A Case Study of Using Metal Scaffold System for Demountable Grandstand: The Opening Ceremony of Hong Kong 2009 East Asian Games

Structural Engineering Branch, ArchSD File code : TemporaryGrandstand.doc

Information Paper on Demountable Grandstand CTW/MKL/CYK/LKN

Issue No./Revision No. :1/- Issue/Revision Date : December 2011

Information Paper

A Case Study of Using Metal Scaffold System for

Demountable Grandstand:

The Opening Ceremony of Hong Kong 2009 East Asian Games

STRUCTURAL ENGINEERING BRANCH

ARCHITECTURAL SERVICES DEPARTMENT

December 2011

Structural Engineering Branch, ArchSD File code : TemporaryGrandstand.doc


Issue No./Revision No. :1/- Issue/Revision Date : December 2011

Table of Contents

1. Introduction ................................................................................................... 1

2. Structural Behaviour and Components ......................................................... 4

3. Design Loading ............................................................................................. 23

4. Dynamic Effects ............................................................................................28

5. Foundation .................................................................................................... 30

6. Construction Supervision .............................................................................. 30

7. Case Study .................................................................................................... 31

8. References ..................................................................................................... 52

Annex A Sample Checking Certificate

Structural Engineering Branch, ArchSD Page 1 of 55 File code : TemporaryGrandstand.doc


Issue No./Revision No. : 1/- Issue/Revision Date : December 2011

1 Introduction

1.1 A grandstand is a structure which provides seating for spectators at entertainment

or sporting events. Grandstands are typically classified into three distinct types:

permanent, demountable and retractable. Structural Engineering Branch (SEB) has

promulgated a set of guidelines in September 2011 - SEB Guidelines SEBGL -

OTH5: Guidelines on the Design for Floor Vibration Due to Human Actions Part

III: Vibration Effect to Grandstands, Sensitive Equipment and Facilities (available:

http://asdiis/sebiis/2k/resource_centre/) – providing guidance the effect of human

induced vibration on permanent grandstands. This paper will focus on the analysis,

design and construction of demountable grandstands by sharing the experience on

the demountable grandstands erected for the Opening Ceremony of Hong Kong

2009 East Asian Games held on 5 December 2009.

1.2 Demountable stands (Photo 1(a)) are lightweight temporary structures whose

trussed appearances are reminiscent of scaffolding systems. These stands are

typically erected for a single specific event (e.g. parade, sports, and show) and

therefore left in place for a short duration to house the large number of spectators.

However, in some events (e.g. in the Opening Ceremony of Hong Kong 2009 East

Asian Games), such demountable grandstands may have occupancies of up to

thousands of people. Demountable grandstands were widely used in the Sydney

2000 and Beijing 2008 Olympics Games, and in the recent Auckland 2011 Ruby

World Cup to increase the seating capacity of the competition venues. Photo 1(b)

shows a large-scale example of demountable grandstands used in the softball centre

of the Sydney 2000 Olympics Games, where 7,000 additional seats were provided

by such demountable grandstands, and Photo 1(c) shows the scaffold system of

another large-scale example of demountable grandstands used in Eden Park

Stadium of the Auckland 2011 Ruby World Cup, where 10,000 additional seats

were provided by such demountable grandstands.

Photo 1(a) Typical Demountable Grandstand (Bellinzona, Switzerland)

http://asdiis/sebiis/2k/resource_centre/




Photo 1(b) Demountable Grandstand in Softball Centre at

the Sydney 2000 Olympics Games

(Source: www.austseat.com.au/)

Photo 1(c) Scaffold System of the Demountable Grandstand

in Eden Park Stadium at the Auckland 2011 Ruby World Club

(Source: www.zimbio.com/pictures/-sUagbFggBA/)

PH 1.3 Unlike permanent structures, demountable grandstands are usually designed to be

repeatedly assembled and disassembled with the use of lightweight components

such as slender steel tubes. The supporting structure and the member connections

for such grandstands are also designed to make the assembly easily, rapidly and

usually with the use of various types of proprietary scaffold system. Usually,

http://www.austseat.com.au/

http://www.zimbio.com/pictures/-sUagbFggBA/




demountable stands are proprietary products designed, supplied in modular units

and installed by specialist contractor employed by the event organizers. Moreover,

to ease installation, the supporting scaffold structures consist of slender tubular

members with short spans between supports, rather than having larger steel sections

with longer spans in permanent grandstands. Ellis and Ji (2000) further note that

because of the short spans and slender tubular scaffolds, sway and front-to-back

vibration in horizontal direction are often the most important modes for

demountable grandstands for human-induced dynamic crowd loads, while vertical

modes are usually not a significant problem.

1.4 Because of the limited time for installation and the incentive to save cost, the

structure of such demountable grandstand will just be able to achieve the minimum

factor of safety. A number of accidents involving the collapse of such demountable

grandstands have occurred overseas resulting in a number of casualties. Typical

causes of these collapses are: overloading, lack of bracing, failure in support,

problems with connections, and synchronized movements of audience (de Brito and

Pimentel 2009). Two serious incidents of collapse of demountable stands occurred

in the UK during 1993 and 1994. The UK Department of the Environment

therefore appointed the Institution of Structural Engineers, who in collaboration

with the Steel Construction Institute, published a guide for clients, contractors,

engineers and suppliers of demountable structures. This guide has then been

updated with latest technological and regulatory changes, and its latest version is

published as Temporary Demountable Structures: Guidance on Procurement,

Design and Use (IStructE 2007).

1.5 The performance of demountable grandstand is the subject of this paper. First,

structural forms and structural components of common types of grandstand are

presented and discussed. Next, the design loading (including the dynamic loads)

for the analysis and design of such structures will be detailed. Finally, the analysis,

design, erection, and inspection process for the demountable grandstands erected

for the Opening Ceremony of Hong Kong 2009 East Asian Games grandstands is

presented and discussed.




2. Structural Behaviour and Components

2.1 Demountable grandstands can be assembled in a variety of shapes and sizes

depending on the client requirement, nature of the event, weather, type of spectator

and terrain. These structures (Photo 2(a) and Photo 2(b)) typically have seats or

benches arranged in tiered rows with access to the seats from aisles that run

perpendicular to the rows of seating (Figure 1). These structures will usually be

dismantled once the event is completed. Most common demountable grandstands

are one that has a scaffold structure with bracing to provide lateral stability to

which a modular floor and seating system is fixed at the top.

Photo 2(a) Demountable Grandstand (Front Elevation)

(Source: www.layher.com)

Photo 2(b) Demountable Grandstand in Australia (Rear Elevation)

(Source: http://www.austseat.com.au/)

http://www.layher.com/





Figure 1 Typical Demountable Grandstand (Plan)

(Source: Crick and Grondin 2008)

2.2 The structural behaviour of these structures is complex due to the presence of

countless components connected by clamps or simply inserted into each other. The

structural system is such that spans are reduced due to a significant number of

vertical supports, and flexural stiffness in a vertical direction would benefit from

that. On the other hand, along the line of seats (or sometimes called “sway”)

direction, stiffness is mainly due to bracing, whereas perpendicular to the line of

seats (or sometimes called “front to back”) direction, apart from bracing, there is

also the presence of frames to support the seats then contributing to stiffen the

structure in this direction. It is thus expected that the flexibility of the structure in

each direction varies significantly, with implications on its static as well as dynamic

behaviour.

2.3 Components of Grandstand

2.3.1 Demountable grandstands are structures generally made of steel and consist of

members, connectors, and planks erected on site. The structural system is a modular

three-dimensional frame, in which height and length of the structure are adjusted

during design to accommodate a specified number of users. Such 3-D frame

consists of proprietary scaffolds (instead of conventional structural steel sections as

the supporting structure for demountable grandstands due to the relatively faster

speed of assemble and lower material and erection costs of scaffold. Structurally,

these scaffolds serves as “support scaffold” (instead of as “access scaffold”), and

are required to carry heavy imposed load similar to falsework used in concreting.




However, those scaffolds used in falsework for concreting are sometimes

manufactured as planar moment-resisting frames (or called the “door-type” scaffold

Figure 2(a)). For the proprietary scaffold used in demountable grandstands

(Figure 2(b)), the joints are usually assumed to be pinned-connection, and its

stability must rely on the brace members. Figure 2(c) shows the structural

components of a typical scaffold used for demountable grandstands. Most

proprietary demountable grandstands have similar components in their structures,

and the various common types of connector and bracing members in the scaffold

will be discussed in the following paragraphs.

Figure 2(a) Frame or Door-Type Figure 2(b) Proprietary Scaffold

Scaffold Demountable Grandstand

Figure 2(c) Structural Components of Typical Scaffold

(Source: Rasmussen and Chandrangsu 2009)

2.3.2 Tubes

To ease erection, all proprietary scaffolds are supplied as a modular system with

tubes and connectors. Structural steel tubes are used to make up of the three

elements of a modular unit: standard (the vertical element), ledger (the horizontal

element), and brace (the diagonal element). The standards are connected to create a

lift via connection tubes in sleeve joints (Photo 3), and are connected to the ledgers




via connectors (Figure 4). Figure 3 shows an example on the size of the

connection tubes, sizes and dimensions of the pins for connecting the connection

tube and standards as extracted from a supplier’s catalogue. Table 1 shows typical

sizes of these three elements in a modular unit (Crick and Grondin 2008). Crick

and Grondin (2008) note that the steel tubes are generally of Canadian Standard

Grade 40.21 300W with minimum yield strength of 44ksi (300MPa). However,

this paper has reservation on the applicability of such general statement, especially

to those tubes used in Hong Kong. Hence, project officer should check the country

of origin of the proprietary scaffold and refer to the catalogues for scaffold to

determine the yield strength of the steel tubes.

Photo 3 Connection tube between upper and lower standards

Figure 3 Connection Tube between Upper and Lower Standards

(Source: http://www.scaffoldgold.com)

Connection Tube

Pin Hole

Details of

Connection Tube

Pin Hole

http://www.scaffoldgold.com/




Table 1 Typical Sizes of Tubes of a Modular Unit for Scaffold

Elements Length Diameter (mm) Thickness (mm)

Standard 0.5m, 1.0m, 1.5m, 2m or

3m 49 3.2

Ledger US: 2.13m or 3.05m

Europe 2.25m or 1.35m 49 3.2

Brace Length to suit standard

and ledger 45 2.3

2.3.3 Connector

There are several systems of connector that can connect the tubes together. In this

paragraph, three common systems, namely couplers, wedge-based connectors and

spigot (Figure 4), will be described.

Right-angle Swivel

(a) Coupler/Clamp (b) Wedge-based connector

(c) Spigot

Figure 4 Common systems of connector

(Source: De Brito and Pimentel 2009, and Labour Department 2001)

2.3.1.1 Coupler/Clamp

Standards are connected to ledgers via right angle couplers (Photo 4(a)), and

braces are connected to the scaffold via swivel couplers (Photo 4(b)) to form tube

and couple (or tube and clamp) scaffold.




Photo 4(a) Right-Angle Couplers/Clamps Photo 4(b) Swivel Couplers/Clamps

(Source: www.tubeandclampscaffold.info) (Source: www.aptsuspensions.co.uk)

Tube and clamp scaffold is commonly used in construction. The ledgers and thus

walking-decks can be placed at any height along the standard, and standards can

be spaced at any distance apart up to the maximum distance allowed by

engineering constraints. Tube and clamp scaffold is also the simplest and

versatile system; but is among the most labour-intensive of all scaffolding

applications, and is therefore generally used only when high capacity, unlimited

adaptability and versatility are required.

2.3.1.2 Wedge-based connector

Kwikform (or KwikStage) scaffold (Photo 5(a)) and allround scaffold are using

wedge-based connector. A distinct feature for such connection system is that the

wedge pin can provide some moment carrying capacity, and hence, unlike the

other systems, braces are sometimes not provided for such scaffold system in light

loading (e.g. access scaffold). Further discussion on the effectiveness of such

connection in carrying moment will be given in Section 2.5. Kwikform scaffold

has metal loops attached to the standard at fixed intervals. The ledger has a

hooked head that fits into the loop and a wedge pin to tighten the connection

(Photo 5(b)). A hammer blow is used to drive the wedge pin between the ledger

head and the loop creating a secure connection. The wedge fixing of the ledgers

gives a simple and fast means of erecting access scaffolding without loose parts,

its rigid 4-way fixing giving a positive location without movement, and wedge

fitting on the standard giving guaranteed vertical alignment. To install braces,

steel tubes (Photo 5(c), 5(d) and 5(e)) with pivoted wedge devices at each end

fitting onto the standards.

http://www.tubeandclampscaffold.info/

http://www.aptsuspensions.co.uk/




(a) (b)

(c) (d) (e)

Photo 5 Kwikform System (Source: www.scaffoldingcn.com and www.rmdkwikform.com)

Allround scaffold (Photo 6(a)) has a rosette (i.e. circular plate with slots) attached

at fixed intervals along the length of the standard. The ledger has a ledger head at

each end that has a horizontal slot that mate with a wedge pin drops down into the

slot on the rosette. The rosettes have 8 slots that allow up to eight members at one

connection. To make a connection, the wedge head is slid over the perforated

rosette. A harmer blow is then used to force the wedge pin into the slot securing

the ledger to standard. Typical connection procedure of the system is shown in

Photo 6(b).

http://www.scaffoldingcn.com/

http://www.rmdkwikform.com/




Photo 6(a) Allround System


Photo 6(b) Connection Procedure of the Allround Scaffold


2.3.1.3 Spigot

Cuplock (or cuplok) scaffold (Photo 7(a)) uses spigot to connect ledger and

standard together. Spigot is a cuplike element fixed to the standard at set intervals

along its length. It allows four members to be connected at one place.






Photo 7(a) Cuplock Scaffold

(Source: www.scaffoldgold.com and www.indiamart.com)

To make a connection, the ledger end is placed into the bottom cup and the top

cup is screwed down on to the top of the ledger end locking it into place. A

hammer blow is used on the top cup to tighten the connection. Thus the top and

bottom of the ledger head is secured against the standard. Typical connection

procedure of the system is shown in Photo 7(b). Holes are intentionally left in

the upper and lower standards so that pins can be inserted so that the standard is

of the correct plumb and to transmit tensile force along the standard.

Photo 7(b) Connection Procedure of Cuplock Scaffold

(Source: www.scaffoldingasia.com)


http://www.indiamart.com/

http://www.scaffoldingasia.com/




Photo 7(c) Special Braces for Cuplock Scaffold

(Source: www.scaffoldgold.com and http://scaffoldsales.com)

A distinct feature for such connection system (like wedge-based connector) is that

it can provide some moment carrying capacity, and hence braces are sometimes

not provided for such scaffold system (especially as access scaffold). Again,

further discussion on the effectiveness of such connection will be given in Section

2.5. To install braces, steel tubes (Photo 7(c)) with couplers at each end for

fitting onto the standards.

2.3.1.4 For demountable temporary grandstands, modular system scaffold such as the

cuplock, Kwikform and allround scaffolds are used most frequently. Scaffold in

the form of tube and clamp scaffold are considered too slow in construction and

labour intensive.

2.4 Bracing

2.4.1 One of the main reasons for the collapse of demountable grandstands is an

insufficient number of bracing members provided (Bolton 1992; Ji and Ellis 1997).

Demountable grandstands must therefore be provided with sufficient bracing

members to resist horizontal loads and wind loads. It is essential that diagonal

bracing be installed at all times. Free-standing individual support towers, and the

start and end bays must have diagonal bracing installed. Moreover, the bracing

elements also have effects on the permissible loadings on the standards. The

following figure shows 5 arrangements of bracing element in a scaffold system,

namely (0) every bay, (A) every 2nd

bay, (B) every 3rd

bay, (C) every 4th

bay, and

(D) every 5th

bay, in the descending order in their permissible loadings:


http://scaffoldsales.com/





2.4.2 Theoretical study

Extensive study of the effect of the bracing on such lightweight scaffold system has

been carried out by Ji and Ellis (1997). The governing principles in providing

bracing are:

1) the load shall take the shortest path to the supports (the “direct force path

principle”); and

2) the internal forces shall be uniformly distributed (the “uniform force

distribution principle”).

Based on these two principles, they listed out the following five criteria for

arranging bracings in an efficient way in order to achieve a larger lateral stiffness:

(a) Bracing members in different storeys should be provided from the top to the

support of the structure.

(b) Bracing members in different storeys should be directly linked where possible.

(c) Bracing members should be linked in a straight line where possible.

(d) Bracing members at the top adjacent bays should be directly linked where

possible.

(e) If extra bracing members are required, they should be used following the

above four criteria.

2.4.3 Table 2 shows six examples of typical bracing arrangement with descriptions on

which criteria as listed above can be fulfilled. Those systems fulfilling all the above

criteria would perform a higher static stiffness when subjected to horizontal loading.





Table 2 Typical Bracing Arrangement for Scaffold

Type Bracing Arrangement Descriptions

1

- Satisfy criteria (a)

- Traditional bracing form

- Load transfer from top through all

members

2

- Satisfy criteria (a) and (b)

- Shorter load path than type 1, higher

static stiffness

3

- Satisfy criteria (a), (b) and (c)

- More straightforward force path

- Higher stiffness than type 1 & 2

4

- Satisfy criteria (a), (b), (c) and (d)

- Highest stiffness amongst the first 4

types

5

- More bracing members used, but not

fully follows the criteria

- Lower stiffness than type 4

6

- Satisfy all 5 criteria

- More uniform inner force distribution

- The highest stiffness among all the

above types

(Source: Ji and Ellis 1997)

2.4.3 Among the six types of bracing arrangement, Type 6, which consists of a pair of

straight cross-bracing from the top to bottom, is the most effective bracing system.

Type 6 in Table 2 only shows the ideal arrangement for a scaffold of two storeys

height. Figure 5(a) shows how to modify Type 6 arrangement for scaffold with

more than two storeys. Such bracing system can satisfy the first three criteria, and

has small number of bracing members. Ji (2003) carried out tests on three models

(frames A, B and C) of bracing system (Figure 5(b)(i)), which were made up of

aluminium members same cross-section of 25 mm by 3 mm with an overall

dimension of 1.025 m×1.025m. The frames were fixed at their supports and a

hydraulic jack was used to apply a horizontal force at the top right-hand joint of the

frame. At a load of 1.07kN, the horizontal displacement were respectively 3.0mm

for frame A, 0.73mm for frame B and 2.2mm for frame C (Figure 5(b)(ii)). They




therefore verified that Type 6 arrangement, which satisfies all five criteria, is the

stiffest.

Figure 5(a) Type 6 Bracing for Multi-Storey Scaffold

Figure 5(b)(i) Models of Bracing Systems

(frames A, B and C being placed from the left to right)

(Source: Ji 2003)

Figure 5(b)(ii) Deflection Curves of Frames A, B and C

(Source: Ji 2003)




2.4.4 Similar remarks have been made earlier in Grant (1975). Grant (1975) notes that

Type 6 bracing arrangement can result in the vertical loads in the standards to resist

the couple created by the horizontal loads to vary proportionately with their

distance from the centre line (Figure 5(c)), whilst Type 1 or 2 bracing arrangement

will create large vertical forces in the two legs adjacent to the standards in the

bracing bay (Figure 5(d)). Grant (1975) further notes that Type 6 bracing

arrangement can effectively redistribute a concentrated vertical load onto the other

standards (Figure 5(e)).

Figure 5(c) Induced vertical loads on standards using Type 6 bracing

(Source: Grant 1975)




Figure 5(d) Induced vertical loads on standards using Type 2 bracing


Figure 5(e) Redistribution of concentrated vertical load using Type 6 bracing





2.4.5 Type 6 bracing arrangement, however, has the following disadvantages:

a) there is no the bracing to the scaffold until its installation is completed, and

hence cannot provide the required stability during erection and dismantle works;

b) the erection of such bracing is much more difficult than the other types, as it is

difficult to align the bracing straight, especially with in the jungle of scaffolds

underneath the seating deck, and Bolton (1997) also raised concerns on the

potential instability of the scaffold system during dismantle when such bracing

has been removed;

c) under the effect of lateral force, the bracing will induce a large tensile force

onto the edge scaffold, which may not be of adequate strength, and anchor or

kentledge may be required at the foundation level to counteract the tensile force;

and

d) such bracing may be required to tie to the scaffold at intermittent levels (Figure

5) by swivel couplers, producing a moment on the thin tubes of the scaffold.

Grant (1975) further comments that in the case that it is not possible to tie the

bracing member to a standard at a node, a ledger is preferred to a standard for

such coupling, as the former is not already heavily loaded;

Hence, in reality, such bracing system will not be adopted by most specialist

contractors, although theoretically such bracing system provides the highest

stiffness. Instead, the common arrangement of bracing system adopted is Type I

(Photo 1(c) and Photo 8).

Photo 8 Typical Bracing System

(Source: http://www.austseat.com.au)

2.4.6 However, even though Type 1 bracing system is usually adopted as bracing system

in such modular scaffold, this paper still recommends that global Type 6 cross-

bracing should be provided around the scaffold system in addition to Type 1 so as

to increase the overall stiffness of the scaffold, especially when the demountable

grandstand is tall. The number of Type 1 bracing may also be reduced. A

suggested arrangement is shown in Figure 6. With such global bracing, there is the

potential for eight braces to interest at one connection, exceeding the maximum of





4 braces even for allround scaffold system. Hence, it is necessary to attach the

braces to the standard or ledger using swivel couplers.

Figure 6 Preferred Bracing Arrangement at End Bays

2.5 Moment Carrying Capacity of Joint

2.5.1 It is generally assumed in the analysis and design that the joints in these scaffolds

are modelled to be pinned connection (i.e. with no moment carrying capacity). In

reality, they have some degree of fixity, especially the cuplike element in cuplock

scaffold and wedge-based connector in Figure 7(a). The cuplock connections

behave as semi-rigid joints, and show looseness with small rotational stiffness at

the beginning of loading. Once the joints lock into place under applied load, the

joints become stiffer (Godley and Beale 1997). Wedge-type joints are generally

more flexible and closer to pinned connections. They also often display substantial

looseness at small rotations (Godley and Beale 2001). As to spigot joints in the

cuplock system, the spigot can create out-of-straightness of the standards, and the

possibility of the joint to open up due to the gap between the standard and the

spigot can produce complexity in modelling (Enright et al 2000). Figure 7(b)

shows typical moment–rotation curves for cuplock and wedge-type joints. It

should further be noted that the relationship for all types of joint is not generally the

same for positive and negative rotations (Godley and Beale 2001), and that the

curves do not show linear relationship.

2.5.2 Although there is moment carrying capacity of the connections (and indeed, the

catalogues of many proprietary scaffold systems also provide their recommended

moment carrying capacity), project officer should note that the uncertainty and

limitations of such connections in carrying moment as discussed in the above

paragraph. This paper still maintains that in the design of demountable grandstand,

the joints should be modelled as pinned connection, and bracing members are

required to provide lateral stability for the scaffold. The moment carrying capacity

of the connections only serves as additional safety margin, especially during the




erection and dismantle processes, where the bracing members have not yet been

installed or has been dismantled.

(i) Wedge-based connector (ii) Cuplock connector

Figure 7(a) Moment carrying capacity in wedge-based and cuplock connector

(Source: Godley and Beale 1997, 2001)

Figure 7(b) Typical moment against rotation graph of joint

in proprietary scaffold

(Source: Chandrangsu and Rasmussen 2006)

2.6 Elastic Critical Load Factor λcr

2.6.1 Eurocode 3 defines the elastic critical load factor λcr as the value of the load factor

by which the loads are to be multiplied to check of buildings for “sway mode”

failures. λcr is therefore an important parameter to classify the scaffold frame into

non-sway, sway or ultra-sway sensitive frame. For both sway and ultra-sway

sensitive frames, the load-carrying capacity of the steel tubes in the scaffold system

decreases with the height of the scaffold, as the effective length of tubes increases




due to the P-- effects. The Code of Practice for the Structural Use of Steel 2005

(the “HK Steel Code”) issued by Buildings Department (as modified by SEI

08/2009: Design Code for Structural Steel) classifies frame using λcr as follows:

a) when λcr 10, the frame is non-sway and the P-- effects can be ignored;

b) when 5≤ λcr<10, the frame is sway and the P--δ effects can be included by

checking the members by either the moment amplification or effective length

methods; and

c) when λcr<5, the frame is ultra-sensitive sway frame, and second order

analysis should be used in the analysis and design to include the P-- effects.

2.6.2 Clause 6.3.2.2 of the HK Steel Code provides two methods to calculate λcr, namely,

the deflection method or the eigenvalue analysis. If eigenvalue analysis is used to

calculate λcr, project officer should first study the form of the buckling mode of the

frame to see if it is a sway buckling mode (Figure 8(a)) or a local column buckling

mode (Figure 8(b)). King (2005) commented that when using eigenvalue analysis

in finding the first sway-mode, “it is important to study the form of each buckling

mode to see if it is a frame mode or a local column mode. In frames where sway

stability is ensured by discrete bays of bracing (often referred to as “braced

frames”), it is common to find that the eigenvalues of the column buckling modes

are lower than the eigenvalue of the first sway mode of the frame. Local column

modes may also appear in unbraced frames at columns hinged at both ends or at

columns that are much more slender than the average slenderness of columns in the

same storey.” Similarly, Rathbone (2002) noted that “where the columns are axial

load predicated, many of the lower buckling modes will be [local] column buckling

modes. It is the [sway] buckling mode of the whole structure that is important” to

include second-order effects.

2.6.3 Therefore, both the deflection method and eigenvalue analysis are applicable to

calculate λcr for the frame with sway buckling mode; but if it is a local column

buckling mode, then the lowest eigenvalue found by the eigenvalue analysis does

not represent the first sway buckling mode λcr. Instead, the eigenvalue analysis

finds the eigenvalue for the column buckling mode. In such case, project officer

should take care in using the eigenvalue analysis and not just use the lowest

eigenvalue which may be local column buckling mode (Figure 8(b)) and it is not

the original intent to use it to define sway sensitivity. Should eigenvalue analysis

be adopted, project officer should therefore scan the output to see which eigenvalue

is the first sway buckling mode. Alternatively, in such case the deflection method

can be used to calculate λcr for the first sway buckling mode.




Figure 8(a) Sway Buckling Mode Figure 8(b) Local Column Buckling Mode

3 Design Loading

3.1 Demountable grandstand should be designed to form a robust and stable three-

dimensional structural arrangement, which will support the design loadings for the

required period with an adequate margin of safety. The loading appropriate for the

design of demountable grandstand comprises dead, imposed, wind and notional

horizontal loads and may require consideration of dynamic loads from crowd.

3.2 Dead Load

Dead load shall include the self-weight of all fixed elements that form part of the

demountable structure.

3.3 Imposed Load

The minimum imposed load on demountable grandstands is controversial. BS

6399: Part 1 (BSI 1996) originally recommended it to be 5.0kPa, which is the same

as that specified in the Code of Practice for Dead and Imposed Loads 2011 (the

“HK Loading Code”) issued by Buildings Department. However, BS 6399: Part 1

has removed the specified minimum imposed load for grandstands in its

amendment in 2002. At the same time, for assembly areas with fixed seating, both

BS 6399: Part 1 and the HK Loading Code specify a minimum imposed load of

4.0kPa. Hence, project officer is advised to exercise judgment on choosing the

minimum imposed load to be adopted in individual case.

3.4 Wind Load

3.4.1 Design Wind Pressure

The design wind pressure acting on such temporary structures to be adopted in the

analysis and design is a controversial issue. The Code of Practice on Wind Effects

in Hong Kong 2004 (the “HK Wind Code”) issued by Buildings Department are

applicable for permanent building structures. It gives the extreme 3-second gust

wind velocity (and hence the wind pressure) for a return period of 50 years.

However, for demountable grandstand, it is generally intended to be used for a

short duration, and will then dismantled. The HK Wind Code states that the basic

wind pressure may be modified by a factor of 0.7 for temporary building with

design life less than one year. However, Buildings Department, when reviewing

the design wind pressure for hoarding in 1999, recognised that the factor of 0.7 is

conservative, as the original intent was that a temporary structure may last for more

than one year though its design life is one year. Buildings Department therefore

issued APP-21 (Demolition Works - Measures for Public Safety) and APP-23 (Hoardings, Covered Walkways and Gantries (including Temporary Access for

Construction Traffic) - Building (Planning) Regulations Part IX) allowed the

design wind pressure to be modified by a factor of 0.37 in hoarding design, as it




was noted that such hoarding would usually last for not more than three years, with

a probability of exceedance of 63.6%.

Moreover, another factor to be considered in choosing the design wind pressure is

the seasonal effect. The design 3-second gust at gradient height of 500m in the HK

Wind Code is specified to be 78.7m/s (i.e. a design wind pressure of 3.72kPa), and

the highest recorded 3-second gust in Hong Kong was 65.0m/s (i.e. a design wind

pressure of 2.54kPa) in 1999 due to Typhoon York at Waglan Island at a height of

90m above the mean sea level. BS 6399: Part 2 (BSI 1997) allows partial factors to

be used by taking account of seasonal variations in wind speeds and if necessary by

altering a probability factor to accept a greater than usual degree of risk that the

design wind speed may be exceeded.

During summer, it is unlikely that when typhoon signal no. 8 or above is hoisted, en

event will not be cancelled or all the workers will still be required to work on the

grandstand. A common situation is that when a typhoon signal no. 3 is hoisted,

workers are still required to carry out erection or dismantle work and an event will

still be held with the seats occupied. For a typhoon signal no. 3 is hoisted, the

maximum hourly mean wind speed is only 62km/h, which corresponds to a 3-

second gust of 23.94m/s. During autumn or winter, easterly monsoon wind prevails

in Hong Kong. The highest recorded monsoon 3-second gust at Hong Kong

Observatory was only 29.9m/s (i.e. a design wind pressure of 0.54kPa), which

occurred in 1934 at 61m above the mean sea level, and this recorded value is

already the maximum 3-second gust measured in autumn or winter in Hong Kong

over the past 70 years. As a compromise, this paper therefore considers that a

design 3-second gust wind velocity of 23.94m/s (i.e. a design wind pressure of

0.41kPa) to be adopted for the design at a height of 0m-10m for temporary

grandstand that is likely to experience typhoon during its service life, which

provides a conservative estimate of the wind pressure when the grandstand is

occupied or when the grandstand is being erected or dismantled. Project officer

should exercise judgment in deciding the appropriate design wind pressure to suit

the particular case.

3.4.2 Force Coefficient

The HK Wind Code gives the force coefficient with different solidity ratios for open

frame structures. The choice of solidity ratio for such open multiple frame

structures is again controversial, especially whether the upstream frame can shield

the frames behind (Figure 9). As any upstream obstruction to a permanent

structure may be demolished during the design life of the permanent structure, the

HK Wind Code states that no allowance shall be made for the general or specific

shielding of other structures or natural features. BS 6399: Part 2 (BSI 1997) also

states that an estimate of the total wind load can be obtained by summing up the

loads on each individual frame; but admits that such summation may be very

conservative, especially when the frames are dense and shielded as for demountable

grandstands of this paper. Choi (1984) noted that the amount of shielding depends




on the solidity ratio of the upstream frame and the spacing between frames, and

recommended to reduce the wind force on the downstream frames by a shielding

factor. BRE Special Digest SD5 (Blackmore 2004) also recommends similar

procedures to calculate the shielding factor to be included in calculating the total

wind load for open frame structures. Figure 10 gives the total force coefficient Cf

as a function of the solidity ratio of the upstream frame, the spacing S, the width

of the structure in the direction of the frame, and the number of frames.

Figure 9 Shielding Effect in Open Frame Structures




Cf

(a) B/S = 2

Cf

(b) B/S = 5

Cf

(c) B/S = 10

Figure 10 Shielding Effect on Force Coefficient Cf for Open Frame Structures

3.4.3 Dickie (1983) mentioned two aspects to be considered in the design of demountable

grandstands to wind loads: possibility of structural damage at high wind speeds

when the grandstand is empty, and increase in load actions at low wind speeds with

spectators present on the structure. �IStructE (2007) recommends three load cases

to be included in the analysis and design:




1) a high load factor of 1.4 for the wind loading when the grandstand is empty

combined with a load factor of 1.0 for dead load to check foundations;

2) equal load factors of 1.2 for wind, dead, live, and nominal horizontal loads to

check foundations; and

3) equal load factors of 1.0 for all four aforementioned loads to check

deflections.

3.5 Dynamic Load

3.5.1 For the design of any structure subject to dynamic loads the avoidance of resonance

effects is important. Demountable grandstands are relatively flexible structures

which will respond dynamically to spectator movements. In an investigation of the

dynamic response of 40 empty demountable grandstands carried out by Littler

(1996), only one grandstand was reported to have a vertical natural frequency

below 9 Hz. On the other hand, the natural frequencies of the grandstands in the

sway direction were much lower, in a range from 1.8 to 6.0 Hz, the majority being

between 3.0 and 5.0 Hz. With regard to the front-to-back direction, in most cases

the natural frequencies were higher than those in the sway direction, ranging from

2.1 to over 10 Hz.

3.5.2 Littler (1996) also carried out tests in some grandstands in use. In one of the tested

grandstands, a reduction from 2.7 to 1.7 Hz was observed in the sway direction due

to the presence of a passive audience. Horizontal natural frequencies above 4 Hz

for empty temporary grandstands were therefore cited as a recommendation to

avoid the range of maximum dancing frequencies. This would probably include an

allowance to take into account the effect of human-structure interaction for

structures in use. Horizontal natural frequencies above 4 Hz for the empty structure

were also mentioned in BS 6399: Part 1 (BSI 1996) as a design strategy to avoid

significant resonance effects.

3.5.3 The significance of the natural frequencies to the design is related to the possibility

of potential resonance of the structure due to excitation produced by the spectators,

e.g. in pop concerts or sports events. In the cases in which the fundamental natural

frequencies of the structure are such that potential resonance can occur and there is

a potential for synchronized and periodic movements, a full dynamic analysis is

recommended instead of applying nominal horizontal loads. For grandstands which

may be subject to synchronised and periodic crowd movements, the easiest

approach would be to estimate the vertical and horizontal natural frequencies and to

ensure avoidance of significant resonance effects. Ji and Ellis (1997) recommend a

vertical frequency greater than 8.4 Hz and the horizontal frequencies greater than

4.0 Hz for an empty structure to avoid resonance effects. If it is not possible to

avoid the resonance effect in this way, design of the grandstand will require a

detailed analysis to assess the effects of dynamic loads arising from the anticipated

resonance effects. Where resonance is unlikely, the use of a nominal horizontal

load approach as per the recommendations of IStructE (2007) can be used. This

approach will be discussed in the Section 4.




4 Notional Horizontal Load Approach

4.1 IStructE (2007) introduces the concept of using notional horizontal loads to account

for, spectator action, and geometrical imperfections of frames (e.g. the lack of

alignment of vertical members). The notional horizontal loads are taken as a

percentage of the imposed vertical load. The notional horizontal loads should be

applied in combination with the wind loads. IStructE (2007) recommends a

simplified approach to include the following notional horizontal load (Table 3) to

design the grandstand for the effect induced by different categories of spectator

action.

Table 3 Notional Horizontal Load for

Different Categories of Spectator Action

Category Spectator activity

Notional

horizontal

load

1

Nominal potential for spectator movement, which

excludes synchronized and periodic crowd

movement. Examples include: lectures/exhibitions,

display/shows, athletic events, golf tournaments and

agricultural shows.

6%

2

Potential for spectator movement more vigorous

than Category 1. Category 2 excludes synchronized

and periodic crowd movement in major musical

concerts, and rugby or football matches.

7.5%

3

Stands with a potential for synchronized and

periodic crowd movement and having vertical and

horizontal fundamental frequencies which avoid

resonance effects. An example is at pop concerts

where strong musical beats are expected.

10%

Notes: For notional horizontal loads, the partial factor should be 1.5 for the load

combination case with factored values of vertical dead and imposed loads.

(Source: IStructE 2007)

4.2 Besides the recommendations in IStructE (2007), the HK Loading Code has

recently introduced a requirement on horizontal imposed loads acting on the

grandstands due to crowd movement. Clause 3.8.2 of the HK Loading Code states

that grandstands, stadiums, assembly platforms, reviewing stands and similar, shall

be designed to withstand minimum horizontal imposed loads due to crowd

movement as follows:

(a) for platforms with seats, the following separate load cases (not applied

simultaneously), applied at floor level at each row of seats:

(i) 0.35 kN/m of seating along the line of seats; or

(ii) 0.15 kN/m of seating perpendicular to the line of the seats.

(b) for platforms without seats, 0.25 kPa of plan area applied in any direction.




4.3 Usually, there will be about 4 rows of seats per 3m width of the structure, and the

notional horizontal loads can therefore be expressed as an equivalent percentage

of the imposed load as follows:

(i) along the line of seats = 4×0.35/3 = 0.47 kPa = 9.3% of the imposed load (5

kPa);

(ii) perpendicular to the line of seats = 4×0.15/3 = 0.2 kPa = 4.0% of the

imposed load (5 kPa).

Therefore, the HK Loading Code has specified a smaller notional horizontal load

(i.e. 4%) perpendicular to the line of seats than that specified in IStructE (2007) (i.e.

6%) for nominal potential of spectator movement, but a larger notional horizontal

load (9.3%) along the line of seats. It should be further noted that for grandstands

without seats, the notional horizontal load is 0.25kPa in any direction, which is

smaller than that along the line of seats for grandstands with seats. However,

where there are no seats, the audiences are expected to erect a larger horizontal

force on the grandstand due to the larger possibility of crowd movement. Moreover,

in the HK Loading Code, the notional horizontal load is set without due

consideration of the different responses of the audiences for difference types of

events, e.g. lectures, football matches or pop concerts. Therefore, project officer is

advised to refer to IStructE (2007), which is more reasonable, and exercise

judgment in choosing the notional horizontal load to suit the particular case,

especially the probability of synchronized crowd movement.




5. Foundation

5.1 Demountable structures are generally lightweight and loaded for relatively short

periods. Hence, the bearing pressures and settlement of the ground are not usually

a problem, unless the structure will be in use for a long period, in which case a full

engineering assessment of the ground should be made. The use of permanent

foundations (e.g. pad footings) will be unlikely. Temporary pad steel bases will

have to provide the reactions to the applied loadings.

5.2 In addition to the dead and imposed load, there may be uplift forces and lateral

loads on the foundation due to its lightweight. It is therefore necessary to check the

methods of transferring such loads to the ground, usually using ground anchors by

anchor bolts at the support points (Photo 9). Occasionally, ground anchors cannot

be used because of the nature of the ground. For example, it may not be permissible

to puncture asphalt or concrete finishes. The structure should then be designed to

accept kentledge of sufficient weight to resist the factored uplift forces.

Photo 9 Use of Ground Anchor at the Support of the Standard

6. Construction Supervision

6.1 Usually, demountable stands are proprietary products designed, supplied and

installed by specialist contractor employed by the event organizers. The event

organizers may seek technical assistance from our Department to provide technical

advice on the design which has been prepared and submitted by the Registered

Structural Engineer employed by the specialist contractor. However, project officer

should note that since these structures are often quickly erected immediately before

an event, there may only be a short period of time available for the Registered

Structural Engineer employed by the specialist contractor to check the design prior

to submission and to inspect the workmanship prior to use.

6.2 Another critical stage for a demountable grandstand is during construction and

dismantle works, where bracing members have not yet been provided or are being

removed respectively. In such stage, stability is only provided by the moment

carrying capacity of the connections. Project officer should ensure that sufficient

bracing members are still provided to ensure their stability.




7. Case Study – East Asian Games 2009

7.1 Background

7.1.1 Held once every four years, the East Asian Games is one of the major events in the

Asian international sports arena. In 2009, Hong Kong hosted the 5th

East Asian

Games (EAG). It was the first time for Hong Kong to host such a large-scale

multi-sport international event, and was an important milestone in holding

international sports event for Hong Kong. An opening ceremony was held on 5

December 2009 at the open space outside the Hong Kong Cultural Centre Piazza

facing the Victoria Harbour. In order to achieve the environmental-friendly and

economical objectives of the event, four temporary demountable grandstands of

different sizes were constructed to provide over 1,500 seats for the guests and

audiences for the opening ceremony. Photo 10 shows the front and side elevations

of the largest grandstand. The structural system including the design and analysis of

the largest demountable grandstand will be discussed in the following paragraphs.

Photo 10 Largest Demountable Grandstand at EAG Opening Ceremony

7.1.2 As such demountable grandstands were proprietary products, the procurement

arrangement was that all demountable grandstands were design-and-build items

designed, erected and dismantled by a specialist contractor employed by the

organizer. The specialist contractor was required to engage a Registered Structural

Engineer (RSE) to prepare the drawings and structural calculation for the

grandstands. The RSE was also required to provide site supervision and to check

that the member sizes, spacing, arrangement of bracings, support details of the

structure had been constructed in accordance with the design drawings and made

necessary arrangement to solve the site problems, e.g. discrepancies between the

site conditions and the design drawings. Upon the completion of the erection of the

grandstands, the RSE was required to certify their safety by issuing a checking

certificate. Our Department was requested to provide technical support to the

organizer on the structural stability of the design, erection and dismantle works.

The specialist contractor was given about one-week time to erect the grandstands,

though the design had been submitted to our Department for comment.




7.2 Structural Layout and Member Capacities

7.2.1 The largest grandstand was of size 28×31m on plan with maximum height of about

10m above the ground level. It provided 1,572 seats for the audiences of the

opening ceremony. The grandstand was assembled by use of portable tiered system

for the seating and proprietary Layher scaffold for the supporting structure. The

Layher scaffold was an allround scaffold system. The standards and ledgers of the

scaffold were circular hollow section of size 49×3.2mm and the bracings of the

scaffold were circular hollow section of size 45×2.3mm. The catalogue of the

supplier contains the safe working load (Table 4 and Figure 11) on them, which

depending the bracing arrangement, their lengths and their bay width.

Table 4 Safe Working Load for Layher Scaffold (Model: Layher Variant II)

Member

Type Member Size

Safe Working

Load (kN) Remark

Standard 49×3.2mm 37.2

(Compression)*

The member capacity provided in

the supplier’s catalogue is the safe

working load for compression.

Member capacity based on the

condition that:

- length of standard = 2m

- one diagonal brace per 3 bays of

standard (i.e. B diagonal brace)

Ledger 49×3.2mm 15.1

(= 22.7/1.5)

The loading provided in the

supplier’s catalogue is the ultimate

capacity. The safe working loads

are obtained by dividing the

ultimate capacity by 1.5.

Safe working loads of ledger and

bracing are lower than that of the

standard as they are controlled by

the connection joint capacity with

the standard

Bracing 45×2.3mm 5.6

(= 8.4/1.5)

Notes: * Tensile capacity of standard is controlled by the connection capacity of

the pins ( = 3/8” approx.) between the upper and lower standards and

the connection tube (see Figure 3).

It should be noted that the safe working load is controlled by the connection joint

capacity between the members and the computer program is difficult to model the

connection joint capacity. Therefore, the safe working load in supplier’s catalogue

instead of that obtained from the computer program will be used in the design.




(a) For Standard

(b) For Ledger and Bracing

Figure 11 Members Capacity for Model Layher Variant II (Source: http://www.layher.com.au)

7.2.2 The standards were spaced at about 1.5m along the line of seats and about 3m

perpendicular to the line of the seats. The ledgers were spaced at about 2m in the

vertical direction. Type I bracings (as defined in Table 2 of section 2.4) were

added to provide the lateral stability of the structure. Figure 12 shows the

structural arrangement of the members of the demountable grandstand.

http://www.layher.com.au/




Figure 12 Typical Section of the Largest Demountable Grandstand at

2009 East Asian Games Opening Ceremony

7.2.2 The standards of the demountable grandstand were either rested on the surface of

the existing structure or ground through pad steel bases in order to provide the

reactions to the applied loads. Photo 11 shows the typical details of the bases of

the standard on the supporting ground.

Photo 11 Typical Base of the Standard

7.3 Structural Analysis and Design by the Specialist Contractor

7.3.1 The RSE employed software SPACE GASS (version 10.50b) in the analysis of the

forces in the scaffold, where the connections between ledger and standard and the

connections between bracing and standard were assumed to pin joints, and due to

symmetry, two 2-D models for respectively along and perpendicular to the line of

seats directions were adopted.

Bracing members to

provide lateral stability Ledger

Standard




7.3.2 In the design submission, the dead load was assumed to be 1 kPa. For the imposed

load, there would altogether 1,572 seats in the largest grandstand of plan size

28m×31m, and the average weight of an audience was 75kgf. Hence, the actual

imposed load on this grandstand would be 1.35kPa. The RSE had therefore taken

the design imposed load as 4 kPa in accordance with the BS 6399: Part 1 (BSI

1997) and Building (Construction) Regulations Table 1 for the usage of assembly

area with fixed seating. Two separate notional horizontal loads (NHL1 and NHL2)

had been considered in the analysis. NHL1 was taken as 1 kPa in accordance with

Table 2.2 of the HK Steel Code, while NHL2 was taken as 6% of the imposed load

in accordance with the recommendations in IStructE (2007).

7.3.3 Wind loads was taken to be the equivalent wind pressure at tropical cyclone

warning signal no.3 and was calculated as 0.41 kPa. However, the wind loads were

assumed to be not to control the design and ignored in the analysis, as the RSE

noted that the notional horizontal force (NHL) due to the dynamic loads would

exceed the wind loads. In addition, the RSE had imposed a restriction that the

occupancy shall be vacant and structures shall be fenced off or dismantled when

typhoon signal of No. 3 or above is hoisted. Therefore, the design case of wind

loads acting together with the notional horizontal force was not required.

Altogether three loading cases have been considered, including DL+LL,

DL+LL+NHL1 and DL+LL+NHL2 for each of the two 2-D models along and

perpendicular to the line of seats.

7.3.4 The design had been carried out by checking the working load against the safe

working load of each member type as provided in the catalogue of the Layher

system (Table 4). The member forces of the steel members were obtained by using

the first order analysis method in the HK Steel Code. The summary of the

maximum member force against the load bearing capacity for each member type is

shown in Table 5. The location of the critical member of standard (i.e. the member

with the highest utilization ratio) is shown in Figure 13.

Notional

horizontal load,

NHL1 = 1 kPa

Notional

horizontal load,

NHL2 = 6% LL




Table 5 Summary of Maximum Member Forces

Member Type Maximum Axial

Force (kN)

Safe Working

Load (kN)

Utilization

Ratio

Standard 33.44 37.20 0.90

Ledger 7.87 15.10 0.52

Bracing 5.52 5.60 0.99

Double Bracing 7.08 11.20 0.63

Figure 13 Critical Members in the 2-D Model

(showing the tallest frame along the line of seats direction)

7.3.5 With the adopted model, the maximum utilization factor was 0.986 at one of the

bracing members. No dynamic analysis was carried out, and notational horizontal

force approach was adopted to cater for the dynamic effect. To cater for the

uncertainty and to guard against any lateral movement, altogether 6 pairs of Type 6

global bracing members (as defined in Table 2 of section 2.4) (Figure 14) were

added at the request of our Department to increase the stiffness of the grandstands

to lateral loads.

Standard

(Maximum axial

force = 33.44 kN,

section utilization

= 0.90)




(a) Perpendicular to the Line of Seats Direction

(b) Along the Line of Seats Direction

Figure 14 Added Global Bracing Members

7.4 Construction Supervision

7.4.1 Due to the tight schedule for erection, some of the connections were not completed

according to the approved drawings, and some bracing members were missing or

were not connected to the nodes between the standard and the ledger (Photo 12(a)

and (b)), and the global bracing members were not provided (Photo 12(c)).

Fortunately, such discrepancies were detected during inspection, and additional

labour and materials were deployed to rectify the deficiencies. The rectification and

final inspection were completed in time such that the checking certificate (at Annex

A) to be issued prior to the opening ceremony. Photo 12(d) shows the grandstand

with all works substantially completed.




Photo 12(a) Misalignment of the Bracing Member with the Nodes

Photo 12(b) Misalignment of the Bracing Member with the Nodes




Photo 12(c) Grandstand with global bracing members not yet installed and with

many defects to be rectified (taken on 2.12.2009)

Photo 12(d) Grandstand Substantially Completed (taken on 4.12.2009)

7.5 Lessons Learnt

7.5.1 The opening ceremony was successfully held on 5 December 2009 evening, and the

grandstands served their function to provide temporary seating for over 1,500

audiences. The scaffold system could carry the design imposed load, and no

significant vibration was noted by the audiences. Despite these facts, this paper

would summarise the experience in providing technical support to their design,

erection and construction supervision in the following paragraphs.




7.5.2 Analysis

7.5.2.1 In the submission from of the specialist contractor, the connections between

ledger and standard and the connections between bracing and standard were

assumed to be pinned joints for simplicity. The elastic critical load factor λcr had

not been checked, and the analysis and design was carried out using 1st order

analysis with no due consideration of the 2nd order (P-) effect. Moreover, 2-D

models were used, and hence the out-of-plane buckling mode was not included in

the analysis. Yet, with our subsequent re-calculations (which are discussed in the

following) the design by the RSE is generally in order.

7.5.2.2 Instead of the 2-D model used by the RSE, we have carried out a 3-D analysis of

the same temporary grandstand by using the same design loading. When

compared with the analysis results of the 3-D model with the 2-D model, it is

found that the location of the critical member of standard (i.e. the standard with

the highest utilization ratio) will be shifted to another member due to the

redistribution of loadings between the members in the 3-D model. However, the

value of the maximum compression force obtained from the 3-D model is

33.68kN which is almost the same as that obtained (33.44kN) from 2-D model

under the same loading case of DL+LL+NHL2 (where LL = 4kPa and NHL2 =

6% LL). This shows that the adoption of 2-D model to obtain the design forces is

generally satisfactory for this temporary grandstand. However, project officer

should note that there may be difference in the analysis results between the 2-D

and 3-D models. The location of the critical member of standard in the 3-D model

is shown in Figure 15.




Figure 15 Critical Members in the 3-D Model

7.5.2.3 We have also carried out an alternative structural analysis of the same temporary

grandstand with a 3-D computer model by using the minimum imposed load of

5.0kPa as stated in the HK Loading Code (though the actual imposed load on this

grandstand being 1.35kPa) to see if there are the difference between the results

from two different analysis and design methods. Moreover, the member forces

are also checked against the safe working loads as stated in the supplier’s

catalogue. The major differences in the analysis and design method between the

RSE’s submission and our subsequent calculations in the following sections are

shown in Table 6.

Standard

(Axial force = 33.68 kN)




Table 6 Model and parameters adopted in alternative analysis and design

Model and

Parameters

Design Submission from

Specialist Contractor

Alternative Analysis in

this Paper

Computer Model 2-D 3-D

Design Method

Working load against safe

working load in supplier’s

catalogue

Working load against safe

working load in supplier’s

catalogue

Design Imposed

Load 4 kPa 5 kPa

Design Notional

Horizontal Load

6% of the design imposed

load (from IStructE 2007)

9.3% of the design imposed

load along the line of seats;

and

6.0% of the design imposed

load perpendicular to the

line of seats

(from Section 4.3)

Design Wind Load

0.41 kPa

(the equivalent wind load

at tropical cyclone warning

signal no. 3)

0.41 kPa (from Section 3.4)

7.5.2.4 A 3-D computer model of the grandstand is shown in Figure 16. Sections of the

computer model are shown in Figure 17. Structural analysis has then been

carried out to evaluate the structural adequacy of the scaffold system with the use

of computer software SAP2000.

Figure 16 3-D Model of the Demountable Grandstand




(a) Perpendicular to the Line of Seats Direction

(b) Along the Line of Seats Direction

Figure 17 Typical Sections of the Demountable Grandstand

7.5.2.5 Section 2.6 states that scaffold system for the demountable grandstands has either

column bucking mode or sway buckling mode depending on the bracing and the

axial load on the standard, and that the lowest eigenvalue found by the eigenvalue

analysis may not represent the first sway buckling mode λcr. In the present case,

the maximum deflection of the grandstand is found to be 8.26mm, and the /H

ratio is 1/1370, and it shows that the scaffold system has sufficient bracing

members to become braced frame. Hence, the dominant buckling mode in the

eigenvalue analysis is the local column buckling mode rather than sway buckling

mode.

7.5.2.6 Both the deflection method and eigenvalue analysis were used to calculate the

elastic critical load factor, λcr, of the structure under each load combination. The

eigenvalue computed using the deflection method is 167, which is much greater

than 10 while the eigenvalue analysis gives smallest eigenvalue of 1.997. From

the deflected shape (shown in Figure 18), the mode of buckling with eigenvalue

of 1.997 is a local column buckling mode. Hence, the eigenvalue of 1.997

obtained from the eigenvalue analysis is due to the large axial load on the

members rather than the deflection due to the horizontal load, and the elastic




critical load factor, λcr, of the structure should be 167. The assumption in the

original submitted design that 2nd order analysis is not required is therefore

justified.

Figure 18 Deflected shape of the scaffold system with eigenvalue of 1.997

7.5.3 Structural Design

7.5.3.1 In the 3-D model design, the dead load was assumed to be 1kPa, and the imposed

load had been taken as 5kPa in accordance with the HK Loading Code for

grandstand, though the actual imposed load on this grandstand was only 1.35kPa.

Moreover, the following notional horizontal loads as stated in Section 7.5.2.3 had

been adopted at floor level at each row of seats:

(i) 9.3% of the design imposed load along the line of seats; or

(ii) 6.0% of the design imposed load perpendicular to the line of the seats.

The design wind pressure was taken as 0.41kPa. In addition, the force coefficient

taking into account the shielding effect should be taken as 0.8 for a solidity ratio

of 0.04 for 16 nos. of frames along the line of seats and 0.7 for a solidity ratio of

0.06 for 11 nos. of frames perpendicular to the line of seats.




7.5.3.2 This paragraph intends to compare the difference in the results from the original

design with those adopted values in the above paragraphs. The structure has been

checked and designed by 1st order analysis with the loadings given in Section

7.5.2.3. Altogether 12 nos. of load combination have been considered. The

section utilization ratios of the members of the structure under each load

combination are shown in Table 8. Load cases 3 and 4 are the two most critical

load combinations resulting in the maximum section utilization ratios of 1.26 for

the standards, 0.41 for the ledgers and 1.24 for the bracings. The utilization ratios

are obtained from comparing the member forces obtained from the computer

program SAP2000 with the safe working load in supplier’s catalogue. As stated in

Section 7.2.1, the safe working load in supplier’s catalogue has been used for

comparison instead of that obtained from the computer program. The critical

members (i.e. members with highest utilization ratios) are shown in Figure 19.

7.5.3.3 From the analysis results, it can be observed that the maximum section utilization

ratio of standard and bracing had increased significantly from 0.90 to 1.26 (40%

increase) and from 0.63 to 1.24 (97% increase) respectively with the adoption of

the loadings given in Section 7.5.2. Under the critical load combination of DL +

LL + NHL(±Y) for the standard and bracing, the changes in member forces of the

critical standard and bracing member are shown in Table 7 for easy reference.

Table 7 Comparison of member force obtained from alternative analysis

Model and Parameters Member

Member Force (kN) Section

Utilization DL + LL NHL(Y) DL + LL

+ NHL(Y)

Design Imposed Load:

4 kPa

Design Notional

Horizontal Load:

6% of the design imposed

load (from IStructE 2007)

Standard 24.45 9.17 33.62 0.90

Bracing

(Double) 0.07 7.04 7.11 0.63

Design Imposed Load:

5 kPa

Design Notional

Horizontal Load:

9.3% of the design

imposed load (from HK

Loading Code)

Standard 28.99 17.74 46.73 1.26

Bracing

(Double) 0.07 13.76 13.83 1.24

7.5.3.4 When compared with the member forces between the two different design

loadings in Table 7, the member force of the standard due to DL + LL has only

increased by 18.6% (from 24.45kN to 28.99kN) but the member force due to

notional horizontal loads in the direction along the line of seats has increased

substantially by 93.4% (from 9.17kN to 17.74kN). For the bracing, the member

force due to DL + LL remains unchanged but the member force due to notional

horizontal loads in the direction along the line of seats has increased substantially




by 95.5% (from 7.04kN to 13.76kN). Therefore, the maximum section utilization

ratios of the standard and bracing have increased significantly from 0.90 to 1.26

(40% increase) and from 0.63 to 1.24 (97% increase) respectively mainly because

of the increase in the adoption of the larger notional horizontal load in the

direction along the line of seats as required by the new HK Loading Code.

7.5.3.5 However, it should be noted that the results are quite conservative because:

a) the imposed load has been taken as 5.0kPa although the actual imposed

load was 1.35kPa; and

b) the wind load has been taken as 0.41kPa, which is the highest wind speed

during the hoisting of typhoon signal no. 3, although the opening

ceremony took place in December 2009 and the RSE had specified that the

grandstand should not be occupied during typhoon signal no. 3 or above

was hoisted.

c) the design notional horizontal load along the line of seats is taken as 9.3%

of the design imposed load instead of 6% as recommended by IStructE

(2007).

Among these factors, it can be noted that the increase in the design notional

horizontal load will increase significantly the member force at the bracing. As

stated in Section 4.3, the notional horizontal load in the Hong Kong Loading

Code is larger than the nominal potential of spectator movement stated in IStructE

(2007) (which is more reasonable in the present opening ceremony), and hence

the inadequacy of the original design to cater for such large notional horizontal

force is expected.

Table 8 Maximum Section Utilization Ratio

Load

Case

Load Combination Section Utilization Ratio

Standard Ledger Bracing

1 DL + LL + NHL (+X) 0.83 0.31 0.56

2 DL + LL + NHL (-X) 0.85 0.27 0.43

3 DL + LL + NHL (+Y) 1.26 0.41 1.23

4 DL + LL + NHL (-Y) 1.28 0.41 1.24

5 DL + LL + WL(+X) + NHL(+X) 0.54 0.22 0.39

6 DL + LL + WL(-X) + NHL(-X) 0.56 0.20 0.36

7 DL + LL + WL(+Y) + NHL(+Y) 0.86 0.40 0.98

8 DL + LL + WL(-Y) + NHL(-Y) 0.86 0.40 0.99

9 DL + WL(+X) 0.12 0.05 0.08

10 DL + WL(-X) 0.13 0.04 0.11

11 DL + WL(+Y) 0.17 0.15 0.27

12 DL + WL(-Y) 0.15 0.15 0.27

where DL= Dead Load, LL = Imposed Load, WL = Wind Load,

NHL = Notional Horizontal Load




(a) Standard

(b) Ledger

Ledger

(Section utilization = 0.41)

Standard

(Section Utilization = 1.26)




(c) Bracing

Figure 19 Critical Members

7.5.4 Dynamic Effect

7.5.4.1 Section 3.5.3 notes that where resonance is unlikely, the use of a nominal

horizontal load approach can be used to include the dynamic effect due to crowd.

A modal analysis should therefore be carried out to find out the fundamental

frequency of the empty structure in order to eliminate the possibility of

resonance due to such synchronized movement, which is one of the most

common failure reasons of such demountable grandstand. The mode shapes of

the structure at fundamental frequency along and perpendicular the line of seats

directions are shown in Figure 20. The fundamental frequency of structure is

found to be 10Hz which is much higher than the recommended minimum

horizontal frequency of 4.0Hz as recommended by Ji and Ellis (1997) as

discussed in Section 4.4. The fundamental frequency of the structure is therefore

considered satisfactory and no further rigorous dynamic analysis is required.

Hence, a check of the fundamental frequency in both sway direction is

required to ensure that such demountable grandstands will not be

susceptible to the resonance due to the dynamic load by spectators’

movement.

Bracing

(Section utilization = 1.24)




7.5.4.2 The added Type 6 global bracing elements have also been included in the

dynamic analysis to investigate their effects, and the results are shown in Table

9. It can be seen that the Type 6 global bracing elements have substantially

improved the stiffness of the scaffold systems, confirming the conclusion from Ji

and Ellis (2001).

(a) Fundamental mode in perpendicular to the line of seats direction

(Frequency = 15 Hz)

(b) Fundamental mode in along the line of seats direction (Frequency = 10 Hz)

Figure 20 Fundamental Mode Shapes of the Grandstands




Table 9 Comparison of Fundamental Frequency With and Without

Type 6 Global Bracing Members

Direction Mode Shape Frequency

Perpendicular

to the line of

seats

15 Hz

20 Hz

Along the

line of seats

10 Hz

12 Hz




7.5.5 Construction

7.5.5.1 It has been checked that all the standards are under compression in the working

load condition with the minimum compression force of 0.66kN under all load

combinations (Figure 21), and hence no ground anchors or kentledge are required

to tie the standard onto the ground. Project officer should note that when there

are any tensile force found in the standards from the structural analysis, it is

necessary to check on site to ensure that all pins are installed to connect

upper with lower standards and that the tensile force due to wind or lateral

loads can be transmitted throughout the allround scaffold by using the pins.

Figure 21 Member with the Minimum Compression Force

7.5.5.2 The RSE should be required to check the erection at the earliest moment,

and should station himself or his supervisory staff on site when the erection

of scaffold commences, such that any deviation from the approved drawings

should be rectified immediately. Sufficient labour should be deployed by the

specialist contractor to carry out the rectification works, and the programme

of works, though tight, should allow the inspection by the RSE.

Standard

Minimum compression = 0.66kN




8 References

Blackmore, P (2004), BRE Special Digest SD 5: Wind Loads on Unclad Structures

(Watford: BRE) (available: www.bre.co.uk, accessed: 24 October 2011).

Bolton, A (1992), “Fatal Mix Caused Stand Fall”, New Civil Engineer, 14 May, pp. 5-6.

Bolton, A (1997), “Correspondence: Effective Bracing Systems for Temporary

Grandstands”, The Structural Engineer, 75(22), pp. 389-9 (available: www.istructe.org,

accessed: 4 November 2011).

BSI (1996), BS 6399: Part 1: 1996 – Loading for Buildings – Code of Practice for Dead

and Imposed Loads (London: BSI).

BSI (1997), BS 6399: Part 2: 1997 – Loading for Buildings – Code of Practice for Wind

Loads (London: BSI).

BSI (2005), BS EN 1993 1-1:2005: Eurocode 3 - Design Of Steel Structures - Part 1-1:

General Rules and Rules for Buildings (London: BSI).

Buildings Department (2004), Code of Practice on Wind Effects in Hong Kong 2004 (Hong

Kong: Buildings Department) (available: www.bd.gov.hk, accessed: 1 September 2011).

Buildings Department (2005), Code of Practice for the Structural Use of Steel 2005 (Hong


Buildings Department (2011), Code of Practice for Dead and Imposed Loads 2011 (Hong


Chandrangsu, T and Rasmussen, K (2006), “Structural Modelling of Support Scaffold

Systems”, Journal of Constructional Steel Research, 67, pp. 866–75.

Choi, E C C (1984), Wind Loading in Hong Kong: Commentary on the Code of Practice on

Wind Effects Hong Kong 1983 (Hong Kong: Hong Kong Institution of Engineers).

Crick, D and Grondin, G Y (2008), Structural Engineering Report No. 275: Monitoring

and Analysis of a Temporary Grandstands (Edmonton, Alberta: Department of Civil and

Environmental Engineering, University of Alberta) (available: www.uofaweb.ualberta.ca,


De Brito, V L and Pimentel, RL (2009), “Cases of Collapse of Demountable Grandstands”,

Journal of Performance of Constructed Facilities, 23, pp. 151-9(available: www.asce.org,


Dickie, J F (1983), “Demountable Grandstand”, The Structural Engineer, 61A(3), pp. 81-6

(available: www.istructe.org, accessed: 4 November 2011).

Ellis, B R and Ji, T (1997), “Effective Bracing Systems for Temporary Grandstands”, The

Structural Engineer, 75(6), pp. 95-100 (available: www.istructe.org, accessed: 4 November

2011).

Enright J, Harriss R and Hancock G J (2000), “Structural Stability of Braced Scaffolding

and Formwork with Spigot Joints”, Proceedings of the Fifteenth International Specialty

Conference on Cold-Formed Steel Structures, pp. 357–76.

Godley, M H R and Beale R G (1997), “Sway Stiffness of Scaffold Structures”, The

Structural Engineer, 75(1), pp. 4–12 (available: www.istructe.org, accessed: 4 November

2011).

http://www.bre.co.uk/

http://www.istructe.org/

http://www.bd.gov.hk/



http://www.uofaweb.ualberta.ca/

http://www.asce.org/







Godley, M H R and Beale R G (2001), “Analysis of Large Proprietary Access Scaffold

Structures”, Structures and Buildings, 146(1), pp. 31–9.

Grant, M (1975), Scaffold Falsework Design to BS 5975 (London: E and F N Spon).

Harung, H S, Lightfoot, E and Duggan, D M (1975), “The Strength of Scaffold Towers

under Vertical Loading”, The Structural Engineer, 53(1), pp. 23-30 (available:

www.istructe.org, accessed: 4 November 2011).

IStructE (2007), Temporary Demountable Structures: Guidance on Procurement, Design

and Use (London: The Institution of Structural Engineers, 3rd

ed).

Ji, T (2003), “Concepts for Designing Stiffer Structures”, The Structural Engineer, 81(21),

pp. 36-42 (available: www.istructe.org, accessed: 12 December 2011).

King, C (2005), NCCI: Calculation of Alpha-cr (Ascot: SCI) (available: www.steelbiz.org,


Labour Department (2001), Code of Practice for Metal Scaffold Safety (Hong Kong:

Labour Department) (available: www.labour.gov.hk, accessed: 4 November 2011).

Littler, J D (1996), “Measuring the Dynamic Response of Temporary Grandstands”,

Proceeding of Structural Dynamics Eurodyn 1996, pp. 907-13.

Peng J L, Chan S L and Wu C L (2007), “Effects of Geometrical Shape and Incremental

Loads on Scaffold Systems”, Journal of Constructional Steel Research, 63, pp. 448–59.

Rasmussen, K and Chandrangsu, T (2009), Research Report No. R905: Review of Past

Research on Scaffold Systems (Sydney: The University of Sydney) (available:

www.civil.usyd.edu.au, accessed: 31 October 2011).

Rathbone, A J (2002), “Second-order Effects – Who Needs Them?” The Structural

Engineer, 80(21), pp. 19-21 (available: www.istructe.org, accessed: 4 November 2011).

Yu, W K (2004), “An Investigation into Structural Behaviour of Modular Steel Scaffolds”,

Steel and Composite Structures, 4(3), pp. 211-26.

Yu, W K, Chung, K F and Chan, S L (2004), “Structural Instability of Multi-storey Door-

type Modular Steel Scaffolds”, Engineering Structures, 26, pp. 867–81.



http://www.steelbiz.org/

http://www.labour.gov.hk/

http://www.civil.usyd.edu.au/





Annex A

Sample Checking Certificate




A Case Study of Using Metal Scaffold System for Demountable Grandstand: The Opening Ceremony of Hong Kong 2009 East Asian Games

Documents

demountable grandstand

b demountable grandstand

demountable stands photo

permanent grandstands

archsd information paper

structural behaviour

file code

olympics games source