The susceptibility of tests on two-dimensional bluff bodies to incident flow variations

Discussion: “An Experimental Evaluation of Drag Coefficient for Rectangular Cylinders Exposed to Grid Turbulence” (Courchesne, J., and Laneville, A., 1982, ASME J. Fluids Eng., 104, pp. 523–527)

“…With the use of the larger models Grid 4, which presumably produces the hightest turbulence levels, is only 5.6 mesh lengths upstream of the model and Grid 3 only 7.5 mesh lengths upstream. Earlier work by this reviewer [16] has shown that at these distances lateral nonuniformities of the mean velocity and turbulence characteristics of the flow will occur behind such grids. Indeed one could go further and say that it is not at all possible to use square mesh grids to simulate the turbulence characteristics appropriate to large buildings in the lower part of the earths boundary layer without the introduction of such lateral nonuniformities into the flow.…”

End effects of nominally two-dimensional thin flat plates

“…For cases where the mean flow does not reattach on the obstacle surfaces, b/c <∼ 2.3 [8], a comparison of results from experimental studies over a wide range of s/c > 10 and different end conditions suggest this expectation holds for 0.4 < b/c < 2.3 [7,9,10,11,12,13,14]. For a given b/c, reported (blockage corrected) values for the midspan mean base pressure, sectional drag and shedding frequency from these studies agree within experimental uncertainty and the flow about the midspan can be considered two-dimensional.…”

“…For rectangular cross-section, 2D bluff bodies with sharp leading edges mounted with a face normal to a uniform stream, the separation points are generally fixed at the leading edges and spanwise end conditions are not expected to affect the vortex formation process [6,7]. Defining the obstacle spanwidth as s, the flow normal dimension c (chord) and the streamwise thickness b.…”