Development and validation of signal processing techniques for quality control using optical vibration measurements

Project Details


In many mechanical engineering applications the vibration behavior at a large number of locations on a structure is wanted. This is in particular the case when using vibration mode shapes to detect damage in structures. Indeed, small defects (like delaminations and cracks) can give rise to localized changes in the vibration pattern. An ideal instrument to perform this task is the scanning laser Doppler vibrometer (SLDV). Because of a number of difficulties, which appear when measuring and processing SLDV signals in industrial environments, applications of the scanning vibrometer are mainly limited to laboratory conditions.
The most important restrictions that limit a more general application of the SLDV are :
- too much (expert) user interaction is needed both to perform the measurement and to process them
- nontreated dark (nonreflecting) surfaces give rise to noisy data with outliers. Classical processing algorithms (e.g. modal
parameter identification algorithms) cannot efficiently handle this poor quality data.
- the measurement time can be excessively high when a large number of spatial locations are to be measured.
- the computational effort to process the large amount of data is very high
The aim of the project is to develop signal processing techniques which solve these problems
Effective start/end date1/01/0431/12/07


  • Laser vibrometrie
  • Trillingsanalyse
  • Systeemidentificatie

Flemish discipline codes

  • Mechanical and manufacturing engineering