Planar shadow image velocimetry for the analysis of the hydrodynamics in bubbly flows

Abstract

In order to allow for more reliable modelling of microscale processes in turbulent bubbly flows, detailed experiments in a laboratory-scale double loop facility and a bubble column with a diameter of 140 mm were performed. The range of bubble mean diameters considered in the measurements was between 2 and 4 mm. The average gas volume fraction was in the range between 0.5 and 5% for different experiments. To allow simultaneous measurements of bubble size, bubble velocity and liquid velocity field, a combined system of planar shadow imaging, particle tracking velocimetry (PTV) and particle image velocimetry (PIV) for online measurements was developed and applied. The measurements of the liquid phase velocities were realized by seeding the flow with polyamide tracer particles having a mean diameter of 65 μm. A background illumination was established by using an array of 551 LEDs with a size of 160 mm by 100 mm. A double image CCD camera with macro optics was used to record the images of tracers and bubbles simultaneously. Because of the small depth of field of the macro camera optics (<4 mm for the bubbles), it was possible to discriminate between bubbles and tracer particles inside and outside the camera's focal plane using the gradient of grey values. A set of digital image filters was applied to perform phase discrimination between bubbles and tracer particles, i.e. to obtain separate double images for bubbles and tracer, respectively. The contour of in-focus bubbles was determined using an edge detecting Sobel filter and a spline interpolation technique. Thereby, the bubble size, shape and orientation could be derived. The bubble velocity was obtained by applying PTV and the continuous phase velocity field was determined by PIV, using a successive refinement of the interrogation area. By recording and evaluating at least 500 double images, it was possible to determine bubble size distributions and mean as well as fluctuating velocities for both phases. Thereby, detailed data on the hydrodynamics of bubble-driven flows are provided.

Development of an optical imaging technique for particle number density

“…Although previous studies were often conducted in the dense bubble concentration field (e.g. 1 -5 % in Broder and Sommerfeld, 2007), present bubble experiments are conducted in dilute bubble concentration cases with the void fraction of 7×10 -4 %. Therefore, there is few overlapping in microscale and mesoscale setups (see Figs.…”

Characterisation of an analogue liquid for hydrodynamic studies of gas-ionic liquid flows

“…The applicability of Sobel filter to bubbly flow is described by Broder and Sommerfeld [54]. The approach was developed further [55][56] by using a normal line Hough transform.…”

Reconstruction of a three-dimensional bubble surface from high-speed orthogonal imaging of dilute bubbly flow

“…It can be computed from a single projection a) as a projected area diameter d projArea of the bubble image, eqn. (3), or b) assuming an ellipsoidal shape or more precisely oblate spheroid, setting the second horizontal axis of the ellipsoid equal to the visible horizontal axis [4], eqn. (4).…”

“…This causes errors in the measured size and velocity distributions of the studied objects. Therefore, overlapping object recognition algorithms [4][5][6][7][8][9][10] are necessary to increase measurement accuracy and sample rate by recognizing individual, in-focus objects. Utilization of the specific pattern recognition algorithms is discussed in [10].…”

“…A more precise estimation is obtained assuming that the bubble has a symmetrical shape along its vertical axis, e.g. ellipsoidal shape, and that the third main axis of the ellipsoid is predicted with the axes visual in the image [4].…”