H. Eloranta

Experiences of turbulence measurement with PIV

The application of particle image velocimetry to turbulence measurement is described. An analysis of physically necessary spatial resolution is presented by using a model spectrum function. A comparison is made with hot-wire anemometry. Some aspects of spatial filtering and two-dimensional sampling are presented with a comparison to large-eddy simulation. The estimation of time mean turbulence quantities from the measured vector fields in a laboratory mixer is used as an example. The measurement results for two different image sizes at the same position in the flow field are compared with different interrogation area sizes. The assumed dependence of the velocity field on the interrogation area size could not be confirmed. The image size seems to produce dependence in all estimated quantities. The measurement errors are critical for the achieved results.

2D Spectral and Turbulence Length Scale Estimation with PIV

A new method to apply spatial two-dimensional power spectral density (2D PSD) analysis to the data measured with Particle Image Velocimetry (PIV) has been introduced. Applying the method to a set of the velocity vector fields characteristic turbulence length scales can be estimated. In this method the computation of 2D PSD has been performed to two kinds of pre-processed data. In the first set, the local average has been spatially subtracted (Spatial decomposition) and in the second set the time-average has been subtracted (Reynolds decomposition). In the computation of 2D PSD the 2D FFT with the variance scaling has been used.Besides 2D spectral analysis this paper uses the distribution analysis of the various turbulence quantities and a structure analysis method to estimate the dimensions of coherent structures in the flow. Another method to analyse turbulence length scales is the estimation of the spatial 2D Auto Correlation Coefficient Function (2D ACCF). All these methods applied side by side to the PIV data increase the understanding of the turbulence, its scales and the nature of the coherent structures.

Effect of Geometrical Parameters on Vortex-Induced Vibration of a Splitter Plate

An experimental study on the effects of various geometrical parameters to the characteristics of vortex-induced vibration (VIV) of a splitter plate is presented. The dynamic response of the fluid-structure system was measured using particle image velocimetry and laser telemetry simultaneously. Combined data of these techniques allow the assessment of the variation in the VIV response due to geometrical parameters, such as channel geometry, aspect ratio (AR), and trailing-edge thickness (d) as well as the imprint of the excited vibration mode on the flow. The effects of AR and d were both investigated with three different plate geometries and the effect of channel convergence was studied with a single plate geometry. Measurements were performed over a range of Reynolds numbers (Re). The results show that the vibrational response of the combined fluid-structure system is affected by the VIV instability in all cases. Within the measured Re range, a characteristic stepwise behavior of the frequency of the dominant vibration mode is observed. This behavior is explained by the synchronization between the vortex shedding frequency (f 0 ) and a natural frequency (f N ) of the fluid-structure system. The results further indicate that this response is modified by geometrical parameters. Channel convergence, i.e., flow acceleration, enhances the vortex shedding, which, in turn, increases the excitation level leading to stronger VIV. Channel convergence does not have a significant effect on f 0 or on the dimensionless vibration amplitude (A/d). An increase of both the number of excited f N s and the level of synchronization was observed with the lowest AR case. The results also suggest that d is the dominant geometrical parameter. It reduces both the Aid of the plate and the number of synchronization regions. This stronger effect on the response of the VIV system is due to the direct effect of d on the excitation mechanism.

Turbulence Control With Particle Image Velocimetry in a Backward-Facing Step

A new concept of particle image velocimetry (PIV) for turbulence control is introduced. The PIV method is used to measure the flow and supply information on the turbulence quantities on-line for a control station connected to an actuator for flow manipulation. With this closed-loop system of PIV and PID controllers, it is possible to control increase or decrease turbulence quantities or length scales. Special techniques for the on-line measurement are developed. The mean velocity is computed with a moving time average operator and the Reynolds decomposition is applied for the calculation of the Reynolds stresses, velocity fluctuations, or other instantaneous turbulence quantities. The control concept is tested in a backward-facing step with a DC-motor based actuator for mixing of the near wall flow. A length-scale estimate similar to the integral scale and the Reynolds shear stress are calculated on-line