What is the flow topology of a convective porous media flow?

What is the minimum domain size we need to simulate to capture the large-scale flow structures?

We have addressed these questions in our recent work published on Journal of Fluid Mechanics. With the aid of massively parallelized numerical simulations, we show that the near-wall, large-scale temperature patterns (supercells) represent the footprint of the flow structure in the core of the domain (megaplumes). We have also analyzed the effect of the domain size (aspect ratio, AR), on the resulting flow topology.

Why teapots always drip

There is an age-old question: How can the so-called “teapot effect” be explained? Our demonstrative experiments illustrating this phenomenon have been discussed in quiz shows by BBC and Servus TV (available from Germany and Austria).

The “teapot effect” has been threatening spotless white tablecloths for ages: if a liquid is poured out of a teapot too slowly, then the flow of liquid sometimes does not detach itself from the teapot, finding its way into the cup, but dribbles down at the outside of the teapot.

This phenomenon has been studied scientifically for decades – now a research team at TU Wien has succeeded in describing the “teapot effect” completely and in detail with an elaborate theoretical analysis and numerous experiments: An interplay of different forces keeps a tiny amount of liquid directly at the edge, and this is sufficient to redirect the flow of liquid under certain conditions.

News by TU Wien and full movie available on YouTube.

Paper published on J. Fluid Mech.

Experiments are performed in the TU Wien Turbulent Water Channel for three values of shear Reynolds number, namely 180, 360 and 720. The paper is open access and available here. This article follows our previous work on the reconstruction and tracking on anisotropic particles in channel flow turbulence.

In this work, we investigate experimentally the dynamics of non-axisymmetric fibres in channel flow turbulence, focusing specifically on the importance of the fibres size relative to the flow scales. To this aim, we maintain the same physical size of the fibres and we increase the shear Reynolds number. Experiments are performed in the TU Wien Turbulent Water Channel for three values of shear Reynolds number, namely 180, 360 and 720. 

Fibres are slender – length to diameter ratio of 120 -, rigid, curved and neutrally buoyant particles and their shape ranges from low curvature – almost straight fibres – to moderate curvature. In all cases, fibres size remains small compared to the channel height (1.5%). Three-dimensional and time-resolved recordings of the laser-illuminated measurement region are obtained from four high-speed cameras and used to infer fibres dynamics. With the aid of multiplicative algebraic reconstruction techniques, fibres position, orientation, velocity and rotation rates are determined. Our measurements span over half channel height, from wall to center, and allow a complete characterization of the fibres dynamics in all the regions of the flow. Specifically, we measure fibre preferential distribution and orientation. We observe that the fibres dynamics is always influenced by their curvature. Through a comparison between measurements of near-wall dynamics of fibres and near-wall dynamics of flow, we identify a causal relationship between fibre velocity and orientation, and the near-wall turbulence dynamics. Finally, we have been able to provide original measurements of the tumbling rate of the fibres, for which we report the influence of fibres curvature. We underline that our measurements confirm previous findings obtained in numerical and experimental works.