based full aperture polishing
Max Schneckenburger 1,*
, Ayaan Shaikh 1, 2, Ufuc Algac 1, 3, and Jan Mewis 1
1 |
ASML Berlin GmbH, Waldkraiburgerstr. 5, 12489 Berlin, Germany |
2 |
Rhine-Waal University of Applied Sciences, Faculty Technology and Bionic, Marie-Curie-Straße 1, 47533 Kleve, Germany |
3 |
Technische Universität Berlin, Straße des 17. Juni 135, 10623 Berlin, Germany |
Received 1 March 2025 / Accepted 22 May 2025
Abstract. Understanding and controlling of the polishing process on conventional NC( Numerical Control) machines is an important step to optimize production, reduce machine time and increase production quality. Nevertheless, due to the high number of process parameters, polishing is not understood and is almost exclusively applied empirically. The work presented in this paper shows a production-ready attachment to a lever arm polishing machine, which can be used to map the removal of material using relative speed. Calculated relative velocity and observed material removal indicate a correlation of 31.5 %. This is a first step towards the complete automation of the polishing process in order to save process times, increase repeatability and avoid handling errors. It is planned to focus other polishing parameters in the future and improve the removal model.
Keywords: Lever arm polishing, Spindle polishing, Automated manufacturing, Model based polishing.
1 Introduction
In the production of high-precision optical components, achieving accurate surface geometries is essential – particularly in applications such as lithography. Even slight deviations from the intended surface shape can lead to optical aberrations that compromise pattern fidelity and overall system performance. As a result, the polishing process plays a key role in ensuring the required surface accuracy. Despite its importance, polishing remains largely empirical and manually controlled [ 1 – 4 ]. Numerous interacting parameters influence the outcome, including mechanical and chemical effects, as well as tribological conditions in the polishing interface. Due to the complexity of these interactions, the process is often carried out iteratively: a surface is polished, measured, and corrected in subsequent steps based on the measurement results. This approach becomes especially time-consuming when in-situ measurement on the polishing machine is not possible. In such cases, the part must be removed, transported, and re-referenced on a separate measurement device— introducing delays and potential risks such as alignment errors or handling damage. Moreover, fo – certain materials, such as germanium, manual polishing can pose health hazards, further motivating efforts to automate the process. This work presents a practical method for capturing and evaluating the polishing behavior on conventional lever arm polishing machines. The goal is to systematically map material removal based
* Corresponding author: max. schneckenburger @ asml. com on relative velocity and establish initial quantitative correlations. This approach lays the foundation for future model-based process control, aiming to reduce process time, increase repeatability, and minimize operator-dependent variation.
A lever arm polishing machine is a simple polishing machine and the principle did not change during the last one hundred years. One part rotates( work piece or the polishing tool) and the other is guided over it in a bananashaped path. A lever arm machine can be built using several set ups: e. g. a rotary disk and a robot or a rotary disk and a Computer Numerical Control( CNC) gantry set up. Wellknown manufacturers are Leico, Loh or Satisloh, Dopa or Stock.
In the shown research, the work piece is always guided and the polishing / lapping dish rotates beneath it. A schematic representation of a lever arm machine can be found in Figure 1.
Such a machine can only polish a convex or concave shape and a combination of both can enable aspheres or flat surface. If more material is removed over time in the center of the work piece than at the outer edge( this depends on the relative speed, which is defined by the position on the polishing tool), the work piece becomes concave. For a convex surface vice versa. Aspheres can be produced by combining them to a limited extent( not all curvature radii are possible). In principle, any axisymmetric optics can be produced with such a machine also cylindric lenses.
In this case, the polishing tool is rotating and the work piece is guided over the polishing dish surface. Since the
This is an Open Access article distributed under the terms of the Creative Commons Attribution License( https:// creativecommons. org / licenses / by / 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.