VORTEX SHEDDING FROM A CYLINDER IN A WATER CHANNEL

1 Aim
To observe the vortex shedding phenomena in flow behind a cylinder at moderate Reynolds
numbers. To study the dependence of the non-dimensional frequency of vortex shedding on
Reynolds number (Re).

2 Theory

Flows with steady boundary conditions can have unsteady solutions; flows around bluff bodies at
intermediate Re are the common examples. For flow around a cylinder, at low Re (< 1, based on
cylinder diameter, D and free stream velocity, U∞) the vorticity generated at the surface of the body
is diffused (not advected) and there is fore and aft symmetry of flow. When Re is increased above
40, the wake begins to become unstable. With increasing Re, the oscillating wake rolls up into two
counter-rotating vortices in two staggered rows. These staggered rows of vortices are known as the
Kárman vortex street. For 40 < Re < 80, the vortex street does not affect the attached vortices on
the cylinder. When Re > 80, the vortex street forms closer to the body and the eddies periodically
break off alternately from the two sides of the cylinder creating an unsteady flow in the wake. The
successive vortices have opposite sense of circulation; the circulation around the cylinder changes
sign periodically resulting in an oscillating lateral force on the cylinder. If the frequency of this
vortex shedding is close to the natural frequency of the cylinder, resonance could occur, resulting
in structural damage on the cylinder. Wires sing because the frequency of this vortex shedding lies
in the acoustic range. Below Re = 200, the wake is laminar, while the wake is turbulent (but the
5
boundary-layer is laminar) for 200 < Re < 3 × 10 . For Re < 2500, the shed vortices are known to
form a regular vortex street beyond four to five diameters from the cylinder. In all the cases where
Re > 80, a definite periodic oscillation of the flow (same as the vortex shedding frequency) is
5
observed near the cylinder till Re = 4 × 10 , when the boundary-layer on the cylinder surface
becomes turbulent. The non-dimensional frequency of vortex shedding is known as the Strouhal
number denoted by S = f /(U∞/D), where f = frequency of vortex shedding, D = diameter of the
cylinder and U∞ = free stream velocity of the fluid. The objective of the experiment is to explore
the dependence of this non- dimensional vortex shedding frequency on the Reynolds number for
5
900 < Re < 1.3 × 10 .

3 Experimental set-up

Figure 1 below depicts a visual of the set-up. It consists of a tank 2.5m × 1.5m with a depth of
150mm, at one end of which are located 2 sets of Aluminum disks which rotate in opposite
directions horizontally and create a flow. The flow is guided to the test section where cylinders of
different diameters can be placed. The flow rate is adjusted by controlling the rate of rotation of the
disks; velocities ranging from 0.01 m/s to 0.2 m/s can be achieved. Water, made black in colour by
dissolving a dye, is used as the fluid and Aluminum powder is used as the tracer. The free stream
velocity is measured by noting the time taken for a floating
particle to traverse a fixed distance in the test section. Photographs of the flow phenomena can be
taken by capturing the reflected light from the aluminium powder by an over-head camera.

FIG 1. Schematic of the Experimental set up.

4 Procedure

1. Fill the tank with water which has a black dye dissolved in it. Add sufficient quantity of
Aluminium powder. Switch on the power to the motor with the variac at its lowest position.

2. Turn the variac clockwise very slowly to start the rotation of the paddles. Adjust the variac to
obtain steady rotation of the disks.

3. Find the velocity of flow by noting the time (t) taken for a floating particle to traverse a fixed
distance (L) in the test region. The floating particle should be away from the walls. Repeat this
measurement a few times and calculate the mean time (tm) to estimate the average free stream
velocity (U∞ = L/tm).

4. Place the cylinder with diameter d = 2 cm centrally in the test section and observe the flow
pattern. Measure the time taken (tv) to shed a specific number (n) of vortices from the
cylinder. Calculate the shedding frequency f = n/tv. Repeat a few times and calculate the mean
vortex shedding frequency. Take a photograph (with slow shutter speed so that streaks are seen) of
the flow phenomena around the cylinder. Repeat the procedure for cylinders with d = 3.5, 5 and 6
cm.

5. Repeat the experiment for higher free stream velocities (say, 0.1, 0.15 and 0.2m/s). Take a
photograph of the phenomena for each cylinder diameter & velocity.

6. Calculate the Strouhal number (St) for each frequency of vortex shedding. Plot St Vs Re.
Comment on the results.

5 Results and Discussion

The following results are to be presented.

1. Hand drawn schematic of the phenomena for each cylinder with increasing Re. Comment
on what happens with increase in free stream velocity (or Re for constant D)

1. Plot St vs Re for all the cylinders and velocities. Discuss the plot yo u have obtained
explaining the physical reasons behind such a plot.
5
2. Compare against empirical relation St = f (Re) = 0.198 – (3.9/Re) valid over 250 < Re < 2x10 .