Performances of Feature Tracking in Turbulent boundary layer investigation (Contributo in atti di convegno)

Type
Label
  • Performances of Feature Tracking in Turbulent boundary layer investigation (Contributo in atti di convegno) (literal)
Anno
  • 2007-01-01T00:00:00+01:00 (literal)
Alternative label
  • M. Miozzi, B. Jacob, A. Olivieri (2007)
    Performances of Feature Tracking in Turbulent boundary layer investigation
    in 7th International Symposium on Particle Image Velocimetry, Rome, Italy, September 11-14, 2007
    (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#autori
  • M. Miozzi, B. Jacob, A. Olivieri (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#pagineTotali
  • 12 (literal)
Note
  • PuM (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#affiliazioni
  • CNR-INSEAN, via di Vallerano 139, 00128 Rome, Italy (literal)
Titolo
  • Performances of Feature Tracking in Turbulent boundary layer investigation (literal)
Abstract
  • The purpose of this paper is to assess the performances of YATS, a Feature Tracking algorithm (Miozzi, 2004), by discussing results obtained from turbulent boundary layer data at moderate Reynolds number (R? = 3200), in the framework of a wider project on drag reducing flows (Olivieri et al., 2005). Propaedeutic tests have been performed on synthetic images in order to characterize the accuracy of the algorithm in terms of bias and rms errors (Miozzi, 2007). YATS is a time-resolved, correlation-based tracking software that solves the optical flow equation in a local framework (Lukas and Kanade, 1981). The algorithm defines its best correlation measure as the minimum of the Sum of Squared Differences (SSD) of intensity values of pixels between the interrogation windows in two consecutive frames. The SSD minimization problem is iteratively solved after linearization, in a least-square approach, by adopting in consecutive steps two different models of motion. In the first step, raw displacement is extracted by imposing a pure translational window motion. In the second step, displacement is refined by allowing an affine window deformation, in which first order accurate image deformation parameters are given directly by the algorithm solution (Miozzi, 2005). Velocity computation is performed only where the solution of YATS linear system exists, i.e. where image intensity gradients are not zero both in x and y directions. This approach maximizes the signal-to-noise ratio and enables the algorithm to investigate challenging situations, like wave impacts (Lugni et al., 2006). In-plane loss-of-pairs is greatly reduced by adopting a pyramidal image representation. Spatial highdensity velocity and velocity gradients are obtained, in a lagrangian fashion, along the trajectory of each feature. The influence of image interpolation in sub-pixel analysis has been tested for classical bicubic method and for BSpline interpolation in the context of generalized interpolation (Thévenaz et al., 2000) of degree 3 and 5. Mean and turbulent statistics distribution have been also evaluated by applying a logical mask, resulting in a better near-wall resolution. The excellent agreement of the results with literature data obtained by means of standard techniques (LDA & HWA) validates the algorithm accuracy. BSpline interpolation scheme is found to better perform in the evaluation of turbulent Reynolds stress, which is underestimate by bicubic classical scheme. (literal)
Prodotto di
Autore CNR

Incoming links:


Prodotto
Autore CNR di
data.CNR.it