A few questions for the folks who are researching into the onset of instability.

Does it take a period of time from the start of simulation, until the unsteady flow structures become evident?

Once formed, do the flow structures remain constant, or do they increase/decrease in effect?

Has anyone simulated Reynold's water-in-pipe experiment? If so, what is observed?

My main area of interest is for lowspeed, viscous, incompressible flows & so flow over a cylinder is also of interest.

For flow over a cylinder, what free-distance to upper & lower boundaries should be used - in terms of cylinder diameters?

Has anyone experimented with, or observed, 'modes' of instability in lowspeed flow?

--------

Thanks very much for your kind assistance.

diaw...

"Does it take a period of time from the start of simulation, until the unsteady flow structures become evident?"

"Once formed, do the flow structures remain constant, or do they increase/decrease in effect?"

"Unsteady" and "remain constant"?

Are you sure this is what you mean?

Jim_Park wrote: "Does it take a period of time from the start of simulation, until the unsteady flow structures become evident?"

"Once formed, do the flow structures remain constant, or do they increase/decrease in effect?"

"Unsteady" and "remain constant"?

Are you sure this is what you mean?

----------

diaw replies:

Hi Jim, thanks for your feedback. Perhaps let me re-phrase my thoughts a little, if that will help.

Numeric observations:

Dependending on how one is observing the 'unsteady' activity, certain structures may be observed. These structures begin to form into cohesive patterns over a period of time, where they basically 'remain constant for a time, before eventually beginning to decay again - over a long time-frame.

The general bulk flow will, of course, not appear to be steady.

I was intereted in other experimenters observations of different variables & perhaps different data interpretations, during such unsteady processes.

I hope that I have clarified what I was trying to say?

Thanks for your input.

diaw...

A classic example of what you are describing is Taylor-Couette flow. Two concentric rotating cylinders are immersed in a viscous fluid. If the differential rotating velocities are great enough, instability develops.

This is my understanding of what happens, but my interpretation may be wrong: At startup, the flow is laminar. Energy is fed into the fluid by the cylinders. Then a boundary layer starts to develop and grow outward. When the shear exceeds some limit, vortices develop near the inner cylinder. The vortices dissipate energy more rapidly than it is fed into the fluid from the rotation, but excess energy is fed into the vortices from the outer region. When the energy in the outer region is sufficiently dissipated, the vortices can no longer be sustained and they collapse. The flow becomes laminar again, and a new boundary layer starts to grow and the process repeats.

Simulations have been done, largely using spectral methods. I have a fantasy of doing this sometime using finite elements.

You said, "Does it take a period of time from the start of simulation, until the unsteady flow structures become evident?" That depends on your initial conditions. If you start with some low-shear laminar flow, it will take a while for the conditions for instability to develop. If you start with an initial flow that is already unstable, the instabilities will develop immediately.

Regarding "experiments:" I live in a house on the top of a ridge in east Tennessee, USA. I have a small garden terraced into the hillside below. I have a water pipe running down the hill to the garden. When I water my garden, sometimes the water comes out of the hose as laminar flow. Then it breaks up and becomes turbulent. Then it becomes laminar again. It is frustrating because the flow decreases significantly during the turbulent flow phases. I don't know what causes this alternating behavior. Maybe there is air in the line, maybe something else like the Taylor-Couette flow described above.

Thanks Jonas - those are two excellent examples.

Taylor-Couete flow:

I quote from a book I have on hand: (refers to 3 figures) "In Couette flow a liquid is contained in the gap between two concentric cylinders one of which has an angular velocity w with respect to the other. At low Reynolds number Re the flow is azimuthal as in (a). As Re increases flow symmetry is lost and vortices develop (b). A further increase of Re develops transverse waves along the lines of vortices (c)". "The Physics of Vibrations and Waves", Pain H.J, 6th ed, Wiley, 2005 - pg 497.

Please note the last section - "develops transverse waves along the lines of vortices..." Could this be a 'structure' in an unsteady flow field?

----------------

Flow in a garden hose...

Postulate wave activity moving up & down the flow field. Mode of unstable flow changes ever-so-slightly... viola...

Excellent.

--------

Has anyone done any simulation of such unsteady flow events?

Take an unsteady flow example & begin looking for small components, or change components.

Thanks Jonas...

diaw...