Session X13: Turbulence: Measurements (10:45am - 11:30am CST)

Time-resolved tomographic PIV measurements of a turbulent shear layer flow impinging on a cavity trailing corner.

Time-resolved tomographic PIV measurement results of three-dimensional velocity and pressure fields for a turbulent shear layer flow over an open cavity at a Reynolds number of 40,000 will be presented. The objective of the research is to investigate the pressure-related turbulence transport phenomena, with a focus on the characterization of the intercomponent turbulence fluctuation energy transfer carried out by the pressure--rate-of-strain tensor in both the shear layer and the impingement regions around the cavity trailing corner. A total of 152,000 tomo-PIV images at a sample rate of 4496 frames per second has been acquired at each measurement station. The pressure is reconstructed from the measured pressure gradient using the parallel ray omni-directional integration method. To ensure the quality of the measurements, the curl-free property of the pressure gradient is used to examine the quality of the measured pressure gradient, and the continuity equation is used to examine the quality of the velocity measurement. The quality of the measured pressure-related terms is also cross-checked with the balance of the Reynolds stress transport budget. With the evaluation of the three-dimensional pressure-related turbulence transport terms enabled by the tomo-PIV measurements, the conjectures raised by Liu and Katz (2018, https://doi.org/10.2514/1.J056168) regarding the magnitude of the spanwise intercomponent energy transfer based on their planar PIV data will be verified.   

PIV Measurements in the Princeton Superpipe

A stereo-PIV system for the Princeton Superpipe facility is designed and tested. The Superpipe uses air at pressures up to 230 bar to study turbulent flows at Reynolds numbers, based on diameter, ranging from 30 x 10\textasciicircum 3 to 35 x 10\textasciicircum 6. The pipe has a diameter of D $=$ 129 mm, a length of 202D, and is contained within the pressurized system. The main challenge for PIV is gaining optical access in this experimental arrangement and to reduce effects of changes in index of refraction. A ray-tracing algorithm was first employed to design the stereo imaging system. It was decided to use mirrors to image the flow with telephoto lenses placed outside the pressure vessel. This configuration avoids the complexity arising from changing refractive index and puts the cameras outside the high-pressure environment. In addition, a high-pressure particle seeding system with Laskin nozzles is developed, and the size distribution of DEHS droplets at up to 200 atm are characterized with a Phase Doppler Anemometry system. The PIV system is used to collect measurements of fully developed pipe flow at various Reynolds numbers. These data are then validated against previous measurements in the Superpipe using NSTAP, a MEMS velocity sensor.   

Exciting atmospheric turbulence on lab-scales by an active grid

Wind tunnel investigations are an important tool for studying flow phenomena and effects on different objects. Therefore, it is crucial to create realistic turbulence in the wind tunnel. Atmospheric turbulence is known to exhibit large integral scales and Reynolds numbers, which are hard to reproduce in a wind tunnel. In this investigation turbulence is generated by an active grid. The active grid shafts are driven by a stochastic process, which is keeping the global blockage constant. The flow is additionally excited by a dynamic variation of the wind tunnel fan speed (also based on a stochastic process). This broad band excitation allows for the generation of a longitudinal integral length scale much larger than the transverse dimension of the wind tunnel ($>$100m) and a four-decade inertial range. The generated turbulence behaves like a slice cut out of a much-larger-scale turbulence. Even though deviations occur on the large scales, on the small scales both longitudinal and transversal components behave in the same way and as if both are part of the same turbulent flow, forced at very large scales. The integral-scale Reynolds number measured in the flow is of the order of $2.2\cdot10^7$, thereby allowing for investigations under realistic atmospheric-like conditions.   

A Novel Cross-Wire Anemometry Data Reduction Method

The high temporal resolution and multiple-component capability of hot-wire anemometry has made it a mainstay of unsteady velocity measurement techniques for more than 70 years. Data-reduction techniques for cross-wire measurements, such as the various table methods, have been used for their accuracy in unsteady flows. These methods require interpolation for each data point which, coupled with high frequency acquisition over long sampling periods, makes these methods computationally intensive. A new method for determining velocity from cross-wire voltages has been developed to process measurements of grid turbulence in a convergent section. As with other methods, calibration data taken at known velocities and angles is required for the approach. The new method uses the calibration data to map the measured voltages to a three-dimensional surface that represents the flow’s velocity. While the development of the surface takes time, the processing time for each data point is reduced. To evaluate the approach, the speed and accuracy of the results is compared to those determined using other cross-wire data-reduction approaches.