Aerodynamic Modifications of Building Shapes | Wind Engineering

Recent studies have concentrated on the effects of relatively minor aerodynamic modifications of building shape in reducing wind induced response of tall buildings.

Effects of Aerodynamic Modifications of Building Shapes on Wind Induced Response of Tall Buildings:

Wind tunnel experiments were conducted by Kwok and Bailey (1987), Kwok et al (1988), and Kwok (1988) to investigate the effects of aerodynamic devices, building edge configuration, and through building opening on the wind induced vibrations of tall buildings.

Aeroelastic models of a square cross-section building with a height to breadth ratio of 9 and the CAARC Standard Tall Building which has a rectangular plan shape, as shown in Fig. 13, were tested in open country and suburban terrains. The building model configurations tested included fins and vented fins at the corners, slotted corners, chamfered corners, and horizontal slots.

The modifications to the building corners ranged from around 9% to 16% of the building breadth. It is evident that aerodynamic modifications which increase the projected area or the effective width of a building, such as fins at corners, would in general not be beneficial.

Horizontal slots, slotted corners and chamfered corners were found to be effective in causing significant reductions in both the dynamic alongwind and crosswind responses of the rectangular cross-section CAARC Standard Tall Building, as shown in Figs. 14 and 15. With chamfered corners, there was up to a 40% reduction in the alongwind response within a reduced wind velocity range of 3 to 20 compared with that for the plain building.

With wind normal to the wide face of the building, the dynamic crosswind response of the modified buildings were found to be up to 30% smaller than that for the plain building at the low range of reduced wind velocities. The critical reduced wind velocity changed from a value of approximately 10 for the plain building and the building with horizontal slots, to about 8 for the building with chamfered corners, and about 9 for the one with slotted corners.

The crosswind response characteristics of those modified buildings changed accordingly, with resonant crosswind response occurring at the modified critical reduced velocity. The resonant crosswind response at the modified critical reduced velocities of 8 for the building with chamfered corners and 9 for the building fitted with slotted corners were found to be larger than those at the corresponding reduced velocities for the plain building.

With the incident wind normal to the narrow face of the building, the crosswind response was found to be essentially wake excited. The long after-body of the building apparently has a disruptive effect on the vortex shedding process and there is no significant response peak associated with a dominant critical velocity effect. Building modifications such as horizontal slots, slotted corners and chamfered corners were found to result in a 30% or more reduction in the crosswind response.

The design wind velocities for modern tall building are at the low range of reduced wind velocities for serviceability requirement and at the mid to high range of reduced velocities for strength considerations. Modifications such as horizontal slots, slotted corners and chamfered corners can offer significant reductions in the magnitude of both the alongwind and crosswind responses compared with a plain building.

The effects of building plan shape on the crosswind response of tall buildings were also investigated by Melbourne and Cheung (1988) and Melbourne (1989). It was concluded from wind tunnel aeroelastic model studies and sample calculations for practical applications that for tall buildings with nominally symmetrical or square plan shapes, modest rounding or chamfering of corners up to 10% does not significantly reduce the crosswind response at low values of reduced wind velocities relevant to the serviceability acceleration criterion for occupant comfort.

However, very significant reductions in the ultimate limit state design loads can be achieved at higher reduced velocities. For more substantial rounding or chamfering of corners such that the building plan shape approaches that of a rough circular shape, for example octagonal or hexagonal shape, the crosswind force spectrum was found to reduce at both the lower and higher ranges of reduced velocities.

A similar characteristic was also observed for tall buildings with reduced upper level plan areas, for examples tapering, by cutting corners or dropping off corners progressively as height increases. However, for building plan shapes which are elliptical or elongated octagonal with a major to minor axis ratio of about 3:2, the critical reduced wind velocities were found to be significantly lower.

The resultant high crosswind response and acceleration level makes it more difficult to meet occupancy comfort criteria. Furthermore, those plan shapes were found to exhibit significant torsional response about the vertical axis.

Dutton and Isyumov (1990) studied the effects of vertical openings on the crosswind response of a square cross-section tall building with a height to breadth ratio of 9 to 1. The crosswind base bending moment spectra obtained from force balance tests and the corresponding computed crosswind tip deflections of the tall building are shown in Figs. 16 and 17 respectively.

Fig. 16 Crosswind base bending moment spectra of buildings with aerodynamic modifications.

Fig. 17 Crosswind response of buildings with aerodynamic modifications.

Alongwind through building openings and in particular combined alongwind and crosswind openings were found to cause a pronounced reduction of the vortex shedding induced forces and hence the crosswind dynamic deflection of the building. The critical reduced wind velocity shifted to a slightly higher value which implied that resonant vibrations of the building would be postponed to a higher wind speed with a longer return period.

However, there was a second spectral peak and hence a corresponding increase in crosswind response at a reduced velocity of around 6 which may make it more difficult to meet occupancy comfort criteria. The level of disruption to the vortex shedding process varied with the width of the opening and large reductions were observed for openings as small as 4% of the building width. Results obtained from a pressure model study suggested the mean base pressure coefficient is a good indicator of the overall aerodynamic, including the vortex shedding process, of a tall building.

The potential beneficial effects of a through building opening in a 390 m high office tower project were studied in a wind tunnel using both force balance and aeroelastic model techniques and reported by Isyumov et al (1992).

The results show that a more sculptured building top, in this case a triangular pyramid shape for the top 15% of the building, can moderate both the alongwind and crosswind responses. A venting or “bleeding” of the building wake provided by a through building opening near the top resulted in additional reductions in these responses.

The effects of building plan shape on the aerodynamic forces and dynamic response of a 600 m high 150 storey super tall building were investigated by Hayashida and Iwasa (1990) and Hayashida et al (1992) using force balance, aeroelastic and pressure model techniques.

The building plan shapes tested include a plain square shape and square shapes with notched or rounded corners, three-point star shapes with or without notched corners, triangular shapes with or without notched corners, and a circular shape with vertical ribs, each having a prototype floor area of 6,400 m². The crosswind force spectra obtained by force balance technique for two typical wind directions are shown in Fig. 18.

Fig. 18 Crosswind force spectra of buildings with different plan shapes.

The critical reduced wind velocity at which resonant vibration occurs, and which is defined by the Strouhal Number, was found to vary according to the shape and modification to the corners. The two three-point star shapes show no distinctive spectral peak associated with the vortex shedding process.

Modifications to the corners were found to cause marked reductions in crosswind force around the vortex shedding frequency and at the low reduced frequencies, particularly for the square shape. The computed crosswind displacement response for the different plan shapes based on the force balance test results shown in Fig. 18 are presented in Fig. 19.

Fig. 19 Maximum crosswind displacement response of buildings with different plan shapes.

For a prototype design wind speed of 64.6 m/s, the maximum crosswind displacement response was ranked in descending order of magnitude as follows:

1) Square shape;

2) Square shape with notched corners;

3) Square shape with rounded corners;

4) Circular shape;

5) Triangular shape with notched corners;

6) Three-point star shape;

7) Triangular shape; and

8) Three-point star shape with notched corners.

The ranking at a lower design wind speed for occupant comfort consideration is different but the plain square shape still has the highest value of crosswind displacement and hence acceleration response. The results obtained from aeroelastic tests were found to be significantly different to those derived from force balance tests.

While the plain square shape was found to generally have a larger alongwind and crosswind displacement responses than the other shapes, the circular shape clearly exhibited aeroelastic lock-in type response at a critical reduced velocity of about 8 around which the crosswind response was considerably larger than the other shapes, including the plain square shape. This highlights that aeroelastic effects can be significant for super tall buildings and force balance technique can severely underestimate the dynamic response.

Miyashita et al (1993) employed force balance technique to study the characteristics of both the wind forces acting on and the response of a square cross-section building shown in Fig. 20 with modified corners or openings. The building studied has a height to breadth ratio of 6.

Fig. 20 Aerodynamics modifications of building shapes.

Modifications to the building included chamfered or notched corners 10% of the breadth, and through building openings 25% of the breadth of the building. Coefficients and spectra of fluctuating wind forces and their correlations were measured, and the displacement response for an ultimate limit state design wind speed of 62 m/s were determined for different angles of wind incidence.

With through building openings, particularly when they were through both the X and Y building axes, substantial reductions in response were found to occur at small angles of wind incidence. However, modifications to the building corners, particularly notched corners, were found to be not particularly effective and can cause an increase in response at low angles of wind incidence.

Design Considerations:

Aerodynamic modifications of building shape, including slotted and chamfered corners, horizontal and vertical through building openings, sculptured building top, tapering, and dropping off corners have been shown to significantly reduce the wind induced loads and response of tall buildings.

Modifications to the building corners such as slotted or chamfered corners need to be applied to the corner region greater than about 10% of the building breadth to be beneficial. Modifications which increase the projected area or the effective breadth of the building would in general not be beneficial.

Reductions of up to 40% in the alongwind response can be achieved by aerodynamic modifications for a wide range of reduced wind velocities which encompasses serviceability and ultimate limit state design wind speeds for most modern tall buildings. Crosswind response at serviceability design wind speed can be reduced by around 30% while very substantial reductions can be achieved at the ultimate limit state design wind speed.

The vortex shedding characteristics of a building are often altered by aerodynamic modifications. If the Strouhal Number is lowered by the modification, the critical reduced wind velocity at which the vortex shedding frequency coincides with the natural frequency of the building will be raised to a higher value with a longer return period.

On the other hand, if the modification causes an increase in the Strouhal Number, resonant response will occur at a lower reduced velocity with a shorter return period which may increase the ultimate limit state design load or make it more difficult to meet serviceability requirements such as occupant comfort criteria. Occasionally secondary spectral peak may appear and the corresponding increase in crosswind response may adversely affect the performance of the building.

The wind load and wind induced response characteristics of a tall building are dependent on the basic plan shape of the building. The potential benefits of a plan shape which has a lower Strouhal Number is a parameter which can offer significant benefit when correctly selected.

In the design of modern tall buildings, particular for those with an unusual plan shape and for super tall buildings, a comprehensive program of wind tunnel model test is always beneficial and provides useful information which can be used to assess the potential benefits of altering the plan shape or employing other forms of aerodynamic modifications to the building.