Optimization of Illumination Systems using TracePro RAYTRACING OPTIMIZES ILLUMINATION An optimization module for the raytracing tool ›TracePro‹ now brings a desired feature to designers of illumination systems: Non-linear optimization. It allows users to optimize on the intensity and/or uniformity parameters of illumination systems. Its macro language approach gives the user the flexibility to modify the analysis functions according to any user's requirements. An optimization module for the raytracing tool ›TracePro‹ now brings a desired feature to designers of illumination systems: Non-linear optimization. It allows users to optimize on the intensity and/or uniformity parameters of illumination systems. Its macro language approach gives the user the flexibility to modify the analysis functions according to any user's requirements. NORBERT HARENDT  he design of illumination systems can be based on a wide range of diverse requirements, but often the most important requirements are maximizing intensity and/or uniformity of light distribution. Since the light sources are becoming smaller and more powerful, the ray forming elements need to follow the same trend and become smaller in size while better facilitating conformity to the required system performance, e.g. light intensity distribution profile. Basically, the system performance requirements are either the maximization and/or improved uniformity of the light intensity distribution on a specified surface or, recently, more frequently requested specific light distribution patterns. The illumination system setup can achieve a different complexity level. On one side, it can be a linear setup comprising of a light source, a ray shaping element and a detector surface. While these systems can be designed containing only one light source, a beam forming element and the specified detector surface, the systems can also contain several light sources, crossed imaging and illumination axis and irregular ray shaping elements or combination elements that are to form the desired illumination pattern. Until the recent introduction of the optimization module, the optics designers were left only with powerful analysis tools and their own experiance. The usual practice was to simulate and analyze the requirements of a potential design setup in raytracing software. The optics designer had the possibility to manually examine the changes in the model setup in order to achieve a better system performance.  The classical optical design software has had an automated optimization for years, which colleagues on the illumination side could only envy, while there was no possibility of including automated performance improvement of the system inside of the non-sequential raytracing software. The underlying variability and complexity of the illumination systems design have hindered the integration of optimization algorithms in the non-sequential raytracing software. The implementation of the optimization module has lead to improved software utilization and more importantly to time saving while the automated optimization of a system performance can now be done in a fraction of the time compared to the manual approach used in the past. The advantage of the newly introduced optimization module is thus obvious: with the shortening of the development time comes increase in productivity. Flexible definition of evaluation criteria  Times have changed and IB/E has developed and introduced a non-linear optimization module for TracePro. Now even the designers of illumination systems can utilize the above stated optimization advantages. The complete integration of the optimization module within TracePro means that the module can utilize any powerful TracePro feature. The module's default setup is an optimization of intensity and/or uniformity of illumination distributions. The main goal in the development of IB/E's module was to maintain optimization versatility. Therefore, the module uses Scheme, TracePro's powerful macro language, and allows users the flexibility to define users' own evaluation criteria. The requested specific illumination distributions can be achieved without any restriction. In addition, the module fulfills more requirements from the optics designer wish list, available are e.g. an arbitrary number of optimizations and the automatic save feature, which saves all the intermediate systems showing an increase in system performance. Finally, the resume feature is available, in case an unexpected breakdown of the system occurs during the analysis. If the operating system does not support the requirements of long and calculation extensive runs, the module allow the user to continue the optimization run exactly at the point where the breakdown has occurred. Application: condenser for fiberoptic light guide  The functionality of the software is clearly demonstrated on the following theoretical example. The basic setup is as follows: a Lambertian emitter with a diameter of 2.0mm is to be focused into a fiberoptic light guide with diameter of 8.0 mm and length of 100.0 mm. The basic approach was carried out with a plano-convex spherical condenser: the realization of a classical imaging of the surfaces into each other will lead to a condenser with 10% performance, which means that 10% of the emitted rays from the light source will be sampled. To quantify this properly, the detection surface was defined as the exit surface on the light guide in order to collect only the propagated light, not the incident one (figure 1). The basic optimization parameters concentrate on the lens and its position between the emitter and light guide. By switching from plano-convex to an aspherical lens, the basic setup leads to 3.5- fold increase in light intensity. With the intensity optimized to the maximum, the irradiance distribution shows significant loss in light intensity near the light guide edges. The approximate 60% of intensity decay in the edges shown in figure 2 is not suitable for every application. Therefore we have switched the weights from the intensity to the uniformity optimization. Resultant is a condenser, which shows a loss of less than 10% in uniformity with nearly the same level of light intensity (figure 3) Practical Examples  In-line illumination integrated in imaging optics bares the risk of significant loss of imaging quality, in particular with high-quality objectives which have been designed diffractionlimited. An example is the family of telecentric measuring objectives TZ5 of IB/E (figure 4). Scalability of the reproduction scale together with a very compact mechanical design demand for a precise and unique location of fiber cuppling, otherwise drops in quality regarding homogenity and intensity of the in-line illumination will occur with some variants of that family. To find a solution to this problem, the merit function to simultaneously evaluate intensity and contrast is required. The optimization module provides flexibility for the user to define these custom merit functions. In our case, we have used a chessboard like grid of black and white squares as a target. The quality of the image of that target is thus described by the quality of the generated black and white squares in the image region. The optimization leads to a unique position for the in-line illumination coupling for the whole family of the compact telecentric measuring optics (figure 5). Creating a light distribution Creation of a specific light distribution often results in increased expenses due to the detailed and repetitious analysis of the formed irradiance distribution, but not with the non-linear optimization module for TracePro. Arbitrary detection surfaces can be defined within the object space. The irradiance distribution to be achieved within the object space can be defined using arrays or functions. One example is the lens array comprised of 14 biaspherical lenses embedded in one plastic device, each of them acting as a light guide for 23 VCSEL elements (figure 6). The optimization module is the main tool, which made the design of such a solution possible. Resume: Flexible optimization The non-linear optimization module for TracePro developed by IB/E makes the optimization of illumination systems possible in TracePro. The combination with a powerful raytracing tool like TracePro enables effective design and analysis in conjunction with an automatic system improvement. The definition of the system performance within the module gives the user maximum flexibility to choose the right parameters to improve system performance, be it intensity and/or uniformity or any other criteria. AUTHOR Dr. NORBERT HARENDT is Head of Machine Vision and Metrology at IB/E in Hutthurm. KONTAKT Ing.-Büro K. Eckerl IB/E,
D-94116 Hutthurm /Germany,
Tel. +49 /85 05 /32 22,
Fax +49 /85 05 /34 00,
www.ib-eckerl.de
###
|
