Civil Engineering Reference
In-Depth Information
7.6.5
Aeroelastic modelling of chimneys
Chimneys and other slender structures of circular cross-sections are vulnerable to cross-
wind excitation by fluctuating pressures due to vortex shedding (Sections 4.6.3 and 11.5).
In the 1950s and 1960s, it was quite common to investigate this behaviour with small-
scale wind-tunnel models. However, the forces from vortex shedding are quite dependent
on Reynolds number (Section 4.2.4), and wind-tunnel tests will severely over-estimate
the cross-wind response of prototype large chimneys (Vickery and Daly, 1984). The
prediction of full-scale response of such structures is better undertaken by the use of
mathematical models of the response (Section 11.5) with input parameters derived from
full-scale measurements at high Reynolds numbers.
7.6.6
Distorted 'dynamic' models
In many cases the resonant response of a structure may be significant, but the prototype
structure may be stiff enough such that aeroelastic forces (i.e. the motion dependent
forces) are not significant. Furthermore, the scaling requirements (Section 7.4.1) and
properties of the available modelling materials may make it difficult, or even impossible,
to simultaneously scale the mass, stiffness and aerodynamic shape of a structure.
In such cases, the mass and stiffness properties of the structure can both be distorted
by the same factor (usually greater than 1.0). Then the correct frequency relationship for
the applied fluctuating wind forces and the structural frequencies is obtained. Internal
forces and moments are correctly modelled (including resonant effects, but neglecting
aeroelastic effects), but the deflections, accelerations and aerodynamic forces are not
scaled correctly.
Such 'distorted' dynamic models have been used on certain open-frame structures,
where aeroelastic 'spine' models were not possible.
7.6.7
Structural loads through pressure measurements
For structures such as large roofs of sports stadiums, or large low-rise buildings, with
structural systems that are well-defined and for which resonant dynamic action is not
dominant, or can be neglected, wind-tunnel pressure measurements on rigid models can
be used effectively to determine load effects such as member forces and bending
moments, or deflections. This method is normally used in conjunction with the area-
averaging pressure technique described in Section 7.5.2. Also required are
influence
coefficients,
representing the values of a load effect under the action of a single uniformly
distributed static 'patch load' acting on the area corresponding to a panel on the wind-
tunnel model. Two methods are possible.
1. Direct on-line weighting of the fluctuating panel pressures recorded in the wind-tunnel
test with the structural influence coefficients, to determine directly fluctuating and
peak values of the load effects (Surry and Stathopoulos, 1977; Holmes, 1988).
2. Measurement of correlation coefficients between the fluctuating pressures on pairs of
panels and calculation of rms and peak load effects by integration (Holmes and Best,
1981; Holmes
et al.,
1997).
