ABSOLUTE INSTABILITY IN CRICULAR CYLINDER WAKES
圆柱尾流的绝对不稳定性

The stability of circular cylinder wake at low Reynolds numbers has been studied experimently. The experiments were separatly carried out in a propeller-driven water channel and in a gravity-driven low turbulence water tunnel. The test section of water channel was 15 cm in width, 10cm in depth and 100 cm in length, and the test section of water tunnel was 4 cm in width.10cm in height and 100 cm in length. The cylinder model used in channel and tunnel were 15 cm long, 0.2 cm in diameter and 4cm long, 1 cm in diameter respectively. The model cylinder was placed horizentally across the test section, both ends in close contact with the vertical walls of the test section. Turbulence intensity of the oncoming flow. was 0.8% in channel and 0.25% in tunnel.The apparatus used both in water channel and water tunnel that applied disturbances to wake flow field was made up of a thin tube, a cylindric container, a round thin aluminium plate and an elastic membrane. The thin tube made of stainless steel, outer diameter 0.12cm, inner diameter 0.08cm, was in inserted into the water tunnel from above, perpendiculely penetrating through the upper wall. The end of the steel tube was less than 0.5 cm deep in water so that no disturbances were imposed to the wake flow by the tube. The end out of the tunnel was connected to the container. The container was sealed by the elastic membrane at the upper end. The round plate, with diameter slightly smaller than the inner diameter of the container, was glued to the elastic membrane, forming a plate that could move up and down in the container. Water from the tunnel filled the container and it was kept at appropriate height to keep the membrane in equilibrium. When an impulsive force was applied upon the membrane, an impulse jet from the thin tube entered the flow field, imposing a disturbance to the wake, then water was sucked back into the container, the membrane backed to the equilibrium. It was tested that one impulse force could cause just one impulse jet, the second time disturbing never appeared. In the test, velocity of the oncoming flow was slowly increased from zero to a certain value and then kept constant there to allow the flow to stabilize. An impulse disturbance was applied to the cylinder wake at some point and the downstream velocity was recorded in long periods before and after the disturbance by a hot film probe located at a point 26 diameters downstream the cylinder.When the Reynolds number is smaller than the critical one, fluctuation of the undisturbed wake velocity is very small. But disturbances applied to the near wake excite vortex shedding that does not decay. It is observed that the disturbance applied must be within certain range behind the cylinder, otherwise, when the disturbance applied to a point some distance downstream that range, it decays. The experimental results indicate that there is an absolutely unstable zone in the near wake. Disturbances within the zone locally amplify and spread,