Post-processing of numerical and experimental data - hardcover - VKI LS 2008-01

Post-processing of numerical and experimental data
February 11-15, 2008, edited by P. Millan & M.L. Riethmuller

Abstract

Recent advances in numerical simulation capability as well as in measurement techniques have enabled the generation of very large amounts of data in the field of Fluid Mechanics. It has therefore become difficult to extract synthetic or phenomenological information from these large quantities of detailed flow data and specific post-processing techniques must be applied to obtain the benefits of these superior tools.

It is the objective of these Lecture Series proceedings to present a survey of advanced processing methods and techniques that allow extracting physical characteristics of flows from sets of data that represent spatial and temporal distributions containing occasionally random aspects.

The notes are organized into four sections. The first one is an introductory course that presents, as a reminder, the different classes of problems encountered in fluid dynamics and put in evidence the needs of experimental and numerical approaches. In the second section, the nature of the data yield by numerical and experimental techniques are analyzed in view of the intricacy of extracting the information required for the modeling of fluid dynamics problems. Some emphasis are given to the very recent progress made in 3D, time dependent PIV techniques and to the processing of the data yielded by this new advanced measurement technique. The different data processing techniques such as spectral analysis, pattern recognition, Proper Orthogonal Decomposition (POD) and wavelet analysis are discussed in the third section. Finally the last section is devoted to the presentation of examples of application of the various post-processing techniques in different domains of fluid dynamics with an emphasis on the analysis of turbulent flow fields.

The proceedings should prove fruitful to both newcomers to post-processing and experienced researchers who will find here a useful update.

Table of contents

• TROPEA, C. – Technische Universität Darmstadt, Germany
  Introduction to the measurement of turbulence in fluid dynamics
• VEYNANTE, D. – CNRS & ECOLE CENTRALE PARIS, France
  Survey of signal processing techniques
• BOUTIER, A. – ONERA, France
  Classification of laser velocimetry techniques
  Optical techniques for velocity measurements
  Particle seeding
  Velocity measurement accuracy in LDV
  Comparison of the data issued from various laser velocimetry techniques: LDV, PIV, DGV
• TROPEA, C. – Technische Universität Darmstadt, Germany
  Requirements for processing data of random nature in fluid dynamics
  Processing of time-dependent data
• CORDIER, L. & BERGMANN, M. - LEA, France
  Proper orthogonal decomposition: an overview
  Two typical applications of POD : coherent structures eduction and reduced order modelling
• SAGAUT, P. – Université Pierre et Marie Curie, France
  Numerical data post-processing : from validation to physical understanding
• BONNET, J.P. & DELVILLE, J. – LEA, Université de Poitiers, France
  Coherent structures in turbulent flows and numerical simulations approaches
• DELVILLE, J. & BONNET, J.P. – LEA, Université de Poitiers, France
  Two-point correlation in fluid dynamics : POD, LES and related methods
• LAMBALLAIS, E. & BONNET, J.P. – LEA, Université de Poitiers, France
  DNS/LES data processing and its relation with experiments
• SCARANO, F. – Delft University of Technology, The Netherlands
  Three-dimensional velocity measurements by tomographic PIV
  Three-dimensional velocity measurements by tomographic PIV. Part II – Application to turbulent shear flows
• SCARANO, F. & VAN OUDHEUSDEN, B.W. – Delft University of Technology, The Netherlands
  Post-processing of time-resolved PIV data unsteady aerodynamic forces and planar pressure imaging
• BILKA, M. & RAMBAUD, P. – von Karman Institute for Fluid Dynamics, Belgium
  Application and limitation of wavelet transform and POD to vertical flows