Flow About a Circular Cylinder in and Near a Turbulent Plane Mixing Layer

1989 ◽  
Vol 111 (2) ◽  
pp. 124-129 ◽  
Author(s):  
M. Kiya ◽  
H. Tamura

This paper deals with an experimental study of the fluid forces, surface pressure, and vortex-shedding frequency of a circular cylinder placed in and near a turbulent plane mixing layer with zero velocity ratio. These characteristics are given as functions of the cylinder diameter d divided by the local mixing-layer width δ and the nondimensional transverse coordinate of the center of the cylinder in a Reynolds number range 5500–46000. For the cylinder d/δ = 2.2–2.3 the r.m.s. lift attained a maximum 40 percent higher than that in the uniform flow when a part of the cylinder surface was near an intermittently turbulent edge on the high-velocity side of the mixing layer; the time-mean resultant force attained a sharp maximum as high as 80 percent of that in the uniform flow when the axis of the cylinder was located at the center of the mixing layer. The former maximum was interpreted in terms of a vortex-body interaction while the latter maximum was found to be associated with a large delay of the boundary-layer separation on the low-velocity side of the cylinder.

Author(s):  
D. Heitz ◽  
G. Arroyo ◽  
P. Marchal ◽  
J. Delville ◽  
J.-H. Garem ◽  
...  

1979 ◽  
Vol 92 (2) ◽  
pp. 269-301 ◽  
Author(s):  
R. E. Britter ◽  
J. C. R. Hunt ◽  
J. C. Mumford

The flow of grid-generated turbulence past a circular cylinder is investigated using hot-wire anemometry over a Reynolds number range from 4·25 × 103 to 2·74 × 104 and a range of intensities from 0·025 to 0·062. Measurements of the mean velocity distribution, and r.m.s. intensities and spectral energy densities of the turbulent velocity fluctuations are presented for various radial and circumferential positions relative to the cylinder, and for ratios of the cylinder radius a to the scale of the incident turbulence Lx ranging from 0·05 to 1·42. The influence of upstream conditions on the flow in the cylinder wake and its associated induced velocity fluctuations is discussed.For all measurements, detailed comparison is made with the theoretical predictions of Hunt (1973). We conclude the following. The amplification and reduction of the three components of turbulence (which occur in different senses for the different components) can be explained qualitatively in terms of the distortion by the mean flow of the turbulent vorticity and the ‘blocking’ or ‘source’ effect caused by turbulence impinging on the cylinder surface. The relative importance of the first effect over the second increases as a/Lx increases or the distance from the cylinder surface increases.Over certain ranges of the variables involved, the measurements are in quantitative agreement with the predictions of the asymptotic theory when a/Lx [Lt ] 1, a/Lx [Gt ] 1 or |k| a [Gt ] 1 (where k is the wavenumber).The incident turbulence affects the gross properties of the flow in the cylinder wake, but the associated velocity fluctuations are probably statistically independent of those in the incident flow.The dissipation of turbulent energy is greater in the straining flow near the cylinder than in the approach flow. Some estimates for this effect are proposed.


1970 ◽  
Vol 42 (3) ◽  
pp. 471-489 ◽  
Author(s):  
S. C. R. Dennis ◽  
Gau-Zu Chang

Finite-difference solutions of the equations of motion for steady incompressible flow around a circular cylinder have been obtained for a range of Reynolds numbers from R = 5 to R = 100. The object is to extend the Reynolds number range for reliable data on the steady flow, particularly with regard to the growth of the wake. The wake length is found to increase approximately linearly with R over the whole range from the value, just below R = 7, at which it first appears. Calculated values of the drag coefficient, the angle of separation, and the pressure and vorticity distributions over the cylinder surface are presented. The development of these properties with Reynolds number is consistent, but it does not seem possible to predict with any certainty their tendency as R → ∞. The first attempt to obtain the present results was made by integrating the time-dependent equations, but the approach to steady flow was so slow at higher Reynolds numbers that the method was abandoned.


Sign in / Sign up

Export Citation Format

Share Document