Professional Documents
Culture Documents
Journal of Engineering Mechanics Volume 137 Issue 2 2011
Journal of Engineering Mechanics Volume 137 Issue 2 2011
reem 2005兲; 共2兲 ACQC, in which both co- and quad-spectra of the
Discussion of “Cross Correlations of Modal generalized forces are considered using the proposed closed-form
Responses of Tall Buildings in Wind- formula; and 共3兲 ECQC, in which the modal correlation coeffi-
Induced Lateral-Torsional Motion” by M. F. cient is determined by direct numerical integrations of the re-
Huang, C. M. Chan, K. C. S. Kwok, and sponse spectra.
P. A. Hitchcock The paper demonstrated the accuracy of the ACQC as com-
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/12/15. Copyright ASCE. For personal use only; all rights reserved.
−0.5
Co−spectra 0
Quad−spectra
−1 −0.2
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/12/15. Copyright ASCE. For personal use only; all rights reserved.
0 0.2 0.4 0.6 0.8 1 0.4 0.5 0.6 0.7 0.8 0.9 1
Reduced frequency fB/U (a) Frequency ratio f /f
H 2 3
Fig. 1. Normalized co- and quad-spectra of the generalized forces for 0.8
1
modal frequencies. Similar normalizations exist for the quad-
spectrum. The format 共a兲 was used in the paper and is regarded as
0.5 an approximation as it is different from the standard normaliza-
tion as given in format 共b兲 used in Chen and Kareem 共2005兲. Fig.
2 of this discussion shows the comparison of these two normal-
0
ization formats for both co- and quad-spectra at f 23 = 0.5共f 2 + f 3兲 in
Co−spectra this tall building example at UH = 20 and 40 m / s. For f 2 / f 3 = 0.4
−0.5 Quad−spectra ⬃1.0, the corresponding reduced frequency f 23B / UH is 0.382
Co−spectra (Appr.) ⬃ 0.546 at 20 m / s, and 0.2191⬃ 0.273 at 40 m / s. As expected,
Quad−spectra (Appr.) the differences of these two normalization formats are negligible.
−1 The closed-form formulation for the modal correlation coefficient
0.4 0.5 0.6 0.7 0.8 0.9 1
(a) Frequency ratio f /f
2 3
is derived based on the fact that the spectra of generalized wind
forces can be regarded as constants around the modal frequencies,
and the modal correlation is noticeable only when two modal
Normalized co− and quad−spectra
1.5
frequencies are closely spaced. Fig. 3 shows the comparison of
1 the modal correlation coefficients with and without the influence
of quad-spectrum. It is seen that the contribution of the quad-
0.5
spectrum is negligible at 20 m / s, while some influence is ob-
0 served at 40 m / s, where the two modal frequencies are closely
spaced and the quad-spectrum is indistinguishable from the
−0.5 Co−spectra cospectrum.
Quad−spectra To further understand the role of quad-spectrum of wind loads
−1 Co−spectra (Appr.) on the modal correlation coefficients of buildings with three-
Quad−spectra (Appr.)
−1.5 dimensional 共3D兲 coupled mode shapes, the square building ex-
0.4 0.5 0.6 0.7 0.8 0.9 1 amined earlier is reconsidered by introducing an eccentricity in
(b) Frequency ratio f /f
2 3 rigidity in the diagonal direction of the building cross section. The
centers of rigidity at different floors are assumed to be on a single
Fig. 2. Comparison of co- and quad-spectra 共uncoupled building兲: 共a兲 vertical line. As a result of eccentricity, the building exhibits 3D
UH = 20 m / s; 共b兲 UH = 40 m / s coupled mode shapes with particularly strong coupled motion in
0 0.2
−0.5 0
Co−spectra
Quad−spectra
−1 −0.2
0.4 0.5 0.6 0.7 0.8 0.9 1
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/12/15. Copyright ASCE. For personal use only; all rights reserved.
0.6
Fig. 4. Normalized co- and quad-spectra of the generalized forces for
23
w/ Co−spectra
1
and quad-spectra of the generalized forces. At the reduced fre-
quency range of interest, the normalized quad-spectra show small
0.5 absolute values as compared to the cospectra. A similar calcula-
tion using the same parameters is carried out for this coupled
building as for the uncoupled case. Figs. 5 and 6 show the results,
0
which again demonstrate a negligible influence of quad-spectrum
Co−spectra on the modal correlation coefficient at 20 m / s, while there is
−0.5 Quad−spectra some influence at 40 m / s for the modes with closely spaced fre-
Co−spectra (Appr.) quencies.
Quad−spectra (Appr.) It should be emphasized that the quad-spectra of the spa-
−1 tiotemporal fluctuations in the surface pressure field over a build-
0.4 0.5 0.6 0.7 0.8 0.9 1
(a) Frequency ratio f /f
2 3
ing surface have no contribution to the PSDs of the overall
integral loads and the generalized loads associated with one- or
three-dimensional mode shapes as a consequence of the quad-
Normalized co− and quad−spectra
1.5 Co−spectra
Quad−spectra spectrum being an odd function 共Kareem 1978兲. The significance
1 of the quad-spectra on the modal correlation coefficients depends
Co−spectra (Appr.)
Quad−spectra (Appr.) on the characteristics of dynamic wind loads and building mode
0.5
shapes. Considering the role of the quad-spectra in the formula-
0 tion presented in this paper as an enhancement of the formulation
given in Chen and Kareem 共2005兲, the authors need to clarify the
−0.5 nature of the XPSD of the generalized forces used in the example
study, as some observations made here suggest a clear departure
−1
from the normal definition. The quad-spectra of wind loads have
−1.5 often been considered negligible as compared to the cospectra
0.4 0.5 0.6 0.7 0.8 0.9 1 based on some wind tunnel testing data on tall buildings 共e.g.,
(b) Frequency ratio f2/f3
Kareem 1992; Tallin and Ellingwood 1985兲. If the quad-spectra
of wind loads significantly impact the response estimates of
Fig. 5. Comparison of co- and quad-spectra of generalized forces for coupled buildings as implied in the paper, it may become essen-
modes 2 and 3 共coupled building兲: 共a兲 UH = 20 m / s; 共b兲 UH tial to carefully model this feature in evaluating wind loads and
= 40 m / s their effects on buildings.
Chen, X., and Kareem, A. 共2005兲. “Coupled dynamic analysis and buildings with similar dynamic properties in orthogonal directions.”
equivalent static wind loads on buildings with three-dimensional Proc., 10th Int. Conf. on Wind Engineering, Copenhagen, Denmark,
modes.” J. Struct. Eng., 131共7兲, 1071–1082. 2389–2396.