Screens, like contractions, reduce
the longitudinal components of turbulence or mean-velocity variation
to a greater extent than the lateral components, so that the number
of screens to be used is determined by the acceptable lateral component
disturbance in the test section. The lateral components of mean
velocity and of the larger turbulent eddies can be reduced more
effectively by a honeycomb: the
mode of action of a honeycomb with cells elongated in the flow direction
is qualitatively obvious but few tests have actually been made,
and all that is certain is that the cell length of the honeycomb
should be at least six or eight times the cell diameter. Stacking
two honeycombs together to get a larger effective length-to-diameter
ratio is acceptable if the two are tightly laced together.
A honeycomb naturally produces some turbulence of its own, with eddy sizes of the same order as the cell diameter, which decays very much more slowly than the small-scale turbulence produced by screens. The early wind tunnels which had a honeycomb but no screens (and usually a very small contraction ratio also) suffered from a very high turbulence intensity in the test section although the mean flow was often quite good. Most modern tunnels have both honeycomb and screens.
The cells in the kraft paper honeycomb used for lightweight doors are not very uniform, but the walls are usually quite accurately perpendicular to the plane of the honeycomb and length-to-diameter ratios of 6 or more can be found. For large or high-performance tunnels, aluminum honeycomb used for aircraft "sandwich" construction is better, but care should be taken to remove burrs: usually fatigue failure can be induced by playing a high-pressure air hose over the honeycomb.