Logo of the Physikalisch-Technische Bundesanstalt
Icon des Fachgebiets 5.10

Complex Measurement Data Analysis in Precision Engineering

Unit 5.10

Profile

This unit supports Division 5 in modeling and analyses of measurement data.

To top

Research/Development

Simulation of Electron Scattering Processes

Electron scatter processes in solids are currently simulated to describe signals of scanning electron microscopy. Synthetically generated electron microscopic images support the quantitative analyses and the estimation of uncertainties in nano particle size. The transmission of electrons through carbon has been simulated for the EMPIR project Black Carbon and for EMPIR project nPSize, the transmission of electrons through gold and titanium oxide has been determined. Titanium oxide does not belong to the materials with inelastic scatter cross sections being generally available. At PTB, an empiric approximation of the double differential cross sections has been carried out, as well for polystyrene. An important application is the calibration of aerosole particle measuring instruments using polystyrene spheres as reference particles. Their sizes are calibrated using electron microscopy. The Opens external link in new windowdetermination of polystyrene particle sizes is based on Monte Carlo simulations using cross sections accordingly. The results have been verified by comparisons with data of international comparisons.

Characterization of Topography Measuring Instruments

To characterize the transfer behavior of topography measuring instruments, in particular 3d microscopes, material measures are used that reveal how the height of structures of varying width is seen, as well as their slopes and curvatures. Currently a large amount of investigations is performed within the framework of the EMPIR project Opens external link in new windowTracOptic being coordinated by Working Group 5.14. Algorithms to analyze cosine waves of varying wave length (Opens external link in new windowchirped sequences with constant wavelength for each single wave) and Initiates file downloadto analyze line structures as 1d gratings of rectangular shape are developed by unit 5.10. Furthermore, 5.10 develops the algorithms to analyze areal roughness and aberration errors.

Determination of Functional Features of Surfaces

Measurement methods and data processing procedures to determine surface features (shape, waviness, roughness) relevant for the function of machine parts are under development. For example, aOpens external link in new window gearing roughness standard has been developed within a cooperation between industry partners, a measurement instrument manufacturer and a manufacturer of gears, as well as departments 5.3 Coordinate Metrology, 5.5 Manufacturing Technology and 5.1 Surface Metrology. This is a test artefact representing a gear with differently textured teeth, each tooth representing a roughness standard. Unit 5.10 supports this project by contributing evaluation methods for filtration and determination of roughness parameters.

To top

Services

Development and Implementations for Calibration Tasks

  • The implementation of the analysis methods of standards on filtration and on the determination of roughness parameters being developed in ISO/TC 213 continuously accompanies the standardization work. The roughness profile analysis software RPTB (https://www.ptb.de/rptb) (Opens external link in new windowwww.ptb.de/rptb) used in PTB working group 5.15 for routine calibration services and also available online is maintained and continued. The new filtration methods and furthermore a new, unambiguous method to determine the parameters Rsm and Rc has been updated in accordance with the newly published standard ISO 21920-2. Furthermore, the implementation of profile element based parameters such as Rz, Rp, Rv has been renewed. A series of parameters being newly specified within ISO 21920-2, such as the autocorrelation length Ral, have been added. The RPTB implementation of all parameters has been validated using Opens external link in new windowsoftgauges by performing a comparison with partly independent implementations of partners. An article on this has been submitted to Metrologia, a draft is Initiates file downloadhere.
     
  • Automation of calibration processes:  Development of software integrating the analysis of measurement data, determination of measurement uncertainties til finally producing calibration certificates without the necessity to transfer any data by hand.

     
  • To analyse areal topography data, in particular those of PTB working group 5.14, implementations of filtration algorithms as well as algorithms to determine parameters specified in ISO 25178-2 and to evaluate power spectral densities have been carried out. Further steps are necessary to improve documentation of the functions and facilitate the use of the routines for many users.

Lectures at TU Braunschweig and IGSM

Dorothee Hüser and Gerd Ehret, Head of Department Opens external link in new window4.2 Imaging and Wave Optics and Head of Working Group 4.22 Flatness Metrology, give lectures on Messdatenauswertung und Messunsicherheit (Data Analysis and Measurement Uncertainty) in German. This is within the scope of Metrologieinitiative Braunschweig at the Institut für Produktionsmesstechnik of the Technical University.

The lecture notes, set with LaTeX, are being developed continously and are available as open source project on the public repository github. This delivers an opportunity to the students and other interested readers to contribute to the manuscript, to improve its content, and to optimize the didactics of the presentation of the contents.

A summary of the principal concepts of the propagation of measurement uncertainty is provided by a presentation of three different approaches: (1) propagation for explicite, univariate indirect measurands representable by linearizable models, (2) Monte-Carlo methods for any kind of models with the assumption that the model by itself has no uncertainty, (3) Utilization of Bayesian statistics with the possibility of having a model prior with finite sized width and having an increase of information by successive measurement campaigns. Two lectures on this topic were given during the summer school of IGSM: Initiates file downloadlecture 1: Review on methods to estimate measurement uncertainties - the various concepts of the "GUM" and its supplements; Initiates file downloadlecture 2: Illustration of three different approaches to estimate measurement uncertainties.

To top