This PDF file contains the front matter associated with SPIE Proceedings Volume 10009, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

During design of optical systems or individual components, customers and designers already predetermine the way through production and the necessary resources. They define the tolerances that are required for their application like shape, micro-roughness and cleanliness, as well as the glass material that is necessary for the targeted optical effect. By the tolerances, they have also an influence on the manufacturer and his decisions on the production environment, measurement technology and the process chain that in his opinion is necessary to meet the specification. In most cases the type of polishing suspension and its parameters are also predetermined by the use of central polishing supply systems.

With this paper, the authors want to give a prospect of a currently running research with the aim of increasing the efficiency of chemo-mechanical polishing processes by a systematic selection and conditioning of the polishing suspension. They want to show that small amounts of well-conditioned and stabilized polishing suspension could be a tool to improve the efficiency in CMP processes under the influence of the global trend to products that are more individual and down to one piece flow production.

Main parts of this research are the identification of influences of polishing suspensions on the quality criteria of optical components and the creation of a method to find well-working suspension parameters in a deterministic and reproducible way.

Although computer controlled polishing (CCP) of aspheres and freeforms is one of the best understood state-of-the-art fab processes today, there are yet some unsolved issues: e.g. compared to bonnet polishing, fluid jet polishing is taking less iteration steps reaching the same form accuracy and ion beam figuring eventually is reaching much higher shape accuracies. This paper is a first move into solving this matter by introducing a novel footprint recording approach for CCP. To that aim, a new method for measuring the impact of a single tool mass acceleration value onto footprint shape is presented, the second derivative footprint recording (SECondo) method. First experimental evidence of the SECondo effect is presented, demonstrating that for bonnet polishing, acceleration of tool mass significantly alters the pressure distribution within the footprint and consequently affects its cross section.

Many processes applied in Computer Controlled Optical Surfacing (CCOS)^1–4 utilise a conformal polishing tool (e.g. the Precessions™ process) which is well suited to correct surface errors that are larger than or similar in size as the contact area between tool and surface, applying a different dwell time as needed. However, due to the conformal nature of the tool it has no significant effect on surface errors with smaller dimensions: the peaks and valleys of the error will be polished equally.

To remove these smaller dimension errors with a conformal tool is impractical. The preferred approach therefore is to apply a sufficiently rigid tool of sufficient size so that it preferentially removes material from the peaks and not from the valleys. A larger tool is capable of smoothing surfaces where the peaks are further apart and also benefits from a larger removal rate, reducing the total process time.

To achieve a uniform contact the form of the tool should be the inverse of the local form of the surface. This is trivial for spherical or plano surfaces, but present problems when the surface is aspherical. The mismatch between a rigid tool and the workpiece increases with the size of the tool for a given asphere, which leads to an upper limit of the tool size that can be used.

The work reported in this article presents a numerical analysis of the mismatch of rigid tools applied on E-ELT prototype segments. It can be readily applied to aspheric or free-form surfaces for which an analytical approach is difficult or impossible and furthermore it provides a detailed analysis of the form of this mismatch, including spatial frequency content. Additionally, an analysis and experimental work is presented to determine the applicability of sub-aperture rigid tools for the polishing of E-ELT segments.

One request for the optical machining of hart and brittle materials such as glass or tungsten carbide is not to exceed the critical depth of cut characterized by the transition point from ductile to brittle material removal.

Bifano et all. demonstrated the possibility to apply ductile cutting mechanism while machining hart and brittle materials by the use of ultra precision machines which enable highest precision in feed control and indentation depths. Based on the investigations a formula relating the material specific properties Youngs-Modulus E, material hardness H and fracture toughness KC was developed [1]. For this, the influence of the cutting conditions and the process environment were not taken into account. But previous investigations have shown a high influence of the cutting fluid while machining binderless nanocrystalline tungsten carbide through ultraprecision diamond turning [2].

Concluding, calculating the transitions point by using the formula shows insufficiently results and cannot be used for setting up a stable and efficient process. To adjust the formula and to allow a better prediction of the cutting process, further investigations are necessary.

This paper focuses in particular on the influence of water with different pH-values (acidic, neutral and alkaline) on the transition point for some brittle materials, binderless tungsten carbide and two different kinds of glass. Basic ruling experiments with a steadily increasing depth of cut have been performed. Measurements by white light interferometry have been done to analyze the influence on the cutting performance of mono crystal diamonds. As a result, a deeper understanding of the chemical influence on the mechanical process of diamond machining is created.

The trend in the optic industry shows, that it is increasingly important to be able to manufacture complex lens geometries on a high level of precision. The most interesting geometries are aspheric and free-form surfaces, which can only be produced with a huge time and cost effort until today. With these geometries it is possible to replace a number of spherical lens with just one aspheric or one free-form lens. As a result, installation space and weight are reduced, as well as scarce resources can be protected.

The goal of the research project ArenA-FOi (Application-oriented analysis of resource-saving and energy-efficient design of industrial facilities for the optical industry) is to create a new processing machine for manufacturing high precision optics. The primary kinematic system of the new machine is a Hexapod, which stands for high precision positioning and repeated accuracy. In addition, a Hexapod system is characterized by high stiffness and the reduction of mass, compared to existing machines. In the first phase of the project, the focus is on the machine concept, based on the required product range. Furthermore, it should be possible to create spiral and raster tool-paths for the process of machining. Especially, the raster tool-path is only applicable by a high radius of curvature which is limited by the maximum inclination of the Hexapod. If the result of the machine concept shows that an additional axis is needed, this axis must ensure synchronous control. Up to now, the processing of aspherical lenses is carried out by using the dwell time and vertical machining for the spherical best fit geometry. In this research project, we want to machine surface normal at every single point of the entries process and we also expect a quality improvement.

In this paper, only the project concept and the project's first phase — the plant construction and the development of aspherical surfaces — are discussed.

The inspection of optical elements for surface imperfections is mostly based on subjective evaluation by human operators. Automatic inspection systems (AIS) may introduce advantages in term of reliability, reproducibility and cycle time. The potential and limits of a camera-based, high-resolution AIS for scratches and digs are discussed. One important aspect is the illumination concept (brightfield or darkfield), regarded in relation to the scattering properties of an imperfection. Another aspect is the achievable spatial resolution of such a system. Different resolution limiting factors are considered, leading to criteria for the choice of digital sensors und imaging optics. Options to overcome a limited depth-of- field are also outlined. Next to technical aspects, the role of related standard specifications and system validations are addressed.

Quality control of optical surfaces specified according to ISO 10110-7 and ISO 14997 today is mainly performed by visual inspection, despite the drawbacks common to all operator-dependent processes, such as a lack in reproducibility and accuracy. In this work we present a detailed validation study of the automated surface inspection system ARGOS, which is capable of classifying flats as well as curved optics. The study reveals that ARGOS detects imperfections as small as 1 μm, can accurately determine imperfection sizes larger than 4 μm, and has an average repeatability of around 1 μm.

Laser interferometer deliver non-contact and zero mass loading measurements. They provide unique opportunities by measuring the movement of macroscopic and microscopic objects with an extraordinary high resolution and precision. Usually laser interferometer measure along the out-of-plane direction, the measurement object is moved along the measurement beam. But there is an increasing need to determine the in-plane motion or vibration characteristics of structures like MEMS or industrial surfaces.

An in-plane laserinterferometric sensor for measuring the lateral displacement of a surface was developed. The measuring principle of the sensor is based on the heterodyne detection of a moving object. Two laser beams with different frequencies are focused to a common point at the surface and the scattered light is collected by a common lens. The Doppler shift due to a lateral movement of the surface is detected. The sensitivity of the detection is a function of the incident light angle. This principle provides high sensitive measurements of the lateral velocity of the object.

The in-plane laserinterferometric sensor consists of a measuring head and a separate controller which includes the laser light sources and the demodulation electronics. The light sources are fiber coupled, so the sensor size is only defined by the size of the optical components used and the reasonable incident angle. Due to the highly integrated design, a small sensor size can be realized by using standard optics.

It can be shown that the in-plane laserinterferometric sensor is able to track surfaces with very low reflectivity as industrially rough surfaces. A short term noise with a standard deviation of 4 nm at 10 kHz bandwidth can be achieved. High precision measurements of objects displacement, speed, acceleration or frequencies are possible.

The vignetting field stop procedure uses a deflectometric approach to acquire big Optical Surfaces — DaOS — and it offers the possibility to measure nearly any shape or form using a scanning routine. The basic physical measurement principle in DaOS is the vignettation of a quasi-parallel light beam emitted by an expanded light source in auto collimation arrangement with a reflecting element. Thereby nearly any curvature of the specimen, is measurable. Due to the fact, that even sign changes in the curvature can be detected, also aspheres and freeform surfaces of any size can be evaluated.

In this publication the vignetting field stop procedure is discussed. Additionally the deflectometric setup is described. Because of some typical influences of beam deflection to the accuracy of angle measurement by using the vignetting principle, suitable methods of calibration for the sensor are examined and the results of these methods are presented.

Furthermore the technical principle of deflectometric measurements using an angle detecting device is explained inclusive of all random and systematic errors generated by the setup.

The last part of this publication shows the actual result of test measurements with calculated absolute deviation of errors with a large lateral dimension as well as the determination of the maximal achievable lateral resolution by detecting mid frequent structures on flat and spherical test parts with a diameter of 300 mm. These measurements are compared critically to reference results which are recorded by interferometry and further scanning methods.

There is an increasing demand for accurate form measurements of large optics with dimensions of up to one meter. For example, reference flats of Fizeau interferometers have to be known in the nanometer range. Synchrotrons or XFEL mirrors are often more curved, but nevertheless have to be manufactured and measured at a comparable or even better level of accuracy. Different approaches like deflectometry or interferometry in combination with stitching methods are typically used for these purposes, but obtaining the low uncertainty levels needed is still a considerable challenge. In this paper, measuring concepts and systems used at PTB for these purposes will be presented. For flatness measurements, a so-called Deflectometric Flatness Reference (DFR) system and a Fizeau interferometer were used. Slightly curved surfaces can be measured with the Traceable Multi Sensor (TMS) method, and CMMs with point or line sensors were also available. We will also present the current measurement capabilities and some measurement examples of form measurements of optical surfaces at PTB. The different setups and their pros and cons will be discussed. Future developments in the field of large-optics measurement will also be shown.

The complexity of optics manufacturing is not only due to the demanding geometrical features of single surfaces, it increases significantly when several surfaces need to be aligned in the correct position within a multi element optical system. Even slight imperfections of this alignment process in the range of a few microns can lead to a substantially reduced quality of the overall system.

The functional orientated wavefront based alignment approach offers an elegant solution to this challenge. Compared to conventional alignment methods it has the potential to reach very high accuracies in the sub-μm-range with an excellent efficiency. A consecutive stepping through the system is not necessary, since after the measurement of the optical performance of the system the correction of all possible degrees of freedom will be calculated at the same time.

In order to transfer this potential into real applications two major challenges related to the available measurement devices have to be met. The first one is the repeatability of the wavefront measurement. Tiny errors in the position lead to also quite small changes in the wavefront properties which need to be detected reliable. The other aspect concerns the limited sensor area of real measurement devices. Depending on the magnitude of the perturbation in the different degrees of freedom the lateral displacement of the resulting spot is generally a lot larger the sensor area available.

Therefore, this publication will initially introduce to the capabilities of the wavefront based alignment approach, illustrating the performance and accuracies as well as some of the parameters decisive to the procedure. Furthermore this paper will elaborate on the mentioned challenges in the transfer of these results to real applications.

In addition to the aspheric or freeform surface a lens or mirror also contains other structures such as edges and fiducials. These surfaces are important for the production process including the assembly of the optical system. It is demonstrated how these structures can be measured with a high precision cylindrical coordinate measuring system involving an optical and a tactile probe system. The measurements reveal a reference coordinate system which defines the position and orientation of the freeform surface which is also a necessary requirement for the measurement of the freeform surface.

Euclid is the name of a space telescope currently being developed in the framework of the ESA Cosmic Vision 2015- 2025 Program [1], [2]. It addresses cosmological questions related to dark matter and dark energy. The lens system of one of the two scientific key instruments on board of Euclid (a combined near-infrared spectrometer and photometer) is designed and tested at the Max Planck Institute for Extraterrestrial Physics.

In this paper, we compare two complementary approaches to determine the imaging quality of the photometer. The first approach is based on a direct camera measurement of the point-spread function (PSF) while the second approach uses a Shack-Hartmann sensor to reconstruct the wave front of the system.

Both methods yield in principle largely overlapping information in terms of e.g. modulation transfer function (MTF), encircled energy (EE) or spot shape. However, the experimental requirements are quite different. Details like the spatial extent and coherence properties of the object have different impact and are essential for the degree of agreement between both approaches. Moreover, the disturbance due to measurement noise is quite unequal and requires different evaluation steps.

The pros and cons of both methods that actually complement one another are investigated and discussed.

Fluid Jet Polishing (FJP) is a novel polishing technology, which experienced growing interest for niche applications over the recent years. This publication suggests FJP for processing of steel moulds for plastic injection moulding. These moulds are typically finished manually by skilled experts to date. FJP is capable of complementing the manual procedure or even replacing it. Fundamental knowledge about the process and its special tool function is provided. An experimental setup is demonstrated along with a polishing strategy and first results on different materials. It is shown, that it is possible to smoothen residual features from grinding, milling and turning below 0.12 mm. Furthermore an extract of current research is illustrated. Lastly a recently designed and built robot cell is presented, which is developed towards automation of the Fluid Jet Polishing process in the near future.

The in-situ etch rate measurement system (ISERM) has not just been developed to bring the automation process of an IBF plant to the next level, but to add another possibility of control for the ion beam and therefore characterize the tool of the plant. Thus far a dummy sample must be used on which footprints and grooves must be generated by the ion beam tool. To determine the etch rate and the shape of the ion beam the sample needs to be measured with a surface interferometer afterward. This is a time intensive and static process and has to be redone in case of changing process parameters, e.g. ion source parameter settings. The ISERM is capable of automatically and repeatedly determining the etch rate and the shape of the ion beam without the need of processing a dummy sample.

This paper builds on previous reported work describing the marriage of robots and CNC polishing machines, both for the pre-processing of parts, and to automate operations hitherto manually conducted on the CNC platforms. This paper reviews strategies for metrology, then takes the work a stage forward by reporting the use of a robot to automate the exchange of a part between CNC machine and metrology station, the probing of the part, and the capture of interferometer data. This constitutes an important step towards realization of an automated manufacturing cell.

Modern production machines collect a variety of data during operation: for example, operating status, information on the process result, operating data of the machine itself and metered value of the environment, of the machine and of the used tools.

These data are needed for the control of the machine. Selected data are displayed at the machine for the operator. In the optical industry there is no automated capture of machine data today. Selectively data are recorded and evaluated manually.

In other industries, for example, electronics manufacturing, it is common to capture data automatically and transfer them to a super ordinate Manufacturing Execution System (MES). The MES links the data of the production machines with quality data, with product data and logistical data.

Based on these date, Key Performance Data (KPI) can be defined. The KPIs represent the progress, the performance of central targets or measure the success of critical factors. The evaluation can be in real-time for online presentation, for specified periods or on request. An overview of the most common KPIs is included in the compilation of VDAM 66412-1: 2009-10.

Today, MES and their databases use cloud computing or internet based systems to link different plants or to integrate customers, suppliers and machine manufacturers in the communication system. The machines are therefore directly or indirectly connected to the Internet (Internet of Things).

Advanced optical systems are more and more demanding in terms of resolution, imaging quality and speed of capture. Controlling mid-spatial frequencies and surface roughness on these optics is critical for mitigating scattering effects such as flare and energy loss. By improving these two frequency regimes, higher contrast images can be collected with improved efficiency and lower distortion.

Classically, Magnetorheological Finishing (MRF) is implemented in production to correct low order errors generated by conventional polishing techniques on planos, spheres, on- and off-axis aspheres and freeform optics achieving figure errors as low as 1nm RMS while using careful metrology setups. MRF is also used routinely to turn a sphere into an asphere or freeform, or to print high resolution wavefront corrective patterns on optical surfaces to compensate for system errors or bulk material inhomogeneity.

Recent advancements enable correction of mid-spatial wavelengths as small as ∼1mm and smoothing of surface roughness to ∼1Å RMS. Using these new developments combined with correction of low order form error have improved MRF performance to manufacture higher precision optics with respect to the mid- and high-spatial frequency regimes.

Efficient mid-spatial frequency corrections utilize optimized process conditions; raster polishing with a small tool size. Furthermore, a novel MRF fluid, called C30, can finish surfaces to ultra-low roughness (ULR) and its low removal rate is optimal for fine figure correction of mid-spatial frequency errors. C30 MRF fluid is able to achieve <1.5Å RMS roughness on Silicon, CaF₂, Fused Silica, glass and other materials. It is best utilized within a fine figure correction process to target mid-spatial frequency errors as well as smooth surface roughness ‘for free’ all in one step.

These expanded capabilities of MRF technology are well suited for producing high precision optics to be used for industrial, medical or semiconductor optics.

Precision Glass Molding (PGM) is a manufacturing method which enables cost efficient production of highly complex glass optics in medium to high quantities. The produced glass optics can be intended for applications in a spectral range from 180 nm ultraviolet to 13 μm thermal infrared. However, it is necessary to choose the glass type according to their intended task and wavelength domain. Typical glass types used for PGM are low Tg optical glasses. Fused silica and chalcogenide glasses expand applications to UV and IR. The use of different glass types entails a variation of the necessary temperature range in which the glass can be processed. The processing temperature of these glasses varies between 200 and 1400 °C. To ensure an economical processing, mold materials and mold manufacturing technologies are varied according to the task at hand. The application of PVD thin film coatings is one of the methods commonly used for prolonging the molds service lifetime. However, the specific coating has to be selected suitable for the glass type and mold material. To illustrate the differences in molding optical grade low Tg glass, fused silica and chalcogenide glass a representative process chain for PGM is described and peculiarities for each glass category is presented in regards to the state of the art. Finally an outlook on ongoing and future issues for applied research in the filed of PGM is given.

The most important part in manufacturing precision optics is a reliable measurement procedure which provides results a few times more accurate than the quality to be produced. In general two specific values are important, the repeatability of several measurements which are done in a row and the absolute accuracy which is mostly defined by the systematical error of the measurement device. The repeatability can be improved relatively simple, by increasing the number of measurements and a following averaging step. To increase the absolute accuracy of a measurement device in the field of precision optics is far more challenging.

In this paper several interferometer absolute testing methods to measure flat surfaces are compared. The main objective was to name a value for the achievable accuracy. Therefor four different methods were analyzed:

1. The three flat test, a method which is already used several decades to determine the quality of a flat surface. As a result, two absolute measured profiles, horizontal and vertical, can be calculated.

2. The multi rotation test, an extension of the classical three flat test. The big advantage of this method is a fully three dimensional map of the systematical error.

3. The systematical error calculated by the SSI-A. Hereby several subapertures are measured over the whole surface. The redundant information’s of the overlapping regions can be used to calculate the systematical error of the system.

4. The rotation of the transmission flat relatively to the interferometer. Thereby the rotation unsymmetrical errors can be calculated and subtracted.

By adaption and coordination of successive process steps the efficiency of producing optics can be improved. In the beginning of this investigation, the focus is on two crucial process steps of the optical manufacturing process.

First, there is the optical and mechanical design of the optical part and system. In this context, it is important to consider what modifications compared to the current standard design are possible respectively useful. Thus, the modifications will only concentrate on parts of the optic that do not interfere with the optical functionality. Furthermore in this part of the study the main aspect is the clamping situation of the optical part in a holder or an optical system. It will be discussed, whether it is useful to add special clamping surfaces respectively a clamping flange in contrast to the standard design. The faultless functioning of the optics has always first priority.

Second, the manufacturing process itself has a great number of single steps. Therefore, a main part of the research is the investigation of current clamping situations in optics manufacturing, with prospect to improve them in order to work perfectly with the design modifications, as mentioned above. This part of the investigation focuses on the clamping system for the lens on each machine necessary for manufacturing.

In the scope of this paper the authors want to summarize characteristics and deficits of current clamping systems and workpiece mounts and give a prospect of the following course of action.

Optical surfaces are usually machined by grinding and polishing. To achieve short polishing times it is necessary to grind with best possible form accuracy and with low sub surface damages. This is possible by using very fine grained grinding tools for the finishing process. These however often show time dependent properties regarding cutting ability in conjunction with tool wear. For a successful grinding process the tools must show a constant chipping performance. A constant, at least predictable wear and cutting behavior is crucial for a deterministic machining. The paper describes the results of a testing series on a test bench for grinding tools. The test bench was presented for the first time on a SPIEConference in San Diego in August 2015 [1]. It is used to investigate the machining behavior of single grinding pellets on different materials under certain conditions, for example load and cutting speed. The bench is equipped with a linear measuring system and allows in-process-analyzing of material removal, tool wear and self-dressing effects. The gained data is used to determine suitable parameters for grinding processes, such as form grinding with ball-shaped tools or cup wheels and also gives information about tool wear and lifetime. This work describes a method to determine the characteristics of pellet grinding tools by tests conducted with a single pellet. The determination of the effective material removal rate and the derivation of the G-ratio are investigated. Especially the change from the newly dressed tool status via the quasi-stationary to the worn status of the tool is described. From the results of these tests the usability of a pellet grinding tool for a specific grinding task can be determined without testing a comparably expensive serial tool. The collected data can be stored and used in an appropriate data base for tool characteristics and be combined with useful applications.

The development of high precision methods for the measurement of plano surfaces became of increasing importance over the last years. Recently accuracies in sub-nanometer range have been achieved on samples up to one meter in diameter. The used measurement method was based on direct deflectometry. The main part of the system was an electronic autocollimator measuring local angular displacements on defined traces along the surface of the plano lens. To stabilize the optical path a pentaprism was used. The measurement accuracy of a similar system was evaluated at the Technologie Campus Teisnach. The used system delivered twenty measurement spots along the profile on one diameter. Four diameters were measured at every 45 degrees. An evaluation algorithm was developed to model the complete threedimensional surface out of a small amount of measurement points. Within this modelling Zernike polynomials were used to reconstruct the surface topography. Two different approaches in using the hierarchy of the polynomials were compared.

At first a reference surface was created by rotating a symmetrical averaged curve of all measured profiles. On the residuals of the original measurements to the symmetrical averaged curve a serial development of the error surface was applied with the help of Zernike polynomials. Different order of Zernike terms were tested because we saw a big influence on the result. This surface was added to the reference surface. The results of the two different approaches were compared. To enable us to compare the results of this measurement method to interferometric and optical 3Dprofilometric measurements the data was converted to xyz-format.

An ideal pitch button blocking process determines the level of workpiece deformation, especially the high-aspect-ratio optics, during the blocking process and process of polishing later. We have studied the pitch button blocking process by Finite Element Analysis (FEA) according to the thermoelastic equation. Meanwhile, the optimized pitch button blocking has been gotten by FEA which includes the thickness and material of blocking plate, as well as the radius, arrangement, elastic modulus and coefficient of thermal expansion of pitch buttons. The numerical simulation of Nd:glass (Ø100 mm×2 mm thickness) shows that the surface figure change (ΔPV) which is induced by the thermal stress during pitch button blocking process is influenced seriously by the thickness of blocking plate.

An important chain link in modeling of three dimensional data (3D data), 3D prototyping, CAD-CAM, computerintegrated manufacturing (CIM), PC gaming, global 3D teleconferencing, future e-commerce, product advertisement and mobile infotainment is the visualization of 3D data and 3D objects. On the one hand, there is an increasing demand for 3D displays providing natural three dimensional viewing experience, but on the other hand there is a lack of available 3D displays capable of generating all depth cues. Available state of the art 3D displays can provide only a few depth cues and a very limited 3D experience. This is due to the inherent physical limits of the different approaches used.

Holographic displays, for instance based on space bandwidth limited wave segment reconstruction, can provide all depth cues and a large viewing volume. They can provide satisfying 3D visualization. But still they are not available. Thick hologram gratings ¹, which also can be referred to as Bragg diffraction based volume gratings, are key components, which enable small form factor holographic 3D displays. Manufacturing large, display size Bragg diffraction based volume gratings is challenging. Collimators are key components within interference lithographic exposure setups. Using off-axis parabolic mirrors (OAPM) as collimating optical elements enables the generation of exact plane waves by using a single reflective surface. Thus, off-axis parabolic mirrors are preferred in order to realize the collimation of large wave fronts. The surface roughness has to be very low. The relationship between holographic 3D displays and specific requirements, which have to be taken into account when manufacturing off-axis parabolic mirrors, will be presented.

Actual optical systems have increasing requirements on accuracy and cost reduction at the same time. This leads to high demands for the process control during production. For high precision lenses it is not unusual that centering accuracies of single lenses and mounted components are the limiting tolerances for the overall system performance. At the same time it is important that also for less demanding systems the most cost efficient production process can be identified. By applying sensitivity analysis methods, typically used in optical design, it is easy to achieve a deeper understanding of the centering process of cemented components. The relevant centering parameters and their respective sensitivity can be calculated for different manufacturing methods and the optimum overall process can be chosen by comparison.