Design and validation of advanced signal processing techniques for processing optical measurements.

  • Vanherzeele, Joris, (Administrative Promotor)

Project Details


1. Aim Both single point as full-field optical measurement techniques like for instance the scanning laser Doppler vibrometer and laser Doppler anemometer, Moiré interferometry and digital image correlation provide an important source of information in many application areas: micro-electronics, materials science, quality caontrol, biomedical sciences etc. However processing this optical data can prove tricky: the data size can be quite vast, and the measurement quality dan typically be rather low for complex applications and/or objects with low reflective properties, which in turn can increase measurement time [1]. During my Ph.D it was demonstrated that processing optical measurement techniques can be done efficiently in the frequency domain using parametric identification techniques [2]. In this project the two-dimensional frequency domain system identification technique, that was developed during my Ph.D will be elaborated to tackle techniques in different application fields: - The processing of Moiré (FTP) fringe patterns. - Digital image correlation (DIC) for the measurement of velocity fields (fluid flow) and strain fields (solids). - Laser Doppler Anemometry (LDA) measurements of periodic phenomena such as acoustic fields or vortex flows in object wakes. 2. Objectives The development of a 2-D system identification algorithm to estimate of the global frequency, phase and amplitude content of a sinusoidal fringe pattern. This base algorithm will be applied in different application fields: - The estimation of the local 2-D delay between two correlated images; phase delay (to be applied in the 'Digital Image Correlation, DIC' and the 'Particle Image Velocimetry, PIV' techniques). - The 1D basic algorithm will be used to synthesize the often irregularly sampled data from LDAmeasurements. This irregular sampling can pose a problem in very strong acoustic fields and fluid flows in general, because the displacement of the tracer particles is larger and therefore seeding is not present at all times in the measurement volume. Furthermore, experiments will be conducted to illustrate the broad applicability of the developed algorithms: - PIV fluid flow measurements (e.g. of the velocity profile around a cylinder in a wind tunnel as this well documented in literature). - Crack propagation measurements of airplane components and composites during fatigue test using DIC. - Profilometry measurement of micro-electromechanical structures using fringe projection methods (FTP). - LDA measurements (e.g. wake behind a cylinder in a uniform flow and sinusoidal acoustic fields).
Effective start/end date1/10/0730/11/08

Flemish discipline codes

  • Other engineering and technology