טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentSabban Lilach
SubjectExperimental Study on the Interaction between Rigid, Heavy
Fibres and Homogeneous, Isotropic Turbulence
DepartmentDepartment of Mechanical Engineering
Supervisor Dr. Rene Van Hout
Full Thesis textFull thesis text - English Version


Abstract

The interaction between fibres and turbulence is important in both industrial and environmental application. Therefore, time resolved, planar particle image velocimetry (TR-PIV) and two-orthogonal view, digital holographic cinematography were used to measure fibre flow interaction and 3D fibre trajectories and orientation dynamics in near homogeneous isotropic air turbulence (HIT). The TR-PIV spatial measurement resolution was of the order of the Kolmogorov length scale and the acquisition frequency was 3 kHz, about six times the Kolmogorov frequency. The Kolmogorov length scale describes the size of the smallest, dissipative scales of turbulence. In the holography measurements the temporal measurement resolution was 2 kHz, about four times the Kolmogorov frequency. The two-view setup significantly reduced processing time since the holograms, acquired by each camera, were digitally reconstructed without determining the “accurate” in-focus position of each fibre.

Two sets of nylon fibres having nominal lengths, 0.5mm (about 3 times the Kolmogorov length scale), and diameters of 13.7 (dtex1.7) and 19.1mm (dtex3.3), were released in a turbulence chamber at a volume fraction of 2?10-6. Eight woofers mounted on the chamber’s corners generated HIT (Taylor scale Reynolds number, Reλ ≈ 130). The fibre Stokes numbers, a non -dimensional number characterizing the particle response to turbulent structures, were of order one. Here, for the first time, the effect of fibre inertia on particle-turbulence coupling was measured in HIT.

Flow characteristics for both unladen and fibre laden cases indicated that the flow was nearly homogenous and isotropic. The number of tracked fibres in the TR-PIV measurements was relatively low due to the highly three dimensional flow field whereas in the holographic cinematography more and much longer tracks were obtained. Therefore, probability density functions (pdf’s) of fibre velocities (based on the TR-PIV data) were more scattered compared to those obtained from the holographic cinematography. However, since the pdf’s based on the TR-PIV and the holography measurements collapsed and were normally distributed, it can be concluded that the TR-PIV measurements sampled the full fibre velocity range. Pdf’s of fibre and air velocities sampled at the fibre centroids also collapsed while instantaneous, relative velocities were distributed symmetrically around zero. Fibre orientations were random as a result of the strong turbulence. However, both dtex1.7 and 3.3 fibres were preferentially located at the periphery of vortices, more so for dtex1.7 fibres that had the lowest Stokes number. In addition, the drag force acting on the fibre was minimized as the magnitude of the relative velocity was minimized and its direction aligned with the fibre major axis. In contrast to fibre translational velocities, distributions of fibre rotation rates were much more peaked. Increasing fibre inertia was positively correlated with a lower decay rate of the Lagrangian, fibre velocity auto-correlation. Furthermore, as a result of inertia, normalized, mean squared fibre rotation rates were lower than for neutrally buoyant fibres having the same aspect ratio. Finally, results indicated turbulence augmentation and turbulent kinetic energy spectra hinted at energy transfer from large turbulence scales to scales in the range of 2 to 7 times the fibre’s length.