Wind Loading Features for Low Buildings | Wind Engineering

Regarding wind loading of low buildings the following overall considerations are appropriate:

i. Atmospheric simulation requirements are generally similar to those for tall buildings.

ii. There is a large variety of low building geometries.

iii. Special tests are usually unjustifiable economically and hence there is more reliance on generic tests.

iv. Aspects special to low buildings are due to their highly variable environment and the nature of their loading.

It was in the 1970’s that more significant emphasis was placed on the evaluation of wind loads on low buildings. An extensive experimental study was carried out at the University of Western Ontario and examined the effect of geometrical parameters (length, height, and roof slope), scale and upstream roughness on a vast variety of flat and gabled-roof buildings.

Data included the time-varying pressures acting over various tributary areas and time- varying moments and forces at critical structural locations, as shown diagrammatically in Fig. 5. Results of the study were expressed in codified format and were included in the Supplement to the National Building Code of Canada since 1980 and the American ANSI/ASCE-7 Standard since 1982.

Typical results of the study shown in Fig. 6 indicate the effect of building height on the most critical values of pressure coefficients (peaks, mean and rms) measured at different points of the building envelope.

The study generally found that the effect of roof slope and height are significant but the dependence of the load coefficients on height can be reduced considerably by referencing them to the velocity pressure at eave or mid-roof height; over the range of lengths studied, for which the length exceeded the width, there was comparatively little change in the measured loads; the dynamic component of all loading effects was dominant over the mean component.

This was particularly true of local pressures, less so for distributed load effects. Finally, marked changes in the terrain roughness affected the pressure coefficients but trends were not always consistent; however, the dynamic component increased consistently with rougher terrain. In the 1980’s and early 1990’s research efforts were made to assess the variability of wind loads on low buildings for different surroundings as well as to evaluate the wind loads on building roofs of different configurations.

For instance, Fig. 7 taken from Meecham (1992) shows a reduction of pressure coefficients on hipped roofs in comparison to those applying on gabled roofs. This justifies the choice of the hipped roof type in several coastal areas frequently hit by violent storms.

Figure 8 shows typical results of peak suction coefficients measured on a trough edge tapping of a multi-gabled (folded) roof type. The plotted data are average values of five 16-s samples; this provides a stable estimate of the instantaneous peak suction coefficient. In general, peak suctions measured on the double-span or multi-span models tend to be larger than those obtained with the single-span model.

Wind pressure coefficients measured on stepped roofs are presented in Fig. 9 in contour form. A significant amount of information is included in these diagrams which are useful for design purposes since the data represent the most critical values of peak pressure and suction coefficients measured for each building configuration regardless of wind direction.

Further to the substantial positive pressures occurring on the low roof of the two- level configuration, much higher suctions have also been measured on the roof corner of the high roof section in comparison with those recorded on the respective corner of the low-roof section. However, these high suctions are comparable in magnitude with those obtained on the corner of the simple flat-roof configuration.

Pressures on sawtooth roofs have also been studied extensively by Holmes (1983) for low roof slopes and Stathopoulos and Saathoff (1992). Figure 10 shows the proposed codified diagram for roof angles between 10° and 30°. Design pressure coefficients have been provided for corner, edge and interior roof zones, as has been customary in the design wind standards for other building geometries. The same definition of the roof edge and corner width (z) used with other roof geometries has been applied.

However, the high corner region is twice as large as the low corner region to reflect the greater area of large suctions measured in the high corner of a sawtooth roof. Simplicity and economy have led to the specification of one curve per zone for positive and negative pressures, with the exception of the roof corner, whose specifications for span A are much higher than for other spans of the sawtooth roof.

Note that the vertical sections of the roof may be designed by using edge/corner specifications of the lower roof edge zone adjacent to them. As with the other geometries, the walls of the sawtooth- roof building may be designed by using coefficients established previously for walls of simple flat-roofed buildings.

The question of actual magnitude of pressure coefficients appropriate for the design of cladding elements on flat roof edges and corners has been researched extensively by Kind (1986) and Stathopoulos (1987).

Figure 11 shows the dramatic effect of the influence of exact pressure tap location on a roof corner. For example, for azimuths 30° to 45° pressure coefficients measured at the closest to the roof corner point have been found 2 to 3 times higher than those measured at ‘usual’ corner points.

These results were obtained with a special flat roof model equipped with an assembly of 224 tappings distributed on one corner and along its concurrent edges. Fortunately, these high suctions affect only a very small portion of the roof corner because of the high gradient of their reduction with distance from the edge.

Furthermore, the effectiveness of these high suctions may be limited in a real building because of the edge profile of the roof which may establish a local Reynolds number modifying the separation of the flow and its characteristics.

On the other extreme, Surry and Stopar (1989) have measured wind-induced suctions on the interior regions of roofs of very large low buildings and have recommended lowering the stipulations of the Supplement to the National Building Code of Canada as far as wind pressure coefficients for design purposes is concerned. Figure 12 compares the variation of peak area load coefficients measured with the current Canadian and Australian Code requirements.

Parapets and Eaves:

The effect of architectural features such as parapets, eaves, rounding the roof edges, surface roughness characteristics etc. have been studied extensively during the last decade or so. For instance, the influence of parapets on the wind loads of low buildings has been examined by Lythe and Surry (1983), Kind (1988), Baskaran and Stathopoulos (1988).

Various attempts have been made to reduce the excessive roof corner suctions caused by some parapet configurations. One such attempt was to discontinue parapets at the corner areas of the roof. Figure 13 shows roof corner pressure traces recorded for an oblique wind direction (0=45°). The high suctions induced by a parapet 0.5 m high are clearly indicated along with their reduction caused by cutting the parapet at the ends of the edges for a length equal to the parapet height.

In spite of the significant reduction of suctions caused by this configuration, suctions are still higher than those measured in the no-parapet case. This has also been shown for a number of parapet heights. Data show that regardless of parapet height, roof corner suctions are approximately equal for the cut-parapet configuration and somewhat higher than the no-parapet case.

Consequently, roofs with low parapets are benefited mostly by this cut configuration. The effect of eaves on the wind loads of low building roofs has been examined by Robertson (1991) and more recently by Stathopoulos and Luchian (1994) for gabled roof buildings of different roof slopes.

Figure 14 shows contours of the most critical peak pressure coefficients measured, regardless of azimuth and eave height, on the upper and lower overhang surfaces for the two different roof slopes tested.

Clearly, the 4:12 sloped overhang is more severely loaded, particularly as far as suction on the upper surface near the gable is concerned. Positive pressures on the upper surface are higher for the 12:12 slope as expected from wind flow considerations. However, lower surfaces are subjected to higher pressures and suctions for the 4:12 slope.

Pressures measured on lower eave surfaces have been compared extensively with values measured on the walls of the models. Although there is a general agreement, it has also been found that the highest positive peaks measured on the lower eave surface are larger than the respective values measured on the wall surfaces.

It is interesting to note that, although wind loads on eave sections have been measured directly, it has not been possible to measure loads on parapet sections themselves due to scaling difficulties. At present, such measurement appears feasible only in full scale.

Internal Pressures:

There has been a fair amount of controversy regarding the appropriate magnitude of internal pressures for design purposes. Additional uncertainty is generated due to the influence of other factors such as envelope porosity and building openings.

Figure 15 shows a typical example of the measured internal pressure in the model building during the second, in which the sudden opening occurs. The short time achieved for the realization of the sudden opening is apparent, as is the magnitude of overshooting, which is lower than the level of the subsequent instantaneous peak values. This has consistently been the case regardless of the method used for creating the sudden opening.