DoE for metrology settings

Richard Thelen (a), Joachim Schulz (a), Pascal Meyer (a), Volker Sailea (a)
(a) Institute for Microstructure Technology, Research Centre Karlsruhe, 76646 Eggenstein, Germany

Abstract

Reproducibility and precision of LIGA structures has been claimed in many publications, founded mainly on brilliant pictures. Because of the poor accessibility to the sidewalls many publications are based on surface measurements without including information about z depending aspects [1] and focus on reproducibility as measured close to the top.
Often this neglects operator’s influence, short time and long time reproducibility, environmental effects on the CMM and others. Tactile optical metrology might help to overcome 2D measurements. Repeatability of tactile optical metrology at IMT was proven to be less than 0,3 μm over some months using ultra fine probes with less than 25 μm diameter. In addition DoE was used to determine the minimum deviation for best possible machine settings. Standard Deviation between 50 and 30 nm was measured. Compared to that, uncertainty remains about 1-2 μm for 3D measurements even with z maximum restricted to 1 mm [2]. Not enough to measure sub-μm product variation that is a typical benefit of LIGA products.
Investigations were started at the Research Centre Karlsruhe to find out more about the effects influencing the measurements to explain why repeatability and capability do not match. Interaction between sample and sensor was the main reason. This was simulated and the results were used to reduce the uncertainty of the system. IMT elaborated a new strategy that improves the capability of a coordinate measurement machine CMM with tactile optical sensor for LIGA parts with sub μm variation.