Nature of sweep and ejection events in transitional and turbulent boundary layers

Abstract

This paper represents an experimental investigation of the physical nature of the turbulence production mechanisms in boundary layers associated with three fundamental phenomena known from previous studies: (i) sweep events, (ii) ejection events, and (iii) coherent vortical structures. The main goal of this study is to clarify connections between all these phenomena in transitional and turbulent flows. The experimental approach is based on a modified version of the hydrogen-bubble (HB) visualization technique called the synchronous visualization method, which is combined with quantitative processing of video images in a PIV-like manner. The results of investigation of sweep/ejection events and their connection with vortical structures are obtained at late and super-late stages of boundary-layer transition, as well as in fully turbulent boundary layers. The developed approach gives us the possibility to obtain detailed synchronous information about instantaneous flow characteristics in a great range of spatial coordinates and in two projections (plane and side views). It is assumed that the results obtained in transitional flow are very useful for analysis of the turbulent production mechanisms occurring in developed turbulent boundary layers due to the existence of the well-known similarity of these mechanisms found in previous investigations. As a result, two types of ejection events and three types of sweep events are found. It is shown that ‘sweep 1’ and ‘ejection 1’ events are produced by well-known Λ- (‘horseshoe’) vortices, while ‘sweep 2’ and ‘ejection 2’ events are generated by ‘ring-like vortices’ (‘typical eddies’) also found previously. The latter turned out to be significantly stronger than the former. The strongest sweep event (called ‘sweep 3’) is observed when ‘sweep 1’ and ‘sweep 2’ events become superimposed in space (in the (x, z)-plane). This superposition gives rise to the appearance (in the visualized plane view) of some very characteristic patterns associated with ‘dark spots’ (‘pockets’) observed previously in the wall regions in the developed turbulent boundary layers.

Keywords:

• boundary layer
• dark spots
• downdraft jets
• laminar-turbulent transition
• late and super-late stages
• sweep-ejection events
• turbulence production
• vortical structures

Acknowledgements

This work is supported by Chinese National Natural Science Foundation (Grant no. 10832001), Russian Foundation for Basic Research (Grant 08-01-91951), and Russian Academy of Sciences.

Notes

1. More detailed descriptions of development of the DNS approach can be found in these papers.

2. The idea of this resemblance itself needs additional development and extension.

3. As far as the disturbances are quasi-periodic, it is advised to look movies continuously as a loop. The junction between the first and the last video frame of the period demonstrates a degree of periodicity of perturbations at this particular stage of transition, wall-normal, and spanwise coordinates.

4. The color map is made different at every next streamwise location for convenience of analysis only.

5. Also complementing them with some quantitative estimates of wall-normal speeds.

6. Note also that the ejection 1 event discussed in detail above is also seen very well in Figure 15(c) at the present HB-wire streamwise coordinate.

7. Some velocity estimates, which prove these statements, are presented below in Section 6.3.

8. It is important to avoid confusing of two aspects of the discussed phenomenon: (i) an instantaneous flow velocity within every barrel, which is higher than the local mean-flow velocity and (ii) the speed of propagation of the barrel as a whole, i.e. of its border, which reminds a head-wave-like disturbance induced by a passing ring vortex and has a speed exceeding that of the surrounding fluid.

9. Only one side of first barrel is show in this figure.

10. This is clearly seen in Figures 16(d) and (e).

11. Note that Figures 29(e) and (f) represent just different realizations (fundamental periods) of the same process occurring at y _w= 4.0 mm and z= 20 mm, and display a degree of periodicity of the transition process at the current stage.

12. It is obvious that this equivalence is purely hypothetical because such amount of wires and light sources is impossible to use simultaneously in reality due to flow distortions and confusing of all flow patterns obtained.

13. Note that our new qualitative method is not in a competition with quantitative PIV approach because none of the present PIV modifications is able to provide synchronized information about the 3D flow patterns existing simultaneously in such extended spatial domains as those studied in the present experiments (about 700 × 200 × 20 mm) providing, simultaneously, a rather good spatial resolution (about one-tenth of a millimeter).

14. This is in general agreement with previous observations.