How Can Boundary Layer Suction Reduce Flow Separation in Pipe Expansion?

1 reply · 3K views · last reply
Timtam
Messages
40
Reaction score
0
I have a question about the following scenario involving a flow separation issue in a pipe expansion

Diagram of a pipe expanding from diameter D1 to D2 with a cone angle theta, showing streamlines and a zone of separation with eddies at the cone wall for a large cone angle. The red-boxed region marks the start of the expansion.


The angle of the expansion is 30* - doubling the diameter from 1D to 2D

We can consider this flow fully developed with a Reynolds of 5000+

Associated with this expansion is a head loss caused by localised flow separation converting flow to eddies and vortices which I understand to then cascade down to the Kolmogorov length scale where they are dispersed as heat.

If I was to zoom in the localisation might look like this

Diagram of flow separation at a 30-degree pipe expansion: a blue sloped wall with dashed red streamlines, with eddies and vortices shown as small dashed circles near the wall, and straight streamlines continuing past the expansion.


The K factor resistant coefficient associated with this configuration is 0.46

Current tables
Table 10.2 of resistance coefficients for gradual pipe enlargement, listing values by diameter ratio D2/D1 (rows) against cone angle θ from 2° to 60° (columns).

In aerodynamics they use boundary layer control suction (removing Boundary Layer Static Pressure) or re-energizing the flow adding Dynamic Pressure) to reduce drag associated with flow separation

Sketch of flow over a sloped wall that drops to a step, with red dotted streamlines, a high pressure zone near the slope, a low pressure zone at the step, and a downward arrow labelled Suction.
I am similarly interested in using boundary layer suction in my pipe (by reducing pressure at the start of the expansion to reduce the head loss**) but I am unsure I understand the actual mechanism that achieves this

My attempt at an explanation Suction creates a localized low pressure zone ( ahead of the localized high pressure zone created by the stagnated flow within the boundary). This creates a new gradient with which this high pressure can disperse . As this gradient is in the streamwise direction this allows the energy to rejoin the flow as DPMy Question

Is this correct ? Would my resistance coefficient be lower than the 0.46 the pressure loss the expansion experiences without suction ?

**I am aware that adding suction will cost energy which will be more than the flow recovered: therefore will be energy deficit. I am only concerned with lowering the Resistance coefficient
 
Attachments
  • Diagram of flow over a downward-sloping wall: dashed red streamlines bend toward the wall, small curled eddies near the surface are labelled High pressure Zone, and a suction duct below the flat section is labelled Low pressure Zone and Suction.
    upload_2016-4-11_17-49-57.png
    49.1 KB · Views: 615
Suction does create a localized low pressure zone near the wall, but the important effect here is that, while there was no wall- normal pressure gradient before (it's a boundary layer, after all), there is now a favorable pressure gradient, however slight, pointing toward the wall. This tends to draw more high-momentum fluid from the free stream down toward the wall, and promotes having a "fuller" boundary layer than would otherwise occur. The result is a boundary layer that, among other things, is more resistant to separation.

I do believe it makes sense that you would experience less flow resistance if you were to add suction as you have described.