Implementation and Evaluation of Optical and Stylus Based Profiling Techniques for Surface Metrology
Analytics
74 views ◎26 downloads ⇓
Abstract
A unifying theme of this thesis is the implementation and characterization of point probes for surface metrology. The implementations include two optical non-contact profiling methods; fiber-based Fabry-Perot interferometry, and confocal microscopy. And a stylus profilometer is metrologically characterized to evaluate its suitability in measuring form of optically smooth surfaces.Fiber-based Fabry-Perot interferometers are typically used to measure displacement in confined spaces. Even though this technique is suitable for measuring surface texture, literature describing its implementation and limitations in its use as a surface profiler are scarce. Therefore, this work attempts to understand the issues with experimental implementation of this technique and the limitations of using a bare fiber probe tip for surface profiling. As outcomes of this part of the research, experimental setup is built and the feasibility in surface height measurements is validated by measuring two sinusoidal reference surfaces with heights of 1 μm and 1.5 μm and wavelengths of 100 μm and 50 μm, respectively. The second part of this thesis is to implement a flexure-based oscillation of an objective lens stack used for confocal microscopy to determine a localized surface height. Current practices include the oscillation of the surface sample, oscillation of the casing that houses the objective lens stack, and MEMS (micro-electromechanical systems) based scanners. The first two approaches limit the achievable bandwidth of oscillation due to the mass of the objects being oscillated, and the latter approach is expensive. Using a flexure mechanism to oscillate the optical lens elements has the potential to increase the bandwidth of oscillation and therefore surface scanning speeds. Since the spot size produced by the objective lens determines the spatial resolution of the measurement, spot size is measured for 40X and 60X commercial objective lenses illuminated by laser and mLED (micro-Light Emitting Diode) light sources. A sinusoidal reference sample is designed and manufactured, which is later measured using a confocal microscope prototype built using a 60X objective lens. A flexure to house the lens stack to provide a surface height scanning range of 10 μm is also designed and fabricated. The third part of this thesis involves characterizing a stylus-based contact profiler for measuring areal form of freeform optics. Typically, for non-contact point probes, the optical axis must be orthogonal to the curvature of the part being measured (probe normal) to reduce errors in the measurement. Additional carriages are required to generate a probe normal scanning path. For many freeform optics, interferometric surface measurement techniques such as Fizeau interferometry requires generating a reference surface for each design prescription. Since the stylus profilers do not have these disadvantages, a commercial stylus profiler (Mahr LD260) is characterized by measuring reference objects such as optical flats, prisms, and spheres. The measurement results are compared with the measurements of these reference objects using a Fizeau interferometer. As a result, the vertical error of the X scanning carriage, side loading on the stylus probe due to the surface slope of the part being measured, and other carriage errors are estimated. A geometrical model of the system including the errors in the motion of carriages is developed, and a Monte Carlo simulation is performed to evaluate the uncertainty in the areal form measurements. The simulation resulted in a measurement uncertainty of less than 200 nm PV for a 100 mm measurement aperture.