See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237240339

WIND LOADS ON BALUSTRADES

Article

CITATIONS READS

0 172

2 authors, including:

A. W. Rofail
Windtech Consultants Pty Ltd
14 PUBLICATIONS   53 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Wind Loads on Structures that exhibit complex behaviour View project

All content following this page was uploaded by A. W. Rofail on 17 April 2014.

The user has requested enhancement of the downloaded file.

 Wind Loads on Balustrades
Christian Mans a, Antonios W. Rofail a
a
Windtech Consultants, 19 Willis Street, Arncliffe, NSW, Australia

ABSTRACT: This paper examines the wind loading on impermeable balustrade members. In
particular, the study focuses on the net pressure coefficients measured across an isolated and a
continuous balustrade member. Comparisons are presented with existing live loading recom-
mendations provided by a number of current international standards. Results suggest that the
highest positive net pressure coefficients (acting in the direction towards the centre of the bal-
cony) occur on the upper corner balcony members, specifically the side wall balustrade located
adjacent to the edge of the building. The highest negative net pressure coefficients were recorded
on balconies situated adjacent to the side edges of the building. For design, the current Austra-
lian, British and Hong Kong standards specify that balustrade members are required to withstand
the larger of either the net wind loading acting across the balustrade or a 1kN/m2 live load. For
the wind loading to dominate over the recommended live load, an eaves height 3-sec gust wind
speed greater than 30m/s is required for an isolated balustrade and 32.5m/s for a continuous bal-
ustrade.

KEYWORDS: Balustrades, Balconies, Wind Loads.

1 INTRODUCTION
Balustrade members are commonly found on low and high-rise residential apartment buildings.
However, possibly due to the relatively small size of the balustrades in comparison to the parent
structure, little published information exists regarding their design wind loading or how a balus-
trade may affect the building cladding pressures. This is primarily due to the difficulties involved
in correctly modelling a balustrade in typical wind tunnel tests. Due to blockage restrictions in a
wind tunnel and the minimum upwind fetch distance required to properly develop the atmos-
pheric boundary layer, the geometric length scale for modelling a high rise building, for exam-
ple, is typically limited to a scale of 1:200 to 1:500. At these small scales, the dimensions of the
balustrade members become very small and, as such, it is not possible to directly instrument and
measure the wind loading on the balustrade. In addition, the balustrade members are typically lo-
cated away from the face of the building, where the magnitude and direction of the local wind
may vary strongly in both time and space. In this case, the traditional pressure measurements on
the nearby building surface do not generally provide any useful information for a balustrade
member, located away from the building surface.
For design, the loading on a balustrade member is generally governed by either the worst de-
sign loading from wind pressures or from impact loading. For example, the current AS/NZS
1170.1-2002 [1] recommends that balustrades in private dwellings shall be designed to resist the
worst loading from either the wind loading or a uniform loading of 1 kN/m2 over the body of the
balustrade. A similar live loading value is also recommended in the current British Standard and
Hong Kong Building Regulations [2,3]. However, neither standard specifically specifies a
method for determining the wind loading on the balustrade.
The purpose of the current study is to directly record the wind loading on balustrade members.
The study explores three areas of concern: the loading on the balustrade itself, the effect of bal-
ustrades on the building surface pressures and the importance of including the balustrade mem-
bers in typical wind tunnel tests.
2 EXPERIMENTAL PROCEDURE
By using a large-scale 1:50 model, it was possible to instrument and measure a number of balus-
trade members on a low and high-rise building. The tests incorporated a model building with
full-scale plan dimensions of 15m and 30m. Tests were repeated for two building heights, with
eaves heights of 18m and 48m. Two balustrade configurations were examined: an isolated balus-
trade and a uniformly distributed balustrade running across the width of the building, where the
individual unit balconies were separated by vertical fins. In both cases, the balustrade dimen-
sions are defined as 3m long with a depth of 2m in full scale. The height of each balustrade wall
is 1.2m with a slab depth of 0.15m.
Tests were also repeated at two locations on the model, along the short building wall and
along the long building wall for the 18m tall building. Tests on the 48m building were limited to
the short building wall. Figure 1 shows photographs of the 18m model in the wind tunnel identi-
fying the two balustrade configurations. The model was placed in atmospheric flow conditions
that modeled a suburban terrain (AS/NZS 1170.2:2002, Terrain Category 3 [4]) and tested over
360 degrees at 15 degree increments. The mean wind speed and turbulence profiles as well as the
normalized power spectral density in the wind tunnel matched the full-scale equivalent values
for the terrain being modeled.

a) Isolated balustrade b) Continuous balustrade

Fig. 1. Photographs of the 1:50 scale model (H=18m) taken in the wind tunnel.

As shown in Figure 1, balustrade members were placed over a single 15m by 15m section of
the building. In total, 90 pressure taps were instrumented over nine balustrade members (i.e. 10
pressure taps for each balustrade). Of these, six pressure taps were positioned on the outer sur-
face of the balustrade and four taps on the inner balustrade surface. As such, the positive pres-
sures from the outer balustrade surface could be simultaneously combined with the negative
pressures from the inner surface to determine the worst net loading acting across the balustrade
member. Figure 2(a) presents the layout of the pressure taps used on the balustrade members.
Pressure taps were positioned on nine balustrade members, labeled A through to I. In analyzing
the pressure results, the balustrades were divided into three main zones: the top corner region
(Balustrades A and C), the edge region (Balustrades B, D, F, G, I) and the mid-wall region (Bal-
ustrades E and H).
As well as directly measuring the loading on the balustrade members, the study also explored
the effect of balustrades on the building surface pressures. Figure 2(b) describes the pressure tap
layout on the building surface. In particular, the study explored the importance of including the
balustrade members in typical wind tunnel pressure cladding tests. Typically, the balustrade it-
self is not included in the model due to the difficulty in correctly modeling the local Reynolds
number effects in small scale models. As such, the balcony slab is generally included on the
model, but not the balustrade itself. To examine this, three varied levels of model detail were
studied: where the balcony slab and the balustrade walls are attached, where only the balcony
slabs were included and thirdly, with no balcony attached. The second approach follows the

common practice in current wind tunnel consultancies for measuring the building surface clad-
ding pressures.

Fig. 2. Pressure tap layouts used in the study. (a) Layout of pressure tap locations measured directly on the balus-
trade. The symbol ‘+’ represents taps positioned on the outer balustrade surface and the ‘o’ symbol represents taps
positioned on the inner balustrade surface; (b) Layout of pressure tap locations measured on the building surface
behind the balcony; (c) Layout of pressure tap locations measured on the building surface of 1:400 model.

Finally, pressure measurements were recorded on a smaller 1:400 scale model, with full-scale
dimensions of 15m by 30m by 48m tall. Following from the traditional methods of modeling
balconies on tall buildings, the balustrade members were not included on the model. Instead, the
thickness of the balcony slab was increased to account for the missing balustrade walls. A total
of 18 pressure taps were included on the model, with a single pressure tap positioned behind
each balcony member. Measurements were repeated for both the isolated and continuous bal-
cony configurations and the no balcony configuration. Figure 2(c) details the pressure tap layout
and the balcony distribution used in the study.
Testing was performed in Windtech’s Blockage Tolerant Boundary Layer Wind Tunnel. All
pressures measured in the wind tunnel were initially referenced to the mean wind speed at a
height located in low turbulence conditions situated well above the model. Pressures were sam-
pled at 1024 samples per second for 64 seconds for the 1:50 model (2048 samples per second for
16 seconds for the 1:400 model). The pressure signal was later low-pass filtered at 500Hz and
digital filtering was applied for the range 0 to 500Hz. Unless otherwise specified, all pressure
coefficients presented in this paper are referenced to the 3-sec gust wind speed at eaves height,
H.

3 DISCUSSION

3.1 Wind loading on balustrade members

Table 1 presents the peak positive and negative net pressure coefficients measured on the balus-
trade members on the 18m tall building, referenced to the 3-sec gust wind speed at roof height,
H. A positive net coefficient defined as acting in the direction towards the centre of the balcony.
Conversely, a negative net coefficient is defined as acting away from the centre of the balcony.
The coefficients are separated into three zones of the building surface: the top corner region,
the edge region and the mid-wall region, as described in the preceding section. This is due to the
unique flow characteristics expected to occur within the three regions. The results are further
separated by examining the pressures measured on the side and main walls on each balcony. For
example, the outer side wall represents the side balustrade wall adjacent to the edge of the build-
ing, while the inner side wall represents the side balustrade wall closest to the centre of the
building. The main balustrade wall is defined as running parallel to the building wall.
Regarding the positive net pressure coefficients, the results suggest that the largest loading
occurred in the top corner region of the building for both sides of the building. The table also
suggests the largest positive loading occurred on the outer side walls of the balconies, for winds
approaching roughly perpendicular to the test section, with a largest pressure coefficient of 1.82.
At this wind angle the side balcony walls are directly inline with the approaching flow, but are
also within the region of separated flow at the corner of the building. Significant positive net
pressure coefficients of up to 1.3 were also recorded on the edge balustrades, while a peak net
positive coefficient of 0.71 was recorded on the centre balustrades. The positive loading on the
continuous balustrade is consistently lower than that recorded on the isolated balustrade, possi-
bly due to the added flow interference from the vertical fins resulting in a stagnation of the flow
behind the balustrade.
The largest negative net pressure coefficients were recorded on the edge balustrade members,
with a largest coefficient of -1.6 recorded on the isolated balustrade configuration. Again, this
was recorded on the outer side wall of the balustrade.
 As the current wind loading standards do not provide any definitive recommendations for
wind loading on balustrade members, comparisons with current live loading recommendations
are presented. As defined by the current AS/NZS 1170.1 2002, balustrade walls are to be de-
signed to withstand a net horizontal live loading of 1kN/m2. A similar value is recommended by
the current British Standard and the Hong Kong Building Regulations.

Configuration Location Cp max Corner Balcony Edge Balcony Midwall Balcony

Outer Inner Main Outer Inner Main Inner Main
Wall Wall Wall Wall Wall Wall Wall Wall

Isolated Short Cp max 1.58 1.27 0.93 1.29 1.05 0.55 0.71 0.59
Balcony Face Tap A1 A6 C7 G1 D6 B7 E1 E7
θ 90 315 0 90 345 15 345 345

Long Cp max 1.82 1.30 1.00 1.07 1.15 0.63 0.51 0.49
Face Tap C6 A6 A7 F6 B6 B7 E6 E7
θ 210 270 270 195 210 285 210 180

Continuous Short Cp max 1.37 1.03 0.93 1.07 1.00 0.81 0.54 0.26
Balcony Face Tap A1 A6 C7 D1 D6 D7 H6 H7
θ 90 75 15 60 0 15 0 0

Long Cp max 1.29 1.07 0.95 1.18 1.02 0.89 0.63 0.27
Face Tap A1 C1 A7 D1 G6 B7 H6 E7
θ 15 165 300 0 285 300 285 285

Configuration Location Cp min Corner Balcony Edge Balcony Midwall Balcony

Outer Inner Main Outer Inner Main Inner Main
Wall Wall Wall Wall Wall Wall Wall Wall

Isolated Short Cp min -0.57 -0.59 -0.45 -1.24 -1.03 -0.56 -1.31 -0.6
Balcony Face Tap C1 C6 C7 I1 I6 I7 E1 E7
θ 75 0 300 0 105 285 90 75

Long Cp min -0.81 -0.61 -0.65 -1.59 -0.78 -0.66 -1.16 -0.63
Face Tap C6 C1 C7 G1 I1 G7 E1 E7
θ 180 0 180 285 60 285 90 285

Continuous Short Cp min -0.41 -0.59 -0.35 -1.36 -1.04 -0.94 -1.16 -0.21
Balcony Face Tap A1 C1 A7 G1 I1 I7 H6 H7
θ 0 315 180 0 315 300 255 75

Long Cp min -0.46 -0.75 -0.42 -1.42 -1.13 -0.85 -0.76 -0.24
Face Tap A1 A6 A7 G1 F1 I7 H6 H7
θ 90 285 180 285 240 225 15 0

Table 1. Summary of maximum and minimum net pressure coefficients recorded across balustrade members on the
1:50 scale 18m building.
 Table 2 presents a summary of the net pressures recorded on the balustrade members for both
the 18m and 48m tall buildings. For simplicity, the highest net pressures are presented, irrespec-
tive of whether loading occurred on the side or main balcony wall and irrespective of the direc-
tion of the net loading. Along with the worst net pressure coefficients, the table presents the
equivalent eave height wind speeds required for the balustrade wind loading to dominate over
the recommended 1kN/m2 live load. Therefore, for design purposes, if the eave height wind
speed exceeds that presented in Table 2, the balustrade should be designed according to the wind
loading coefficients provided in Table 2. Conversely, if the eaves height wind speed is below
that presented in Table 2, the balustrade member should be designed according to the 1kN/m2
live loading.
Regarding the isolated balustrade configuration, the table suggests that balustrades should be
designed to withstand wind loading for eave height wind speeds greater than 30m/s and 32.5m/s
for the continuous balustrade.

Building Height Balustrade Configuration Corner Edge Mid-wall

Balcony Balcony Balcony

H = 18m Isolated Balcony Cp 1.82 1.59 1.31

H/B = 0.6 VH 3-sec gust (m/s) 30.1 32.2 35.5

Continuous Balcony Cp 1.37 1.42 1.16

VH 3-sec gust (m/s) 34.7 34.1 37.7

H = 48m Isolated Balcony Cp 1.65 1.75 1.56

H/B = 1.6 VH 3-sec gust (m/s) 31.6 30.7 32.5

Continuous Balcony Cp 1.05 1.17 0.91

VH 3-sec gust (m/s) 39.6 32.6 42.5

Table 2. Summary of worst net pressure coefficients measured across the balustrade members and the equivalent
eaves height 3-sec gust wind speeds required to exceed the AS/NZS 1170.1:2002 recommended 1kN/m2 live load-
ing.

3.2 Effect of balustrades on building surface pressures

Previous research has suggested that balconies may tend to increase the windward wall pressures
by up to 60% at the top corner of the building, but decrease the minimum pressures on the side
and rear walls [5]. Table 3 summarizes the worst positive and negative pressure coefficients re-
corded on the short wall surface for the three tested balcony configurations: where the balcony
slab and the balustrade walls are modeled, where only the balcony slabs are modeled and thirdly,
with no balcony attached. A description of the alternative configurations is included in the Table
heading.

Corner Balcony Edge Balcony Mid-wall Balcony

Config. Cp max 1:50 1:400 1:50 1:400 1:50 1:400
H=18m H=48m H=48m H=18m H=48m H=48m H=18m H=48m H=48m
A Cp max 0.94 0.73 0.87 1.03 0.82 0.92 1.03 0.70 0.95
Tap A7 C7 C F3 F5 F E2 H4 E
θ 45 345 330 0 45 330 0 345 15
B Cp max 0.54 0.71 0.94 0.60 0.73 1.00 0.51 0.68 0.92
Tap C7 C7 C B4 B7 D E4 E3 E
θ 315 315 345 315 15 30 0 0 0
C Cp max 1.06 0.77 1.02 0.75 0.92 0.75
Tap C2 A8 - I4 10 - E1 E7 -
θ 345 15 345 270 15 15
D Cp max 1.10 0.81 1.01 0.76 0.97 0.78
Tap A3 C4 - B6 D4 - E6 E3 -
θ 330 30 315 30 0 30
E Cp max 0.94 0.73 0.87 0.95 1.24 0.99 0.93 0.93 0.92
Tap C8 A8 C B10 A1 F E3 E8 H
θ 330 15 345 0 75 345 15 300 15

Corner Balcony Edge Balcony Interior Balcony

Config. Cp max 1:50 1:400 1:50 1:400 1:50 1:400
H=18m H=48m H=48m H=18m H=48m H=48m H=18m H=48m H=48m
A Cp max -1.31 -1.15 -1.20 -1.14 -1.23 -1.20 -0.93 -1.05 -1.20
Tap C3 A1 C F2 B7 C H4 H3 E
θ 270 90 285 270 75 285 270 75 90
B Cp max -0.72 -1.13 -1.20 -0.64 -1.20 -1.35 -0.39 -0.72 -1.14
Tap C3 A1 C F7 C3 G E3 E2 E
θ 285 75 285 285 270 90 75 90 90
C Cp max -1.53 -1.36 -1.38 -1.46 -1.0 -1.16
Tap C2 C3 - G3 A2 - E8 H5 -
θ 300 270 75 75 60 90
D Cp max -1.10 -0.95 -1.07 -1.34 -0.77 -0.74
Tap C8 A7 - F4 A1 - E3 H3 -
θ 270 180 270 90 75 75
E Cp max -1.67 -1.76 -1.27 -1.62 -2.06 -2.67 -1.37 -1.21 -1.69
Tap C3 C6 A F7 C4 G H1 E2 H
θ 270 285 75 285 195 90 90 195 90

Table 3. Summary of maximum and minimum cladding pressure coefficients measured behind the balcony.
Configuration A: Isolated Balcony with slab and balustrade attached.
Configuration B: Continuous Balcony with slab and balustrade attached.
Configuration C: Isolated Balcony with slab only attached.
Configuration D: Continuous Balcony with slab only attached.
Configuration E: No balcony attached.
 The results suggest that the isolated balconies cause higher positive and negative wall pres-
sure coefficients than the continuous balcony configuration. For the 18m building, largest posi-
tive coefficients of approximately 1.0 were recorded behind the isolated balcony in comparison
to coefficients of 0.6 behind the continuous balcony. Similar reductions were also found to occur
for the negative coefficients, with a largest negative coefficient of -1.31 recorded behind the iso-
lated balcony and -0.72 behind the continuous balustrade. This can be attributed to the added in-
terference of the flow by the vertical fins, creating pockets of stagnated flow behind the balus-
trade elements and breaking up the side vortices.
Regarding the isolated balcony configuration, reasonably accurate estimations of the cladding
pressures behind the balcony can be achieved by modeling the balcony slab and ignoring the bal-
ustrade walls. Overestimations of around 15% were obtained for positive pressure coefficients
and approximately 20% for negative coefficients.
Similar comparisons from the continuous balcony configuration were not as encouraging,
where modeling without the balustrade walls produced coefficients almost twice as large as
those recorded with the balustrade walls in place. It is believed the differences in the positive co-
efficients may be attributed to a slight shift in the stagnation point of the flow on the front sur-
face of the building. The reduction in the negative coefficients is likely due to the balustrades
aiding in disrupting the flow around the building and hence sheltering the wall surface from the
surrounding flow field and from the turbulent component of the flow. Although not shown here,
similar results were observed from measurements recorded on the long building surface.
Comparisons between the 48m tall 1:400 and 1:50 scale models are also presented in Table 3.
The results show there is similar agreement between the pressures measured on the 1:400 model
building, where only the balcony slab was included but with a height of 1.35m, and the 1:50
model building, where the balcony slab and balustrade members were modeled. It is believed
that the current methodology of modeling the balcony slab with an extended depth, rather than
including the balustrade members on the model, is a relatively accurate assumption. Results from
the 1:400 model indicate that the mid-wall balcony pressures are significantly greater than those
measured on the 1:50 model, however the pressures are only slightly lower than those measured
behind the corner and edge balconies. The corner and edge balconies show good agreement with
the exception of the maximum positive pressures in the continuous balcony configuration, which
were overestimated by approximately 35%.

4 CONCLUSIONS
A wind tunnel study was performed to directly measure the wind loading on balustrade mem-
bers. Highest positive net peak pressure coefficients of up to 1.8, when referenced to the 3-sec
gust wind speed at eaves height, were recorded on the top corner balconies. Highest negative net
peak pressure coefficients of up to -1.6 were recorded on balconies adjacent to the edge of the
building. The results suggest, for design purposes, balustrade members should be designed for
wind loading if the 3-sec gust wind speed exceeds 30m/s, otherwise the balustrade may be de-
signed for the 1kN/m2 live load.
For small scale models, in most cases the building cladding pressures measured behind the
balcony can be accurately inferred by including the balcony slab with an extended slab thick-
ness.

5 REFERENCES
1 AS/NZS 1170.1:2002, Structural Design Actions, Part 1: Permanent, Imposed and other Actions, Standards
Australia, Homebush, 2002.
2 BS 6399-1:1996, Loading for buildings, Part 1: Code of practice for dead and imposed loads, British Standards
Institution, London, 2002.
 3 Building (Construction) Regulations, Chapter 123B: Imposed loads, Department of Justice, Hong Kong, 1991.
4 AS/NZS 1170.2:2002, Structural Design Actions, Part 2: Wind actions, Standards Australia, Homebush, 2002.
5 A. Rofail, Wind pressures on the cladding of building facades, MEng. Thesis, University of Sydney, Australia,
1991.

View publication stats