1. Introduction
Façade design tends to be diverse with increasing requirements for architectural functions in modern high-rise buildings. Building surface-mounted appurtenances such as ribs, balconies, mullions, and sunshades protruding from the façade are widely used in architectural design [
1]. Conventionally, the influence of these elements on wind loads and corresponding structural responses of the building is overlooked due to their small size compared with the whole building scale. However, larger façade appurtenances in the scale of meters such as vertical ribs have become prevalent in high-rise buildings in recent years, not only for functional requirements but also for aesthetic reasons. In this situation, the aerodynamic roughness condition of the building surface varies and is highly influenced by the vertical ribs. The wind-induced responses of the structure may be dramatically changed, especially in high-rise buildings under higher wind velocity. Thus, it is highly desirable to examine the relationship between the façade roughness components and their potential effects on the wind-induced responses.
Protruding appurtenances have great potential for optimizing a building’s aerodynamic performance without interfering with architectural design and living conditions. In contrast, some aerodynamic optimizations such as tapering or setback have some inherent drawbacks such as decreasing the building’s space utilization and conflicting with the architectural concept [
2]. Therefore, aerodynamic optimizations should balance aerodynamic efficiency and architectural design aspects, and therefore, façade roughness components are promising in mitigating the wind-induced responses of high-rise buildings [
3].
Previous studies mainly examined the local wind pressure changes of buildings caused by façade appurtenances. Stathopoulos and Zhu concluded that the local pressures reduce significantly in the upper zone of the windward wall and the lower zone of the side and leeward wall with increasing surface roughness, while remaining almost unchanged in the other zones [
4]. Chand et al. found that the provision of balconies alters the wind pressure distribution on the windward wall of a high-rise building but does not introduce significant changes on the leeward side [
5]. Maruta et al. investigated the effects of different balcony widths on the local wind pressure and concluded that the transmission of disturbances is restrained by increasing surface roughness due to the elimination of fluctuating pressures induced by the separation bubbles [
6]. Shen et al. investigated the influence of outer pierced ornamental components on the wind pressure distributions on a twisted high-rise building [
7]. They found that the existence of ornament components can obviously reduce the peak negative pressure in the middle of each side of the building surface, which is conducive to the wind-resistant design of the cladding. A Computational Fluid Dynamics (CFD) simulation conducted by Montazeri and Blocken also indicated that building balconies lead to significant changes in the wind pressure distribution on building surfaces because of flow separation and recirculation across the façade in multiple areas [
8]. Quan et al. reported that vertical ribs significantly decrease the most unfavorable suction coefficients in the corner recession regions and edge regions of façades [
9]. Yuan et al. used thin horizontal splitter plates to simulate appurtenance configurations and found that appurtenances can reduce the negative local peak pressure of the higher leading corner on a building’s side face [
10]. Hu et al. found that vertical openings on the external skin of a double-skin façade effectively reduce the wind pressures on the side and leeward surfaces [
11]. Liu et al. examined the flow fields around their models with the high-frequency particle velocimetry technique and found that vertical ribs significantly attenuate the turbulence intensity in the separated shear layer and near the wake region, and the reduction of turbulence eventually reduces the fluctuating wind pressure on the side and leeward surfaces [
12].
In addition to the local wind pressure, there are also some studies that investigated the influence of appurtenances on the overall wind loads and wind-induced responses of high-rise buildings. Shen et al. employed two different numbers of rough strips and five thicknesses to investigate the influence of model surface roughness on the wind loads of large hyperbolic cooling towers and concluded that the base shears of the cooling towers increase with more surface roughness elements [
13]. Huang et al. examined several sandpapers and rough strips with various spacing and thickness sets and concluded that increased roughness can reduce the degree of vortex shedding as well as the vibration amplitude of super-tall buildings [
14]. Wang et al. conducted two aeroelastic model tests on tall buildings with and without rough strips and found that the existence of rough strips reduces 30% of resonant amplitude in the cross-wind direction of a building [
15]. Hui et al. investigated horizontal splitter plates’ effects on wind loads of high-rise buildings through wind tunnel tests and found that discontinuous appurtenances reduce the cross-wind base moments, with the largest decrement being of about 5% [
16]. Yang et al. found that continuous and stagger arrangements of vertical ribs significantly reduce fluctuating cross-wind layer force, with the reduction amplitude achieving up to 57.3% [
17]. In the studies by Hui et al. and Yang et al., the maximum width of the ribs of 3.75 m in the prototype was used, accounting for about 12.5% of the building width. However, ribs with such great width on a high-rise building might not be practical. Nevertheless, for high-rise buildings with attached vertical ribs, a comprehensive study remains lacking to determine wind loads and wind-induced responses accurately, considering the uncertain structural dynamic properties.
Regarding high-rise buildings with attached vertical ribs, this paper aims to clarify the effects of rib configurations, including rib width and rib distribution pattern, on the wind loads and wind-induced responses of a building. A series of wind tunnel tests in association with HFFB were conducted with four models and two exposure types. The static and fluctuating wind forces are compared among all the models, and the key parameters of the vertical ribs that can exert beneficial effects in reducing the wind-induced responses, especially the cross-wind responses, are determined. The spectra of the base overturning moments are demonstrated to explain the related reduction mechanism. Parametric analysis of wind-induced accelerations and displacements atop the building and base overturning moments with respect to the natural frequency and approaching wind velocity is also conducted. Finally, the results of two exposure conditions, i.e., the suburban and open terrains, are compared to investigate the influence of terrain conditions on the reduction effect of the rib configurations.
4. Effects in Two Exposures
The cross-wind responses of high-rise buildings are very sensitive to the intensity of turbulence and therefore associated with the terrain exposure type. Two exposures, i.e., open and suburban exposures, were tested in the wind tunnel. According to the results in the open exposure, the models F4 and H4 are effective in mitigating the wind-induced responses. Thus, the testing cases with 4% relative width in the suburban exposure were examined. For the suburban exposure, the power law index α is 0.22.
The mean overturning moment coefficients with various rib configurations in the open and suburban exposures are demonstrated in
Figure 14. It can be found that the absolute mean moment coefficients of the model S0 in the suburban exposure are smaller than those in the open exposure, indicating that the mean force will be weakened by the stronger turbulence. Similar results can also be found for the models H4 and F4, in which the weakening degree is not as large as the model S0 due to the existence of the surface-attached ribs. At the azimuth 0°, the mean overturning moment coefficient of the model S0 in the open exposure decreases by 15.2% compared with that in the suburban exposure, and almost the same decreasing amplitude can be found for the models H4 and F4.
The standard deviations of overturning moment coefficients with various rib configurations in the open and suburban exposures are demonstrated in
Figure 15. As expected, in the azimuths of 0° to 70°, the fluctuating moment coefficients in the suburban exposure are larger than those in the open exposure, indicating that a rougher exposure will result in stronger turbulence and therefore a larger fluctuating value. However, at azimuth 90°, which corresponds to cross-wind direction, the fluctuating moment coefficient of the model S0 in the suburban exposure is larger than that in the suburban exposure. This is mainly due to the fact that the fluctuating values are not only related to the oncoming flow turbulence but also associated with the signature turbulence induced by the vortex shedding. In a rougher exposure, the signature turbulence induced by the vortex shedding may be suppressed by the vortex-induced shedding and result in a lower fluctuating moment coefficient, which can explain the fluctuating results of the model S0 at azimuth 90°.
Figure 16 shows the base overturning moment spectra in the two exposures, varied with the reduced frequency parameter
fB/U. For the along-wind spectra at azimuth 0°, the spectrum amplitudes of models in the suburban exposure are generally larger than those in the open exposure, which confirms the larger fluctuating values in the rougher exposure shown in
Figure 16. For the cross-wind spectra at azimuth 90°, the peak of the moment spectrum of model S0 in the suburban exposure is smaller compared with that in the open exposure, indicating again that more turbulence will lead to the suppressing effect induced by the vortex shedding.
Figure 17 illustrates the normalized cross-wind accelerations atop the building in the open and suburban exposures at azimuth 0°. The amplitudes of the wind-induced accelerations in the suburban exposure are all smaller than those in the open exposure in the reduced velocity range of 9.0–11.5. This reveals that a rougher exposure has a beneficial effect of disrupting the regular shedding of vortices and causing the cross-wind acceleration to be appreciably smaller than that of a smooth exposure. Similar results can also be found in the works of Yang et al. [
24].
Results in the two exposures indicate that the models H4 and F4 are significantly efficient in reducing the wind-induced accelerations compared with the model S0. In order to have a clear view of the reduction amplitude, the ratio factor “
RFa” for the acceleration is thereby defined to quantify the degree of mitigation.
where
is the normalized acceleration of the model S0.
The
RFa of the models H4 and F4 in the two exposures in the azimuth 0° are shown in
Figure 18. For the two distribution patterns, the
RFas are almost identical in the two exposures in the reduced velocity range of 9.0–12.5, although the accelerations in the two exposures differ apparently. The
RFa of the models H4 and F4 reaches around 0.5 in the reduced velocities around 10.5, where the accelerations arrive at their maximum values.
5. Conclusions
This study experimentally investigated the effects of vertical ribs on the wind-induced responses of high-rise buildings in a boundary layer wind tunnel. The building models with four ribs configurations were tested in open and suburban exposures. The base overturning moments and corresponding responses are demonstrated and compared between models and exposures. The main conclusions of this study are as follows.
(1) In the along-wind direction, the model with vertical ribs of 4% relative width shows a beneficial effect in reducing the mean force coefficients compared with the models with vertical ribs of 2% relative width and without ribs. In the cross-wind direction, the model with vertical ribs of 4% relative width demonstrates its ability to reduce the fluctuating base overturning moments and peak of base moment spectra compared with the model without ribs, while the model with vertical ribs of 2% relative width shows no enhancements. This is mainly due to the fact that the vertical ribs with 4% relative width have a considerable benefit in disrupting the regular shedding of vortices.
(2) The base moment spectra and wind-induced responses, including the accelerations and base moments between the full-distributed models and half-distributed models, are not obviously distinguished in the cross-wind direction, indicating that the ribs in the corner play a vital role in suppressing the intensity of vortex shedding.
(3) In the cross-wind direction with a rougher exposure, the signature turbulence induced by the vortex shedding may be suppressed by the vortex-induced shedding and result in a lower fluctuating moment coefficient. Additionally, the acceleration and base moment in the two exposures also differ apparently. However, the ratio factors of the acceleration and base moment are almost identical in the two exposures in the reduced velocity range of 9.0–12.5.