Florian Mauch

Improved signal model for confocal sensors accounting for object depending artifacts

The conventional signal model of confocal sensors is well established and has proven to be exceptionally robust especially when measuring rough surfaces. Its physical derivation however is explicitly based on plane surfaces or point like objects, respectively. Here we show experimental results of a confocal point sensor measurement of a surface standard. The results illustrate the rise of severe artifacts when measuring curved surfaces. On this basis, we present a systematic extension of the conventional signal model that is proven to be capable of qualitatively explaining these artifacts.

Advantages of chromatic-confocal spectral interferometry in comparison to chromatic confocal microscopy

Chromatic confocal microscopy (CCM) and spectral interferometry (SI) are established and robust sensor principles. CCM is a focus-based measurement principle, whose lateral and axial resolutions depend on the sensors numerical aperture (NA), while the measurement range is given by the spectral bandwidth and the chromatic dispersion in the axial direction. Although CCM is a robust principle, its accuracy can be reduced by self-imaging effects or asymmetric illumination of the sensor pupil. Interferometric principles based on the evaluation of the optical path difference, e.g., SI, have proven to be robust against self-imaging. The disadvantage of SI is its measurement range, which is limited by the depth of focus. Hence, the usable NA and the lateral resolution are restricted. Chromatic-confocal spectral interferometry (CCSI) is a combination of SI and CCM, which overcomes these restrictions. The increase of robustness of CCSI compared to CCM due to the interferometric gain has been demonstrated before. In this contribution the advantages of CCSI in comparison to CCM concerning self-imaging artifacts will be demonstrated. Therefore, a new phase-evaluation algorithm with higher resolution concerning classical SI-based evaluation algorithms is presented. For the comparison of different sensor systems, a chirp comparison standard is used.

Open-source graphics processing unit-accelerated ray tracer for optical simulation

Abstract. Ray tracing still is the workhorse in optical design and simulation. Its basic principle, propagating light as a set of mutually independent rays, implies a linear dependency of the computational effort and the number of rays involved in the problem. At the same time, the mutual independence of the light rays bears a huge potential for parallelization of the computational load. This potential has recently been recognized in the visualization community, where graphics processing unit (GPU)-accelerated ray tracing is used to render photorealistic images. However, precision requirements in optical simulation are substantially higher than in visualization, and therefore performance results known from visualization cannot be expected to transfer to optical simulation one-to-one. In this contribution, we present an open-source implementation of a GPU-accelerated ray tracer, based on nVidias acceleration engine OptiX, that traces in double precision and exploits the massively parallel architecture of modern graphics cards. We compare its performance to a CPU-based tracer that has been developed in parallel.

Laterally chromatically dispersed, spectrally encoded interferometer

We present a single-shot line sensor based on spectral interferometry. Light of a broadband laser source is chromatically dispersed by a grating and focused onto a line on the surface such that each focal point on this line is formed by another wavelength. The entire height profile is obtained by applying a phase evaluation algorithm to the registered interference signal, followed by a model-based approach. The sensor concept is finally verified by experimental results.

Optical metrology for process control: modeling and simulation of sensors for a comparison of different measurement principles

To increase the quality of future products and decrease the manufacturing cost at the same time a systematic control of the fabricated objects is necessary. A promising approach for inline quality control of surface and form parameters is the use of optical measurement systems. This is due to the non-destructive nature of the optical measurement techniques. But in the production environment there are many challenges to overcome for optical sensors. Examples are temperature fluctuation, vibrations, fluids on the object surface and rough surfaces. Therefore, it is likely that not all optical measurement methods are suitable for that task. Hence, a classification of the different principles is necessary with the objective to identify the most appropriate measurement approach for a particular inspection task. In this contribution we start with a systematic approach for a review of sensors within production systems. Then we concentrate on the most robust class of optical sensors, the point sensors. In order to minimize the effect of mechanical vibrations it is desirable to employ measurement techniques that are able to measure the height of an object point in a very short time. Therefore, we focus in this work on chromatic-confocal microscopy and spectral interferometry. The aim is to compare these measurement methods for their ability to cope with the challenges given by the production environment in general. To this end we will develop simulation models for the mentioned techniques and compare two exemplarily sensors for their capability to be used for process control.

Model-based assistance system for confocal measurements of rough surfaces

Confocal sensors are well established in optical surface metrology and their performance has been thoroughly studied both experimentally and theoretically. However most of the theoretical work has been based upon the assumption of locally flat or point like measurement objects. As confocal sensors have become increasingly popular in industrial inspection of rough surfaces in recent years, severe measurement artifacts have been observed in certain situations. The physical reason for these artifacts was not fully understood and therefore a systematic procedure to choose a set of sensor parameters, that minimizes the impact of these artifacts, has been missing. In fact planning measurements of rough surfaces has been a formidable task that even highly experienced experts approached on a trial and error basis. To make things even worse, different confocal measurement systems, e.g. from different manufacturers, and different sensor parameters, e.g. different numerical aperture objectives, typically give substantially differing results. A reliable interpretation of these results let alone a sound judgement of the remaining uncertainty in the measurement results is very difficult. Starting from a quick review of a recently developed signal model, we therefore present an attempt to systematically guide the user of confocal sensors through the planning of an inspection task. In order to support our proposal, we present measurements of two roughness calibration standards, that were conducted with varying numerical aperture objectives on a custom build confocal microscope with rotating micro lenses. The uncertainty in these measurements is then compared to the predictions of our assistance system.

Model-based approach for planning and evaluation of confocal measurements of rough surfaces

Confocal sensors, along with white light interferometers, have been considered promising candidates for replacing traditional tactile sensors in the inspection of rough surfaces for several years now. However, despite their obvious advantages concerning measurement speed and a non-contact working principle, their acceptance for routine measurement tasks in industry is still unsatisfactory. This is presumably because the mechanisms that limit the resolution of such sensors are more complex than those of tactile sensors. The planning for the measurements, e.g. choosing the proper objective lens, is still a very challenging task. Therefore, many confocal measurements produce erroneous results because of improperly chosen sensor parameters. At the same time, the behavior of improperly configured sensors is not well understood, rendering the recognition of erroneous measurements a hard task as well. Current national as well as international technical standards suggest an approach based on the spatial frequencies of the surface profile and a cut-off frequency that is characteristic for the sensor. While this is in perfect analogy to optical two-dimensional imaging devices, we present theoretical as well as experimental reasoning against this approach. Instead, we propose an approach based on local surface curvatures and the radii of curvature of the illuminating wavefronts.

Object Depending Measurement Uncertainty of Confocal Sensors

Confocal gating is a very efficient tool to restrict light reaching the detector of an optical sensor to a certain volume. In fact, it was invented to facilitate imaging inside biological, strongly scattering media [1]. From there on a number of different sensor implementations relying on confocal gating were developed and so called confocal microscopes have been used extensively in many kinds of applications [2]. Meanwhile their popularity is so high, that confocal microscopy has been put on Natures list of milestones in light microscopy [3]. In classic confocal sensors, imaging is achieved by mechanically scanning the measurement object in axial direction relative to one or several optical measurement spots in parallel [4, 5]. This mechanical, axial scan has been replaced later by a spectral distribution of the measurement spot in axial direction both for point sensors [6] and areal sensor configurations [7]. More recently this chromatic confocal measurement principle has been combined with spectral interferometry in order to achieve a more constant lateral resolution over the measurement range while retaining an interferometric axial resolution [8]. Also it was realized recently that even in measurement systems, that utilize coherence gating [9], confocal gating is inherent and has to be taken into account [10]. Furthermore, it was shown that by properly aligning the properties of the coherence and the confocal gating, resolving power of coherence gating sensors can be improved [11].

Object depending artifacts in confocal measurements

Confocal sensors are well established in optical surface metrology. Especially when measuring rough surfaces, their robustness is widely appreciated. However, it was shown lately that certain object features can produce severe artifacts in confocal measurements that are hard to identify as false measurements. Experimental evidence of these artifacts is given with a measurement of a suitable surface conducted with a chromatic confocal point sensor. Furthermore various simulations are presented that identify a self-imaging property of the surface features as the root of the artifacts. These simulations also pave the way to a more precise yet still intuitive signal model for confocal measurements.

GPU-accelerated ray tracing for optical simulation and design

Using an open source tool based on nVidias OptiX engine, the applicability of GPU-accelerated ray tracing to optical simulation is investigated for a variety of applications including coherent ray tracing and nonsequential stray light analysis.